Datasets:

Infi-MM
/

InfiMM-WebMath-40B

Dataset card Data Studio Files Files and versions Community

Split (1)

train · ~22.8M rows (showing the first 514k)

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.

URL stringlengths 15 1.68k	text_list sequencelengths 1 199	image_list sequencelengths 1 199	metadata stringlengths 1.19k 3.08k
https://yellow-erp.com/page/guides/practical-dev-guide83/lesson_17/understanding_calculation_registers/	[ "# 1C:Enterprise 8.3. Practical Developer’s Guide. Lesson 17 (1:00). Charts of calculation types and calculation registers. Understanding calculation registers\n\n## Understanding calculation registers\n\nThe Calculation register configuration object is intended to describe the structure of accumulated data resulting from calculations. Based on the Calculation register configuration object, the platform creates database tables that store accumulated data generated by various database objects.\n\nCalculation registers are not designed to be edited directly by users. A developer can allow users to edit a calculation register, if required. However, normally calculation registers are modified by algorithms of other database objects, rather than through a direct user intervention.\n\nJust like any other register, a calculation register has resources where it stores numerical data, it has dimensions used for getting register resource values by those dimensions, and it has attributes, which describe each calculation register record.\n\nWhat makes calculation registers different is their periodicity, the ability to use the algorithms of displacement by action period and dependency by base period, and the connection with a chart of calculation types. We will discuss each of these features in the following sections.\n\n### Periodicity\n\nThe periodicity of a calculation register can have one of the following values:\n\n• Day\n• Month\n• Quarter\n• Year\n\nThe calculation register periodicity defines the time period where each register record belongs.\n\nIf the periodicity is set to Day, each record in the register belongs to some day; if periodicity is set to Month, each register record belongs to some month, and so on.\n\nTo reflect the fact that a record belongs to a certain period, the register has the RegistrationPeriod system attribute of the Date type. When writing data to the register, the platform always sets the value of that attribute to the beginning of the period where the data belongs.\n\nFor example, if data is written to a calculation register with periodicity set to Month, and the RegistrationPeriod for that data is 4/8/2014, the register stores the data with the RegistrationPeriod value set to 4/1/2014 (fig. 17.6).", null, "Fig. 17.6. Recording document data to a calculation register\n\nIf you set the register periodicity to Year in this scenario, the stored registration period value changes to 1/1/2014 (fig. 17.7).", null, "Fig. 17.7. Recording document data to a calculation register\n\n### Displacement by action period\n\nAnother important feature of a calculation register is the ability to displace certain records with other records by action period.\n\nIn doing so, each calculation register record generates an actual action period which, in general, represents the sum total of several periods contained within a single action period (fig. 17.8).", null, "Fig. 17.8. Absence calculation record displaces a Salary calculation record by action period\n\nAfter adding a salary accrual record, the structure of calculation register table records looks as shown in tables 17.1 and 17.2.\n\nTable 17.1. Calculation register table\n\n ... Beginning of action period End of action period Calculation type ... ... ... ... ... ... ... 04/01/2014 00:00:00 04/30/2014 23:59:59 Salary ... ... ... ... ... ...\n\nTable 17.2. Actual action period table\n\n ... Beginning of action period End of action period Calculation type ... ... ... ... ... ... ... 04/01/2014 00:00:00 04/30/2014 23:59:59 Salary ... ... ... ... ... ...\n\nAfter adding records from the Absence calculation type, which displaces Salary calculations by action period, the records related to salary calculation look as shown in tables 17.3 and 17.4.\n\nTable 17.3. Calculation register table\n\n ... Beginning of action period End of action period Calculation type ... ... ... ... ... ... ... 04/01/2014 00:00:00 04/30/2014 23:59:59 Salary ... ... 04/04/2014 00:00:00 04/10/2014 23:59:59 Absence ... ... ... ... ... ...\n\nTable 17.4. Actual action period table\n\n ... Beginning of action period End of action period Calculation type ... ... ... ... ... ... ... 04/01/2014 00:00:00 04/03/2014 23:59:59 Salary ... ... 04/11/2014 00:00:00 04/31/2014 23:59:59 Salary ... ... ... ... ... ...\n\n### Dependency by base period\n\nAnother calculation register feature is support of register record dependency by base period. This feature is intended for calculating dependent (secondary) register records based on the result of primary record calculation.\n\nA calculation register can support two types of dependency by base period: dependency by action period and dependency by registration period.\n\n#### Dependency by action period\n\nDependency by action period means that when the platform analyzes base records, it selects the records where the actual action period and the specified base period overlap.\n\nFor example, at the beginning of April, March payroll is calculated. Bonuses for March are calculated based on March salaries. In this scenario, as a rule, developers use dependency by action period (fig. 17.9).", null, "Fig. 17.9. Dependency by action period\n\nThe Beginning of base period and End of base period fields are applicable only for the records with calculation types that have dependencies by base period specified (in this case, for calculation of bonuses).\n\nThe base value received from a specific record that influences a calculation generally is not equal to the result stored within that record. The base value is calculated based on the ratio between the part of the actual period of the influencing record that overlaps the base period and the full actual period of the influencing record, and also on the schedule data linked to record.\n\n#### Dependency by registration period\n\nDependency by registration period means that when the platform analyzes base records, it selects the records whose Registration period field value falls within the specified base period.\n\nLet us consider an example of penalty calculation for March salary accruals. Absence records registered in March serve as a base for calculating the penalty amounts (these can include March absence records as well as February absence records. In this scenario, as a rule, developers use dependency by registration period (fig. 17.10).", null, "Fig. 17.10. Dependency by registration period\n\nThe final important feature of a calculation register is its connection with the chart of calculation types. This connection enables displacement by action period and dependency by base period, since a chart of calculation types describes the interrelationship between calculation types.\n\nA calculation register can have subordinate Recalculation objects. They are used to register the fact of adding register records that affect the calculation results of existing register records. A Recalculation configuration object can have multiple dimensions, each specifying a connection between the dimensions of this calculation register and the dimensions of calculation registers that influence it. In some scenarios it can be the same register.\n\nThe platform stores the list of register records that are subject to recalculations in database tables based on Recalculation configuration objects. Recalculation tables are populated automatically based on calculation register records that are affected by leading calculation types, as well as on calculation register records that have their actual periods changed. Based on this data, a developer can decide whether the register records require recalculation.\n\nThe last observation that we need to make regarding calculation registers concerns the ability to link a calculation register to a schedule. The schedule should be represented by an information register (nonperiodic, with a mandatory dimension of Date type, and a resource of Number type), which contains time-based source data used in the calculations. The dimensions of that schedule can be, for example, a work schedule (a reference to a catalog) and a date, and its resource can store the number of working hours for that date. In this scenario you can link a calculation register record to a specific work schedule (by specifying a reference to a work schedule catalog as a record attribute) and then use 1C:Enterprise script tools to get the number of working hours in the action period, actual period, or registration period for each record.\n\nLearn more! For details on the structure of 1C:Enterprise script objects intended for calculation register operations, see section Quick developer reference. Calculation registers." ]	[ null, "https://yellow-erp.com/uploads/images/guides/practicaldevguide83/pict_17-6.png", null, "https://yellow-erp.com/uploads/images/guides/practicaldevguide83/pict_17-7.png", null, "https://yellow-erp.com/uploads/images/guides/practicaldevguide83/pict_17-8.png", null, "https://yellow-erp.com/uploads/images/guides/practicaldevguide83/pict_17-9.png", null, "https://yellow-erp.com/uploads/images/guides/practicaldevguide83/pict_17-10.png", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.85948455,"math_prob":0.9848972,"size":8447,"snap":"2023-40-2023-50","text_gpt3_token_len":1801,"char_repetition_ratio":0.25962335,"word_repetition_ratio":0.13192818,"special_character_ratio":0.238783,"punctuation_ratio":0.20927395,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9525796,"pos_list":[0,1,2,3,4,5,6,7,8,9,10],"im_url_duplicate_count":[null,5,null,5,null,5,null,5,null,5,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2023-12-08T21:36:42Z\",\"WARC-Record-ID\":\"<urn:uuid:53b89d2b-1427-413b-8e0c-4b70499c12b1>\",\"Content-Length\":\"64393\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:3e559bb6-8c39-43ef-a0fe-e95618927028>\",\"WARC-Concurrent-To\":\"<urn:uuid:5a247ae9-9fdf-4267-aa55-390b26b25047>\",\"WARC-IP-Address\":\"195.35.34.9\",\"WARC-Target-URI\":\"https://yellow-erp.com/page/guides/practical-dev-guide83/lesson_17/understanding_calculation_registers/\",\"WARC-Payload-Digest\":\"sha1:ZAQUXXVB4UEVED6Z636O4SIEM2I3NTDN\",\"WARC-Block-Digest\":\"sha1:DMFQVUDYTMKXXUZEIARF4CR2J2XP4MYN\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2023/CC-MAIN-2023-50/CC-MAIN-2023-50_segments_1700679100779.51_warc_CC-MAIN-20231208212357-20231209002357-00588.warc.gz\"}"}
https://nctoolkit.readthedocs.io/en/v0.4.5/api.html	[ "# API Reference¶\n\n## Session options¶\n\n `options`(\\\\kwargs) Define session options.\n\n## Opening/copying data¶\n\n `open_data`([x, checks]) Read netCDF data as a DataSet object `open_url`([x, ftp_details, wait, file_stop]) Read netCDF data from a url as a DataSet object `open_thredds`([x, wait, checks]) Read thredds data as a DataSet object Read geotiff and convert to nctoolkit dataset Convert an xarray dataset to an nctoolkit dataset This will first save the xarray dataset as a temporary netCDF file. `DataSet.copy`(self) Make a deep copy of an DataSet object Note: This will not make disk copies of the temporary files underlying datasets, so it will be disk-space efficient.\n\n## Merging or analyzing multiple datasets¶\n\n `merge`(\\datasets[, match]) Merge datasets `cor_time`([x, y]) Calculate the temporal correlation coefficient between two datasets This will calculate the temporal correlation coefficient, for each time step, between two datasets. `cor_space`([x, y]) Calculate the spatial correlation coefficient between two datasets This will calculate the spatial correlation coefficient, for each time step, between two datasets.\n\n## Accessing attributes¶\n\n `DataSet.variables` List variables contained in a dataset `DataSet.contents` Detailed list of variables contained in a dataset. `DataSet.times` List times contained in a dataset `DataSet.years` List years contained in a dataset `DataSet.months` List months contained in a dataset `DataSet.levels` List levels contained in a dataset `DataSet.size` The size of an object This will print the number of files, total size, and smallest and largest files in an DataSet object. `DataSet.current` The current file or files in the DataSet object `DataSet.history` The history of operations on the DataSet `DataSet.start` The starting file or files of the DataSet object `DataSet.calendar` List calendars of dataset files `DataSet.ncformat` List formats of files contained in a dataset\n\n## Plotting¶\n\n `DataSet.plot`(self[, vars, autoscale, out])\n\n## Variable modification¶\n\n `DataSet.assign`(self[, drop]) Create new variables Existing columns that are re-assigned will be overwritten. :param drop: Set to True if you want existing variables to be removed once the new ones have been created. Defaults to False. `DataSet.rename`(self, newnames) Rename variables in a dataset `DataSet.set_missing`(self[, value]) Set the missing value for a single number or a range `DataSet.sum_all`(self[, drop]) Calculate the sum of all variables for each time step\n\n## netCDF file attribute modification¶\n\n `DataSet.set_longnames`(self[, name_dict]) Set the long names of variables `DataSet.set_units`(self[, unit_dict]) Set the units for variables\n\n## Vertical/level methods¶\n\n `DataSet.top`(self) Extract the top/surface level from a dataset This extracts the first vertical level from each file in a dataset. `DataSet.bottom`(self) Extract the bottom level from a dataset This extracts the bottom level from each netCDF file. `DataSet.vertical_interp`(self[, levels]) Verticaly interpolate a dataset based on given vertical levels This is calculated for each time step and grid cell `DataSet.vertical_mean`(self[, thickness, …]) Calculate the depth-averaged mean for each variable This is calculated for each time step and grid cell Calculate the vertical minimum of variable values This is calculated for each time step and grid cell Calculate the vertical maximum of variable values This is calculated for each time step and grid cell Calculate the vertical range of variable values This is calculated for each time step and grid cell Calculate the vertical sum of variable values This is calculated for each time step and grid cell `DataSet.vertical_integration`(self[, …]) Calculate the vertically integrated sum over the water column This calculates the sum of the variable multiplied by the cell thickness Calculate the vertical sum of variable values This is calculated for each time step and grid cell Invert the levels of 3D variables This is calculated for each time step and grid cell Create a mask identifying the deepest cell without missing values.\n\n## Rolling methods¶\n\n `DataSet.rolling_mean`(self[, window]) Calculate a rolling mean based on a window `DataSet.rolling_min`(self[, window]) Calculate a rolling minimum based on a window `DataSet.rolling_max`(self[, window]) Calculate a rolling maximum based on a window `DataSet.rolling_sum`(self[, window]) Calculate a rolling sum based on a window `DataSet.rolling_range`(self[, window]) Calculate a rolling range based on a window\n\n## Evaluation setting¶\n\n `DataSet.run`(self) Run all stored commands in a dataset\n\n## Ensemble creation¶\n\n `create_ensemble`([path, recursive]) Generate an ensemble\n\n## Arithemetic methods¶\n\n `DataSet.abs`(self) Method to get the absolute value of variables `DataSet.add`(self[, x, var]) Add to a dataset This will add a constant, another dataset or a netCDF file to the dataset. `DataSet.assign`(self[, drop]) Create new variables Existing columns that are re-assigned will be overwritten. :param drop: Set to True if you want existing variables to be removed once the new ones have been created. Defaults to False. `DataSet.exp`(self) Method to get the exponential of variables `DataSet.log`(self) Method to get the natural log of variables `DataSet.log10`(self) Method to get the base 10 log of variables `DataSet.multiply`(self[, x, var]) Multiply a dataset This will multiply a dataset by a constant, another dataset or a netCDF file. :param x: An int, float, single file dataset or netCDF file to multiply the dataset by. If multiplying by a dataset or single file there must only be a single variable in it, unless var is supplied. The grids must be the same. :type x: int, float, DataSet or netCDF file :param var: A variable in the x to multiply the dataset by :type var: str. `DataSet.power`(self[, x]) Powers of variables in dataset :param x: An int or float to take the variables to the power of :type x: int, float `DataSet.sqrt`(self) Method to get the square root of variables `DataSet.square`(self) Method to get the square of variables `DataSet.subtract`(self[, x, var]) Subtract from a dataset This will subtract a constant, another dataset or a netCDF file from the dataset. :param x: An int, float, single file dataset or netCDF file to subtract from the dataset. If a dataset or netCDF is supplied this must only have one variable, unless var is provided. The grids must be the same. :type x: int, float, DataSet or netCDF file :param var: A variable in the x to use for the operation :type var: str. `DataSet.divide`(self[, x, var]) Divide the data This will divide the dataset by a constant, another dataset or a netCDF file. :param x: An int, float, single file dataset or netCDF file to divide the dataset by. If a dataset or netCDF file is supplied, this must have only one variable, unless var is provided. The grids must be the same. :type x: int, float, DataSet or netCDF file :param var: A variable in the x to use for the operation :type var: str.\n\n## Ensemble statistics¶\n\n `DataSet.ensemble_mean`(self[, nco, ignore_time]) Calculate an ensemble mean `DataSet.ensemble_min`(self[, nco, ignore_time]) Calculate an ensemble min `DataSet.ensemble_max`(self[, nco, ignore_time]) Calculate an ensemble maximum `DataSet.ensemble_percentile`(self[, p]) Calculate an ensemble percentile This will calculate the percentles for each time step in the files. Calculate an ensemble range The range is calculated for each time step; for example, if each file in the ensemble has 12 months of data the statistic will be calculated for each month. Calculate an ensemble standard deviation Calculate an ensemble sum The sum is calculated for each time step; for example, if each file in the ensemble has 12 months of data the statistic will be calculated for each month. Calculate an ensemble variance\n\n## Subsetting operations¶\n\n `DataSet.crop`(self[, lon, lat, nco, nco_vars]) Crop to a rectangular longitude and latitude box `DataSet.select`(self, \\\\kwargs) A method for subsetting datasets to specific variables, years, longitudes etc. `DataSet.drop`(self, \\\\kwargs) Remove variables This will remove stated variables from files in the dataset.\n\n## Time-based methods¶\n\n `DataSet.set_date`(self[, year, month, day, …]) Set the date in a dataset You should only do this if you have to fix/change a dataset with a single, not multiple dates. `DataSet.shift`(self, \\\\*kwargs) Shift method.\n\n## Interpolation and resampling methods¶\n\n `DataSet.regrid`(self[, grid, method, recycle]) Regrid a dataset to a target grid `DataSet.to_latlon`(self[, lon, lat, res, …]) Regrid a dataset to a regular latlon grid `DataSet.resample_grid`(self[, factor]) Resample the horizontal grid of a dataset `DataSet.time_interp`(self[, start, end, …]) Temporally interpolate variables based on date range and time resolution `DataSet.timestep_interp`(self[, steps]) Temporally interpolate a dataset to given number of time steps between existing time steps `DataSet.fill_na`(self[, n]) Fill missing values with a distance-weighted average. `DataSet.box_mean`(self[, x, y]) Calculate the grid box mean for all variables This is performed for each time step. `DataSet.box_max`(self[, x, y]) Calculate the grid box max for all variables This is performed for each time step. `DataSet.box_min`(self[, x, y]) Calculate the grid box min for all variables This is performed for each time step. `DataSet.box_sum`(self[, x, y]) Calculate the grid box sum for all variables This is performed for each time step. `DataSet.box_range`(self[, x, y]) Calculate the grid box range for all variables This is performed for each time step.\n\n `DataSet.mask_box`(self[, lon, lat]) Mask a lon/lat box\n\n## Anomaly methods¶\n\n `DataSet.annual_anomaly`(self[, baseline, …]) Calculate annual anomalies for each variable based on a baseline period The anomaly is derived by first calculating the climatological annual mean for the given baseline period. `DataSet.monthly_anomaly`(self[, baseline]) Calculate monthly anomalies based on a baseline period The anomaly is derived by first calculating the climatological monthly mean for the given baseline period.\n\n## Statistical methods¶\n\n `DataSet.tmean`(self[, over]) Calculate the temporal mean of all variables `DataSet.tmin`(self[, over]) Calculate the temporal minimum of all variables `DataSet.tmedian`(self[, over]) Calculate the temporal median of all variables :param over: Time periods to average over. `DataSet.tpercentile`(self[, p, over]) Calculate the temporal percentile of all variables `DataSet.tmax`(self[, over]) Calculate the temporal maximum of all variables `DataSet.tsum`(self[, over]) Calculate the temporal sum of all variables `DataSet.trange`(self[, over]) Calculate the temporal range of all variables `DataSet.tstdev`(self[, over]) Calculate the temporal standard deviation of all variables Calculate the temporal cumulative sum of all variables `DataSet.tvar`(self[, over]) Calculate the temporal variance of all variables `DataSet.cor_space`(self[, var1, var2]) Calculate the correlation correct between two variables in space This is calculated for each time step. `DataSet.cor_time`(self[, var1, var2]) Calculate the correlation correct in time between two variables The correlation is calculated for each grid cell, ignoring missing values. Calculate the area weighted spatial mean for all variables This is performed for each time step. Calculate the spatial minimum for all variables This is performed for each time step. Calculate the spatial maximum for all variables This is performed for each time step. `DataSet.spatial_percentile`(self[, p]) Calculate the spatial sum for all variables This is performed for each time step. Calculate the spatial range for all variables This is performed for each time step. `DataSet.spatial_sum`(self[, by_area]) Calculate the spatial sum for all variables This is performed for each time step. Calculate the spatial range for all variables This is performed for each time step. Calculate the spatial range for all variables This is performed for each time step. `DataSet.centre`(self[, by, by_area]) Calculate the latitudinal or longitudinal centre for each year/month combination in files. This applies to each file in an ensemble. by : str Set to ‘latitude’ if you want the latitiduinal centre calculated. ‘longitude’ for longitudinal. by_area : bool If the variable is a value/m2 type variable, set to True, otherwise set to False. Calculate the zonal mean for each year/month combination in files. Calculate the zonal minimum for each year/month combination in files. Calculate the zonal maximum for each year/month combination in files. Calculate the zonal range for each year/month combination in files. Calculate the meridonial mean for each year/month combination in files. Calculate the meridonial minimum for each year/month combination in files. Calculate the meridonial maximum for each year/month combination in files. Calculate the meridonial range for each year/month combination in files.\n\n## Merging methods¶\n\n `DataSet.merge`(self[, join, match]) Merge a multi-file ensemble into a single file 2 methods are available.\n\n## Splitting methods¶\n\n `DataSet.split`(self[, by]) Split the dataset Each file in the ensemble will be separated into new files based on the splitting argument.\n\n## Output and formatting methods¶\n\n `DataSet.to_nc`(self, out[, zip, overwrite]) Save a dataset to a named file This will only work with single file datasets. `DataSet.to_xarray`(self[, decode_times]) Open a dataset as an xarray object `DataSet.to_dataframe`(self[, decode_times]) Open a dataset as a pandas data frame `DataSet.zip`(self) Zip the dataset This will compress the files within the dataset. `DataSet.format`(self[, ext]) Zip the dataset This will compress the files within the dataset. This works lazily. :param ext: New format. Must be one of “nc”, “nc1”, “nc2”, “nc4” and “nc5” . netCDF = nc1 netCDF version 2 (64-bit offset) = nc2/nc netCDF4 (HDF5) = nc4 netCDF4-classi = nc4c netCDF version 5 (64-bit data) = nc5 :type ext: str.\n\n## Miscellaneous methods¶\n\n `DataSet.na_count`(self[, over]) Calculate the number of missing values `DataSet.na_frac`(self[, over]) Calculate the number of missing values `DataSet.distribute`(self[, m, n]) Split the dataset into multiple evenly sized horizontal and vertical new files Collect a dataset that has been split using distribute `DataSet.cell_area`(self[, join]) Calculate the area of grid cells. `DataSet.first_above`(self[, x]) Identify the time step when a value is first above a threshold This will do the comparison with either a number, a Dataset or a netCDF file. :param x: An int, float, single file dataset or netCDF file to use for the threshold(s). If comparing with a dataset or single file there must only be a single variable in it. The grids must be the same. :type x: int, float, DataSet or netCDF file. `DataSet.first_below`(self[, x]) Identify the time step when a value is first below a threshold This will do the comparison with either a number, a Dataset or a netCDF file. :param x: An int, float, single file dataset or netCDF file to use for the threshold(s). If comparing with a dataset or single file there must only be a single variable in it. The grids must be the same. :type x: int, float, DataSet or netCDF file. `DataSet.last_above`(self[, x]) Identify the final time step when a value is above a threshold This will do the comparison with either a number, a Dataset or a netCDF file. :param x: An int, float, single file dataset or netCDF file to use for the threshold(s). If comparing with a dataset or single file there must only be a single variable in it. The grids must be the same. :type x: int, float, DataSet or netCDF file. `DataSet.last_below`(self[, x]) Identify the last time step when a value is below a threshold This will do the comparison with either a number, a Dataset or a netCDF file. :param x: An int, float, single file dataset or netCDF file to use for the threshold(s). If comparing with a dataset or single file there must only be a single variable in it. The grids must be the same. :type x: int, float, DataSet or netCDF file. `DataSet.cdo_command`(self[, command, ensemble]) Apply a cdo command `DataSet.nco_command`(self[, command, ensemble]) Apply an nco command `DataSet.compare`(self[, expression]) Compare all variables to a constant `DataSet.gt`(self, x) Method to calculate if variable in dataset is greater than that in another file or dataset This currently only works with single file datasets `DataSet.lt`(self, x) Method to calculate if variable in dataset is less than that in another file or dataset This currently only works with single file datasets Reduce dimensions of data This will remove any dimensions with only one value. `DataSet.reduce_grid`(self[, mask]) Reduce the dataset to non-zero locations in a mask :param mask: single variable dataset or path to .nc file. The mask must have an identical grid to the dataset. :type mask: str or dataset. `DataSet.set_precision`(self, x) Set the precision in a dataset `DataSet.check`(self) Check contents of files for common data problems. Check if files are corrupt A quick hack to change the grid file in North West European shelf Nemo grids. `DataSet.set_gridtype`(self, grid) Set the grid type. Create a mask identifying the shallowest cell without missing values. `DataSet.strip_variables`(self[, vars]) Remove any variables, such as bnds etc., from variables.\n\n## Ecological methods¶\n\n `DataSet.phenology`(self[, var, metric, p]) Calculate phenologies from a dataset Each file in an ensemble must only cover a single year, and ideally have all days." ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.7825389,"math_prob":0.9568636,"size":14374,"snap":"2022-40-2023-06","text_gpt3_token_len":3163,"char_repetition_ratio":0.23535143,"word_repetition_ratio":0.44153845,"special_character_ratio":0.22304161,"punctuation_ratio":0.091647334,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.99656713,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2022-09-29T03:50:00Z\",\"WARC-Record-ID\":\"<urn:uuid:628939b2-a0c2-4da4-81bf-53856d134376>\",\"Content-Length\":\"107309\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:31053c2e-474e-4808-b802-74b369eaacc6>\",\"WARC-Concurrent-To\":\"<urn:uuid:95fbfd6f-bc5c-41df-8020-387f22bca6a5>\",\"WARC-IP-Address\":\"104.17.33.82\",\"WARC-Target-URI\":\"https://nctoolkit.readthedocs.io/en/v0.4.5/api.html\",\"WARC-Payload-Digest\":\"sha1:GSIILEUOYD4HF2H72SZ5WK2DYWNSBPRK\",\"WARC-Block-Digest\":\"sha1:NVEXCC7QDRQNGPZRMBFCCV6I3NQMT2M3\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2022/CC-MAIN-2022-40/CC-MAIN-2022-40_segments_1664030335304.71_warc_CC-MAIN-20220929034214-20220929064214-00605.warc.gz\"}"}
https://www.r-bloggers.com/2020/07/puzzling-regression-anatomy/	[ "[This article was first published on Economics and R - R posts, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)\nWant to share your content on R-bloggers? click here if you have a blog, or here if you don't.\n\nConsider a regression model with the following causal structure:", null, "The variable `x1` affects `y` directly and also indirectly via `x2`. The following R code implements the model and simulates a corresponding data set.\n\n```set.seed(1)\nn = 10000\nbeta1 = 1; beta2=1\nx1 = rnorm(n,0,1)\nx2 = x1+rnorm(n,0,1)\ny = beta1x1 + beta2x2 + rnorm(n,0,1)\n```\n\nAssume we want to consistently estimate the direct linear effect `beta1` from `x1` on `y`. To do so, we can simply estimate a multiple linear regression where we add `x2` as a control variable:\n\n```coef(lm(y~x1+x2))\n\n## (Intercept) x1 x2\n## 0.007553765 0.998331323 1.000934100\n```\n\nBut what does it intuitively mean to add `x2` as control variable? The Frisch-Waugh-Lovell Theorem implies that we get the same estimator for `beta1` as in the multiple regression above by the following procedure:\n\n```# y.tilde is residual of regression\n# y on x2\ny.tilde = resid(lm(y~x2))\n\n# x1.tilde is residual of regression\n# x1 on x2\nx1.tilde = resid(lm(x1~x2))\n\n# Regression y.tilde on x1.tilde\n# we get the same estimate for beta1\n# as in the multiple regression with x1 and x2\ncoef(lm(y.tilde ~ x1.tilde))\n\n## (Intercept) x1.tilde\n## -5.104062e-17 9.983313e-01\n```\n\nHence, controlling for `x2` means that we essentially regress the residual variations of `y` and `x1` that cannot be linearly explained by `x2` on each other. So far this seems intuitive.\n\nThe interesting thing is that one gets the same estimate for `beta1` also with one of the following two regressions below (but only the regression above also yields correct standard errors):\n\n```# Approach A\nlm(y.tilde ~ x1)\n# Approach B\nlm(y ~ x1.tilde)\n```\n\nApproach A regresses the residual variation of `y` that cannot be linearly predicted by `x2` on `x1`. Approach B regresses `y` on the residual variation of `x1` that cannot be linearly predicted by `x2`.\n\nOnly one approach yields a consistent estimate of `beta1`. Make a guess which one…\n\nLet’s check:\n\n```# A: Inconsistent\ncoef(lm(y.tilde ~ x1))\n\n## x1\n## 0.4860465\n\n# B: Consistent\ncoef(lm(y ~ x1.tilde))\n\n## x1.tilde\n## 0.9983313\n```\n\nSo only approach B works. Angrist and Pischke (2009) refer to it as regression anatomy. For me that result was a bit puzzling for a long time because my intuitive interpretation of what it means to control for `x2` was more in line with approach A. I first want to shed light on that intuition and explain why approach A does not work. Afterward I want to give some intuition for the working approach B.\n\n## An intuition for control variables and why approach A fails\n\nI have different intuitions what controlling for `x2` means in the linear regression:\n\n`y = beta0 + beta1x1 + beta2x2 + eps`\n\nOne of my intuitions is the following:\n\n“By controlling for `x2`, we essentially subtract the variation that can be linearly explained by `x2` from `y`, i.e. up to an estimation error we subtract `beta2x2`.”\n\nThis interpretation suggests that approach A should work, but that approach fails to get a consistent estimate for `beta1`. So is the intuition above wrong? Not completely, but the qualification “up to an estimation error” causes trouble for approach A. Consider the following code.\n\n```# Modified approach A\ny.tilde2 = y - beta2x2\ncoef(lm(y.tilde2 ~x1))\n\n## x1\n## 0.9992699\n```\n\nIt is a modified version of approach A. It computes the residual variation `y.tilde2` by directly subtracting `beta2x2` from `y`. Now we get a consistent estimator of `beta1` when regressing `y.tilde2` on `x1`.\n\nBut approach A differs because we subtract `beta2.hatx2` from `y` where `beta2.hat` is estimated in the first stage regression:\n\n```# Same result as original approach A\nbeta2.hat = coef(lm(y~x2))\nbeta2.hat # inconsistent estimate of beta2\n\n## x2\n## 1.510801\n\ny.tilde = y-beta2.hatx2\ncoef(lm(y.tilde ~x1)) # inconsistent estimate of beta1\n\n## x1\n## 0.4860465\n```\n\nThe problem with approach A is that we don’t estimate `beta2.hat` consistently in the regression of `y` on `x2`. Instead, since `x1` and `x2` are correlated, `beta2.hat` also captures some of the direct effect of `x1` on `y`. This means in `y.tilde` we have already removed some of the effect from `x1` on `y` that we want to estimate. Therefore approach A yields an estimator for `beta1` that is biased towards 0.\n\nRemark: In the original computation of approach A, we also subtract the estimated constant from the initial regression when computing `y.tilde`, but that has no effect on the slope coefficient in the second stage regression.\n\nInterestingly, in some empirical papers an approach similar to approach A is performed, i.e. one first computes residuals of `y` from a first regression and then regresses those residuals on another set of explanatory variables. But the computation above shows that one should really be careful with this approach, since it only works if the first regression yields consistent estimates.\n\nLet us consider an example where such an approach would work. Consider the following modified model:", null, "We now have an additional variable `z` that affects `x2` but is uncorrelated with `x1`.\n\n```z = rnorm(n,0,1)\nx2 = x1+z+rnorm(n,0,1)\ny = beta1x1 + beta2*x2 + rnorm(n,0,1)\n```\n\nWe now conduct a variation of approach A where `y.tilde3` are the residuals of an instrumental variable regression of `y` on `x2` using `z` as instrument:\n\n```library(AER)\nreg1 = ivreg(y~x2\|z)\ncoef(reg1) # consistent beta2.hat\n\n## x2\n## 1.006464\n\ny.tilde3 = resid(reg1)\ncoef(lm(y.tilde3 ~ x1)) # consistent beta1.hat\n\n## x1\n## 0.9756005\n```\n\nWe now see that regressing `y.tilde3` on `x1` yields a consistent estimator of `beta1`.\n\n### Why does approach B work\n\nLet us now discuss why approach B works. Given our causal structure I find it more intuitive to first discuss why a similar approach works to consistently estimate `beta2`.\n\n```# x2.tilde is residual from regression\n# of x2 on x1\nx2.tilde = resid(lm(x2~x1))\n# consistent estimate of beta2\ncoef(lm(y ~ x2.tilde))\n\n## x2.tilde\n## 0.996786\n```\n\nHere I have the following intuition why it works. Intuitively, to consistently estimate the causal effect of `x2` on `y` we need to distill variation of `x2` that is uncorrelated with `x1`. If we regress `x2` on `x1`, the residuals `x2.tilde` of this regression are by construction uncorrelated with `x1`. They describe the variation of `x2` that cannot be linearly predicted by `x1`. That is exactly the variation of `x2` needed to consistently estimate `beta2`.\n\nThe equivalent procedure also works to estimate `beta1` consistently:\n\n```x1.tilde = resid(lm(x1~x2))\ncoef(lm(y ~ x1.tilde)) # consistent\n\n## x1.tilde\n## 0.98519\n```\n\nSo even though `x2` does not influence `x1` we can similarly distill in `x1.tilde` the relevant variation in `x1` that is uncorrelated with `x2`. For the regression anatomy it is irrelevant which causal direction has generated the correlation between `x1` and `x2`.\n\n### Final remarks\n\nI find it amazing that over many years I still often learn new intuitions for basic econometric concepts like multiple linear regression. Currently, I think introducing multiple regression via the Frisch-Waugh-Lovell theorem and the regression anatomy can be much more helpful to build intuition in an applied empirical course than covering the matrix algebra. (Of course, it is a different story if you want to prove econometric theorems.) For an example of such a course, you can check out the open online material (videos, quizzes, interactive R exercises) of my course Market Analysis with Econometrics and Machine Learning.\n\nAngrist, Joshua D., and Jörn-Steffen Pischke. 2009. Mostly Harmless Econometrics: An Empiricist’s Companion.", null, "" ]	[ null, "https://skranz.github.io/images/puzzling_anatomy.svg", null, "https://skranz.github.io/images/puzzling_anatomy2.svg", null, "https://feeds.feedburner.com/~r/skranz_R/~4/MjXH00Uef1I", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.83303195,"math_prob":0.94818723,"size":7742,"snap":"2022-27-2022-33","text_gpt3_token_len":2069,"char_repetition_ratio":0.15986043,"word_repetition_ratio":0.039308175,"special_character_ratio":0.2601395,"punctuation_ratio":0.11140065,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9978922,"pos_list":[0,1,2,3,4,5,6],"im_url_duplicate_count":[null,4,null,4,null,4,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2022-07-03T06:02:19Z\",\"WARC-Record-ID\":\"<urn:uuid:a89dfa8b-b451-46f2-8b45-217a3825c999>\",\"Content-Length\":\"100474\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:40d179a6-b8c5-4cba-a41b-995f8b588ea4>\",\"WARC-Concurrent-To\":\"<urn:uuid:e45df9ab-6320-4c71-8624-2b2605c7293e>\",\"WARC-IP-Address\":\"172.64.104.18\",\"WARC-Target-URI\":\"https://www.r-bloggers.com/2020/07/puzzling-regression-anatomy/\",\"WARC-Payload-Digest\":\"sha1:VWGUSGF53M5YSQUTFD236Q2M4LBUZMU2\",\"WARC-Block-Digest\":\"sha1:SXVWFBKQAMRHICA3WJKYY6VM3I54T2NF\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2022/CC-MAIN-2022-27/CC-MAIN-2022-27_segments_1656104215790.65_warc_CC-MAIN-20220703043548-20220703073548-00188.warc.gz\"}"}
https://www.hindawi.com/journals/sp/2021/9944358/	[ "#### Abstract\n\nWireless sensors localization is still the main problem concerning wireless sensor networks (WSN). Unfortunately, range-free node localization of WSN results in a fatal weakness–, low accuracy. In this paper, we introduce kernel regression to node localization of anisotropic WSN, which transfers the problem of localization to the problem of kernel regression. Radial basis kernel-based G-LSVR and polynomial-kernel-based P-LSVR proposed are compared with classical DV-Hop in both isotropic WSN and anisotropic WSN under different proportion beacons, network scales, and disturbances of communication range. G-LSVR presents the best localization accuracy and stability from the simulation results.\n\n#### 1. Introduction\n\nCurrently, wireless sensors localization is still the main problem concerning wireless sensor networks (WSN). The localization algorithms of WSN can be classified into the range-based measurement method and the range-free measurement method. The former one can get high accuracy but need range information , while the latter one gets low accuracy without any range information.\n\nIn order to improve the accuracy of range-free node localization, machine learning is introduced to the localization of WSN . Preciously, artificial neural network (ANN) was used in range-free localization algorithms, and its accuracy and performance greatly promoted when compared with other traditional algorithms . Besides, Phoemphon et al. [13, 14] have used a fuzzy logic to range-free localization in WSNs. This algorithm mainly focuses on the heterogeneous scenarios with lower complexity. A graph-based localization algorithm was presented using conventional neural networks (CNN) and support vector machine (SVM) in the paper. Another localization method using SVM for large-scale WSNs was proposed in . A fuzzy c-means training sample reduction (FCMTSR) method was used to reduce the training overhead and learning calculation. FCMTSR-SVM algorithm not only improves the localization accuracy but also reduces the training samples time. Based on SVM, a new algorithm called support vector machine-based range-free localization (RFSVM) has been presented in . A transmit matrix was introduced to show the relation between hops and distances and to train the system, while SVM was used to find the unknown nodes in WSNs. Another range-free localization method has been developed using SVM classifier . Recently, kernel-based approach is first proposed in isotropic and anisotropic WSN . Active AVR is used in the localization and the simulating results show an improved performance in localization .\n\nIn this paper, in order to improve the accuracy of node localization in anisotropic WSN, we introduce kernel regression approach to node localization firstly. The contribution in this paper can be summarized to (1) systematically review the literature regarding the approaches for wireless sensors localization and identity the existing issues in this field and (2) to propose a novel graph-based localization approach and introduce the design and implementation of the proposed approach in detail and (3) a set of experiments comparing proposed approach with existing ones are performed and discussed. The rest of the paper is organized as follows. Section 2 discusses the related theories. The proposed algorithm is discussed and algorithm simulation and analysis are presented in Section 3. The conclusions and future scope are highlighted in Section 4.\n\n#### 2. Materials and Methods\n\n##### 2.1. Relative Theories\n\nSVM is a universal machine learning method based on statistical learning theory framework. Vapnik and Lerner began to study machine learning problem with limited samples from 1960s and then Vapnik and Chervonenkis proposed VC theory and statistical learning theory, which were constructed on the principle of structural risk minimization. Later in 1982, Cortes and Vapnik further proposed structural risk minimization (SRM) principle. It is pockmarking because it provides a same framework for solving the learning problem of limited samples and helps to solve a series of problems, including selecting neural network structure and local minimum [25, 26].\n\nThe methods based on SVM are widely used in pattern recognition, regression estimate, etc. . The better efficiency and accuracy are achieved in pattern recognition field, handwritten numeral recognition, and image recognition .\n\nSVM includes support vector classifier and support vector regression. It was first proposed to solve the problem of classification at the beginning, including linear classification and nonlinear classification. To solve the problem of nonlinear classification, SVM transforms the training set into a high-dimensional Hilbert space through nonlinear mapping. To solve the problem of linear classification, linear division is used to construct classification hyperplane in the space and to obtain decision function.\n\nLater, SVM was extended from classification problem to regression problem [32, 33] and a new regression algorithm, support vector regression (SVR), was then proposed.\n\nFigure 1 presents the transformation procedures from localization problem to regression problem.\n\n###### 2.1.1. Support Vector Regression\n\nTo achieve regression using samples in Figure 2, SVR transfers the training points , to training points , in high-dimensional Hilbert space by mapping first. Secondly, the training set after mapping does linear regression in the space. Nonlinear regression function can be represented as formula (1) after kernel function , replacing the inner product in the target function of dual problem.\n\n###### 2.1.2. Kernel Function\n\nKernel function is an important component of SVM and the performance of SVM varies with kernel function to a large extent. Selecting a kernel function is very important to SVM model. The kernel function can be described as the vector in inputting space. It is mapped to a high-dimensional space through a nonlinear transformation, and the linear classification can be obtained in the high-dimensional space through an optimal classification hyperplane, which can improve the nonlinear processing capability and reduce the dimension.\n\nDefinition 1. (kernel function ). Supposing is a subset of . defined in is a kernel function if there is a mapping function from to a Hilbert space H to make , , indicating the inner product in space H.\nMercer’s theorem shows the requirements of kernel function, which is also the requirement of a symmetric function correspondent to the inner product of a feature space.\n\nTheorem 1. (Mercer’s theorem ). Supposing is a compact set of . K is a symmetric function with continuous real value in . Integral operator , thenSupposing are the eigenfunctions of when eigenvalue and , then(1)(2) and (3) is workable to all and is uniform convergence to all sequencesKernel function meeting the conditions of Mercer’s theorem is called Mercer kernel. Mercer’s theorem indicates that is a kernel function if function K meets the following formula:Function can be written as an inner product in an eigenspace, as shown in the following formula:Therefore,where is the Hilbert space.\nAnother equal condition of Mercer kernel is defined by Gram matrix based on definite-sample-definition kernel function. The condition is easy to validate when compared with Mercer kernel.\n\nDefinition 2. (Gram matrix). Matrix K () with line l and column l (element in line i and column j is ) is called Gram matrix of function K with for a given function and .\n\nTheorem 2. Necessary and sufficient condition of integral operator positive semidefinite is that Gram matrix of K is positive semidefinite with any , supposing is a compact set in Rn and is a continuous and symmetric function in .\nSome properties of kernel function can be got from the abovementioned two theorems.\nRatiocination 1: the following functions are kernel functions, supposing and are kernel functions in , constant , and is polynomial with all coefficients positive.Some kernel functions frequently used can be got according to the above theorems.(1)Polynomial-kernel function , where q is the order of polynomial(2)Radial basis kernel function , where is the width parameter of radial basis function (RBF)(3)Sigmoid-kernel function In addition, there are Fourier series kernel, B-spline kernel, tensor-kernel functions, etc.\n\n##### 2.2. Range-Free Node Localization Based on Kernel Function \n\nSupposing there is WSN with beacons and nodes and the location of each node can be expressed as follows,here, the location of M beacons is known and the location of other N nodes is unknown. and .\n\nThe only measurement information in the algorithm is approximate information that is indicated by hop number between all the nodes. The geometric distance between and can be defined as follows:\n\nThe approximate information between and indicates the hop number between two nodes. Thus, the localization problem can be described as giving , , and , estimating .\n\nHere, , , . The basic idea of the approach is to transfer the localization problem to kernel regression problem. The location of WSN node can be calculated by the following formula with the aid of approximate information, ,where is the kernel function and .\n\nThe algorithm consisted of the following three steps:Step 1. Measurement. Each beacon and each node exchange their hop counting information and each beacon sends its location to other sensor nodes.Step 2. Training (SVR). It estimates the distance model of each node and broadcasts it. All other sensor nodes will receive M distance models.Step 3. Localization. It calculates the distance of the node to all the beacons by using the distance models in the above step and then calculates its location distributive by multilateral measurement method.\n\nThe communication cost of the algorithm is data packets.\n\n#### 3. Results and Discussion\n\nThe localization errors of RBF-kernel-based algorithm, polynomial-kernel-based algorithm, and DV-Hop algorithm are compared. DV-Hop algorithm is a classical algorithm in all the range-free algorithms . Localization error is defined as the distance difference between the real location and estimated location of sensor node and RBF-kernel is the Gauss kernel as follows:\n\nGauss kernel-based localization through SVR can be denoted as G-LSVR. The extrapolation ability of a polynomial is good when the parameter of the polynomial-kernel function q = 2 or 4 in . Let the power of polynomial q = 2 and the polynomial-kernel be\n\nObjective function in formula (12) includes and . indicates confidence interval, and demonstrates empiric risk. C is a constant, which balances empiric risk and confidence interval.\n\nPolynomial-based localization through SVR can be denoted as P-LSVR.\n\nSpecifically, equations (10), (11), and (12) are valid only under the following conditions:(1)Only the connectivity between nodes and location of beacons is needed, while other range information, such as received signal strength indicator, angle, and time, is negligible.(2)The communication range of all sensor nodes and beacons is the same. The sensor nodes and beacons are distributed in either isotropic network or anisotropic network with a square area of 100 × 100, as shown in Figures 3 and 4.(a)Isotropic network: all the sensor nodes are evenly distributed in the whole network and the node density in the network is the same.(b)Anisotropic network: all the sensor nodes are discrete distributed in the whole network (taking Type-C network as an example).\n\nThe following three simulation environments are adopted and the communication range of all the sensor nodes is R:(1)Different node density:(a)Sparse network: 20 beacons and 100 sensor nodes;(b)Medium network: 20 beacons and 200 sensor nodes;(c)Dense network: 100 beacons and 300 sensor nodes.(2)The number of sensor nodes in the network is 300 with the node density of beacons varying from 10% to 30%. The communication radius of all the sensor nodes is R.(3)Communication transmission model of nodes is irregular if more environmental factors concerned, and the number of sensor nodes in the network is 300 with the node density of beacons varying from 10% to 30%.\n\nThe parameters have priority than simulations due to the overfitting of kernel-based algorithms. Regularized parameter C in formula (12) controls the balance between complexity and accuracy in G-LSVR for a better generalization. Parameter of Gauss kernel function is a constant and is crucial to the balance between definiteness and sensitivity. Thus, the value of C and should be determined first by simulations. Tables 1 and 2 show the localization error using algorithm G-LSVR with different C and in the WSN with 100 sensor nodes and 20 beacons. These data are calculated by the mean number of ten individual simulations. It is notable that the minimum localization error is obtained when C = 20 and = 20 in both isotropic WSN and anisotropic WSN.\n\nIn simulation environment (1), the communication range R of all the nodes is 20. The localization performance of the three different algorithms in isotropic WSN is shown in Table 3 and the localization performance in anisotropic WSN is shown in Table 4. It is clear that the three algorithms can get high localization accuracy in isotropic WSN and the performance of P-LSVR and G-LSVR is better than that of DV-Hop. At the same time, the localization error of DV-Hop is very large and the localization error of P-LSVR and G-LSVR is smaller than that of DV-Hop in anisotropic WSN obviously.\n\nThe localization error in dense network is minimum to any localization algorithms. It is the balance between node density and localization performance. But sometimes more beacons will cost more resource.\n\nIn simulation environment (2), the amount of sensor nodes is set to 300 and the number of beacons vary from 10% to 30%. The communication range of all the nodes is 20. The comparison of localization performance is shown in Figures 5 and 6 in isotropic WSN and anisotropic WSN, respectively. It is noticeable that the accuracy of kernel-based localization algorithm is proportional to the number of beacons. The localization performance of G-LSVR is better than that of P-LSVR. DV-Hop can get better performance in isotropic WSN, while worse performance in anisotropic WSN, in which DV-Hop uses a big hop number to denote long distance in isotropic.\n\nIn simulation environment (3), more realistic factors are concerned. Irregular wireless transmitting model , as shown in Figure 7, is used to simulate disturbance of communication range R because of multiaccess channels, disturbance, and other environmental factors. Degree of irregularity (DOI) is used to express the irregularity. Thus, the communication range can be denoted as\n\nFigures 8 and 9 present the localization performances of the three algorithms in isotropic and anisotropic WSN, supposing that the number of sensor nodes in the whole WSN is 300, the proportion of beacons varies from 10% to 30%, communication range is 20, and DOI is set to 0.1∼0.3. It is clearly shown that the kernel-based localization algorithms P-LSVR and G-LSVR are more stable with better performance and smaller error.\n\nThe localization error of P-LSVR and G-LSVR is inversely proportional to the number of beacons in both isotropic WSN and anisotropic WSN. Owning to the sparse of SVR from the point of machine learning, the error of P-LSVR and G-LSVR is convergence. The error of DV-Hop is small in isotropic WSN, while it is big in anisotropic WSN. Therefore, DV-Hop is rarely used in practical applications.\n\n#### 4. Conclusions\n\nThis paper proposed a kernel-based range-free localization algorithm to transfer the problem of localization to the problem of kernel regression. The algorithm only needs simple approximate information and each sensor node estimates its own location distributivity. Radial basis kernel-based G-LSVR and polynomial-kernel-based P-LSVR proposed in this paper are compared with DV-Hop in both isotropic WSN and anisotropic WSN under different proportions of beacons, network scales, and disturbances of communication range. The kernel-based algorithms are better in localization accuracy and stability.\n\n#### Data Availability\n\nThe data used to support the study are available from Dr. He via email, [email protected].\n\n#### Conflicts of Interest\n\nThe authors declare that they have no conflicts of interest.\n\n#### Acknowledgments\n\nThis research was supported by the Natural Science Foundation of Zhejiang Province under grant no. LGF21F020015." ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.8777486,"math_prob":0.9351446,"size":24812,"snap":"2023-40-2023-50","text_gpt3_token_len":5743,"char_repetition_ratio":0.16986457,"word_repetition_ratio":0.07680127,"special_character_ratio":0.22972755,"punctuation_ratio":0.18034826,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9842428,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2023-10-01T08:18:19Z\",\"WARC-Record-ID\":\"<urn:uuid:a7a1b8ab-5a64-47fc-89b9-d4d226be288d>\",\"Content-Length\":\"723480\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:579bcfa8-92c4-45be-8201-945e3081355d>\",\"WARC-Concurrent-To\":\"<urn:uuid:0f0a0441-42fc-488e-ba93-9a8a62962bcf>\",\"WARC-IP-Address\":\"104.18.40.243\",\"WARC-Target-URI\":\"https://www.hindawi.com/journals/sp/2021/9944358/\",\"WARC-Payload-Digest\":\"sha1:2MW5LQBBBBZQ6RSB3TC2SMRD25TYUYC7\",\"WARC-Block-Digest\":\"sha1:5ZEIKVQUOA7NVGAT4DUUWX463CF2ZTQY\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2023/CC-MAIN-2023-40/CC-MAIN-2023-40_segments_1695233510810.46_warc_CC-MAIN-20231001073649-20231001103649-00229.warc.gz\"}"}
http://uoj.ac/problem/261	[ "# #261. 【NOIP2016】天天爱跑步\n\n### 样例一\n\n#### input\n\n6 3\n2 3\n1 2\n1 4\n4 5\n4 6\n0 2 5 1 2 3\n1 5\n1 3\n2 6\n\n\n\n#### output\n\n2 0 0 1 1 1\n\n\n\n### 样例二\n\n#### input\n\n5 3\n1 2\n2 3\n2 4\n1 5\n0 1 0 3 0\n3 1\n1 4\n5 5\n\n\n\n#### output\n\n1 2 1 0 1\n\n\n\n### 限制与约定\n\n2\n3$=992$$=992$$W_j=0$\n4\n5$=993$$=993 6=99994$$=99994$树退化成一条链，其中 $1$ 与 $2$ 有边，$2$ 与 $3$ 有边，$\\dots$，$n-1$ 与 $n$ 有边\n7\n8\n9$=99995$$=99995所有的 S_i=1 10 11 12 13=99996$$=99996$所有的 $T_i=1$\n14\n15\n16\n17$=99997$$=99997 18 19 20=299998$$=299998$" ]	[ null ]	{"ft_lang_label":"__label__zh","ft_lang_prob":0.9623435,"math_prob":0.9999008,"size":1813,"snap":"2019-26-2019-30","text_gpt3_token_len":1632,"char_repetition_ratio":0.12216695,"word_repetition_ratio":0.015151516,"special_character_ratio":0.47324875,"punctuation_ratio":0.002915452,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9991132,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2019-07-17T17:27:51Z\",\"WARC-Record-ID\":\"<urn:uuid:62f51a88-dc9f-4b2f-b1c7-f04a23d5d5c7>\",\"Content-Length\":\"16911\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:b4ed16f5-edc9-40ac-94c2-acc37fd2f5c7>\",\"WARC-Concurrent-To\":\"<urn:uuid:dfb0e8dc-d9ee-434c-b1e0-82aa7d4fa712>\",\"WARC-IP-Address\":\"114.215.147.61\",\"WARC-Target-URI\":\"http://uoj.ac/problem/261\",\"WARC-Payload-Digest\":\"sha1:ZBV23OBRBGWXNZLD46WR6BSPFVETLK3H\",\"WARC-Block-Digest\":\"sha1:NQEVIJXUPX7JHEM7KPPRM4NZMJYA6Y2F\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2019/CC-MAIN-2019-30/CC-MAIN-2019-30_segments_1563195525355.54_warc_CC-MAIN-20190717161703-20190717183703-00271.warc.gz\"}"}
https://kukuruku.co/post/erlang-for-beginners-data-types-variables-lists-and-tuples/	[ "# Erlang for Beginners. Data Types, Variables, Lists and Tuples\n\nErlang", null, "It’s the first article of the series. For many of you it may seem terribly trite as I’ll review the very basis of the subject. But this tutorial is going to be really useful for Erlang beginners. I’ll also dwell on some interesting things that aren’t obvious.\n\nIn order to begin your work with Erlang, execute the following in the terminal:\n\n``````\\$ sudo apt-get install erlang\n\\$ erl\n``````\n\nThat will start the Erlang interpreter. You should carry out examples from the article in it. In Erlang, in contrast to most languages, we put a period at the end of expression. Comments begin with % symbol and go on till the end of the line.\n\nSo, let’s start.\n\n### Numbers\n\nErlang supports two types of numeric variables: integer and float. We can perform the following mathematical operations: addition, subtraction, multiplication and division.\n\n``````1> 7 + 3.\n10\n2> 12 - 4.\n8\n3> 5 * 2.\n10\n4> 12 / 6\n2.0\n5> 7 div 3.\n2\n6> 7 rem 3.\n1\n``````\n\nPlease note, that a floating-point number was the result of division. Erlang is clever enough to automatically cast numeric variables to the necessary type.\n\nErlang also allows performing several operations at a time. Mathematical operations conform to the standard priority rules. Therefore, the result of 2 + 3 * 4. expression will be 14, since multiplication has higher priority than addition. Use the brackets to explicitly define the order of calculations:\n\n``````1> (2 + 3) * 4.\n20\n2> -(10 + 3).\n-13\n``````\n\nBesides, you don’t have to restrict yourself to decimal notation only. You can use numbers with any base from 2 to 36. For that purpose, you should indicate the number in the form of Base#Value.\n\n``````1> 2#11011.\n27\n2> 10#1198.\n1198\n3> 16#A04F.\n41295\n4> 36#1TA.\n2350\n``````\n\nThere’s more to come. In Erlang you can use numbers with different bases in one expression:\n\n``````1> 2#11011 + 36#1TA.\n2377\n2> 10#1198 - 16#A04F.\n-39841\n3> 3#201 * 4#321.\n1083\n``````\n\n### Atoms\n\nAtoms are analogues of constant variables from other languages. The atom value conforms precisely to its name. Roughly speaking, an atom is a string that can not be changed.\n\nAtoms should begin with a lowercase letter and can contain lowercase and uppercase characters, digits, underscore (_) and the “at” sign (@). We can also enclose an atom in single quotes. In this case it will be able to contain any characters. An atom can’t coincide with the reserved word. Therefore, the following atom names are invalid: after and andalso band begin bnot bor bsl bsr bxor case catch cond div end fun if let not of or orelse query receive rem try when xor.\n\n``````1> atom.\natom\n2> otherAtom.\notherAtom\n3> atom_with_underscore.\natom_with_underscore\n4> one@more@atom.\none@more@atom\n5> 'Atom with whitespace'.\n'Atom with whitespace'\n6> ' Atom with special charactes #^&?'.\n' Atom with special charactes #^&?'\n7> Atom.\n* 1: variable 'Atom' is unbound\n8> after.\n* 1: syntax error before: 'after'\n``````\n\n### Boolean Data Types and Comparison operators\n\nBoolean data types in Erlang are the two reserved atoms: true and false. There’s an interesting and not obvious fact connected with it. I’ll tell you about it later.\n\nThe language has all the basic logical operations, such as and, or, xor and not:\n\n``````1> true or false.\ntrue\n2> true and false.\nfalse\n3> true xor false.\ntrue\n4> not false.\ntrue\n5> (not (true xor true)) or (false and true).\ntrue\n``````\n\nand and or operators always calculate expression values from both sides. So by executing (1 > 2) or (3 < 4). we’ll find the values for both expressions, though after calculating the right expression we already know the result. In order to avoid this, use andalso and orelse operators.\n\nUse the following operators to compare values: ==, =:=, =/=, /=, <, =<, > and >=. If your usual language uses == and != to test for and against equality, Erlang uses =:= and =/=.\n\n``````1> 2 == 2.0.\ntrue\n2> 2 =:= 2.0.\nfalse\n3> 3 /= 3.0.\nfalse\n4> 3 =/= 3.0.\ntrue\n5> 5 > 5.\nfalse\n6> 5 =< 5.\ntrue\n``````\n\nIf you’ve programmed in other languages before, you’re most likely got used that true is equal to 1 and false is equal to . This rule isn’t working in Erlang:\n\n``````1> true == 1.\nfalse\n2> false == 0.\nfalse\n3> false > 19. %% !!!\ntrue\n``````\n\nHave you paid attention to the third line? A bit strange, isn’t it? The thing is that Erlang allows comparing the values of different types. It does that by the following rule:\n\nnumber < atom < reference < fun < port < pid < tuple < list < bit string\n\nWe’ve mentioned above that true and false are atoms. Now we can see that atoms are greater than numbers. Therefore false > 19..\n\n### Variables\n\nThere are no variables in pure functional languages (such as Haskell). But Erlang does allow us to create variables, though there’s one restriction: we can assign a value to a variable just once. If we try to do it again (with different value) it will lead to en error.\n\nThe variable name should begin with a capital letter or an underscore. The variable name can consist of one underscore only. But a variable with such name doesn’t memorize the value. Such variables can be used to match them with the pattern (find the information on the matter below).\n\n``````1> Variable = 10 - 7.\n3\n2> OtherVariable = 3 * 4.\n12\n3> _result = Variable + OtherVariable.\n15\n4> _result.\n15\n5> Variable = 5.\n** exception error: no match of right hand side value 5\n``````\n\n### Tuples\n\nSometimes it’s more comfortable to have a group of variables that are connected somehow. For that purpose Erlang provides such construction as tuple. The tuple looks like {Value1, Value2, …, ValueN}. It can contain any amount of variables.\n\nLet’s take a look how we can use tuples. The example is quite simple. We need to store the information about a point in the coordinate space (X and Y coordinates). We could create two different variables and store the coordinates in them. But it’s easier to store them together.\n\n``````1> MyPoint = {2,5}.\n{2,5}\n``````\n\nAs I’ve already mentioned, the size of a tuple isn’t limited by two values. A tuple can also contain values of different types, including other tuples.\n\n``````1> MyTuple = {1,myAtom,true,{1,false}}.\n{1,myAtom,true,{1,false}}\n``````\n\n### Pattern Matching\n\nIn order to extract values from the tuple (not for this purpose only) we utilize pattern matching. First of all, let’s consider the way the matching operator (=) works. The = operator has the role of comparing values and complaining if they’re different. If they’re the same, it returns the value. What this operator does when mixed with variables is that if the left-hand side term is a variable and it is unbound (has no value associated to it), Erlang will automatically bind the right-hand side value to the variable on the left-hand side. The comparison will consequently succeed and the variable will keep the value in memory.\n\nPattern matching means that instead of a single variable we indicate a template that has to conform the data. If the data correspond to the template, we’ll confront variables of this template with the appropriate values.\n\n``````1> {X,Y} = {1,2}.\n{1,2}\n2> X.\n1\n3> Y.\n2\n4> Z = {Y,X}.\n{2,1}\n5> {A,B,C} = {myAtom,true,Z}.\n{myAtom,true,{2,1}}\n6> A.\nmyAtom\n7> B.\ntrue\n8> C.\n{2,1}\n``````\n\nPattern matching is one of the most powerful tools of functional languages. We can use it not only to extract data from tuples. We’ll apply this approach quite often.\n\nSometimes we don’t need all the data. For example, we may need just the second value of a three-tuple. Not to “produce” useful entities we can use a special variable we’ve mentioned before: _. In that way we are going to indicate that there can be several such variables in this template part. Quite convenient, isn’t it?\n\n``````1> {_,X,_} = {1,2,3}.\n{1,2,3}\n2> X.\n2\n``````\n\n### Lists\n\nA list is an analogue of arrays in imperative languages. A list looks like the following: [Value1, Value2, …, ValueN]. Its elements shouldn’t necessarily be of one type. One list can contain numbers, atoms, tuples, other lists, etc.\n\n``````1> [1,2,true,atom,{5,4},[true,false]].\n[1,2,true,atom,{5,4},[true,false]].\n``````\n\nWhen working with lists in Erlang, there’s the following strange thing:\n\n``````1> [100,101,102,103].\n\"defg\"\n``````\n\nErlang printed a list as a string. But you shouldn’t worry. This refers to its displaying in terminal only. Our list actually contains numbers. Such behavior is connected with peculiarities of the language. Initially, there were no strings in Erlang. So lists containing symbol numbers were used for working with it. Fortunately, the language is developing. So today we can work with strings.\n\nWe can add (++) and subtract lists from each other (–). You should remember that these operators are also right associative.\n\n``````1> [1,2,3,4,5] ++ [6,7].\n[1,2,3,4,5,6,7]\n2> [1,2,3,4,5] -- [2,3].\n[1,4,5]\n3> [1,2,3] ++ [].\n[1,2,3]\n4> [1,2,3,4,5] -- [1,2,3] -- .\n[3,4,5]\n5> [1,2,3] ++ [4,5,6] -- [4,5].\n[1,2,3,6]\n``````\n\nWe can also compare the lists. There are standard comparison operators for that purpose. At first we compare list heads. If they are equal, we compare heads of the tails, etc. The lists are compared by the first differing elements. In the example provided below the first list is «greater», because the first element differing from the corresponding element of the second list is greater (4 > 1).\n\n``````1> [1,2,3,4,0] > [1,2,3,1,1000,2000,6589].\ntrue\n``````\n\nLists are divided into two parts: head and tail. A head is the first element of the list, a tail is everything else. A tail also has a head and a tail. When matching with the pattern we use \| operator to indicate the boundary between the head and the tail\n\n``````1> [Head\|Tail] = [1,2,3,4,5].\n[1,2,3,4,5]\n1\n3> Tail.\n[2,3,4,5]\n4> [Second\|_] = Tail.\n[2,3,4,5]\n5> Second.\n2\n``````\n\n#### List Generator\n\nOf course, we won’t always define lists manually. It’s quite tiresome and isn’t interesting at all. Fortunately, the language developers share this opinion. Therefore, Erlang has a tool for automatic list creation. It’s better to take a look at it by an example. First of all, let’s write a code that will create a list automatically. The list will contain numbers from 1 to 10, multiplied by 3.\n\n``````1> [X3 \|\| X <- [1,2,3,4,5,6,7,8,9,10]].\n[3,6,9,12,15,18,21,24,27,30]\n``````\n\nOur expression is [Expr \|\| Item < — SourceList]. Erlang takes elements from SourceList one by one and replaces each element to Expr expression instead of Item variable. We’ll add the expression result to the resulting list. Quite simple, isn’t it?\n\nBut the generator in its current form is almost useless. Let’s complicate the task. We’ll make the generator work only with even numbers (that are greater than 5) from the source list.\n\n``````1> [X3 \|\| X <- [1,2,3,4,5,6,7,8,9,10], X rem 2 =:= 0, X > 5].\n[18,24,30]\n``````\n\nNow our generator is as follows: [Expr \|\| Item < — SourceList, Condition1, Condition2, …, Condition2]. It operates exactly the same way as the previous variant. But now Erlang controls that each element of the source list fits the stated requirements. If the element doesn’t meet at least one requirement – we skip it.\n\nBut that’s not the whole story. There can be several source lists. Let’s write a generator as yet another example. It will return all possible combinations of even numbers from 1 to 5 and odd numbers from 6 to 10. Let’s represent the combination by a two-tuple – a pair.\n\n``````1> [{X,Y} \|\| X <- [1,2,3,4,5], Y <- [6,7,8,9,10], X rem 2 =:= 0, Y rem 2 =:= 1].\n[{2,7},{2,9},{4,7},{4,9}]\n``````\n\nIn the most general case the generator looks like [Expr \|\| Item1 < — SourceList1, Item2 < — SourceList2, …, ItemN < — SourceListN, Condition1, Condition2, …, ConditionN]. In this case Erlang will return the Cartesian product of the source lists (or rather their elements meeting requirements).\n\nList generator is an extremely powerful tool provided by the language and we’ll use it quite often.\n\n## Summary\n\nIn the article we’ve reviewed the very basic parts of the language. We’ve considered the data types present in the language as well as the way they interact. We’ve also reviewed such fundamental notions as pattern matching and list generators.\n\nIn the next article we’ll consider the way we can use existing functions and create new ones. We’ll also take a look how to work with Erlang modules.\n\nThank you for reading the article. Have a nice code.", null, "" ]	[ null, "https://s3.amazonaws.com/kukuruku-co/uploads/images/00/00/01/2014/06/06/04d1d661d1.png", null, "https://s3.amazonaws.com/kukuruku-co/images/avatars/avatar_light.png", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.83061355,"math_prob":0.95008844,"size":12025,"snap":"2023-40-2023-50","text_gpt3_token_len":3279,"char_repetition_ratio":0.112137094,"word_repetition_ratio":0.0038295835,"special_character_ratio":0.2964657,"punctuation_ratio":0.17725752,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9512311,"pos_list":[0,1,2,3,4],"im_url_duplicate_count":[null,3,null,null,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2023-11-28T12:46:19Z\",\"WARC-Record-ID\":\"<urn:uuid:db8d420d-a55e-41f4-9bd1-715d1704a2de>\",\"Content-Length\":\"31933\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:7e1811cd-a497-4dce-a77b-c116393e2c9d>\",\"WARC-Concurrent-To\":\"<urn:uuid:cc5b6c6e-1f1d-4f56-a942-424f92cf6307>\",\"WARC-IP-Address\":\"172.67.203.55\",\"WARC-Target-URI\":\"https://kukuruku.co/post/erlang-for-beginners-data-types-variables-lists-and-tuples/\",\"WARC-Payload-Digest\":\"sha1:YVI5ECLJP3C5SWEC3MOEOWEN32P6I4ZK\",\"WARC-Block-Digest\":\"sha1:LWEZ2JW2UNQRSMKWZAJAI7UBPQBAMTZY\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2023/CC-MAIN-2023-50/CC-MAIN-2023-50_segments_1700679099514.72_warc_CC-MAIN-20231128115347-20231128145347-00178.warc.gz\"}"}
https://www.colorhexa.com/343b52	[ "# #343b52 Color Information\n\nIn a RGB color space, hex #343b52 is composed of 20.4% red, 23.1% green and 32.2% blue. Whereas in a CMYK color space, it is composed of 36.6% cyan, 28% magenta, 0% yellow and 67.8% black. It has a hue angle of 226 degrees, a saturation of 22.4% and a lightness of 26.3%. #343b52 color hex could be obtained by blending #6876a4 with #000000. Closest websafe color is: #333366.\n\n• R 20\n• G 23\n• B 32\nRGB color chart\n• C 37\n• M 28\n• Y 0\n• K 68\nCMYK color chart\n\n#343b52 color description : Very dark desaturated blue.\n\n# #343b52 Color Conversion\n\nThe hexadecimal color #343b52 has RGB values of R:52, G:59, B:82 and CMYK values of C:0.37, M:0.28, Y:0, K:0.68. Its decimal value is 3423058.\n\nHex triplet RGB Decimal 343b52 `#343b52` 52, 59, 82 `rgb(52,59,82)` 20.4, 23.1, 32.2 `rgb(20.4%,23.1%,32.2%)` 37, 28, 0, 68 226°, 22.4, 26.3 `hsl(226,22.4%,26.3%)` 226°, 36.6, 32.2 333366 `#333366`\nCIE-LAB 25.159, 3.51, -14.871 4.503, 4.467, 8.607 0.256, 0.254, 4.467 25.159, 15.279, 283.279 25.159, -4.182, -18.078 21.135, 1.042, -9.35 00110100, 00111011, 01010010\n\n# Color Schemes with #343b52\n\n• #343b52\n``#343b52` `rgb(52,59,82)``\n• #524b34\n``#524b34` `rgb(82,75,52)``\nComplementary Color\n• #344a52\n``#344a52` `rgb(52,74,82)``\n• #343b52\n``#343b52` `rgb(52,59,82)``\n• #3c3452\n``#3c3452` `rgb(60,52,82)``\nAnalogous Color\n• #4a5234\n``#4a5234` `rgb(74,82,52)``\n• #343b52\n``#343b52` `rgb(52,59,82)``\n• #523c34\n``#523c34` `rgb(82,60,52)``\nSplit Complementary Color\n• #3b5234\n``#3b5234` `rgb(59,82,52)``\n• #343b52\n``#343b52` `rgb(52,59,82)``\n• #52343b\n``#52343b` `rgb(82,52,59)``\n• #34524b\n``#34524b` `rgb(52,82,75)``\n• #343b52\n``#343b52` `rgb(52,59,82)``\n• #52343b\n``#52343b` `rgb(82,52,59)``\n• #524b34\n``#524b34` `rgb(82,75,52)``\n• #161923\n``#161923` `rgb(22,25,35)``\n• #202533\n``#202533` `rgb(32,37,51)``\n• #2a3042\n``#2a3042` `rgb(42,48,66)``\n• #343b52\n``#343b52` `rgb(52,59,82)``\n• #3e4662\n``#3e4662` `rgb(62,70,98)``\n• #485171\n``#485171` `rgb(72,81,113)``\n• #525d81\n``#525d81` `rgb(82,93,129)``\nMonochromatic Color\n\n# Alternatives to #343b52\n\nBelow, you can see some colors close to #343b52. Having a set of related colors can be useful if you need an inspirational alternative to your original color choice.\n\n• #344352\n``#344352` `rgb(52,67,82)``\n• #344052\n``#344052` `rgb(52,64,82)``\n• #343e52\n``#343e52` `rgb(52,62,82)``\n• #343b52\n``#343b52` `rgb(52,59,82)``\n• #343952\n``#343952` `rgb(52,57,82)``\n• #343652\n``#343652` `rgb(52,54,82)``\n• #353452\n``#353452` `rgb(53,52,82)``\nSimilar Colors\n\n# #343b52 Preview\n\nText with hexadecimal color #343b52\n\nThis text has a font color of #343b52.\n\n``<span style=\"color:#343b52;\">Text here</span>``\n#343b52 background color\n\nThis paragraph has a background color of #343b52.\n\n``<p style=\"background-color:#343b52;\">Content here</p>``\n#343b52 border color\n\nThis element has a border color of #343b52.\n\n``<div style=\"border:1px solid #343b52;\">Content here</div>``\nCSS codes\n``.text {color:#343b52;}``\n``.background {background-color:#343b52;}``\n``.border {border:1px solid #343b52;}``\n\n# Shades and Tints of #343b52\n\nA shade is achieved by adding black to any pure hue, while a tint is created by mixing white to any pure color. In this example, #06070a is the darkest color, while #fdfdfe is the lightest one.\n\n• #06070a\n``#06070a` `rgb(6,7,10)``\n• #0e1016\n``#0e1016` `rgb(14,16,22)``\n• #161822\n``#161822` `rgb(22,24,34)``\n• #1d212e\n``#1d212e` `rgb(29,33,46)``\n• #252a3a\n``#252a3a` `rgb(37,42,58)``\n• #2c3246\n``#2c3246` `rgb(44,50,70)``\n• #343b52\n``#343b52` `rgb(52,59,82)``\n• #3c445e\n``#3c445e` `rgb(60,68,94)``\n• #434c6a\n``#434c6a` `rgb(67,76,106)``\n• #4b5576\n``#4b5576` `rgb(75,85,118)``\n• #525e82\n``#525e82` `rgb(82,94,130)``\n• #5a668e\n``#5a668e` `rgb(90,102,142)``\n• #626f9a\n``#626f9a` `rgb(98,111,154)``\n• #6d79a2\n``#6d79a2` `rgb(109,121,162)``\n• #7984aa\n``#7984aa` `rgb(121,132,170)``\n• #858fb2\n``#858fb2` `rgb(133,143,178)``\n• #919ab9\n``#919ab9` `rgb(145,154,185)``\n• #9da5c1\n``#9da5c1` `rgb(157,165,193)``\n• #a9b0c8\n``#a9b0c8` `rgb(169,176,200)``\n• #b5bbd0\n``#b5bbd0` `rgb(181,187,208)``\n• #c1c6d8\n``#c1c6d8` `rgb(193,198,216)``\n• #cdd1df\n``#cdd1df` `rgb(205,209,223)``\n• #d9dce7\n``#d9dce7` `rgb(217,220,231)``\n• #e5e7ee\n``#e5e7ee` `rgb(229,231,238)``\n• #f1f2f6\n``#f1f2f6` `rgb(241,242,246)``\n• #fdfdfe\n``#fdfdfe` `rgb(253,253,254)``\nTint Color Variation\n\n# Tones of #343b52\n\nA tone is produced by adding gray to any pure hue. In this case, #3e4048 is the less saturated color, while #002086 is the most saturated one.\n\n• #3e4048\n``#3e4048` `rgb(62,64,72)``\n• #393e4d\n``#393e4d` `rgb(57,62,77)``\n• #343b52\n``#343b52` `rgb(52,59,82)``\n• #2f3857\n``#2f3857` `rgb(47,56,87)``\n• #2a365c\n``#2a365c` `rgb(42,54,92)``\n• #253361\n``#253361` `rgb(37,51,97)``\n• #1f3067\n``#1f3067` `rgb(31,48,103)``\n• #1a2d6c\n``#1a2d6c` `rgb(26,45,108)``\n• #152b71\n``#152b71` `rgb(21,43,113)``\n• #102876\n``#102876` `rgb(16,40,118)``\n• #0b257b\n``#0b257b` `rgb(11,37,123)``\n• #062280\n``#062280` `rgb(6,34,128)``\n• #002086\n``#002086` `rgb(0,32,134)``\nTone Color Variation\n\n# Color Blindness Simulator\n\nBelow, you can see how #343b52 is perceived by people affected by a color vision deficiency. This can be useful if you need to ensure your color combinations are accessible to color-blind users.\n\nMonochromacy\n• Achromatopsia 0.005% of the population\n• Atypical Achromatopsia 0.001% of the population\nDichromacy\n• Protanopia 1% of men\n• Deuteranopia 1% of men\n• Tritanopia 0.001% of the population\nTrichromacy\n• Protanomaly 1% of men, 0.01% of women\n• Deuteranomaly 6% of men, 0.4% of women\n• Tritanomaly 0.01% of the population" ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.50864863,"math_prob":0.80352086,"size":3684,"snap":"2022-40-2023-06","text_gpt3_token_len":1673,"char_repetition_ratio":0.125,"word_repetition_ratio":0.011070111,"special_character_ratio":0.56677526,"punctuation_ratio":0.23463687,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9911449,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2022-10-06T08:23:13Z\",\"WARC-Record-ID\":\"<urn:uuid:acce7a53-29f5-496d-afb9-76cefa17815e>\",\"Content-Length\":\"36119\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:ba5dc43d-1548-40c7-9379-0ef95fc312fc>\",\"WARC-Concurrent-To\":\"<urn:uuid:028b8f8c-c9e5-4472-9e8f-29e87f7a747c>\",\"WARC-IP-Address\":\"178.32.117.56\",\"WARC-Target-URI\":\"https://www.colorhexa.com/343b52\",\"WARC-Payload-Digest\":\"sha1:5FUCJIRNQCAF7V2UCHBFHPY6Z2O35WA3\",\"WARC-Block-Digest\":\"sha1:WDYJH4FHO2WJXUF7MVEJVFDDGNZXNKTW\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2022/CC-MAIN-2022-40/CC-MAIN-2022-40_segments_1664030337731.82_warc_CC-MAIN-20221006061224-20221006091224-00039.warc.gz\"}"}
https://www.polytechforum.com/control/about-difference-control-rule-10337-.htm	[ "Hi, friends:\nI am doing a research in a discrete time system with time delay. This system is implied in networked haptic display. this system has two haptic\ndisplays in the remote situation.\nI use position control rule, position: Y1=x1-(1/z)x2 and position: Y2=x2-(1/z)x1 as input to virtual environment, where x1 and x2 are represent position of two haptic system seperatedly, force of operators is input, (1/z) is time delay, when it touch the virtul environment,I hope to generate the virtual feedback force. Now, I only think about the simpliest virtual envirnoment, thus, I use virtual spring with paramater k to represent virtual environment. I use Y1-Y2=(1+(1/z))(x1-x2) and Y2-Y1=(1+1/z)(x2-x1) as virtual feedback force, but it does't work well, anybody can give me some hint, normally, when we use difference position control rules, how to get the force feedback! Thank you very much!" ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.8818589,"math_prob":0.8280065,"size":1742,"snap":"2021-43-2021-49","text_gpt3_token_len":449,"char_repetition_ratio":0.13521288,"word_repetition_ratio":0.9298893,"special_character_ratio":0.24971297,"punctuation_ratio":0.1284153,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9506832,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2021-10-18T07:46:31Z\",\"WARC-Record-ID\":\"<urn:uuid:2993274d-8b2a-48a5-b1c6-7293cfcfc2cc>\",\"Content-Length\":\"32402\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:e7560f91-48dd-45a3-b02c-0ed2e4fcb1ff>\",\"WARC-Concurrent-To\":\"<urn:uuid:dc824a36-20e2-4136-b6eb-c0a0a48c5e61>\",\"WARC-IP-Address\":\"67.222.8.252\",\"WARC-Target-URI\":\"https://www.polytechforum.com/control/about-difference-control-rule-10337-.htm\",\"WARC-Payload-Digest\":\"sha1:BR27SKIHWW76U6GOPXRJMGZSTOKY5FQ3\",\"WARC-Block-Digest\":\"sha1:FRZDR7POIQYMFQASSYCLGEM4WESGDXWR\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2021/CC-MAIN-2021-43/CC-MAIN-2021-43_segments_1634323585199.76_warc_CC-MAIN-20211018062819-20211018092819-00303.warc.gz\"}"}
https://math.stackexchange.com/questions/780433/a-question-about-convergence-interval-of-power-series	[ "A question about convergence interval of power series\n\nCould you give me some hint how to solve this problem:\n\nSuppose $a_n$ is sequence defined as $a_1=\\frac12,a_{n+1}=\\frac12\\left({a_n}^2+a_n\\right)$. I managed to prove that $a_n$ is decreasing sequence, $a_n\\to0$ and radius of convergence of power series $\\sum_{n\\ge1}a_nx^n$ is 2.\n\nThis is all quite standard staff, but how to find convergence interval ?\n\nI could not decide about convergence of $\\sum_{n\\ge1}2^na_n$ because: 1)ratio test is inconclusive; 2)root test=$2\\sqrt[n]{a_n}$ and I could not estimate $\\sqrt[n]{a_n}$; 3)I tried comparison test,knowing that $\\sum_{n\\ge1}2^na_{2^n}$ but could not compute the $\\lim_{n\\to\\infty}\\frac{a_n}{a_{2^n}}$, $a_n$ is decreasing, therefore from some n$a_n\\ge a_{2^n}$ but does $\\frac{a_n}{a_{2^n}}$ converge to finite number ?\n\nThanks.\n\nHint: Note that\n\n$$a_{n + 1} > \\frac 1 2 a_n$$\n\nfor every $n$, so a brief argument (perhaps with induction) shows that\n\n$$a_{n + 1} > \\frac 1 {2^n} a_n$$\n\nHence $2^{n + 1} a_{n + 1} > 2 a_n$. Consider something similar at $-2$.\n\n• Could you please add more how you last statement helps conclude about convergence ? – user97484 May 4 '14 at 5:15\n\nHint: The radius of convergence is defined as $1/\\beta$, where $\\beta = \\lim\\sup(\|a_n\|^{1/n})$.\n\n• So $lim_{n\\to \\infty}\\sqrt[n]{a_n}\\le limsup{\\sqrt[n]{a_n}}=\\frac12$ and thus $2\\lim_{n\\to \\infty}\\sqrt[n]{a_n}\\le 1$. This is still inconclusive. – user97484 May 4 '14 at 7:47" ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.82739925,"math_prob":0.99984634,"size":779,"snap":"2019-26-2019-30","text_gpt3_token_len":261,"char_repetition_ratio":0.11612903,"word_repetition_ratio":0.0,"special_character_ratio":0.33247754,"punctuation_ratio":0.09090909,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":1.0,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2019-06-16T21:14:12Z\",\"WARC-Record-ID\":\"<urn:uuid:f061feeb-dcab-4f09-8dfa-42bd55dc1fc6>\",\"Content-Length\":\"135115\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:4df1875b-ea3d-45de-94cf-14a4212a5068>\",\"WARC-Concurrent-To\":\"<urn:uuid:dcb1e42a-52b0-4a04-895e-7de3038c93e4>\",\"WARC-IP-Address\":\"151.101.1.69\",\"WARC-Target-URI\":\"https://math.stackexchange.com/questions/780433/a-question-about-convergence-interval-of-power-series\",\"WARC-Payload-Digest\":\"sha1:TRNU2C64PEKF7ZAVY2Q4DZCEAVPZZV7P\",\"WARC-Block-Digest\":\"sha1:SB5PTRBSU47LSKF2F2WW76LP2RPTX5LO\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2019/CC-MAIN-2019-26/CC-MAIN-2019-26_segments_1560627998298.91_warc_CC-MAIN-20190616202813-20190616224813-00019.warc.gz\"}"}
https://www.jobilize.com/course/section/the-collection-export-file-complete-version-by-openstax?qcr=www.quizover.com	[ "# 0.1 Importing and exporting to connexions (Page 3/3)\n\n Page 3 / 3\n\nThere are two options available for exporting modules:\n\n• Plain CNXML allows you to download the CNXML file for the module, which contains all of the module text and CNXML markup. The exported file is titled \" <moduleid> -plain.cnxml\", where <moduleid> is the module's ID number.\n• Zip File allows you to download the module's CNXML file (titled \"index.cnxml\") along with all attached resource files (such as embedded images, downloadable handouts, etc.).\nYou can also access the CNXML source code for any published module by appending \"source\" to the module's URL (e.g. (External Link) ).\n\n## Exporting a collection\n\nYou can use the collection export feature to download a copy of a collection for external use. There are two versions of the collection available:\n\n• The complete version , which includes information about the collection structure as well as the complete contents of all component modules.\n• The CollXML-only version , which includes only the structural information for the collection and does not include component modules.\nWhen viewing the published version of a collection online, you can export both versions of the collection from the metadata page - simply scroll to the bottom of the collection home page and click on the 'Metadata' link, and locate the appropriate link at the bottom of the 'Metadata' section.\n\nCollection authors can also download both versions of the collection export file from their workgroup after publishing a collection, and can download the CollXML-only version from any checked-out collection.\n\nThere is currently no support for importing collections, including those exported through this feature. At present, the collection export feature is provided primarily for developers interested in taking advantage of existing content for use with external platforms supporting CollXML/CNXML documents.\n\n## The collection export file (complete version)\n\nThe complete exported collection is packaged as a ZIP file titled \" <collectionid> _ <version> _complete.zip\", where <collectionid> and <version> are the collection's ID and version number, respectively. Once expanded, this version of the exported collection contains the following:\n\n• A CollXML document (titled \" <collectionid> _ <version> _collection.xml\") describing the collection's structure.\n• For each component module, a folder titled \" <moduleid> \" (the module's ID). Each of these folders contains:\n• The CNXML document for the module (titled \"index.cnxml\").\n• Any resource files, such as embedded images or downloadable handouts, that are attached to the module.\n\n## The collection export file (collxml-only version)\n\nThe structure-only version of the exported collection is available as a downloadable CollXML document titled \" <id> _ <version> _collection.xml\", where <id> and <version> are the collection's ID and version number, respectively. The CollXML file contains information about the collection including references to component modules, the order in which they are presented in the collection, and chapter/section information, along with a copy of the collection metadata.\n\nFor more information regarding the contents of the collection export file, please see the CollXML help page .\n\nexplain and give four Example hyperbolic function\n_3_2_1\nfelecia\n⅗ ⅔½\nfelecia\n_½+⅔-¾\nfelecia\nThe denominator of a certain fraction is 9 more than the numerator. If 6 is added to both terms of the fraction, the value of the fraction becomes 2/3. Find the original fraction. 2. The sum of the least and greatest of 3 consecutive integers is 60. What are the valu\n1. x + 6 2 -------------- = _ x + 9 + 6 3 x + 6 3 ----------- x -- (cross multiply) x + 15 2 3(x + 6) = 2(x + 15) 3x + 18 = 2x + 30 (-2x from both) x + 18 = 30 (-18 from both) x = 12 Test: 12 + 6 18 2 -------------- = --- = --- 12 + 9 + 6 27 3\nPawel\n2. (x) + (x + 2) = 60 2x + 2 = 60 2x = 58 x = 29 29, 30, & 31\nPawel\nok\nIfeanyi\non number 2 question How did you got 2x +2\nIfeanyi\ncombine like terms. x + x + 2 is same as 2x + 2\nPawel\nxx=2\nfelecia\n2+2x=\nfelecia\nMark and Don are planning to sell each of their marble collections at a garage sale. If Don has 1 more than 3 times the number of marbles Mark has, how many does each boy have to sell if the total number of marbles is 113?\nMark = x,. Don = 3x + 1 x + 3x + 1 = 113 4x = 112, x = 28 Mark = 28, Don = 85, 28 + 85 = 113\nPawel\nhow do I set up the problem?\nwhat is a solution set?\nHarshika\nfind the subring of gaussian integers?\nRofiqul\nhello, I am happy to help!\nAbdullahi\nhi mam\nMark\nfind the value of 2x=32\ndivide by 2 on each side of the equal sign to solve for x\ncorri\nX=16\nMichael\nWant to review on complex number 1.What are complex number 2.How to solve complex number problems.\nBeyan\nyes i wantt to review\nMark\nuse the y -intercept and slope to sketch the graph of the equation y=6x\nhow do we prove the quadratic formular\nDarius\nhello, if you have a question about Algebra 2. I may be able to help. I am an Algebra 2 Teacher\nthank you help me with how to prove the quadratic equation\nSeidu\nmay God blessed u for that. Please I want u to help me in sets.\nOpoku\nwhat is math number\n4\nTrista\nx-2y+3z=-3 2x-y+z=7 -x+3y-z=6\ncan you teacch how to solve that🙏\nMark\nSolve for the first variable in one of the equations, then substitute the result into the other equation. Point For: (6111,4111,−411)(6111,4111,-411) Equation Form: x=6111,y=4111,z=−411x=6111,y=4111,z=-411\nBrenna\n(61/11,41/11,−4/11)\nBrenna\nx=61/11 y=41/11 z=−4/11 x=61/11 y=41/11 z=-4/11\nBrenna\nNeed help solving this problem (2/7)^-2\nx+2y-z=7\nSidiki\nwhat is the coefficient of -4×\n-1\nShedrak\nthe operation is x * y =x + y/ 1+(x × y) show if the operation is commutative if x × y is not equal to -1\nA soccer field is a rectangle 130 meters wide and 110 meters long. The coach asks players to run from one corner to the other corner diagonally across. What is that distance, to the nearest tenths place.\nJeannette has \\$5 and \\$10 bills in her wallet. The number of fives is three more than six times the number of tens. Let t represent the number of tens. Write an expression for the number of fives.\nWhat is the expressiin for seven less than four times the number of nickels\nHow do i figure this problem out.\nhow do you translate this in Algebraic Expressions\nwhy surface tension is zero at critical temperature\nShanjida\nI think if critical temperature denote high temperature then a liquid stats boils that time the water stats to evaporate so some moles of h2o to up and due to high temp the bonding break they have low density so it can be a reason\ns.\nNeed to simplify the expresin. 3/7 (x+y)-1/7 (x-1)=\n. After 3 months on a diet, Lisa had lost 12% of her original weight. She lost 21 pounds. What was Lisa's original weight?\nGot questions? Join the online conversation and get instant answers!\n\n#### Get Jobilize Job Search Mobile App in your pocket Now!", null, "By Angela Eckman", null, "By Courntey Hub", null, "By Sam Luong", null, "By Madison Christian", null, "By Robert Murphy", null, "By OpenStax", null, "By OpenStax", null, "By Christine Zeelie", null, "By Katy Pratt", null, "By OpenStax" ]	[ null, "https://farm8.staticflickr.com/7338/11279010265_84fd771fb6_t.jpg", null, "https://farm9.staticflickr.com/8796/17127874488_9ce17b2f7d_t.jpg", null, "https://www.jobilize.com/quiz/thumb/power-engineering-4a-08-plant-fire-protec-pa08-quiz.png;jsessionid=4fvrk93rObJiYwgX0XIe6G3X9sd-BQ_FAW4UgXak.web001", null, "https://www.jobilize.com/quiz/thumb/microbiology-chapter-10-11-unit-3-test-3-by-madison.png;jsessionid=4fvrk93rObJiYwgX0XIe6G3X9sd-BQ_FAW4UgXak.web001", null, "https://www.jobilize.com/quiz/thumb/lessons-for-the-young-economist-by-dr-robert-murphy-mises.png;jsessionid=4fvrk93rObJiYwgX0XIe6G3X9sd-BQ_FAW4UgXak.web001", null, "https://www.jobilize.com/quiz/thumb/sociology-01-an-introduction-to-sociology-mcq-quiz-openstax.png;jsessionid=4fvrk93rObJiYwgX0XIe6G3X9sd-BQ_FAW4UgXak.web001", null, "https://www.jobilize.com/quiz/thumb/human-body-anatomy-physiology-mcq-quiz.png;jsessionid=4fvrk93rObJiYwgX0XIe6G3X9sd-BQ_FAW4UgXak.web001", null, "https://www.jobilize.com/quiz/thumb/grade-10-module-2-1-it-quiz-part-1-by-christine-zeelie.png;jsessionid=4fvrk93rObJiYwgX0XIe6G3X9sd-BQ_FAW4UgXak.web001", null, "https://www.jobilize.com/quiz/thumb/social-work-midterm-exam-by-katy-pratt.png;jsessionid=4fvrk93rObJiYwgX0XIe6G3X9sd-BQ_FAW4UgXak.web001", null, "https://www.jobilize.com/quiz/thumb/human-body-anatomy-physiology-mcq-quiz.png;jsessionid=4fvrk93rObJiYwgX0XIe6G3X9sd-BQ_FAW4UgXak.web001", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.8521057,"math_prob":0.8894213,"size":1483,"snap":"2021-04-2021-17","text_gpt3_token_len":289,"char_repetition_ratio":0.21298175,"word_repetition_ratio":0.09009009,"special_character_ratio":0.20296696,"punctuation_ratio":0.09361702,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.951982,"pos_list":[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20],"im_url_duplicate_count":[null,null,null,null,null,1,null,1,null,1,null,1,null,2,null,1,null,1,null,2,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2021-01-17T12:34:44Z\",\"WARC-Record-ID\":\"<urn:uuid:9b871172-a2ad-4080-8149-9d9bbb4de61e>\",\"Content-Length\":\"121267\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:f5075613-a160-4e11-923b-985d94d24d3a>\",\"WARC-Concurrent-To\":\"<urn:uuid:ac4b036d-e309-4ce6-857b-13e24a58f83b>\",\"WARC-IP-Address\":\"207.38.87.179\",\"WARC-Target-URI\":\"https://www.jobilize.com/course/section/the-collection-export-file-complete-version-by-openstax?qcr=www.quizover.com\",\"WARC-Payload-Digest\":\"sha1:JJWAEGRZFGKRAXCJLHVIV4JIXHCUZOIM\",\"WARC-Block-Digest\":\"sha1:D66FSVTIAC6HGJRZ7F5CXNWN5GTBPYNM\",\"WARC-Identified-Payload-Type\":\"application/xhtml+xml\",\"warc_filename\":\"/cc_download/warc_2021/CC-MAIN-2021-04/CC-MAIN-2021-04_segments_1610703512342.19_warc_CC-MAIN-20210117112618-20210117142618-00146.warc.gz\"}"}
https://www.emathzone.com/tutorials/geometry/area-of-a-segment.html	[ "# Area of a Segment\n\nA segment is a portion of a circle which is cut off by a straight line not passing through the center. The segment smaller than the semi-circle is called the minor segment and the segment larger than the semi-circle is called the major segment.", null, "(a) The area of the minor segment when angle $\\theta$ and radius $r$ are given:\n\nArea of segment $=$ area of sector $AOBC$ $\\pm$ area of $\\Delta AOB$\n$= \\frac{1}{2}{r^2}\\theta \\pm \\frac{1}{2}{r^2}\\sin \\theta$\n$= \\frac{1}{2}{r^2}(\\theta – \\sin \\theta )$", null, "Now the area of the major segment $=$ area of circle $–$ area of the minor segment\n$= \\frac{1}{2}{r^2}(2\\pi – \\theta + \\sin \\theta )$\n\nExample:\n\nA chord $AB$ of a circle of radius $15$cm makes an angle of ${60^ \\circ }$at the center of the circle. Find the area of the major and minor segment.\n\nSolution:", null, "Given that $\\angle AOB = {60^ \\circ }$, radius, $r = 15$cm\n\n$\\therefore$ area of the sector $OAB$ $= \\frac{\\theta }{{360}} \\times \\pi {r^2}$\n$= \\frac{{60}}{{360}} \\times 3.1415 \\times 15 \\times 15 = 117.75$ square cm\n$\\therefore$ area of $\\Delta OAB$ $= \\frac{1}{2}{r^2}\\sin \\theta$\n$= \\frac{1}{2} \\times 15 \\times 15 \\times \\sin {60^ \\circ } = \\frac{{225 \\times 1.73}}{4} = 97.31$ square cm\nArea of the minor segment $=$ area of sector $OAB$$–$ area of $\\Delta OAB$\n$= 117.75 – 97.31 = 20.44$ square cm\nArea of the circle $= \\pi {r^2}$\n$= 3.1415 \\times {(15)^2} = 3.1415 \\times 225 = 706.5$ square cm\nArea of the major segment $=$ area of the circle $–$ area of the minor segment\n$= 706.5 – 20.4 = 686.1$ square cm.\n\n(b) The area of a segment when the height and length of the chord of the segment are given:", null, "Let $r =$ be the radius of the circle\n$h =$ be the height of the segment\n$c =$ be the length of the chord\n\nWe note that $ODB$ is a right triangle; the hypotenuse is $OB = r$ and the other two sides are $OD = r – h$ and $BD = c/2$\n$\\therefore$ by Pythagorean Theorem\n\n${\\left( {\\frac{c}{2}} \\right)^2} + {\\left( {r – h} \\right)^2} = {r^2}$\n${\\left( {\\frac{c}{2}} \\right)^2} + {r^2} – 2rh + {h^2} = {r^2}$\n${\\left( {\\frac{c}{2}} \\right)^2} – 2rh + {h^2} = 0$\n\nSolving for $r$, $c$ and $h$, we obtain the following formulas:\n\n$r = \\frac{{{{\\left( {\\frac{c}{2}} \\right)}^2} + {h^2}}}{{2h}}$ — (1)\n$h = r \\pm \\sqrt {{r^2} – {{\\left( {\\frac{c}{2}} \\right)}^2}}$ — (2)\n$c = 2\\sqrt {h\\left( {2r – h} \\right)}$ — (3)\n\nNote:\n$r + \\sqrt {{r^2} – {{\\left( {\\frac{c}{2}} \\right)}^2}}$ gives the height of the major segment\n$r – \\sqrt {{r^2} – {{\\left( {\\frac{c}{2}} \\right)}^2}}$ gives the height of the minor segment\n\nMany formulas are given for finding the approximate area of a segment. Two of the common methods are:\n\nMethod-I: $A = \\frac{{2hc}}{3} + \\frac{{{h^2}}}{{2c}} = \\frac{h}{{6c}}\\left( {3{h^2} + 4{c^2}} \\right)$\nNote: If the height of the segment is less $\\frac{1}{{10}}$ than the radius of the circle, then $A = \\frac{{2hc}}{3}$.\n\nMethod-II: $A = \\frac{{4{h^2}}}{3}\\sqrt {\\frac{{2r}}{h} – 0.608}$\n\nExample:\n\nIf the chord of the segment of a circle is $66$ cm and the height of the segment is $10$ cm, find the radius of the circle.\n\nSolution:\n\nGiven that, $c = 66$cm, $h = 10$cm\n$\\therefore$ $r = \\frac{{{{\\left( {\\frac{c}{2}} \\right)}^2} + {h^2}}}{{2h}}$\n$= \\frac{{{{(33)}^2} + {{(10)}^2}}}{{2 \\times 10}} = \\frac{{1089 + 100}}{{20}} = 59.45$cm\n\nExample:\n\nFind the area of the segment whose chord is $10$ cm and whose height is $1.5$ cm.\n\nSolution:\n\nGiven that $h = 1.5$cm, $c = 10$cm\n$\\therefore$ $A = \\frac{{2hc}}{3} + \\frac{{{h^2}}}{{2c}}$\n$= \\frac{2}{3}(1.5)(10) + \\frac{{{{(1.5)}^2}}}{{2 \\times 1}} = 10.17$ square cm." ]	[ null, "https://www.emathzone.com/wp-content/uploads/2015/01/segment-01.gif", null, "https://www.emathzone.com/wp-content/uploads/2015/01/segment-02.gif", null, "https://www.emathzone.com/wp-content/uploads/2015/01/segment-03.gif", null, "https://www.emathzone.com/wp-content/uploads/2015/01/segment-04.gif", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.7181188,"math_prob":1.0000093,"size":3722,"snap":"2019-51-2020-05","text_gpt3_token_len":1405,"char_repetition_ratio":0.18612157,"word_repetition_ratio":0.13517666,"special_character_ratio":0.48387963,"punctuation_ratio":0.07438017,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":1.0000095,"pos_list":[0,1,2,3,4,5,6,7,8],"im_url_duplicate_count":[null,6,null,6,null,6,null,6,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2019-12-06T20:24:11Z\",\"WARC-Record-ID\":\"<urn:uuid:dc957536-987e-44e0-82a6-af0953212fb0>\",\"Content-Length\":\"44086\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:4fc3a7bc-f469-4ba4-8899-1d7a4443d556>\",\"WARC-Concurrent-To\":\"<urn:uuid:58fd07e6-173c-490d-8765-d37b160df5eb>\",\"WARC-IP-Address\":\"151.139.241.10\",\"WARC-Target-URI\":\"https://www.emathzone.com/tutorials/geometry/area-of-a-segment.html\",\"WARC-Payload-Digest\":\"sha1:FIJW6LKJXRKWJTGP4K7B3ZC5VEYW62VW\",\"WARC-Block-Digest\":\"sha1:T5GJHA64SISQCXFWYYC2W4437S5LSQ4O\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2019/CC-MAIN-2019-51/CC-MAIN-2019-51_segments_1575540490972.13_warc_CC-MAIN-20191206200121-20191206224121-00433.warc.gz\"}"}
https://proofwiki.org/wiki/Double_Negation_with_Erroneous_Conjunction	[ "# Double Negation with Erroneous Conjunction\n\nJump to navigation Jump to search\n\n## Source Work\n\nChapter $1$: Informal statement calculus\n$1.2$. Truth functions and truth tables: Example $1.6 \\ \\text{(c)}$\n\n## Mistake\n\n$\\paren {p \\leftrightarrow \\paren {\\land \\paren {\\sim p} } }$ is a tautology.\n\n## Correction\n\nAs it stands, this statement is meaningless, as $\\land$ is a binary operator.\n\nThe most obvious assumption is that $\\land$ is a typo for $\\sim$, and that:\n\n$\\paren {p \\leftrightarrow \\paren {\\sim \\paren {\\sim p} } }$ is a tautology.\n\nis meant.\n\nSee Double Negation/Formulation 2 for an analysis of this.\n\nNote that in Alan G. Hamilton: Logic for Mathematicians (2nd ed.):\n\n$\\sim$ is the symbol used for $\\neg$, the logical negation operator\n$\\leftrightarrow$ is the symbol used for $\\iff$, the biconditional operator." ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.7383418,"math_prob":0.99767125,"size":1082,"snap":"2021-04-2021-17","text_gpt3_token_len":303,"char_repetition_ratio":0.097402595,"word_repetition_ratio":0.22023809,"special_character_ratio":0.30406654,"punctuation_ratio":0.20772947,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9981847,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2021-04-12T07:41:59Z\",\"WARC-Record-ID\":\"<urn:uuid:89721b4a-1320-4e7a-873b-ea073118c901>\",\"Content-Length\":\"36270\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:38b0ac06-90a2-41e8-a762-dc951533ed52>\",\"WARC-Concurrent-To\":\"<urn:uuid:398c58e3-56a5-4dec-ab36-320a5f228ed7>\",\"WARC-IP-Address\":\"104.21.84.229\",\"WARC-Target-URI\":\"https://proofwiki.org/wiki/Double_Negation_with_Erroneous_Conjunction\",\"WARC-Payload-Digest\":\"sha1:ESOLTWY6SWAJGYW5HHAM5ZFYAT4GUZYA\",\"WARC-Block-Digest\":\"sha1:IVVJAMFHEG4JDUGFUAD2YI6GTHPWKXNW\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2021/CC-MAIN-2021-17/CC-MAIN-2021-17_segments_1618038066613.21_warc_CC-MAIN-20210412053559-20210412083559-00352.warc.gz\"}"}
https://calculus.subwiki.org/wiki/Logarithmic_derivative	[ "# Logarithmic derivative\n\n## Definition\n\n### At a point\n\nSuppose", null, "$f$ is a function and", null, "$x_0$ is a point in the interior of the domain of", null, "$f$ such that the derivative", null, "$f'(x_0)$ exists, and", null, "$f(x_0) \\ne 0$ (note that", null, "$f'(x_0)$ may be zero or nonzero). The logarithmic derivative of", null, "$f$ at", null, "$x_0$ is defined as:", null, "$\\! \\frac{f'(x_0)}{f(x_0)}$\n\nAlternatively, it can be defined as:", null, "$\\! \\frac{d}{dx}(\\ln\|f(x)\|)\|_{x = x_0}$\n\n### As a function\n\nSuppose", null, "$f$ is a function of one variable. The logarithmic derivative of", null, "$f$ is the function:", null, "$\\! x \\mapsto \\frac{f'(x)}{f(x)}$\n\nAlternatively, it is the function:", null, "$\\! x \\mapsto \\frac{d}{dx}(\\ln\|f(x)\|)$\n\nThe domain of this function is precisely the set of points", null, "$x$ for which", null, "$f(x) \\ne 0$ and the derivative", null, "$f'$ exists.\n\nMORE ON THE WAY THIS DEFINITION OR FACT IS PRESENTED: We first present the version that deals with a specific point (typically with a", null, "$\\{ \\}_0$ subscript) in the domain of the relevant functions, and then discuss the version that deals with a point that is free to move in the domain, by dropping the subscript. Why do we do this?\nThe purpose of the specific point version is to emphasize that the point is fixed for the duration of the definition, i.e., it does not move around while we are defining the construct or applying the fact. However, the definition or fact applies not just for a single point but for all points satisfying certain criteria, and thus we can get further interesting perspectives on it by varying the point we are considering. This is the purpose of the second, generic point version.\nDIVISION BY ZERO: The formulation of this definition or fact involves a denominator that could a priori be zero. To avoid this, we stipulate conditions in our formulation that exclude the points or situations where the denominator is zero. It may or may not be possible to use concepts of limits to extend the definition somewhat to some of the excluded points." ]	[ null, "https://calculus.subwiki.org/w/images/math/8/f/a/8fa14cdd754f91cc6554c9e71929cce7.png ", null, "https://calculus.subwiki.org/w/images/math/0/b/2/0b21a666a81629962ade8afd967826ed.png ", null, "https://calculus.subwiki.org/w/images/math/8/f/a/8fa14cdd754f91cc6554c9e71929cce7.png ", null, "https://calculus.subwiki.org/w/images/math/1/6/5/165e320cd480fcaabf49a13eb4ac4424.png ", null, "https://calculus.subwiki.org/w/images/math/d/7/0/d70d83f0a3665cf9315368cec8ecefd9.png ", null, "https://calculus.subwiki.org/w/images/math/1/6/5/165e320cd480fcaabf49a13eb4ac4424.png ", null, "https://calculus.subwiki.org/w/images/math/8/f/a/8fa14cdd754f91cc6554c9e71929cce7.png ", null, "https://calculus.subwiki.org/w/images/math/0/b/2/0b21a666a81629962ade8afd967826ed.png ", null, "https://calculus.subwiki.org/w/images/math/5/1/4/514ec5e2236dea4bc29b069858b66ef9.png ", null, "https://calculus.subwiki.org/w/images/math/b/9/6/b961672bafaf4ed6cc5ea4010d921e69.png ", null, "https://calculus.subwiki.org/w/images/math/8/f/a/8fa14cdd754f91cc6554c9e71929cce7.png ", null, "https://calculus.subwiki.org/w/images/math/8/f/a/8fa14cdd754f91cc6554c9e71929cce7.png ", null, "https://calculus.subwiki.org/w/images/math/c/b/7/cb7a892eaf9482d92c8f9f27ff8eb1a0.png ", null, "https://calculus.subwiki.org/w/images/math/4/3/c/43cb5b42b4f97aa13aedfcc437a0e0ed.png ", null, "https://calculus.subwiki.org/w/images/math/9/d/d/9dd4e461268c8034f5c8564e155c67a6.png ", null, "https://calculus.subwiki.org/w/images/math/5/9/1/5915ef67690859969c90907b6ec384ff.png ", null, "https://calculus.subwiki.org/w/images/math/6/b/d/6bda8af54c40bc23ed858e9e9f5c11d2.png ", null, "https://calculus.subwiki.org/w/images/math/d/d/2/dd2e2b7e15a19ea54599b73f52f77bed.png ", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.9052319,"math_prob":0.99879366,"size":1680,"snap":"2019-51-2020-05","text_gpt3_token_len":343,"char_repetition_ratio":0.14797136,"word_repetition_ratio":0.013840831,"special_character_ratio":0.19404761,"punctuation_ratio":0.09259259,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9996588,"pos_list":[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36],"im_url_duplicate_count":[null,null,null,null,null,null,null,null,null,2,null,null,null,null,null,null,null,1,null,1,null,null,null,null,null,1,null,1,null,null,null,1,null,null,null,null,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2019-12-06T05:48:43Z\",\"WARC-Record-ID\":\"<urn:uuid:ea337526-8115-47d9-a750-6cf96c7a8679>\",\"Content-Length\":\"23397\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:b243a4e2-bf6b-4d7e-9961-1b41a0bf4731>\",\"WARC-Concurrent-To\":\"<urn:uuid:e4696f8c-ee36-43b8-b07b-5aad1319066b>\",\"WARC-IP-Address\":\"96.126.114.7\",\"WARC-Target-URI\":\"https://calculus.subwiki.org/wiki/Logarithmic_derivative\",\"WARC-Payload-Digest\":\"sha1:AO24UXXYHXF6MIW2HS7CLBIKGKAKQ3KS\",\"WARC-Block-Digest\":\"sha1:W4M2V6CWWBLYOISZ2WGOEKGTIUMUHNDU\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2019/CC-MAIN-2019-51/CC-MAIN-2019-51_segments_1575540484815.34_warc_CC-MAIN-20191206050236-20191206074236-00185.warc.gz\"}"}
https://www.geeksforgeeks.org/trimorphic-number/?ref=lbp	[ "", null, "Open in App\nNot now\n\n# Trimorphic Number\n\n• Last Updated : 09 Aug, 2022\n\nGiven a number N, the task is to check whether the number is Trimorphic number or not. A number is called Trimorphic number if and only if its cube ends in the same digits as the number itself. In other words, number appears at the end of its cube.\n\nExamples:\n\n```Input : 5\nOutput : trimorphic\nExplanation: 555=125\n\nInput : 24\nOutput : trimorphic\nExplanation: 242424=13824\n\nInput:10\nOutput: not trimorphic\nExplanation: 101010=1000 ```\nRecommended Practice\n\nApproach:\n\n```1. Store the cube of given number.\n2. Loop until N becomes 0 as we have to match\nall digits with its cube.\ni) Check if (n%10 == cube%10) i.e. last digit\nof number = last digit of cube or not\na) if not equal, return false.\nii) Otherwise continue i.e. reduce number and\ncube i.e. n = n/10 and cube = cube/10;\n3- Return true if all digits matched.```\n\nBelow is the implementation of the approach.\n\n## C++\n\n `// C++ program to check if a``// number is Trimorphic``#include ``using` `namespace` `std;` `// Function to check Trimorphic number``bool` `isTrimorphic(``int` `N)``{`` ``// Store the cube`` ``int` `cube = N * N * N;` ` ``// Start Comparing digits`` ``while` `(N > 0)`` ``{`` ``// Return false, if any digit`` ``// of N doesn't match with`` ``// its cube's digits from last`` ``if` `(N % 10 != cube % 10)`` ``return` `false``;` ` ``// Reduce N and cube`` ``N /= 10;`` ``cube /= 10;`` ``}` ` ``return` `true``;``}` `// Driver code``int` `main()``{`` ``int` `N = 24;` ` ``isTrimorphic(N) ? cout << ``\"trimorphic\"` `:`` ``cout << ``\"not trimorphic\"``;` ` ``return` `0;``}`\n\n## Java\n\n `// Java program to check if a``// number is Trimorphic`` ` `class` `GFG {`` ` ` ``// Function to check Trimorphic number`` ``static` `boolean` `isTrimorphic(``int` `N)`` ``{`` ``// Store the cube`` ``int` `cube = N * N * N;`` ` ` ``// Start Comparing digits`` ``while` `(N > ``0``) {`` ` ` ``// Return false, if any digit`` ``// of N doesn't match with`` ``// its cube's digits from last`` ``if` `(N % ``10` `!= cube % ``10``)`` ``return` `false``;`` ` ` ``// Reduce N and cube`` ``N /= ``10``;`` ``cube /= ``10``;`` ``}`` ` ` ``return` `true``;`` ``}`` ` `// Driver code``public` `static` `void` `main(String[] args)`` ``{`` ``int` `N = ``24``;`` ` ` ``if` `(isTrimorphic(N) == ``true``)`` ``System.out.println(``\"trimorphic\"``);`` ``else`` ``System.out.println(``\"not trimorphic\"``);`` ``}``}`` ` `//This article is contributed by prerna saini.`\n\n## Python3\n\n `# Python 3 program to check``# if a number is Trimorphic` ` ` `# Function to check``# Trimorphic number``def` `isTrimorphic(N) :`` ` ` ``# Store the cube`` ``cube ``=` `N ``` `N ``` `N`` ` ` ``# Start Comparing digits`` ``while` `(N > ``0``) :`` ` ` ``# Return false, if any digit`` ``# of N doesn't match with`` ``# its cube's digits from last`` ``if` `(N ``%` `10` `!``=` `cube ``%` `10``) :`` ``return` `False`` ` ` ``# Reduce N and cube`` ``N ``=` `N ``/``/` `10`` ``cube ``=` `cube ``/``/` `10`` ` ` ``return` `True` `# Driver code``N ``=` `24``if``(isTrimorphic(N)) :`` ``print``(``\"trimorphic\"``)``else` `:`` ``print``(``\"not trimorphic\"``)`` ` ` ` `# This code is contributed by Nikita tiwari.`\n\n## C#\n\n `// C# program to check if``// a number is Trimorphic``using` `System;` `class` `GFG {` `// Function to check Trimorphic number`` ``static` `bool` `isTrimorphic(``int` `N)`` ``{`` ``// Store the cube`` ``int` `cube = N * N * N;` ` ``// Start Comparing digits`` ``while` `(N > 0) {` ` ``// Return false, if any digit`` ``// of N doesn't match with`` ``// its cube's digits from last`` ``if` `(N % 10 != cube % 10)`` ``return` `false``;` ` ``// Reduce N and cube`` ``N /= 10;`` ``cube /= 10;`` ``}` ` ``return` `true``;`` ``}` `// Driver code``public` `static` `void` `Main()`` ``{`` ``int` `N = 24;` ` ``if` `(isTrimorphic(N) == ``true``)`` ``Console.Write(``\"trimorphic\"``);`` ``else`` ``Console.Write(``\"not trimorphic\"``);`` ``}``}` `// This article is contributed``// by Smitha Dinesh Semwal.`\n\n## PHP\n\n ` 0)`` ``{`` ` ` ``// Return false, if any digit`` ``// of N doesn't match with`` ``// its cube's digits from last`` ``if` `(``\\$N` `% 10 != ``\\$cube` `% 10)`` ``return` `-1;` ` ``// Reduce N and cube`` ``\\$N` `/= 10;`` ``\\$cube` `/= 10;`` ``}` ` ``return` `1;``}` `// Driver code`` ``\\$N` `= 24;` ` ``\\$r` `= isTrimorphic(``\\$N``) ? ``\"trimorphic\"` `:`` ``\"not trimorphic\"``;`` ``echo` `\\$r``;` `// This code is contributed by aj_36``?>`\n\n## Javascript\n\n ``\n\nOutput:\n\n`trimorphic`\n\nFind nth Trimorphic Number\nExamples:\n\n```Input : 10\nOutput : 51\n\nInput : 11\nOutput : 75```\n\n## C++\n\n `// CPP program to find nth``// Trimorphic number``#include ``using` `namespace` `std;` `# define INT_MAX 2147483647``// Functions to find nth``// Trimorphic number``bool` `checkTrimorphic(``int` `num)``{`` ``int` `cube = num * num * num;`` ``// Comparing the digits`` ``while``(num > 0)`` ``{`` ``// Return false, if any digit`` ``// of num doesn't match with`` ``// its cube's digits from last`` ``if` `(num % 10 != cube % 10)`` ``return` `false``;` ` ``// Reduce num and cube`` ``num /= 10;`` ``cube /= 10;`` ``}`` ``return` `true``;``}``int` `nthTrimorphic(``int` `n)``{`` ` ` ``int` `count = 0;`` ``// Check in max int size`` ``for``(``int` `i = 0; i < INT_MAX; i++)`` ``{`` ``//check number is Trimorphic or not`` ``if``(checkTrimorphic(i))`` ``count++;`` ` ` ``// if counter is equal to the`` ``// n then return nth number`` ``if``(count == n)`` ``return` `i;`` ``}``}` `// Driver code``int` `main()``{`` ``int` `n = 9;`` ``cout<< nthTrimorphic(n);`` ``return` `0;``}` `// This code is contributed by jaingyayak.`\n\n## Java\n\n `// Java program to find nth``// Trimorphic number` `class` `GFG``{``static` `int` `INT_MAX = ``2147483647``;` `// Functions to find nth``// Trimorphic number``static` `boolean` `checkTrimorphic(``int` `num)``{`` ``int` `cube = num * num * num;`` ` ` ``// Comparing the digits`` ``while``(num > ``0``)`` ``{`` ``// Return false, if any`` ``// digit of num doesn't`` ``// match with its cube's`` ``// digits from last`` ``if` `(num % ``10` `!= cube % ``10``)`` ``return` `false``;` ` ``// Reduce num and cube`` ``num /= ``10``;`` ``cube /= ``10``;`` ``}`` ``return` `true``;``}` `static` `int` `nthTrimorphic(``int` `n)``{`` ` ` ``int` `count = ``0``;`` ` ` ``// Check in max int size`` ``for``(``int` `i = ``0``; i < INT_MAX; i++)`` ``{`` ``// check number is`` ``// Trimorphic or not`` ``if``(checkTrimorphic(i))`` ``count++;`` ` ` ``// if counter is equal to`` ``// the n then return nth number`` ``if``(count == n)`` ``return` `i;`` ``}`` ``return` `-``1``;``}` `// Driver code``public` `static` `void` `main(String[] args)``{`` ``int` `n = ``9``;`` ``System.out.println(nthTrimorphic(n));``}``}` `// This code is contributed by mits.`\n\n## Python3\n\n `# Python3 program to find``# nth Trimorphic number``import` `sys``# Functions to find nth``# Trimorphic number``def` `checkTrimorphic(num):` ` ``cube ``=` `num ``` `num ``` `num`` ` ` ``# Comparing the digits`` ``while``(num > ``0``):`` ``# Return false, if any digit`` ``# of num doesn't match with`` ``# its cube's digits from last`` ``if` `(num ``%` `10` `!``=` `cube ``%` `10``):`` ``return` `False` ` ``# Reduce num and cube`` ``num ``=` `int``(num ``/` `10``)`` ``cube ``=` `int``(cube ``/` `10``)`` ``return` `True` `def` `nthTrimorphic(n):`` ``count ``=` `0`` ` ` ``# Check in max int size`` ``for` `i ``in` `range``(sys.maxsize):`` ``# check number is`` ``# Trimorphic or not`` ``if``(checkTrimorphic(i)):`` ``count``+``=``1`` ` ` ``# if counter is equal to`` ``# the n then return nth`` ``# number`` ``if``(count ``=``=` `n):`` ``return` `i` `# Driver code``if` `__name__``=``=``'__main__'``:`` ``n ``=` `9`` ``print``(nthTrimorphic(n))` `# This code is contributed``# by mits.`\n\n## C#\n\n `// C# program to find nth``// Trimorphic number``using` `System;` `class` `GFG``{``static` `int` `INT_MAX = 2147483647;` `// Functions to find nth``// Trimorphic number``static` `bool` `checkTrimorphic(``int` `num)``{`` ``int` `cube = num * num * num;`` ` ` ``// Comparing the digits`` ``while``(num > 0)`` ``{`` ``// Return false, if any`` ``// digit of num doesn't`` ``// match with its cube's`` ``// digits from last`` ``if` `(num % 10 != cube % 10)`` ``return` `false``;` ` ``// Reduce num and cube`` ``num /= 10;`` ``cube /= 10;`` ``}`` ``return` `true``;``}` `static` `int` `nthTrimorphic(``int` `n)``{`` ` ` ``int` `count = 0;`` ` ` ``// Check in max int size`` ``for``(``int` `i = 0; i < INT_MAX; i++)`` ``{`` ``// check number is`` ``// Trimorphic or not`` ``if``(checkTrimorphic(i))`` ``count++;`` ` ` ``// if counter is equal to`` ``// the n then return nth number`` ``if``(count == n)`` ``return` `i;`` ``}`` ``return` `-1;``}` `// Driver code``static` `int` `Main()``{`` ``int` `n = 9;`` ``Console.Write(nthTrimorphic(n));`` ``return` `0;``}``}` `// This code is contributed by mits.`\n\n## PHP\n\n ` 0)`` ``{`` ``// Return false, if any digit`` ``// of num doesn't match with`` ``// its cube's digits from last`` ``if` `(``\\$num` `% 10 != ``\\$cube` `% 10)`` ``return` `false;` ` ``// Reduce num and cube`` ``\\$num` `= (int)(``\\$num` `/ 10);`` ``\\$cube` `= (int)(``\\$cube` `/ 10);`` ``}`` ``return` `true;``}` `function` `nthTrimorphic(``\\$n``)``{`` ``\\$count` `= 0;`` ` ` ``// Check in max int size`` ``for``(``\\$i` `= 0;`` ``\\$i` `< PHP_INT_MAX; ``\\$i``++)`` ``{`` ``// check number is`` ``// Trimorphic or not`` ``if``(checkTrimorphic(``\\$i``))`` ``\\$count``++;`` ` ` ``// if counter is equal to`` ``// the n then return nth`` ``// number`` ``if``(``\\$count` `== ``\\$n``)`` ``return` `\\$i``;`` ``}``}` `// Driver code``\\$n` `= 9;``echo` `nthTrimorphic(``\\$n``);` `// This code is contributed``// by mits.``?>`\n\n## Javascript\n\n ``\n\nOutput:\n\n`49`\n\nTime complexity: O(INT_MAX*logN)\n\nspace complexity: O(1)\n\nMy Personal Notes arrow_drop_up" ]	[ null, "https://media.geeksforgeeks.org/gfg-gg-logo.svg", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.51949644,"math_prob":0.97461236,"size":9671,"snap":"2022-40-2023-06","text_gpt3_token_len":3110,"char_repetition_ratio":0.18351091,"word_repetition_ratio":0.49796334,"special_character_ratio":0.35601282,"punctuation_ratio":0.12901376,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9996884,"pos_list":[0,1,2],"im_url_duplicate_count":[null,null,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2023-02-06T17:14:13Z\",\"WARC-Record-ID\":\"<urn:uuid:89a1f44d-b1eb-473c-9e77-2863add5195f>\",\"Content-Length\":\"252524\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:6d00c787-4f11-4eb1-8cc5-64008f46cbec>\",\"WARC-Concurrent-To\":\"<urn:uuid:78269672-cf02-49aa-84a8-04b62a7c84fb>\",\"WARC-IP-Address\":\"23.46.238.224\",\"WARC-Target-URI\":\"https://www.geeksforgeeks.org/trimorphic-number/?ref=lbp\",\"WARC-Payload-Digest\":\"sha1:4UX4TOPLCEFER5CWZVRETPXT3YJGPEOM\",\"WARC-Block-Digest\":\"sha1:DOU76QEH5EI4RDA4VO3FIPC6V7EBUEYA\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2023/CC-MAIN-2023-06/CC-MAIN-2023-06_segments_1674764500356.92_warc_CC-MAIN-20230206145603-20230206175603-00860.warc.gz\"}"}
https://appdividend.com/2021/05/12/how-to-convert-python-char-to-int/	[ "AppDividend\nLatest Code Tutorials\n\n# How to Convert Python char to int\n\nIn Python, all characters have an integer ASCII value. A char can either be represented as an ASCII value or, if the char comprises digits only, as the number it represents.\n\nThe conversion from char to int is an explicit type conversion in which the data type is manually modified by the user as they want.\n\n## Python char to int\n\nTo convert Python char to int, use the ord() method. The ord() function accepts a character and returns the ASCII value of that character.\n\nThe ord() is a built-in function that returns the number representing the Unicode code of a specified character.\n\n### Example\n\nLet’s define a character and convert that character into an integer using the ord() function.\n\n```charData = \"K\"\n\nprint(ord(charData))```\n\n#### Output\n\n`75`\n\nThat means the ASCII value of the K character is 75.\n\nThe ASCII value of capital K is 75, but the ASCII value of small k is different.\n\n```charData = \"k\"\n\nprint(ord(charData))```\n\n#### Output\n\n`107`\n\nYou can see that it’s case-sensitive. Captial and small characters have different ASCII values.\n\nLet’s take another character and convert it into an integer.\n\n```charData = \"B\"\n\nprint(ord(charData))\n```\n\n`107`\n\n## Converting Python int to char\n\nTo convert int to char in Python, use the chr() function. The chr() is a built-in function that takes an integer as an argument and returns a string representing a character at that code point.\n\nIn the above section, we converted chr to int, and in this section, we will convert into int to chr data type.\n\n```intData = 75\n\nprint(chr(intData))\n```\n\n#### Output\n\n`K`\n\nIn this example, we pass the ASCII value to the chr() function, converting it into a character. The character value of 75 is K because the chr() function takes an ASCII value as an argument and converts it into a character value.\n\nThis is different from converting String to an integer since we are here dealing with characters and not strings. To convert string to int, use the int() function, and converting int to string, use the str() function.\n\nThat is it for converting char to int in Python tutorial.\n\nThis site uses Akismet to reduce spam. Learn how your comment data is processed." ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.8109881,"math_prob":0.9636453,"size":2027,"snap":"2021-21-2021-25","text_gpt3_token_len":447,"char_repetition_ratio":0.17597628,"word_repetition_ratio":0.0057306592,"special_character_ratio":0.22397631,"punctuation_ratio":0.085642315,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9911523,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2021-06-25T07:34:55Z\",\"WARC-Record-ID\":\"<urn:uuid:59a16d29-d858-4339-b30d-312c53f4b9ca>\",\"Content-Length\":\"105481\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:a040b7b4-6af1-4edd-a03c-b7670ae563bd>\",\"WARC-Concurrent-To\":\"<urn:uuid:d9133ed8-d33d-4d07-aaa8-ac1b30799460>\",\"WARC-IP-Address\":\"50.16.49.81\",\"WARC-Target-URI\":\"https://appdividend.com/2021/05/12/how-to-convert-python-char-to-int/\",\"WARC-Payload-Digest\":\"sha1:XQREB5YN7N3TRBV2STMFANST2ZMZTUDU\",\"WARC-Block-Digest\":\"sha1:HRUODH7I54OTHL6VYAWKFXTVGFPJFEHK\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2021/CC-MAIN-2021-25/CC-MAIN-2021-25_segments_1623487622113.11_warc_CC-MAIN-20210625054501-20210625084501-00375.warc.gz\"}"}
https://help.scilab.org/docs/6.1.0/pt_BR/scicos_simulate.html	[ "Scilab Home page \| Wiki \| Bug tracker \| Forge \| Mailing list archives \| ATOMS \| File exchange\nChange language to: English - Français - 日本語 - Русский\nAjuda do Scilab >> Xcos > batch_functions > scicos_simulate\n\n# scicos_simulate\n\nFunction for running xcos simulation in batch mode\n\n### Syntax\n\n```Info=scicos_simulate(scs_m)\nInfo=scicos_simulate(scs_m,Info)\nInfo=scicos_simulate(scs_m,context)\nInfo=scicos_simulate(scs_m,flag)\nInfo=scicos_simulate(scs_m ,Info [, context] [,flag])```\n\n### Arguments\n\nscs_m: A diagram data structure (see scs_m structure).\n\nInfo: A list. It must be set to `list()` at the first call, then use output `Info` as input `Info` for the next calls. `Info` contains compilation and simulation information and is used to avoid recompilation when not needed.\n\nContextValues: A Scilab struct containing values of symbolic variables used in the context and xcos blocks.\n\nflag: A string. If it equals 'nw' (no window), then blocks using graphical windows are not executed. Note that the list of such blocks must be updated as new blocks are added.\n\n### Description\n\nThis function is used to simulate xcos diagrams in batch mode. It requires the scs_m structure which can be obtained by loading in Scilab the `.zcos` file (see importXcosDiagram ).\n\nNote that before being able to simulate you should first load the block library using `loadXcosLibs()`.\n\nThe `ContextValues` may be used to change the main parameters value set in the main diagram context. example: if the variable `A` is set to 1 in the main context of the diagram. One can change the `A` value for a `scicos_simulate` simulation by setting\n\n`ContextValues.A=2`\n\nIt is also possible to use variables defined inside Scilab directly without using the `ContextValues` argument, but for such a use the context definition must allow this: example, if one wants to allow simulation use the `A` Scilab variable value for the parameter `A` the diagram context definition should contain\n\n`if ~exists('A') then A=1,end`\nNote that this second solution is fragile because it rely on the current value of `A` in Scilab.\n\n### File content\n\n• SCI/modules/scicos/macros/scicos_auto/scicos_simulate.sci\n\n### Examples\n\nThe xcos diagram in SCI/modules/xcos/demos/batch_simulation.zcos.", null, "```// load the blocks library and the simulation engine\n\nimportXcosDiagram(\"SCI/modules/xcos/demos/batch_simulation.zcos\")\n\ntypeof(scs_m) //The diagram data structure\n\n//This diagram uses 3 context variables :\n// Amplitude : the sin function amplitude\n// Pulsation : the sin function pulsation\n// Tf : the final simulation time\nscs_m.props.context; //the embedded definition\n\n//first batch simulation with the parameters embedded in the diagram\nscicos_simulate(scs_m);\n// Change the final time value\nContext.Tf=10;\nscicos_simulate(scs_m,Context);\n// without display\nContext.Tf=10;\nContext.Pulsation=9;\nscicos_simulate(scs_m,list(),Context,'nw');\n//get the variable created by the \"from workspace block\"\ncounter```" ]	[ null, "https://help.scilab.org/docs/6.1.0/pt_BR/batch_simulation.zcos.png", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.57485765,"math_prob":0.6670363,"size":3393,"snap":"2021-04-2021-17","text_gpt3_token_len":823,"char_repetition_ratio":0.16760106,"word_repetition_ratio":0.062240664,"special_character_ratio":0.2086649,"punctuation_ratio":0.11304348,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.99072456,"pos_list":[0,1,2],"im_url_duplicate_count":[null,1,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2021-01-24T02:58:43Z\",\"WARC-Record-ID\":\"<urn:uuid:aab2bdbe-01e7-48be-b82e-d4ee947e4907>\",\"Content-Length\":\"31444\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:83ae2541-7906-4454-9364-e920febcd8bb>\",\"WARC-Concurrent-To\":\"<urn:uuid:a7f378fe-b920-419b-ad79-481e6252dcfc>\",\"WARC-IP-Address\":\"176.9.3.186\",\"WARC-Target-URI\":\"https://help.scilab.org/docs/6.1.0/pt_BR/scicos_simulate.html\",\"WARC-Payload-Digest\":\"sha1:SQYCZFUV4A76ZHI7DI5QJXITOD2GSJ4T\",\"WARC-Block-Digest\":\"sha1:FFDAJIKINTEZFDEG7PX3IRWGIHKGCSLN\",\"WARC-Identified-Payload-Type\":\"application/xhtml+xml\",\"warc_filename\":\"/cc_download/warc_2021/CC-MAIN-2021-04/CC-MAIN-2021-04_segments_1610703544403.51_warc_CC-MAIN-20210124013637-20210124043637-00488.warc.gz\"}"}
https://www.monlivre.net/exploring-mathematics-an-engaging-introduction-to-proof-cambridge-mathematical-textbooks/	[ "Home Mathématiques Exploring Mathematics: An Engaging Introduction to Proof (Cambridge Mathematical Textbooks)\n\n# Exploring Mathematics: An Engaging Introduction to Proof (Cambridge Mathematical Textbooks)", null, "", null, "Exploring Mathematics gives students experience with doing mathematics – interrogating mathematical claims, exploring definitions, forming conjectures, attempting proofs, and presenting results – and engages them with examples, exercises, and projects that pique their interest. Written with a minimal number of pre-requisites, this text can be used by college students in their first and second years of study, and by independent readers who want an accessible introduction to theoretical mathematics.\n\nCore topics include proof techniques, sets, functions, relations, and cardinality, with selected additional topics that provide many possibilities for further exploration. With a problem-based approach to investigating the material, students develop interesting examples and theorems through numerous exercises and projects. In-text exercises, with complete solutions or robust hints included in an appendix, help students explore and master the topics being presented. The end-of-chapter exercises and projects provide students with opportunities to confirm their understanding of core material, learn new concepts, and develop mathematical creativity.\n\nTélécharger" ]	[ null, "https://www.monlivre.net/wp-content/uploads/2018/11/images-10.jpg", null, "https://www.monlivre.net/wp-content/uploads/2018/11/images-10-208x300.jpg", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.9348772,"math_prob":0.61814284,"size":1156,"snap":"2019-51-2020-05","text_gpt3_token_len":191,"char_repetition_ratio":0.125,"word_repetition_ratio":0.0,"special_character_ratio":0.1583045,"punctuation_ratio":0.13333334,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.98600054,"pos_list":[0,1,2,3,4],"im_url_duplicate_count":[null,1,null,1,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2019-12-16T07:30:23Z\",\"WARC-Record-ID\":\"<urn:uuid:eec8a7a1-ceae-42a1-8a31-ac41f97592c9>\",\"Content-Length\":\"136667\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:6c635c36-616c-49db-824d-9ed94501bde2>\",\"WARC-Concurrent-To\":\"<urn:uuid:242e9109-ca52-4f55-b27d-10d3491e3d4d>\",\"WARC-IP-Address\":\"144.91.71.213\",\"WARC-Target-URI\":\"https://www.monlivre.net/exploring-mathematics-an-engaging-introduction-to-proof-cambridge-mathematical-textbooks/\",\"WARC-Payload-Digest\":\"sha1:WXC2OM5F56TVNSWJD5IYVVVOMLWE3OPE\",\"WARC-Block-Digest\":\"sha1:TF2WJJOT7QMDFHY52KR64HW632Q2V345\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2019/CC-MAIN-2019-51/CC-MAIN-2019-51_segments_1575541318556.99_warc_CC-MAIN-20191216065654-20191216093654-00462.warc.gz\"}"}
https://www.oldquestionpapers.net/2010/11/engineering-assistants-question-papers-in-doordarshan-and-air.html	[ "# Engineering Assistants Question Papers in Doordarshan and AIR 2016\n\nEngineering Assistants Question Papers in Doordarshan and AIR 2016.\n\nEngineering Assistants Question Papers in Doordarshan and AIREngineering Assistants Question Papers in Doordarshan and AIR Model Question Papers Solved\n\n26. Lamps used for street lighting are connected in\n(a) parallel\n(b) series and parallel both\n(c) series\n(d) none of the above\nAns:a\n\n27. A. C. can be measured with the help of:\n(a) moving coil galvanometer\n(b) hot wire ammeter\n(c) tangent galvanometer\n(d) galvanometer\nAns:b\n\n28.A p-n junction is said to be forward biased, when\n(a) the positive pole of the battery is joined to the p-semiconductor and negative pole to the nsemiconductor\n(b) the positive pole of the battery is joined to the n-semiconductor and negative pole of the\nbattery is joined to the p-semiconductor\n(c) the positive pole of the battery is connected to n- semiconductor and p- semiconductor\n(d) a mechanical force is applied in the forward direction\nAns :a\n\n29. At absolute zero, Si acts as ?\n(a) non-metal\n(b) metal\n(c) insulator\n(d) none of these\nAns :a\n\n30. When N-type semiconductor is heated ?\n(a) number of electrons increases while that of holes decreases\n(b) number of holes increases while that of electrons decreases\n(c) number of electrons and holes remain same\n(d) number of electron and holes increases equally.\nAns :d\n\n31. Radiowaves of constant amplitude can be generated with\n(a) FET\n(b) filter\n(c) rectifier\n(d) oscillator\nAns :d\n\n32. In a common base amplifier the phase difference between the input signal voltage and the\noutput\nvoltage is ?\n(a) 0\n(b) pi/4\n(c)pi/2\n(d) p\nAns :d\n\n33. When a triode is used as an amplifier the phase difference between the input signal voltage\nand the output is ?\n(a) 0\n(b)pi\n(c) pi/2\n(d)pi/4\nAns :b\n\n34. The depletion layer in the P-N junction region is\ncaused by?\n(a) drift of holes\n(b) diffusion of charge carriers\n(c) migration of impurity ions\n(d) drift of electrons\nAns: b\n\n35. The following truth table corresponds to which Logic gate\nA B Output\n0 0 0\n0 1 1\n1 0 1\n1 1 1\n(a)NAND\n(b) OR\n(c)AND\n(d) XOR\nAns :b\n\n36. To use a transistor as an amplifier\n(a) The emitter base junction is forward biased and the base collector junction is reversed biased\n(b) no bias voltage is required\n(c) both junction are forward biased\n(d) both junctions are reversed biased.\nAns :a\n\n37. Which one of the following is the weakest kind of the bonding in solids\n(a) ionic\n(b) metallic\n(c) Vander Walls\n(d) covalent\nAns :c\n\n38.For amplification by a triode, the signal to be amplified is given to\n(a) the cathode\n(b) the grid\n(c) the glass-envelope\n(d) the anode\nAns :b\n\n39. A piece of copper and other of germanium are cooled from the room temperature to 80K,\nthen\n(a) resistance of each will increase\n(b) resistance of copper will decrease\n(c) the resistance of copper will increase while that of germanium will decrease\n(d) the resistance of copper will decrease while that of germanium will increase\nAns:d\n\n40. Diamond is very hard because ?\n(a) it is covalent solid\n(b) it has large cohesive energy\n(c) high melting point\n(d) insoluble in all solvents\nAns:b\n\n41. The part of the transistor which is heavily doped to produce large number of majority carriers\nis\n(a) emitter\n(b) base\n(c) collector\n(d) any of the above depending upon the nature of transistor\nAns:a\n\n42. An Oscillator is nothing but an amplifier with\n(a) positive, feedback\n(b) negative feedback\n(c)large gain\n(d) no feedback\nAns:a\n\n43. When a P-N junction diode is reverse biased the flow of current across the junction is mainly\ndue\nto\n(a) diffusion of charge\n(b) drift charges\n(c) depends on the nature o material\n(d) both drift and diffusion of charges\nAns:b\n\n44. Which of the following gates corresponds to the truth table given below?\nA B Output\n1 1 0\n0 1 1\n1 0 1\n0 0 1\n(a) NAND\n(b) OR\n(c) AND\n(d) XOR\nAns:a\n\n45. Which of the following, when added as an impurity, into the silicon, produces n-type semi\nconductor\n(a)Phosphorous\n(b) Aluminium\n(c)Magnesium\n(d) both ‘b’ and ‘c’\nAns:a\n\n46. The current gain for a transistor working as common-base amplifier is 0.96. If the emitter\ncurrent is 7.2 mA, then the base current is\n(a)0.29mA\n(b)0.35mA\n(c)0.39 mA\n(d) 0.43 mA\nAns:a\n\n47. When a n-p-n transistor is used as an amplifier then ?\n(a) the electrons flow from emitter to collector\n(b) the holes flow from emitter to collector\n(c) the electrons flow from collector to emitter\n(d) the electrons flow from battery to emitter\nAns:a\n\n48. When arsenic is added as an impurity to silicon, the resulting material is?\n(a) n-type semiconductor\n(b) p-type semiconductor\n(c) n-type conductor\n(d) Insulator\nAns:a\n\n49. To obtain a p-type germanium semiconductor,it must be doped with ?\n(a)arsenic\n(b)antimony\n(c)indium\n(d)phosphorus\nAns:a\n\n50. A semi-conducting device is connected in a series circuit with a battery and a resistance. A\ncurrent is found to pass through the circuit. If the polarity of the battery is reversed, the current\ndrops to almost zero. The device may be\n(a) A p-n junction\n(b) An intrinsic semi-conductor\n(c) A p-type semi-conductor\n(d) An n-type semi-conductor\nAns:a\n\nSimilar Pages" ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.7964662,"math_prob":0.950808,"size":5587,"snap":"2023-14-2023-23","text_gpt3_token_len":1502,"char_repetition_ratio":0.136844,"word_repetition_ratio":0.09484536,"special_character_ratio":0.2521926,"punctuation_ratio":0.07913669,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.97394305,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2023-06-01T05:41:48Z\",\"WARC-Record-ID\":\"<urn:uuid:64dba7ca-2d4f-4e81-b3b4-5b3daa1da9e1>\",\"Content-Length\":\"74066\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:4aa2772f-5467-41d5-90b5-9817496f5122>\",\"WARC-Concurrent-To\":\"<urn:uuid:df6df645-0850-4025-9038-6d05f3b4c57a>\",\"WARC-IP-Address\":\"103.174.102.29\",\"WARC-Target-URI\":\"https://www.oldquestionpapers.net/2010/11/engineering-assistants-question-papers-in-doordarshan-and-air.html\",\"WARC-Payload-Digest\":\"sha1:VMDGVTIYCSXUTQ5BTRSDXGX7XQQBUKKR\",\"WARC-Block-Digest\":\"sha1:63GKKHY7YZ3I7LFQPCVTNP5FDIPQI6UO\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2023/CC-MAIN-2023-23/CC-MAIN-2023-23_segments_1685224647614.56_warc_CC-MAIN-20230601042457-20230601072457-00452.warc.gz\"}"}
http://monardo.info/math-worksheets-on-multiplication/	[ "Math Worksheets On Multiplication\n\nMath Worksheets On Multiplication math worksheets on multiplication multiplication worksheets dynamically created multiplication printable. math worksheets on multiplication multiplication worksheet for math drills free also has divisions printable. math worksheets on multiplication multiplication worksheets for 5th grade multiplication worksheets templates. Math Worksheets On Multiplication math worksheets on multiplication free math worksheets download excel template. math worksheets on multiplication multiplication worksheets free easier to grade customizable templates. math worksheets on multiplication grade 4 multiplication worksheets free printable k5 learning ideas. Math Worksheets On Multiplication math worksheets on multiplication missing factor multiplication worksheets school ideas pinterest template.", null, "Math Worksheets On Multiplication Multiplication Worksheets Dynamically Created Multiplication Printable", null, "Math Worksheets On Multiplication Multiplication Worksheet For Math Drills Free Also Has Divisions Printable", null, "Math Worksheets On Multiplication Multiplication Worksheets For 5th Grade Multiplication Worksheets Templates", null, "Math Worksheets On Multiplication Free Math Worksheets Download Excel Template", null, "Math Worksheets On Multiplication Multiplication Worksheets Free Easier To Grade Customizable Templates", null, "Math Worksheets On Multiplication Grade 4 Multiplication Worksheets Free Printable K5 Learning Ideas", null, "Math Worksheets On Multiplication Missing Factor Multiplication Worksheets School Ideas Pinterest Template\n\nMath Worksheets On Multiplication math worksheets on multiplication multiplication worksheet for math drills free also has divisions printable. math worksheets on multiplication multiplication worksheets for 5th grade multiplication worksheets templates. math worksheets on multiplication free math worksheets download excel template. math worksheets on multiplication multiplication worksheets free easier to grade customizable templates. math worksheets on multiplication grade 4 multiplication worksheets free printable k5 learning ideas." ]	[ null, "http://monardo.info/wp-content/uploads/2018/11/math-worksheets-on-multiplication-multiplication-worksheets-dynamically-created-multiplication-printable.png", null, "http://monardo.info/wp-content/uploads/2018/11/math-worksheets-on-multiplication-multiplication-worksheet-for-math-drills-free-also-has-divisions-printable.png", null, "http://monardo.info/wp-content/uploads/2018/11/math-worksheets-on-multiplication-multiplication-worksheets-for-5th-grade-multiplication-worksheets-templates.png", null, "http://monardo.info/wp-content/uploads/2018/11/math-worksheets-on-multiplication-free-math-worksheets-download-excel-template.jpg", null, "http://monardo.info/wp-content/uploads/2018/11/math-worksheets-on-multiplication-multiplication-worksheets-free-easier-to-grade-customizable-templates.png", null, "http://monardo.info/wp-content/uploads/2018/11/math-worksheets-on-multiplication-grade-4-multiplication-worksheets-free-printable-k5-learning-ideas.gif", null, "http://monardo.info/wp-content/uploads/2018/11/math-worksheets-on-multiplication-missing-factor-multiplication-worksheets-school-ideas-pinterest-template.png", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.66486955,"math_prob":0.7391986,"size":1365,"snap":"2019-26-2019-30","text_gpt3_token_len":184,"char_repetition_ratio":0.3614989,"word_repetition_ratio":0.8076923,"special_character_ratio":0.12967034,"punctuation_ratio":0.069767445,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.999721,"pos_list":[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14],"im_url_duplicate_count":[null,1,null,1,null,1,null,1,null,1,null,1,null,1,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2019-06-17T19:04:47Z\",\"WARC-Record-ID\":\"<urn:uuid:949b9b0f-db49-4223-acf5-16abcc267ca8>\",\"Content-Length\":\"48592\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:65a1e750-6f1b-45c2-a0ad-aa0c897c85a1>\",\"WARC-Concurrent-To\":\"<urn:uuid:bbe1663c-8974-4fd4-93a6-8b7ccb97db3c>\",\"WARC-IP-Address\":\"104.28.28.20\",\"WARC-Target-URI\":\"http://monardo.info/math-worksheets-on-multiplication/\",\"WARC-Payload-Digest\":\"sha1:4VYYYSUPJMYNYXBQOHEC4AHGGW7OLHZ6\",\"WARC-Block-Digest\":\"sha1:KYMCRBJLDSMNXMTR4NEKQDCJVUSFHFC4\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2019/CC-MAIN-2019-26/CC-MAIN-2019-26_segments_1560627998558.51_warc_CC-MAIN-20190617183209-20190617205209-00248.warc.gz\"}"}
https://divided-by.com/43-by-3/	[ "# 43 divided by 3 equals 14.33\n\n43 is not divisible by 3.\n\n43 divided by 3 equals 14.333333333333.\n\n43 divisible by 1, and 43.\n\n/\n=\n14.33\nThe Divisible tool below performs three tasks to calculate 43 divided by 3. It checks if 43 is divisible by 3, it divides the two numbers, and it can show you all the numbers that 43 is divisible by.\n\n## What is 43 divided by 3?\n\nWhen searching for this answer, there are many different ways that you can phrase this question, including:\n\n• What is 43 divided by 3?\n• How much is 43 divided by 3?\n• What does 43 divided by 3 equal?\n\nThe result of 43/3 is 14.333333333333.\n\nWe can also show the result of this division in other ways.\n\n• 43 divided by 3 in decimal = 14.333333333333\n• 43 divided by 3 in fraction = 43/3\n• 43 divided by 3 in percentage = 1433.3333333333%\n\n## The dividend and divisor of 43 divided by 3\n\nWhen performing division, you need a dividend and a divisor.\n\nA dividend is the number that is being divided and a divisor is the number that it is being divided by.\n\n43 is the dividend (the number we are dividing), 3 is the divisor (the number we are dividing by) and 14.333333333333 is the quotient (the result of the division)\n\n## What is the quotient and remainder of 43 divided by 3?\n\nOur calculator will also tell you what the quotient and remainder is of any calculation you perform.\n\nAs we explained above, the quotient is the result of dividing the dividend by the divisor. In lots of cases, the result of this calculation will be an integer (a whole number), meaning that a number can be divided fully without anything left over.\n\nWhen a number is unable to be divided fully, the amount left over is what we call the remainder.\n\nThe quotient and remainder of 43 divided by 3 = 14 R 1\n\nThe individual parts of this calculation are:\n\n• 43 is the dividend\n• 3 is the divisor\n• 14 is the quotient\n• 1 is the remainder\n\nThank you for using our division calculator. If you want to perform a new calculation, simply input two new values in the boxes and hit the calculate button. Alternatively, look at at the similar divisions below that relate to the dividend you selected." ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.91879827,"math_prob":0.99741113,"size":1891,"snap":"2023-40-2023-50","text_gpt3_token_len":478,"char_repetition_ratio":0.22575517,"word_repetition_ratio":0.06497175,"special_character_ratio":0.29402432,"punctuation_ratio":0.0974359,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9999145,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2023-11-30T04:52:59Z\",\"WARC-Record-ID\":\"<urn:uuid:d5bfc70a-6e47-40be-9a67-951ea530edb9>\",\"Content-Length\":\"47313\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:5cc36aad-14b5-46fa-83ac-2251f5481460>\",\"WARC-Concurrent-To\":\"<urn:uuid:9de3b51f-9e52-409f-a95d-ffe6435e4e20>\",\"WARC-IP-Address\":\"104.21.14.93\",\"WARC-Target-URI\":\"https://divided-by.com/43-by-3/\",\"WARC-Payload-Digest\":\"sha1:EU52PHJSGN6I64RGAGIVSQSLKTTZSXSW\",\"WARC-Block-Digest\":\"sha1:IYJQA7AHHXSLFM7WGYMOFVLZXXOJSIIZ\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2023/CC-MAIN-2023-50/CC-MAIN-2023-50_segments_1700679100164.87_warc_CC-MAIN-20231130031610-20231130061610-00791.warc.gz\"}"}
http://jinjinenglish.com/278369032.html	[ "2019年03月02日\n\nY691、Z162、Z165线安全生命防护工程\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`\n\n```\n\n```\n`\t`" ]	[ null ]	{"ft_lang_label":"__label__zh","ft_lang_prob":0.8987413,"math_prob":0.992402,"size":1909,"snap":"2020-34-2020-40","text_gpt3_token_len":2107,"char_repetition_ratio":0.1527559,"word_repetition_ratio":0.5714286,"special_character_ratio":0.29753798,"punctuation_ratio":0.04032258,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.99396956,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2020-09-26T02:04:19Z\",\"WARC-Record-ID\":\"<urn:uuid:3a96085c-ada1-4488-8acf-365afbc323a0>\",\"Content-Length\":\"215890\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:b550048b-58fd-46e6-a4c9-af47004914fc>\",\"WARC-Concurrent-To\":\"<urn:uuid:74378088-1d8f-4d5e-8fe0-85480e1de640>\",\"WARC-IP-Address\":\"45.12.128.137\",\"WARC-Target-URI\":\"http://jinjinenglish.com/278369032.html\",\"WARC-Payload-Digest\":\"sha1:53MKSPDPXA6DY7OMMKUQX77E3VRH4QJL\",\"WARC-Block-Digest\":\"sha1:736HSN4OEBUT6BNOB676XKPZXAYDJNQ2\",\"WARC-Identified-Payload-Type\":\"application/xhtml+xml\",\"warc_filename\":\"/cc_download/warc_2020/CC-MAIN-2020-40/CC-MAIN-2020-40_segments_1600400232211.54_warc_CC-MAIN-20200926004805-20200926034805-00362.warc.gz\"}"}
https://scholarsarchive.byu.edu/etd/3722/	[ "## All Theses and Dissertations\n\n#### Abstract\n\nGiven a simple graph G with vertex set V(G)={1,2,...,n} define S(G) to be the set of all real symmetric matrices A such that for all i not equal to j, the ijth entry of A is nonzero if and only if ij is in E(G). The range of the ranks of matrices in S(G) is of interest and can be determined by finding the minimum rank. The minimum rank of a graph, denoted mr(G), is the minimum rank achieved by a matrix in S(G). The maximum nullity of a graph, denoted M(G), is the maximum nullity achieved by a matrix in S(G). Note that mr(G)+M(G)=\|V(G)\| and so in finding the maximum nullity of a graph, the minimum rank of a graph is also determined. The minimum rank problem for a graph G asks us to determine mr(G) which in general is very difficult. A simple graph is planar if there exists a drawing of G in the plane such that any two line segments representing edges of G intersect only at a point which represents a vertex of G. A planar drawing partitions the rest of the plane into open regions called faces. A graph is outerplanar if there exists a planar drawing of G such that every vertex lies on the outer face. We consider the class of outerplanar graphs and summarize some of the recent results concerning the minimum rank problem for this class. The path cover number of a graph, denoted P(G), is the minimum number of vertex-disjoint paths needed to cover all the vertices of G. We show that for all outerplanar graphs G, P(G)is greater than or equal to M(G). We identify a subclass of outerplanar graphs, called partial 2-paths, for which P(G)=M(G). We give a different characterization for another subset of outerplanar graphs, unicyclic graphs, which determines whether M(G)=P(G) or M(G)=P(G)-1. We give an example of a 2-connected outerplanar graph for which P(G) ≥ M(G).A cover of a graph G is a collection of subgraphs of G such that the union of the edge sets of the subgraphs is equal to the E(G). The rank-sum of a cover C of G is denoted as rs(C) and is equal to the sum of the minimum ranks of the subgraphs in C. We show that for an outerplanar graph G, there exists an edge-disjoint cover of G consisting of cliques, stars, cycles, and double cycles such that the rank-sum of the cover is equal to the minimum rank of G. Using the fact that such a cover exists allows us to show that the minimum rank of a weighted outerplanar graph is equal to the minimum rank of its underlying simple graph.\n\nPhD\n\n#### College and Department\n\nPhysical and Mathematical Sciences; Mathematics\n\n2013-07-05\n\nDissertation\n\n#### Handle\n\nhttp://hdl.lib.byu.edu/1877/etd6430\n\n#### Keywords\n\nouterplanar graph, minimum rank, maximum nullity, path cover number, partial 2-path, edge-disjoint cover, weighted graph\n\nEnglish\n\nCOinS" ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.90014255,"math_prob":0.9864563,"size":2976,"snap":"2019-26-2019-30","text_gpt3_token_len":738,"char_repetition_ratio":0.16184388,"word_repetition_ratio":0.027944112,"special_character_ratio":0.23219086,"punctuation_ratio":0.1015873,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.99905175,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2019-07-19T08:42:28Z\",\"WARC-Record-ID\":\"<urn:uuid:7a7fea96-65fc-4cf1-9120-504e2bc0c48b>\",\"Content-Length\":\"40538\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:446d82e8-b3cc-4ceb-b9aa-ff90fc6856be>\",\"WARC-Concurrent-To\":\"<urn:uuid:5c7109c3-2050-4bca-89b4-ca6b86370cb7>\",\"WARC-IP-Address\":\"72.5.9.223\",\"WARC-Target-URI\":\"https://scholarsarchive.byu.edu/etd/3722/\",\"WARC-Payload-Digest\":\"sha1:ROTIIQLJKYPU2RRGBWHQKHXBEXBRCGJZ\",\"WARC-Block-Digest\":\"sha1:IUOHEFLYXNRVKT42WSQYL5LCJPAWTUWL\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2019/CC-MAIN-2019-30/CC-MAIN-2019-30_segments_1563195526153.35_warc_CC-MAIN-20190719074137-20190719100137-00511.warc.gz\"}"}
https://lifewithdata.com/2023/06/01/how-to-test-for-normality-in-python/	[ "# How to Test for Normality in Python\n\nTesting for normality is a crucial step in many statistical analyses because several statistical techniques assume that data is normally distributed. In Python, several libraries, including SciPy and statsmodels, offer functions to test for normality. In this guide, we will discuss various methods to perform a normality test in Python, including visual methods (like Q-Q plots) and statistical tests (like the Shapiro-Wilk test, the Anderson-Darling test, and the Kolmogorov-Smirnov test).\n\n## Required Libraries\n\nEnsure you have the following libraries installed:\n\n1. NumPy: A library for numerical operations.\n2. pandas: A library for data manipulation and analysis.\n3. SciPy: A library for scientific computing and technical computing.\n4. statsmodels: A library for estimating statistical models.\n5. Matplotlib and Seaborn: Libraries for data visualization.\n\nYou can install these libraries using pip:\n\npip install numpy pandas scipy statsmodels matplotlib seaborn\n\n## Importing the Libraries\n\nAfter installing, import these libraries:\n\nimport numpy as np\nimport pandas as pd\nfrom scipy import stats\nimport statsmodels.api as sm\nimport matplotlib.pyplot as plt\nimport seaborn as sns\n\n## Generating a Sample Dataset\n\nFor this guide, we’ll generate a dataset from a normal distribution:\n\n# Set the seed for reproducibility\nnp.random.seed(0)\n\n# Generate a dataset from a normal distribution\ndata = np.random.normal(loc=0, scale=1, size=1000)\n\nThis generates 1000 data points from a normal distribution with a mean (loc) of 0 and a standard deviation (scale) of 1.\n\n## Visual Methods for Checking Normality\n\n### Histogram\n\nThe histogram provides a graphical representation of data distribution:\n\nsns.histplot(data, kde=True)\nplt.title('Histogram')\nplt.show()\n\nIn a histogram of normally distributed data, you’d expect to see a bell curve shape.\n\n### Q-Q Plot\n\nA Q-Q (quantile-quantile) plot is a graphical tool to help us assess if a dataset comes from a certain distribution:\n\nsm.qqplot(data, line='45')\nplt.title('Q-Q Plot')\nplt.show()\n\nIn a Q-Q plot, if the data is normally distributed, the points will fall along the 45-degree line.\n\n## Statistical Methods for Checking Normality\n\n### Shapiro-Wilk Test\n\nThe Shapiro-Wilk test is one of the most powerful normality tests:\n\nW, p_value = stats.shapiro(data)\nprint(f'Shapiro-Wilk Test: W={W}, p-value={p_value}')\n\nThe null hypothesis of the Shapiro-Wilk test is that the data is normally distributed. So if the p-value is less than the chosen alpha level (typically 0.05), then the null hypothesis is rejected and there is evidence that the data is not from a normally distributed population.\n\n### Anderson-Darling Test\n\nThe Anderson-Darling test is another powerful statistical test for checking normality:\n\nresult = stats.anderson(data)\nprint(f'Anderson-Darling Test: statistic={result.statistic}, critical_values={result.critical_values}, significance_level={result.significance_level}')\n\nFor the Anderson-Darling test, if the returned statistic is larger than these critical values then for the corresponding significance level, the null hypothesis that the data come from the chosen distribution can be rejected.\n\n### Kolmogorov-Smirnov Test\n\nThe Kolmogorov-Smirnov test can be applied to any continuous distribution to compare the sample distribution with the theoretical one:\n\nD, p_value = stats.kstest(data, 'norm')\nprint(f'Kolmogorov-Smirnov Test: D={D}, p-value={p_value}')\n\nLike the other tests, the null hypothesis of the K-S test is that the data is normally distributed. If the p-value is less than alpha (typically 0.05), the null hypothesis is rejected.\n\n## Conclusion\n\nIn this article, we discussed how to perform a normality test in Python using both visual and statistical methods. Remember, though, no statistical test is fully decisive, and they should be used in conjunction with graphical methods and ideally, domain knowledge. Even when a test suggests that data is normally distributed, this doesn’t guarantee that the data is normal—it simply provides strong evidence. Conversely, these tests are sensitive to sample size; with large data, they can detect insignificant deviations from normality, leading to rejection of the null hypothesis when the data is close to normal for practical purposes." ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.7637411,"math_prob":0.83702534,"size":4197,"snap":"2023-40-2023-50","text_gpt3_token_len":905,"char_repetition_ratio":0.13641784,"word_repetition_ratio":0.027072757,"special_character_ratio":0.20085776,"punctuation_ratio":0.1259947,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9963905,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2023-09-21T19:45:13Z\",\"WARC-Record-ID\":\"<urn:uuid:e12a6120-0141-49d3-aac2-5ef39c4c3612>\",\"Content-Length\":\"108510\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:af306a7e-fac6-4c8a-bb36-6ab5630bb8c6>\",\"WARC-Concurrent-To\":\"<urn:uuid:81dd4ae5-07a4-4744-b61f-469a9bef354a>\",\"WARC-IP-Address\":\"192.0.78.249\",\"WARC-Target-URI\":\"https://lifewithdata.com/2023/06/01/how-to-test-for-normality-in-python/\",\"WARC-Payload-Digest\":\"sha1:5UGE6RPZ5Q7LII5KXVLUFRK2RENWSZLX\",\"WARC-Block-Digest\":\"sha1:G6WETOLIX5FPD7HSZ3DD5GVYWMO5LXNS\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2023/CC-MAIN-2023-40/CC-MAIN-2023-40_segments_1695233506029.42_warc_CC-MAIN-20230921174008-20230921204008-00876.warc.gz\"}"}
https://physics.stackexchange.com/questions/489689/how-do-resistors-generate-different-heat-if-we-make-the-current-fixed-and-change	[ "# How do resistors generate different heat if we make the current fixed and changed the voltage and resistance? Notice the flow of charge is constant\n\nConsider having a circuit which consists of a battery and one resistor.\n$$V = 10$$ volts, $$R = 5$$ ohms, so $$I = 2$$ Amperes, and $$P = 20$$ watts.\n\nIf we double the voltage and resistance, the current will be the same and the power will be equal to 40 watts, hence the resistor will be hotter than the first case.\n\nNow here is the silly question.\n\nThe current in each of the two cases is constant and equals the current of the other. The speed which charges move across the wire is constant. So why is more heat generated while the speed of charges is constant? I know that the potential is doubled, but the potential is potential, and we can't make use of potential energy unless it is converted to kinetic energy. How can the resistor make use of this (potential) energy, the mechanism which the resistor turns potential energy into heat.\n\n• Where does it say that heat is produced by charge alone? – user207421 Jul 5 '19 at 9:47\n\nYour initial circuit is like this:", null, "So you get 2A flowing and a power of 20W dissipated in the resistor.\n\nThen you double the voltage and the resistance:", null, "The two batteries add up to single source of 20V. The two resistors add up to a total resistance of 10$$\\Omega$$. So, as you correctly state, the current is the same as before (2A). Therefore the power is now 40W\n\nBut this power is shared between the two resistors: 20W each, exactly as before.\n\nYou might also note that the potential at the point between the resistors is 10V, so each resistor has 10V across it, exactly as before.\n\nReally all you've done is doubled up the circuit so that you have twice of what you had before.\n\n• Think of sucking liquid through a straw. It flows at a certain speed.\n• Now squeeze the straw (increase the resistance). The flow (volume per second) slows down.\n• In order to keep the flow-speed high, you must suck harder. That's the voltage (think of it as an electric \"pressure\"). This dent in the straw is now \"using up\" all of the pressure, which now is much higher.\n\nBasically, the dent in the straw does cause energy loss and thus would have caused the flow to slow down. You just don't notice because you replenish the flow right away by having the pressure higher.\n\nThis is directly analogous to electric circuits: The larger resistance does cause energy loss and thus would have caused the current (electron flow) to slow down since it \"sucks out\" their kinetic energy. But you don't notice because you counteract this loss with a higher voltage that \"forces\" the electrons forward to keep the same speed nevertheless.\n\nThink of the resistor as being like a pipe packed full of magically immoveable (so they won't get swept away by the flow) pebbles(). The electric current is like water flowing through this pipe. Different resistors have different pebble densities, and the more tightly packed the pebbles are with fewer and more convoluted void spaces between them, the harder it is for the water to get through. The electric field in the wire pushes the water through the pebble bed, like a source of water pressure, such as a pump or downhill slope does with real water.\n\nNow, suppose one pipe has a relatively coarse and gap-filled pebble fill (lower $$R$$), while the other has a finer and more densely-packed fill (higher $$R$$). Both are subjected to a flow of water, and those flows are tuned so that both maintain the same perfusion (rate of water making it through) of the pebble bed. Do you think equal driving force, and so pressure, will result in identical perfusion? If not, which do think will need more and which less?\n\nAnd then, if you're driving it harder, that means a greater force is being exerted and more work is being done in any given time to get the same result (flow or rate of perfusion), hence the faster also the rate at which energy is being dissipated into heat. Which, then, has the largest dissipation and hence gets the hottest?\n\n() NB to other readers: Technically, the \"pebbles\" in a real resistor are moveable and, in fact, flying around all over (for a suitably quantum-mechanically fuzzy notion of 'flying'), but I am using this to keep the illustration simple.\n\nHere's a mental experiment you can do.\n\nTake a resistor which generates a certain amount of heat Q while a voltage is applied to it, resulting in a current. Now, cut the resistor into two equal parts perpendicular to the direction of the current. Quite obviously, if the whole resistor generated Q, each part should generate Q/2. Yet, the current in each part is exactly the same, so by your logic each half should still generate the same amount of heat as the initial resistor.\n\n• Thank you a lot . – Mahmoud Amin Jul 6 '19 at 8:41" ]	[ null, "https://i.stack.imgur.com/0YW0z.png", null, "https://i.stack.imgur.com/f4dtt.png", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.9396232,"math_prob":0.977164,"size":832,"snap":"2020-10-2020-16","text_gpt3_token_len":192,"char_repetition_ratio":0.14855072,"word_repetition_ratio":0.0,"special_character_ratio":0.23317307,"punctuation_ratio":0.10119048,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.99414474,"pos_list":[0,1,2,3,4],"im_url_duplicate_count":[null,7,null,7,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2020-02-25T04:42:36Z\",\"WARC-Record-ID\":\"<urn:uuid:aa3e6630-3e1f-4427-b8bb-5d7cc9151005>\",\"Content-Length\":\"163360\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:dcc0133e-921d-4c47-ac30-b866b5d7ba5f>\",\"WARC-Concurrent-To\":\"<urn:uuid:88fec1fc-a4e6-4ca4-b9ae-a7d1c58fd4d6>\",\"WARC-IP-Address\":\"151.101.129.69\",\"WARC-Target-URI\":\"https://physics.stackexchange.com/questions/489689/how-do-resistors-generate-different-heat-if-we-make-the-current-fixed-and-change\",\"WARC-Payload-Digest\":\"sha1:KWQJJ7Y7BPO6V2MC2EPCHZWU43KOUWLW\",\"WARC-Block-Digest\":\"sha1:FVBDCYEBBR4LOIONB7SRJ4GVCEYR3UHW\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2020/CC-MAIN-2020-10/CC-MAIN-2020-10_segments_1581875146004.9_warc_CC-MAIN-20200225014941-20200225044941-00229.warc.gz\"}"}
https://support.numxl.com/hc/en-us/articles/217084526-Statistical-Testing	[ "# Statistical Testing\n\nStatistical/Hypothesis testing is a common method of drawing inferences about a population based on statistical evidence from a sample.\n\nAll hypothesis tests share the same basic terminology and structure:\n\n1. Null Hypothesis is an assertion about a population that you would like to test. The null hypothesis may be formalized by asserting that a population parameter, or a combination of population parameters, has a certain value. For instance, to test the mean of a given population equals to $\\mu_o$, we form the hypothesis as follow: $$H_o: \\mu = \\mu_o$$\n2. Alternative Hypothesis is a contrasting assertion about the population that can be tested against the null hypothesis. For our example above, we construct the following alternatives:\n1. $H_1: \\mu \\gt \\mu_o -$ population average is greater than $\\mu_o$ (right-tail test)\n2. $H_1: \\mu \\lt \\mu_o -$ population average is smaller than $\\mu_o$ (right-tail test)\n3. $H_1: \\mu \\neq \\mu_o -$ population average is different from $\\mu_o$ (two-tailed test)\n3. The p-value of a test is the probability, under the null hypothesis, of obtaining a value of the test statistic as extreme or more extreme than the value computed from the sample.\n4. The significance level of a test is a threshold of probability $\\alpha$ agreed to before the test is conducted. A typical value of $\\alpha$ is 5%. If the p-value of a test is less than $\\alpha$, the test rejects the null hypothesis. If the p-value is greater than $\\alpha$, there is insufficient evidence to reject the null hypothesis. Note that the lack of evidence for rejecting the null hypothesis is not evidence for accepting the null hypothesis. Also, note that the substantive \"significance\" of an alternative cannot be inferred from the statistical significance of a test.\n\nResults of hypothesis tests are often communicated with a confidence interval. A confidence interval is an estimated range of values with a specified probability of containing the true population value of a parameter. Upper and lower bounds for confidence intervals are computed from the sample estimate of the parameter and the known (or assumed) sampling distribution of the estimator." ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.8901301,"math_prob":0.9897785,"size":2197,"snap":"2020-45-2020-50","text_gpt3_token_len":463,"char_repetition_ratio":0.15777473,"word_repetition_ratio":0.031518623,"special_character_ratio":0.21574874,"punctuation_ratio":0.0839895,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9994247,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2020-11-28T10:38:55Z\",\"WARC-Record-ID\":\"<urn:uuid:c2ca98e9-1ad6-48ab-82e5-b00fc2778b9d>\",\"Content-Length\":\"33086\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:34b1e6be-ef2a-42f7-bbd6-a90540fc0c40>\",\"WARC-Concurrent-To\":\"<urn:uuid:0464fcef-2681-4a0e-8e85-9a251a2c45ef>\",\"WARC-IP-Address\":\"104.16.51.111\",\"WARC-Target-URI\":\"https://support.numxl.com/hc/en-us/articles/217084526-Statistical-Testing\",\"WARC-Payload-Digest\":\"sha1:FCYIXD3SKZERW2TGRB3I57NVE3DDQLWF\",\"WARC-Block-Digest\":\"sha1:LZTVF5BDLXRT3JVIN6BSIDWPKMKV7HTL\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2020/CC-MAIN-2020-50/CC-MAIN-2020-50_segments_1606141195417.37_warc_CC-MAIN-20201128095617-20201128125617-00628.warc.gz\"}"}
http://laxbytes.com/binwomstats19/PRIatac001018002.php	[ "``` EXPLANATION\nRANK PERCENTILE VALUE * WEIGHT = NET\nGames Played 18 * 0.20 = 3.60\nGames Started 18 * 0.30 = 5.40\nGoals 39 98 55 * 0.65 = 35.75\nAssists >100 95 18 * 0.55 = 9.94\nTotal Points 52 98 73 * 0.13 = 9.49\nDraw Controls >100 96 57 * 0.45 = 25.65\nTotal Shots >100 96 90 * 0.10 = 9.00\nShots on Goal 89 97 75 * 0.10 = 7.50\nMissed Shots Off Goal >100 89 15 * -0.05 = -0.75\nShot Percentage >100 94 0.61 * 1.00 = 0.61\nShots-on-goal Percentage >100 84 0.83 * 1.00 = 0.83\nGround Balls >100 88 26 * 0.45 = 11.70\nTurnovers >100 90 25 * -0.60 = -15.00\nCaused Turnovers >100 91 16 * 2.00 = 32.00\nFree Position Goals >100 95 8 * 0.10 = 0.80\nFree Position Misses >100 88 6 * -0.10 = -0.60\nGoals Saved -- -- 0 * 0.80 = 0.00\nGoals Allowed -- -- 0 * -0.80 = -0.00\nShots Faced -- -- 0 * 0.01 = 0.00\nSave Percentage -- -- 0.00 * 5.00 = 0.00\n\nPIR RAW (NET/GamesPlayed) 6.27\nDefensive Factor (DF) 0.84\nStrength Schedule (SOS) 0.70\nBasis Points 2.50\n-------\nPIR = (PRI RAW)DFSOS + Basis Points 6.14\n\nList of all players for Coastal Carolina\nList of all players for Division I\n\n```" ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.5292462,"math_prob":0.99174553,"size":1056,"snap":"2019-43-2019-47","text_gpt3_token_len":444,"char_repetition_ratio":0.13878328,"word_repetition_ratio":0.009090909,"special_character_ratio":0.5823864,"punctuation_ratio":0.1764706,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9761254,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2019-11-15T00:18:08Z\",\"WARC-Record-ID\":\"<urn:uuid:b0c30b77-eef6-42d6-a065-d80de489ffd1>\",\"Content-Length\":\"52640\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:b72d8070-1460-4f20-8da6-e06b3f89d1b0>\",\"WARC-Concurrent-To\":\"<urn:uuid:a9f61381-a9a6-430f-875f-1062a78a3c40>\",\"WARC-IP-Address\":\"50.63.209.1\",\"WARC-Target-URI\":\"http://laxbytes.com/binwomstats19/PRIatac001018002.php\",\"WARC-Payload-Digest\":\"sha1:7VHVZFB35XWKVA4BCCXVS6SXIVRLNCXP\",\"WARC-Block-Digest\":\"sha1:72WPYUIZTKN5HIHIZQVYOI3FEPECSF3Q\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2019/CC-MAIN-2019-47/CC-MAIN-2019-47_segments_1573496668544.32_warc_CC-MAIN-20191114232502-20191115020502-00387.warc.gz\"}"}
https://www.cut-the-knot.org/Probability/TwoCoins.shtml	[ "# Two Coins: One Fair, one Biased\n\n### Problem", null, "### Solution 1\n\nAssuming the fair coin is the one that was tossed, then the probability of it showing heads and of the other coin showing two heads is given by\n\n$\\displaystyle \\frac{1}{2}\\cdot{3\\choose 2}\\left(\\frac{2}{3}\\right)^2\\left(\\frac{1}{3}\\right)^{1}=\\frac{2}{9}.$\n\nOn the other hand, if the biased coin was tossed once then the probability of the observed result is\n\n$\\displaystyle \\frac{2}{3}\\cdot{3\\choose 2}\\left(\\frac{1}{2}\\right)^2\\left(\\frac{1}{2}\\right)^{1}=\\frac{2}{8},$\n\ntelling us that more likely it was the biased coin that was tossed once.\n\n### Solution 2\n\nWithout any a-priori knowledge, let us assume that the first coin is equally likely to be either of the two coins. Let $D$ denote the observation and $B$ the event that the first coin is biased. Thus, $P(B)=P(\\overline{B})=1/2$,\n\n\\displaystyle\\begin{align} P(D\|B)&=\\left[\\frac{2}{3}\\right]\\cdot \\left[C(3,2)\\left(\\frac{1}{2}\\right)^2\\left(\\frac{1}{2}\\right)\\right]=\\frac{1}{4}, \\\\ P(D\|\\overline{B})&=\\left[\\frac{1}{2}\\right]\\cdot \\left[C(3,2)\\left(\\frac{2}{3}\\right)^2\\left(\\frac{1}{3}\\right)\\right]=\\frac{2}{9}. \\end{align}\n\nUsing Bayes theorem,\n\n\\displaystyle\\begin{align} P(B\|D)&=\\frac{P(D\|B)P(B)}{P(D\|B)P(B)+P(D\|\\overline{B})P(\\overline{B})} \\\\ &=\\frac{1/4}{1/4+2/9}=\\frac{9}{17}>\\frac{1}{2}. \\end{align}\n\nThus, the observation makes the first coin more likely to be biased.\n\n### Solution 3\n\nPrior (relative) odds are even, while likelihood ratio is\n\n\\displaystyle\\begin{align} LR&=\\frac{Pr(E\|B,F)}{Pr(E\| F,B)}= \\frac{\\displaystyle\\frac{2}{3}\\cdot\\left(\\frac{1}{2}\\right)^3}{\\displaystyle\\frac{1}{2}\\cdot\\left(\\frac{2}{3}\\right)^2\\cdot\\frac{1}{3}}= \\frac{\\displaystyle\\frac{2}{3}\\cdot\\frac{1}{8}}{\\displaystyle\\frac{1}{2}\\cdot\\frac{4}{27}}\\\\ &= \\frac{9}{8}. \\end{align}\n\nOdds in favor of the first coin being biased and the second coin being fair are $9:8,$ i.e. $\\displaystyle prob=\\frac{9}{17}.$\n\n### Acknowledgment\n\nThis is problem 1989-4 from the A Friendly Mathematics Competition by Rick Gillman (ed.)\n\nSolution 2 is by Amit Itagi; Solution 3 is by Joshua B. Miller.", null, "" ]	[ null, "https://www.cut-the-knot.org/Probability/TwoCoins.jpg", null, "https://www.cut-the-knot.org/gifs/tbow_sh.gif", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.6938518,"math_prob":0.99977833,"size":2166,"snap":"2019-35-2019-39","text_gpt3_token_len":759,"char_repetition_ratio":0.19657724,"word_repetition_ratio":0.009259259,"special_character_ratio":0.35457063,"punctuation_ratio":0.07349666,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.99997044,"pos_list":[0,1,2,3,4],"im_url_duplicate_count":[null,3,null,null,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2019-09-19T21:38:44Z\",\"WARC-Record-ID\":\"<urn:uuid:77909c82-3896-475a-93b3-ae22cde64bf2>\",\"Content-Length\":\"13476\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:e82e4875-691d-4a60-985b-4efc265b5f46>\",\"WARC-Concurrent-To\":\"<urn:uuid:1247d7dd-3879-4126-8efb-88ddb759b82d>\",\"WARC-IP-Address\":\"107.180.50.227\",\"WARC-Target-URI\":\"https://www.cut-the-knot.org/Probability/TwoCoins.shtml\",\"WARC-Payload-Digest\":\"sha1:7LQTC4G26NN3QVV3OTJEXCKQ2ZQSYDN4\",\"WARC-Block-Digest\":\"sha1:COKLRNSZ4PXVMV4B2HJTXORQW5B7FEQZ\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2019/CC-MAIN-2019-39/CC-MAIN-2019-39_segments_1568514573735.34_warc_CC-MAIN-20190919204548-20190919230548-00515.warc.gz\"}"}
https://www.fmaths.com/square-root/often-asked-what-is-square-root-of-25.html	[ "# Often asked: What Is Square Root Of 25?\n\n## How do you find the square of 25?\n\nIn the case of 25 we find that 52= 25, so 5 is a square root of 25.\n\n## Is the square root of 25 a perfect square?\n\nSince 25 is a natural number and the square root of 25 is a natural number (5), 25 is a perfect square. Since 102.01 is a rational number and the square root of 102.01 is a rational number (10.1), 102.01 is a perfect square.\n\n## What is the perfect square of 25?\n\nIn mathematics, a square is a product of a whole number with itself. For instance, the product of a number 2 by itself is 4. In this case, 4 is termed as a perfect square. Example 1.\n\nInteger Perfect square\n23 x 23 529\n24 x 24 576\n25 x 25 625\n\n## What are the positive and negative square roots of 4?\n\nBut the roots could be positive or negative or we can say there are always two roots for any given number. Hence, root 4 is equal to ±2 or +2 and -2 ( positive 2 and negative 2). You can also find square root on a calculator. Square Root From 1 to 50.\n\nYou might be interested: Quick Answer: How To Put Square Root Symbol In Excel?\nNumber Square Root Value\n4 2\n5 2.236\n6 2.449\n7 2.646\n\n## What is Square Root 26 simplified?\n\nExplanation: 26 =2⋅13 has no square factors, so √ 26 cannot be simplified.\n\n## Is the square root of negative 25 a real number?\n\n√- 25 =? If you enter SQRT (- 25 ) into a spreadsheet on your computer, you will get an error message saying something like “argument must be greater or equal to 0” because the square root of negative 25 is not possible. There is no real number multiplied by itself that will equal – 25.\n\n## IS 400 a perfect square?\n\n400 is a perfect square. Because 20 * 20 = 400.\n\n## IS 144 a perfect square?\n\nA perfect square is an integer whose square root is always an integer. For example, 9, 36, 144 etc are perfect squares. As we know, 144 is a perfect square.\n\n## Is 72 a perfect square?\n\n72 is not a perfect square. It is represented as √ 72. The square root of 72 can only be simplified.\n\n## What is the square of 3?\n\nList of Perfect Squares\n\nNUMBER SQUARE SQUARE ROOT\n3 9 1.732\n4 16 2.000\n5 25 2.236\n6 36 2.449\n\n## Which is the square root of 144?\n\nThe value of the square root of 144 is equal to 12. In radical form, it is denoted as √ 144 = 12.\n\n## Is 20 a perfect square?\n\nA: No, the number 20 is not a perfect square.", null, "" ]	[ null, "data:image/svg+xml,%3Csvg%20xmlns='http://www.w3.org/2000/svg'%20viewBox='0%200%20110%20110'%3E%3C/svg%3E", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.8746784,"math_prob":0.9982644,"size":2648,"snap":"2021-21-2021-25","text_gpt3_token_len":769,"char_repetition_ratio":0.25869894,"word_repetition_ratio":0.1509091,"special_character_ratio":0.35234138,"punctuation_ratio":0.12580645,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.99853283,"pos_list":[0,1,2],"im_url_duplicate_count":[null,null,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2021-05-11T13:04:31Z\",\"WARC-Record-ID\":\"<urn:uuid:4f003c66-58a0-4c30-8d4b-8ceecd13c945>\",\"Content-Length\":\"36088\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:1a1c6c2f-8af1-4bfd-a942-2e7bc4633740>\",\"WARC-Concurrent-To\":\"<urn:uuid:9688bd10-302e-4867-8961-4079b5a513f0>\",\"WARC-IP-Address\":\"104.21.5.247\",\"WARC-Target-URI\":\"https://www.fmaths.com/square-root/often-asked-what-is-square-root-of-25.html\",\"WARC-Payload-Digest\":\"sha1:GPFCTS6GHIFMRIW7YKDX7GHK72QNEQYL\",\"WARC-Block-Digest\":\"sha1:4JZBGXH6NPBV6FK34S43V77R3S4WB5L3\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2021/CC-MAIN-2021-21/CC-MAIN-2021-21_segments_1620243989614.9_warc_CC-MAIN-20210511122905-20210511152905-00446.warc.gz\"}"}
http://werner.yellowcouch.org/log/a-reading-of-a-theano-compiled-graph/	[ "# A reading of a Theano compiled graph\n\nI’m trying to understand a compiled theano function, which I printed with theano.printing.debugprint. The full monster is printed below, yet it is only with a couple of lines that I have some problems.\n\nThis code first computes a random yes/no vector in node 6. Node 5 is used to create the appropriate shape (resembling x)\n\n```Gemm{inplace} [id A] <TensorType(float32, matrix)> '(dcost/dW)' 23\n\|Dot22 [id B] <TensorType(float32, matrix)> '' 22\n\| \|InplaceDimShuffle{1,0} [id C] <TensorType(float32, matrix)> 'x_tilde.T' 12\n\| \| \|Elemwise{Mul}[(0, 0)] [id D] <TensorType(float32, matrix)> 'x_tilde' 8\n\| \| \|RandomFunction{binomial}.1 [id E] <TensorType(float32, matrix)> '' 6\n\| \| \| \|<RandomStateType> [id F]\n\| \| \| \|MakeVector{dtype='int64'} [id G] <TensorType(int64, vector)> '' 5\n\| \| \| \| \|Shape_i{0} [id H] <TensorType(int64, scalar)> '' 1\n\| \| \| \| \| \|x [id I] <TensorType(float32, matrix)>\n\| \| \| \| \|Shape_i{1} [id J] <TensorType(int64, scalar)> '' 0\n\| \| \| \| \|x [id I] <TensorType(float32, matrix)>\n\| \| \| \|TensorConstant{1} [id K] <TensorType(int8, scalar)>\n\| \| \| \|TensorConstant{0.75} [id L] <TensorType(float32, scalar)>\n\| \| \|x [id I] <TensorType(float32, matrix)>\n\| \|Elemwise{Composite{((i0 - i1) * i2 * i1)}}[(0, 2)] [id M] <TensorType(float32, matrix)> '' 21\n\| \|TensorConstant{(1, 1) of 1.0} [id N] <TensorType(float32, (True, True))>\n\| \|Elemwise{Composite{scalar_sigmoid((i0 + i1))}}[(0, 0)] [id O] <TensorType(float32, matrix)> 'reduced' 15\n\| \| \|Dot22 [id P] <TensorType(float32, matrix)> '' 11\n\| \| \| \|Elemwise{Mul}[(0, 0)] [id D] <TensorType(float32, matrix)> 'x_tilde' 8\n\| \| \| \|W [id Q] <TensorType(float32, matrix)>\n\| \| \|InplaceDimShuffle{x,0} [id R] <TensorType(float32, row)> '' 2\n\| \| \|B [id S] <TensorType(float32, vector)>\n\| \|Dot22 [id T] <TensorType(float32, matrix)> '(dcost/dreduced)' 20\n\| \|Elemwise{Composite{((i0 * (i1 - Composite{scalar_sigmoid((i0 + i1))}(i2, i3)) * Composite{scalar_sigmoid((i0 + i1))}(i2, i3) * (i4 - Composite{scalar_sigmoid((i0 + i1))}(i2, i3))) / i5)}}[(0, 2)] [id U] <TensorType(float32, matrix)> '' 18\n\| \| \|TensorConstant{(1, 1) of -2.0} [id V] <TensorType(float32, (True, True))>\n\| \| \|x [id I] <TensorType(float32, matrix)>\n\| \| \|Dot22 [id W] <TensorType(float32, matrix)> '' 17\n\| \| \| \|Elemwise{Composite{scalar_sigmoid((i0 + i1))}}[(0, 0)] [id O] <TensorType(float32, matrix)> 'reduced' 15\n\| \| \| \|InplaceDimShuffle{1,0} [id X] <TensorType(float32, matrix)> 'W.T' 3\n\| \| \| \|W [id Q] <TensorType(float32, matrix)>\n\| \| \|InplaceDimShuffle{x,0} [id Y] <TensorType(float32, row)> '' 4\n\| \| \| \|B_Prime [id Z] <TensorType(float32, vector)>\n\| \| \|TensorConstant{(1, 1) of 1.0} [id N] <TensorType(float32, (True, True))>\n\| \| \|Elemwise{mul,no_inplace} [id BA] <TensorType(float32, (True, True))> '' 16\n\| \| \|InplaceDimShuffle{x,x} [id BB] <TensorType(float32, (True, True))> '' 14\n\| \| \| \|Subtensor{int64} [id BC] <TensorType(float32, scalar)> '' 10\n\| \| \| \|Elemwise{Cast{float32}} [id BD] <TensorType(float32, vector)> '' 7\n\| \| \| \| \|MakeVector{dtype='int64'} [id G] <TensorType(int64, vector)> '' 5\n\| \| \| \|Constant{1} [id BE]\n\| \| \|InplaceDimShuffle{x,x} [id BF] <TensorType(float32, (True, True))> '' 13\n\| \| \|Subtensor{int64} [id BG] <TensorType(float32, scalar)> '' 9\n\| \| \|Elemwise{Cast{float32}} [id BD] <TensorType(float32, vector)> '' 7\n\| \| \|Constant{0} [id BH]\n\| \|W [id Q] <TensorType(float32, matrix)>\n\|TensorConstant{1.0} [id BI] <TensorType(float32, scalar)>\n\|InplaceDimShuffle{1,0} [id BJ] <TensorType(float32, matrix)> '' 19\n\| \|Elemwise{Composite{((i0 * (i1 - Composite{scalar_sigmoid((i0 + i1))}(i2, i3)) * Composite{scalar_sigmoid((i0 + i1))}(i2, i3) * (i4 - Composite{scalar_sigmoid((i0 + i1))}(i2, i3))) / i5)}}[(0, 2)] [id U] <TensorType(float32, matrix)> '' 18\n\|Elemwise{Composite{scalar_sigmoid((i0 + i1))}}[(0, 0)] [id O] <TensorType(float32, matrix)> 'reduced' 15\n\|TensorConstant{1.0} [id BI] <TensorType(float32, scalar)>\nRandomFunction{binomial}.0 [id E] '' 6\n```\n\nThen at the second marked bold section, we see that node 16 performs the multiplication of two subnodes 14 and 13; which both are very similar except for a constant 0 or 1.\n\nThe code for node 14 reuses the same vector as node 5, but then casts it to float32 (which is node 7). And then the magic, the thing I do not understand happens. From that casted vector, a subtensor with constant 0 is selected. What does this do ?\n\nThe questions:\n\n1. Does the SubTensor (node 10 or node 9) select the first element of the tensor from node 7 ?\n2. Node 7 is merely a casted version of node 5. Does that vector contain the random data generated in node 6 ?\n3. Once the subtensors of node 7 are selected they are allowed to broadcast (node 13 and 14) to finally be multiplied with each other in node 16. Is it correct to say that node 16 thus computes the multiplication of the first random element with the second random element (from a random vector that might be quiet somewhat larger) ?\n4. When I print out the types of the nodes, we see that the output of the subtensor is indeed a scalar (as expected), yet the type of InPlaceDimensionShuffle and the ElementWise{mul} is (True,True). What for type is that ?\n5. If ElementWise{Mul} is not specifying ‘inplace’, (as happens in node 6), which of the two children is then modified ? Is it the node associated with the RandomFunction (thus node5) or does the randomFunction (node 6) provide us with another copy that can be modified ?\n\nThe image graph of the above function is given below. The yellow box is the one that confuses me because, if I read that right its output is merely a broadcastable 0.", null, "After a day bashing my head against this nonsense I fugerd out the following:\n\nMakevector in node 5 uses the shape as input parameters and concatenates them into a small vector (See http://deeplearning.net/software/theano/library/tensor/opt.html#theano.tensor.opt.MakeVector).\n\nThus its output is a vector with two values: the dimensions of x. The random generator then uses that as input to generate a correctly sized vector. And the yellow box in the computation merely multiplies the two dimensions with each other to calculate the element-count of the input. Answering each of the subquestions:\n\n1. yes the subtensor selects respectively the 0th and 1st element of the intput vector.\n2. node 7, does indeed contain the casted data from node 5. However node 5 does not contain the random data.\n3. it is wrong to say that node 16 computes the multiple of the two first random values. Right is: node 16 computes the multiplication of the dimension sizes of the input vector x.\n4. The (True,True) type merely tells us the broadcasting pattern. See section http://deeplearning.net/software/theano/library/tensor/basic.html#all-fully-typed-constructors\n5. Without inplace an elementwise multiplication destroys one of its inputs. The inputs that are destroyed are marked in red. documentation on color coding http://deeplearning.net/software/theano/library/printing.html#theano.printing.pydotprint" ]	[ null, "http://werner.yellowcouch.org/log/wp-content/uploads/2016/06/gradient_marked.png", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.64800346,"math_prob":0.96878844,"size":6922,"snap":"2023-40-2023-50","text_gpt3_token_len":2170,"char_repetition_ratio":0.23793004,"word_repetition_ratio":0.18436293,"special_character_ratio":0.33646345,"punctuation_ratio":0.12429378,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.998811,"pos_list":[0,1,2],"im_url_duplicate_count":[null,5,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2023-10-02T16:55:34Z\",\"WARC-Record-ID\":\"<urn:uuid:031771f9-741f-4796-a063-c6e450b81b38>\",\"Content-Length\":\"40602\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:4ddcbfe1-7105-4680-b816-c3916ebb469c>\",\"WARC-Concurrent-To\":\"<urn:uuid:788363f3-3a96-44b1-b88f-0b1389ebf9fe>\",\"WARC-IP-Address\":\"188.40.140.211\",\"WARC-Target-URI\":\"http://werner.yellowcouch.org/log/a-reading-of-a-theano-compiled-graph/\",\"WARC-Payload-Digest\":\"sha1:Q4CQTHBVDG3ZL7CVDVOFVX6WI3EHUGDU\",\"WARC-Block-Digest\":\"sha1:37JZ3AHOYIHJ4BUW5MP46EDTYRCPFTRK\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2023/CC-MAIN-2023-40/CC-MAIN-2023-40_segments_1695233511002.91_warc_CC-MAIN-20231002164819-20231002194819-00702.warc.gz\"}"}
https://thetechtrick.com/how-to-count-cells-in-google-sheets/	[ "Counting cells in Google Sheets is a useful feature that allows you to quickly determine the number of cells containing data within a specific range. Whether you need to count the total number of cells, count cells with specific values, or count cells based on certain criteria, Google Sheets provides various functions and methods to accomplish this task efficiently. In this guide, we will explore different techniques to count cells in Google Sheets, enabling you to analyze and summarize your data effectively.\n\n### Using the COUNT function in Google Sheets to count cells", null, "Google Sheets is a powerful tool that allows users to create and manipulate spreadsheets online. One of the most useful features of Google Sheets is the ability to count cells. Whether you need to count the number of cells that contain a specific value or simply want to know how many cells are in a range, Google Sheets has you covered.\n\nTo count cells in Google Sheets, you can use the COUNT function. The COUNT function is a built-in function that counts the number of cells in a range that meet a specified condition. This function is extremely versatile and can be used in a variety of ways.\n\nTo use the COUNT function, you first need to select the range of cells that you want to count. This can be a single cell, a range of cells, or even an entire column or row. Once you have selected the range, you can enter the COUNT function into a cell and specify the condition that you want to count.\n\nFor example, let’s say you have a spreadsheet that contains a list of students and their grades. You want to know how many students received a grade of “A”. To do this, you would select the range of cells that contains the grades and enter the following formula into a cell: =COUNTIF(A2:A10,”A”). This formula tells Google Sheets to count the number of cells in the range A2:A10 that contain the value “A”.\n\nIn addition to counting cells that meet a specific condition, you can also use the COUNT function to count cells that are not empty. This can be useful if you want to know how many cells in a range contain data. To count non-empty cells, you can use the COUNTA function. The COUNTA function works in a similar way to the COUNT function, but instead of counting cells that meet a condition, it counts cells that are not empty.\n\nTo use the COUNTA function, you simply need to select the range of cells that you want to count and enter the COUNTA function into a cell. For example, if you have a range of cells that contains student names and you want to know how many students are listed, you would enter the following formula into a cell: =COUNTA(A2:A10). This formula tells Google Sheets to count the number of cells in the range A2:A10 that are not empty.\n\nIn conclusion, counting cells in Google Sheets is a simple and powerful feature that can be used in a variety of ways. Whether you need to count cells that meet a specific condition or simply want to know how many cells are in a range, Google Sheets has you covered. By using the COUNT function, you can easily count cells and gain valuable insights from your data. So next time you find yourself needing to count cells in Google Sheets, remember to use the COUNT function and let Google Sheets do the work for you.\n\n### Counting cells based on specific criteria in Google Sheets\n\nGoogle Sheets is a powerful tool that allows users to create and manipulate spreadsheets online. One of the most common tasks in spreadsheet management is counting cells based on specific criteria. Whether you need to count the number of cells that meet a certain condition or count cells that contain specific text, Google Sheets provides several functions that can help you accomplish this task efficiently.\n\nOne of the most commonly used functions for counting cells in Google Sheets is the COUNTIF function. This function allows you to count the number of cells in a range that meet a specific condition. To use the COUNTIF function, you need to specify the range of cells you want to count and the condition that must be met. For example, if you want to count the number of cells in column A that contain the text “apple,” you would use the formula =COUNTIF(A:A, “apple”).\n\nIn addition to the COUNTIF function, Google Sheets also provides the COUNTIFS function, which allows you to count cells based on multiple criteria. This function is particularly useful when you need to count cells that meet more than one condition. To use the COUNTIFS function, you need to specify the ranges of cells you want to count and the conditions that must be met for each range. For example, if you want to count the number of cells in column A that contain the text “apple” and the number of cells in column B that contain the text “orange,” you would use the formula =COUNTIFS(A:A, “apple”, B:B, “orange”).\n\nAnother useful function for counting cells in Google Sheets is the COUNTA function. Unlike the COUNTIF and COUNTIFS functions, which count cells based on specific criteria, the COUNTA function counts all non-empty cells in a range. This can be particularly helpful when you need to count the total number of cells in a range, regardless of their content. To use the COUNTA function, you simply need to specify the range of cells you want to count. For example, if you want to count the total number of cells in column A, you would use the formula =COUNTA(A:A).\n\nIn addition to these functions, Google Sheets also provides several other functions that can be used to count cells based on specific criteria. The SUMIF function, for example, allows you to sum the values in a range that meet a specific condition. The AVERAGEIF function, on the other hand, allows you to calculate the average of the values in a range that meet a specific condition. These functions can be particularly useful when you need to perform calculations on cells that meet certain criteria.\n\n### Using formulas to count cells with text or numbers in Google Sheets\n\nGoogle Sheets is a powerful tool that allows users to create and manipulate spreadsheets online. One of the most common tasks in spreadsheet management is counting cells. Whether you need to count the number of cells containing text or numbers, Google Sheets provides several formulas that can help you accomplish this task efficiently.\n\nTo count cells with text in Google Sheets, you can use the COUNTIF formula. This formula allows you to specify a range of cells and a criteria to count the number of cells that meet that criteria. For example, if you want to count the number of cells in column A that contain the word “apple,” you can use the formula =COUNTIF(A:A, “apple”). This formula will return the total count of cells in column A that contain the word “apple.”\n\nSimilarly, if you need to count cells with numbers in Google Sheets, you can use the COUNT formula. The COUNT formula counts the number of cells in a range that contain numbers. For instance, if you want to count the number of cells in column B that contain numbers, you can use the formula =COUNT(B:B). This formula will give you the total count of cells in column B that contain numbers.\n\nIn some cases, you may need to count cells that meet multiple criteria. Google Sheets provides the COUNTIFS formula to handle such situations. The COUNTIFS formula allows you to specify multiple ranges and criteria to count cells that meet all the specified conditions. For example, if you want to count the number of cells in column C that contain the word “apple” and have a value greater than 10, you can use the formula =COUNTIFS(C:C, “apple”, C:C, “>10”). This formula will return the count of cells in column C that meet both conditions.\n\nFurthermore, Google Sheets offers the COUNTA formula, which counts the number of cells in a range that are not empty. This formula is useful when you want to count the total number of cells in a range, regardless of their content. For instance, if you want to count the number of cells in column D that are not empty, you can use the formula =COUNTA(D:D). This formula will give you the total count of non-empty cells in column D.\n\nIn addition to these formulas, Google Sheets provides other functions that can be used to count cells based on specific criteria. The SUMIF formula, for example, allows you to sum the values of cells that meet a certain condition. By using the SUMIF formula with a condition that evaluates to TRUE for each cell you want to count, you can achieve the desired result.\n\nIn conclusion, counting cells in Google Sheets is a fundamental task in spreadsheet management. Whether you need to count cells with text or numbers, Google Sheets offers a variety of formulas and functions that can help you accomplish this task efficiently. By using formulas like COUNTIF, COUNT, COUNTIFS, and COUNTA, you can easily count cells based on specific criteria and obtain the desired results. So, next time you find yourself needing to count cells in Google Sheets, remember these formulas and make your spreadsheet management a breeze.\n\nGoogle Sheets is a powerful tool that offers a wide range of features for data analysis and manipulation. One of the most basic tasks in spreadsheet analysis is counting cells. While the process may seem simple, there are actually several advanced techniques that can be used to count cells in Google Sheets. In this article, we will explore some of these techniques and how they can be applied to different scenarios.\n\nThe most straightforward way to count cells in Google Sheets is by using the COUNT function. This function allows you to count the number of cells in a range that meet specific criteria. For example, if you want to count the number of cells in a range that contain a certain value, you can use the formula =COUNTIF(range, criteria). The “range” parameter specifies the range of cells to be counted, while the “criteria” parameter defines the condition that must be met for a cell to be included in the count.\n\nAnother useful function for counting cells in Google Sheets is COUNTA. This function counts the number of cells in a range that are not empty. Unlike the COUNT function, which only counts cells that meet specific criteria, COUNTA counts all cells in a range, regardless of their content. This can be particularly useful when you want to count the total number of cells in a range, regardless of whether they contain specific values or not.\n\nIn addition to these basic counting functions, Google Sheets also offers more advanced techniques for counting cells. One such technique is using the FILTER function in combination with the COUNT function. The FILTER function allows you to create a new range of cells based on specific criteria. By using this function together with the COUNT function, you can count the number of cells in a range that meet certain conditions. For example, if you want to count the number of cells in a range that are greater than a certain value, you can use the formula =COUNT(FILTER(range, range > value)).\n\nAnother advanced technique for counting cells in Google Sheets is using the SUMPRODUCT function. This function allows you to perform calculations on arrays of values. By using the SUMPRODUCT function together with logical operators, you can count the number of cells in a range that meet multiple conditions. For example, if you want to count the number of cells in a range that are greater than a certain value and less than another value, you can use the formula =SUMPRODUCT((range > value1) * (range < value2)).\n\n## Q&A\n\n1. How can I count cells with specific text in Google Sheets?\nTo count cells with specific text in Google Sheets, you can use the COUNTIF function. For example, to count the number of cells containing the text “apple” in a range A1:A10, you can use the formula “=COUNTIF(A1:A10, “apple”)”.\n\n2. How can I count cells with numbers in Google Sheets?\nTo count cells with numbers in Google Sheets, you can use the COUNT function. For example, to count the number of cells with numbers in a range A1:A10, you can use the formula “=COUNT(A1:A10)”.\n\n3. How can I count cells with a specific color in Google Sheets?\nUnfortunately, Google Sheets does not have a built-in function to count cells based on their color. However, you can use a script or an add-on like “Color Code+” to achieve this functionality.\n\n4. How can I count cells with a specific condition in Google Sheets?\nTo count cells with a specific condition in Google Sheets, you can use the COUNTIF or COUNTIFS function. COUNTIF is used for a single condition, while COUNTIFS is used for multiple conditions. For example, to count the number of cells in a range A1:A10 that are greater than 5, you can use the formula “=COUNTIF(A1:A10, “>5″)”.In conclusion, counting cells in Google Sheets can be done using the COUNT function or by manually selecting the range of cells to be counted. The COUNT function is a simple and efficient way to count cells based on specific criteria, while manually selecting cells allows for more flexibility in counting non-contiguous ranges. Both methods provide accurate results and can be used according to the user’s preference and requirements." ]	[ null, "https://thetechtrick.com/wp-content/uploads/2023/08/7c93ccbcb79c9147c2dd2ea6807954ad.jpeg", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.89821345,"math_prob":0.8437772,"size":14400,"snap":"2023-40-2023-50","text_gpt3_token_len":2921,"char_repetition_ratio":0.23471798,"word_repetition_ratio":0.38321167,"special_character_ratio":0.20125,"punctuation_ratio":0.09705248,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.97939306,"pos_list":[0,1,2],"im_url_duplicate_count":[null,1,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2023-09-30T19:38:07Z\",\"WARC-Record-ID\":\"<urn:uuid:7045289c-ce59-48fd-bc2c-78e7507f4882>\",\"Content-Length\":\"90487\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:70bd4e3e-83f8-4e5d-9ffd-3e9b2382dfa1>\",\"WARC-Concurrent-To\":\"<urn:uuid:b26cc01d-fc85-4783-ab6a-e8eb58564860>\",\"WARC-IP-Address\":\"89.58.2.243\",\"WARC-Target-URI\":\"https://thetechtrick.com/how-to-count-cells-in-google-sheets/\",\"WARC-Payload-Digest\":\"sha1:GSNE3RZWX4MHKMGNFCPQTB4N7QBKFTCO\",\"WARC-Block-Digest\":\"sha1:75EIHLXMPCXCV4QJNACM3ZTYIL7MZC2J\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2023/CC-MAIN-2023-40/CC-MAIN-2023-40_segments_1695233510707.90_warc_CC-MAIN-20230930181852-20230930211852-00388.warc.gz\"}"}
https://chemistry.stackexchange.com/questions/40314/what-is-the-difference-between-tic-and-ticc	[ "# What is the difference between TIC and TICC?\n\nI have picked up a textbook on mass spectrometry, and I am finding it did not clearly distinguish between total ion chromatogram (TIC) and total ion current chromatogram (TICC) in the section (1.3 Ion Chromatograms) I just read. I suspect there was some sort of typo, but I don't have the experience to tell yet.\n\nHere is an abridged excerpt from the textbook that has confused me:\n\nThe total ion current (TIC) can either be measured by a hardware TIC monitor before mass analysis (nA to uA range), or its equivalent can be reconstructed or extracted after mass analysis. [...] Thus, the TIC represents a measure of the overall intensity of ion production of ion production or of mass spectral output as a function of time, respectively. The TIC obtained by means of data reduction, [...], is termed total ion chromatogram (TIC). The term total ion current chromatogram (TICC) refers to a chromatogram obtained by plotting the total ion detected in each of a series of mass spectra recorded as a function of retention times of the chromatographically separated components of a mixture (which is essentially implicated by: TIC).\n\nI would like to an explanation of the difference between total ion chromatogram (TIC), and total ion current chromatogram (TICC)." ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.93963844,"math_prob":0.77718693,"size":1255,"snap":"2019-35-2019-39","text_gpt3_token_len":270,"char_repetition_ratio":0.17266187,"word_repetition_ratio":0.019512195,"special_character_ratio":0.2063745,"punctuation_ratio":0.10344828,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9655245,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2019-09-16T10:24:16Z\",\"WARC-Record-ID\":\"<urn:uuid:4babdd45-51ec-4875-8980-2e70b5f63aff>\",\"Content-Length\":\"133535\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:07f2f6e8-722b-4277-9bfe-ced4773abd8c>\",\"WARC-Concurrent-To\":\"<urn:uuid:5bab3add-97c8-4472-b8c1-87f3ac2fac58>\",\"WARC-IP-Address\":\"151.101.193.69\",\"WARC-Target-URI\":\"https://chemistry.stackexchange.com/questions/40314/what-is-the-difference-between-tic-and-ticc\",\"WARC-Payload-Digest\":\"sha1:PRMQEHNKRVXVJFU3JHYMF6M5O7U7B7AH\",\"WARC-Block-Digest\":\"sha1:GCGSKSUU62RCRI6JK4EBHAFC2HTXVEUS\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2019/CC-MAIN-2019-39/CC-MAIN-2019-39_segments_1568514572517.50_warc_CC-MAIN-20190916100041-20190916122041-00215.warc.gz\"}"}
https://calculatorpack.com/whp-water-horsepower-calculator/	[ "# WHP (Water Horsepower) Calculator\n\nAre you interested in calculating the power of your water flow? The Water Horsepower Calculator is the perfect tool for you! This calculator is designed to help you determine the amount of power that can be generated by a water flow system, providing you with detailed and accurate calculations in just a few clicks. Whether you're a professional engineer or a DIY enthusiast, this calculator is accessible and user-friendly for anyone looking to harness the power of water. With the Water Horsepower Calculator, you can easily estimate the horsepower of your water flow and make informed decisions about your hydroelectric projects or water-powered machinery. Let's dive in and explore the incredible potential of water power!\n\n## WHP Calculator\n\nCalculate the water horsepower required to generate a certain amount of power\n\ngpm\nft\nEnter efficiency as a percentage (0-100)\nlbm/ft3\nin\nWHP Calculator Results\nFlow Rate0\nEfficiency0\nDensity0\nPump Type0\nImpeller Diameter0\nGear Ratio0\nWater Horsepower0\nBrake Horsepower0\nMotor Horsepower0\n\nTo understand the power output of engines, the whp water horsepower calculator is a useful tool. It complements our wheel horsepower calculator, allowing you to explore engine performance thoroughly.\n\n## How to Use the WHP Calculator\n\nThe WHP Calculator is a powerful tool designed to calculate the water horsepower required to generate a certain amount of power. It is commonly used in engineering and fluid mechanics to determine the necessary water horsepower for various applications such as pumps and motors. This calculator enables users to make informed decisions regarding power generation and equipment selection.\n\n## Instructions for Utilizing the Calculator\n\nTo effectively utilize the WHP Calculator, follow these steps:\n\n1. Flow Rate: Enter the flow rate in gallons per minute (gpm). This represents the volume of water flowing through the system per minute.\n2. Head: Specify the head in feet. Head refers to the height or pressure difference between the water source and the point of discharge.\n3. Efficiency: Input the efficiency as a percentage. Efficiency represents the effectiveness of the system in converting input energy to useful work. Enter a value between 0 and 100.\n4. Density: Enter the density of the water in pounds per cubic foot (lbm/ft³). Density represents the mass per unit volume of the water.\n5. Pump Type: Select the appropriate pump type from the available options: centrifugal or positive displacement. This choice depends on the specific characteristics and requirements of the system.\n\na. For centrifugal pumps:\n\n• Impeller Diameter: If using a centrifugal pump, enter the impeller diameter in inches (in). The impeller is a key component responsible for fluid circulation.\n\nb. For positive displacement pumps:\n\n• Gear Ratio: If using a positive displacement pump, input the gear ratio. The gear ratio represents the relationship between the sizes of the driving and driven gears.\n\nAfter entering the required values, click the Calculate WHP button to initiate the calculation. The calculator will determine the water horsepower, brake horsepower, and motor horsepower based on the provided input.\n\n## WHP Calculator Formula\n\nThe WHP Calculator employs the following formulas to calculate the respective values:\n\nWater Horsepower = (Flow Rate × Head) / (3960 × Efficiency × (Impeller Diameter<sup>4</sup>)) (for centrifugal pumps)\n\nWater Horsepower = (Flow Rate × Head × Density × 62.4 × Gear Ratio) / 550 (for positive displacement pumps)\n\nBrake Horsepower = Water Horsepower / Efficiency\n\nMotor Horsepower = Brake Horsepower / 0.85\n\nThese formulas enable the calculator to determine the water horsepower, brake horsepower, and motor horsepower based on the given input parameters.\n\n## Illustrative Example\n\nLet's consider an example to illustrate how the WHP Calculator works. Suppose we have a centrifugal pump with the following specifications:\n\n• Flow Rate: 500 gpm\n• Efficiency: 85%\n• Density: 62.4 lbm/ft³\n• Impeller Diameter: 10 inches\n\nAfter entering these values into the calculator, it will provide the following results:\n\n• Flow Rate: 500 gpm\n• Efficiency: 85%\n• Density: 62.4 lbm/ft³\n• Pump Type: CENTRIFUGAL\n• Impeller Diameter: 10 in\n• Water Horsepower: 4.65\n• Brake Horsepower: 5.47\n• Motor Horsepower: 6.43\n\nBased on the given parameters, the WHP Calculator determines that the water horsepower required is 4.65, resulting in a brake horsepower of 5.47 and a motor horsepower of 6.43.\n\n## Illustrative Table Example\n\nHere is a table showcasing multiple rows of example data and their corresponding results obtained from the WHP Calculator:\n\nFlow Rate\n\nEfficiency\n\nDensity\n\nPump Type\n\nImpeller Diameter\n\nGear Ratio\n\nWater Horsepower\n\nBrake Horsepower\n\nMotor Horsepower\n\n500 gpm100 ft85%62.4 lbm/ft³CENTRIFUGAL10 in-4.655.476.43\n800 gpm150 ft90%61.2 lbm/ft³CENTRIFUGAL12 in-7.568.9010.47\n1000 gpm200 ft80%60.8 lbm/ft³POSITIVE DISPLACEMENT-2.55.276.207.29\n\nThis table demonstrates how the WHP Calculator handles different combinations of input values, providing the corresponding water horsepower, brake horsepower, and motor horsepower.\n\nThe WHP Calculator is a valuable tool for engineers, fluid mechanics professionals, and individuals involved in power generation applications. By utilizing this calculator, you can accurately determine the water horsepower required to generate a certain amount of power. Additionally, the calculator provides insights into brake horsepower and motor horsepower, aiding in equipment selection and power system design.", null, "" ]	[ null, "https://calculatorpack.com/media/authors/aariz-ahmed.jpg", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.77806664,"math_prob":0.9309397,"size":5587,"snap":"2023-40-2023-50","text_gpt3_token_len":1285,"char_repetition_ratio":0.17427906,"word_repetition_ratio":0.08158508,"special_character_ratio":0.22391266,"punctuation_ratio":0.11649484,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9784341,"pos_list":[0,1,2],"im_url_duplicate_count":[null,null,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2023-11-30T01:01:31Z\",\"WARC-Record-ID\":\"<urn:uuid:ae92b086-5b37-4752-86f3-e725e96cdc16>\",\"Content-Length\":\"36822\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:f5e6505c-1e99-4337-ac0c-e8f77760bd0f>\",\"WARC-Concurrent-To\":\"<urn:uuid:3ac30480-145e-4304-94b4-68dd6043f05b>\",\"WARC-IP-Address\":\"104.21.93.233\",\"WARC-Target-URI\":\"https://calculatorpack.com/whp-water-horsepower-calculator/\",\"WARC-Payload-Digest\":\"sha1:PCINFS5TVZFLVC3LGQK7FB5DI7UZARUN\",\"WARC-Block-Digest\":\"sha1:T24AXX6DEGT2QRHLIVDHCVSWS6ULO45Q\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2023/CC-MAIN-2023-50/CC-MAIN-2023-50_segments_1700679100164.15_warc_CC-MAIN-20231130000127-20231130030127-00225.warc.gz\"}"}
https://discourse.odriverobotics.com/t/axis-error-0x01-when-put-into-closed-loop/1244	[ "# Axis error 0x01 when put into closed loop\n\nI purchased 2 odrive boards and 4 D6374 motors, and 4 CUI AMT102 encoders. We attached the encoders to our motors using a mount we designed. The odrive will successfully execute AXIS_STATE_FULL_CALIBRATION_SEQUENCE without an error. I then set odrv0.axis0.encoder.config.pre_calibrated = True and odrv0.axis0.motor.config.pre_calibrated = True, saved the configuration then rebooted. When the odrive is moved to AXIS_STATE_CLOSED_LOOP_CONTROL the odrv0.axis0.error = 0x01. Ive tried this sequence a number of times swapping out each piece of hardware (since I purchased multiple sets) with no resolution (making sure I erase the configuration and reboot at every swap). Needless to say I would think the should have been very easy since the controller and the motor come from odrive.\n\nThis the results:\n\nodrv0.axis0.motor returns:\n\n``````error = 0x0000 (int)\narmed_state = 0 (int)\nis_calibrated = True (bool)\ncurrent_meas_phB = -0.013124942779541016 (float)\ncurrent_meas_phC = 0.08837532997131348 (float)\nDC_calib_phB = -1.839632272720337 (float)\nDC_calib_phC = -2.2229766845703125 (float)\nphase_current_rev_gain = 0.02500000037252903 (float)\ncurrent_control:\np_gain = 0.022648924961686134 (float)\ni_gain = 31.99142074584961 (float)\nv_current_control_integral_d = 0.0 (float)\nv_current_control_integral_q = 0.0 (float)\nIbus = 0.0 (float)\nfinal_v_alpha = 0.0 (float)\nfinal_v_beta = 0.0 (float)\nIq_setpoint = 0.0 (float)\nIq_measured = 0.0 (float)\nmax_allowed_current = 71.99999237060547 (float)\ngate_driver:\ndrv_fault = 0 (int)\ntiming_log:\nTIMING_LOG_GENERAL = 0 (int)\nTIMING_LOG_MEAS_R = 0 (int)\nTIMING_LOG_MEAS_L = 0 (int)\nTIMING_LOG_ENC_CALIB = 0 (int)\nTIMING_LOG_IDX_SEARCH = 0 (int)\nTIMING_LOG_FOC_VOLTAGE = 0 (int)\nTIMING_LOG_FOC_CURRENT = 0 (int)\nconfig:\npre_calibrated = True (bool)\npole_pairs = 7 (int)\ncalibration_current = 10.0 (float)\nresistance_calib_max_voltage = 1.0 (float)\nphase_inductance = 2.2648924641544e-05 (float)\nphase_resistance = 0.031991422176361084 (float)\ndirection = 1 (int)\nmotor_type = 0 (int)\ncurrent_lim = 70.0 (float)\nrequested_current_range = 70.0 (float)\ncurrent_control_bandwidth = 1000.0 (float)\nset_current_control_bandwidth(current_control_bandwidth: float)\n``````\n\nodrv0.axis0.encoder returns:\n\n``````error = 0x0000 (int)\nindex_found = False (bool)\ncount_in_cpr = 3 (int)\ninterpolation = 0.5 (float)\nphase = 0.6743001937866211 (float)\npos_estimate = 3.9967904090881348 (float)\npll_vel = 0.0 (float)\npos_cpr = 3.765063762664795 (float)\nhall_state = 3 (int)\nvel_estimate = 0.0 (float)\nconfig:\nmode = 0 (int)\nuse_index = False (bool)\npre_calibrated = True (bool)\nidx_search_speed = 10.0 (float)\ncpr = 8192 (int)\noffset = 4558 (int)\noffset_float = 1.049887776374817 (float)\nbandwidth = 1000.0 (float)\ncalib_range = 0.019999999552965164 (float)\n``````\n\nodrv0.axis0.controller returns:\n\n``````pos_setpoint = 0.0 (float)\nvel_setpoint = 0.0 (float)\nvel_integrator_current = 0.0 (float)\ncurrent_setpoint = 0.0 (float)\nconfig:\ncontrol_mode = 2 (int)\npos_gain = 20.0 (float)\nvel_gain = 0.0005000000237487257 (float)\nvel_integrator_gain = 0.0010000000474974513 (float)\nvel_limit = 20000.0 (float)\nset_pos_setpoint(pos_setpoint: float, vel_feed_forward: float, current_feed_forward: float)\nctrl_reset()\nset_vel_setpoint(vel_setpoint: float, current_feed_forward: float)\nset_current_setpoint(current_setpoint: float)\nstart_anticogging_calibration()\n``````\n\nfinally, just the axis part from odrv0.axis0:\n\n``````error = 0x0001 (int)\nenable_step_dir = False (bool)\ncurrent_state = 1 (int)\nrequested_state = 0 (int)\nloop_counter = 1263250 (int)\nconfig:\nstartup_motor_calibration = False (bool)\nstartup_encoder_index_search = False (bool)\nstartup_encoder_offset_calibration = False (bool)\nstartup_closed_loop_control = False (bool)\nstartup_sensorless_control = False (bool)\nenable_step_dir = False (bool)\ncounts_per_step = 2.0 (float)\nramp_up_time = 0.4000000059604645 (float)\nramp_up_distance = 12.566370964050293 (float)\nspin_up_current = 10.0 (float)\nspin_up_acceleration = 400.0 (float)\nspin_up_target_vel = 400.0 (float)\n``````\n\nI assume your goal is to be able to start your motor control without doing calibration. To do that, please follow these instructions, specifically the “Encoder with Index signal” part.\n\nNote that you must search for the index on every startup before you can do closed loop control.\n\nIs there a way to not use the index? Thats what I was trying to do by doing the offset cal, then setting the pre-cal flag to True.\n\nSorry, not with these encoders." ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.6159201,"math_prob":0.99440336,"size":4107,"snap":"2022-05-2022-21","text_gpt3_token_len":1286,"char_repetition_ratio":0.20740922,"word_repetition_ratio":0.004048583,"special_character_ratio":0.37375212,"punctuation_ratio":0.16608392,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.99776924,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2022-05-28T17:36:46Z\",\"WARC-Record-ID\":\"<urn:uuid:29eb3e22-1daa-40bf-bf03-e896917a14c8>\",\"Content-Length\":\"24060\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:182638bf-b915-499e-badf-9a9952a3894c>\",\"WARC-Concurrent-To\":\"<urn:uuid:f996d431-b889-4e1c-aeff-4d15303ba7ff>\",\"WARC-IP-Address\":\"138.197.107.143\",\"WARC-Target-URI\":\"https://discourse.odriverobotics.com/t/axis-error-0x01-when-put-into-closed-loop/1244\",\"WARC-Payload-Digest\":\"sha1:SYPTAP7FYGEKISFTILA4RDNYJFMXK4SF\",\"WARC-Block-Digest\":\"sha1:AH2ATTFJXITMXANYYPAQZMIGBSZX5QPA\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2022/CC-MAIN-2022-21/CC-MAIN-2022-21_segments_1652663016949.77_warc_CC-MAIN-20220528154416-20220528184416-00786.warc.gz\"}"}
https://www.tutorialspoint.com/find-the-ratio-in-which-the-line-2-xplus3-y-5-0-divides-the-line-segment-joining-the-points-8-9-and-2-1-also-find-the-coordinates-of-the-point-o	[ "# Find the ratio in which the line $2 x+3 y-5=0$ divides the line segment joining the points $(8,-9)$ and $(2,1)$. Also find the coordinates of the point of division.\n\nGiven:\n\nThe line $2 x+3 y-5=0$ divides the line segment joining the points $(8,-9)$ and $(2,1)$.\n\nTo do:\n\nWe have to find the ratio in which the line $2 x+3 y-5=0$ divides the line segment joining the points $(8,-9)$ and $(2,1)$ and the coordinates of the point of division.\n\nSolution:\n\nLet the line $2 x+3 y-5=0$ divides the line segment joining the points $A(8,-9)$ and $B(2,1)$ in the ratio $k: 1$ at point $P$.\n\nUsing section formula, we get,\n\n$(x,y)=(\\frac{mx_{2}+nx_{1}}{m+n}, \\frac{my_{2}ny_{1}}{m+n})$\n\nCoordinates of $P=(\\frac{k(2)+1(8)}{2+1}, \\frac{k(1)+1(-9)}{2+1})$\n\n$=(\\frac{2k+8}{k+1}, \\frac{k-9}{k+1})$\n\nPoint $P$ lies on the line $2 x+3y-5=0$.\n\nThis implies,\n\n$2(\\frac{2k+8}{k+1})+3(\\frac{k-9}{k+1})-5=0$\n\n$2(2k+8)+3(k-9)-5(k+1)=0$\n\n$4k+16+3k-27-5 k-5=0$\n\n$2k-16=0$\n\n$k=8$\n\n$\\Rightarrow k: 1=8: 1$\n\nTherefore, the point $P$ divides the line in the ratio $8: 1$.\n\nThe point of division $P=(\\frac{2(8)+8}{8+1}, \\frac{8-9}{8+1})$\n\n$=(\\frac{16+8}{9},\\frac{-1}{9})$\n\n$=(\\frac{24}{9}, \\frac{-1}{9})$\n\n$=(\\frac{8}{3}, \\frac{-1}{9})$\n\nHence, the required point of division is $(\\frac{8}{3}, \\frac{-1}{9})$." ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.7415097,"math_prob":1.0000097,"size":4143,"snap":"2023-14-2023-23","text_gpt3_token_len":1411,"char_repetition_ratio":0.25175163,"word_repetition_ratio":0.27530363,"special_character_ratio":0.3968139,"punctuation_ratio":0.12238148,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":1.0000094,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2023-05-29T16:34:07Z\",\"WARC-Record-ID\":\"<urn:uuid:cb0b0542-df87-44cc-a130-15130ddbf0a5>\",\"Content-Length\":\"45379\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:983ac536-9ce9-41c4-8c8c-96b791d1447b>\",\"WARC-Concurrent-To\":\"<urn:uuid:94c5126b-03c5-45aa-bbe0-8d8023bbab99>\",\"WARC-IP-Address\":\"192.229.210.176\",\"WARC-Target-URI\":\"https://www.tutorialspoint.com/find-the-ratio-in-which-the-line-2-xplus3-y-5-0-divides-the-line-segment-joining-the-points-8-9-and-2-1-also-find-the-coordinates-of-the-point-o\",\"WARC-Payload-Digest\":\"sha1:MZIBPVE2BNV5MX6OAFTSYXZXJUOOUAFP\",\"WARC-Block-Digest\":\"sha1:ZOKZT6DCWXOI3PH7HSR2KASH2VAYUYM5\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2023/CC-MAIN-2023-23/CC-MAIN-2023-23_segments_1685224644867.89_warc_CC-MAIN-20230529141542-20230529171542-00708.warc.gz\"}"}
https://se.mathworks.com/help/instrument/get.html	[ "# get\n\nInstrument object properties\n\n## Syntax\n\n```get(obj) out = get(obj) out = get(obj,'PropertyName') ```\n\n## Arguments\n\n `obj` An instrument object or an array of instrument objects. `'``PropertyName``'` A property name or a cell array of property names. `out` A single property value, a structure of property values, or a cell array of property values.\n\n## Description\n\n`get(obj)` returns all property names and their current values to the command line for `obj`. The properties are divided into two sections. The base properties are listed first and the object-specific properties are listed second.\n\n`out = get(obj)` returns the structure `out` where each field name is the name of a property of `obj`, and each field contains the value of that property.\n\n`out = get(obj,'PropertyName')` returns the value `out` of the property specified by `PropertyName` for `obj`. If `PropertyName` is replaced by a 1-by-n or n-by-1 cell array of character vectors containing property names, then `get` returns a 1-by-n cell array of values to `out`. If `obj` is an array of instrument objects, then `out` will be an m-by-n cell array of property values where m is equal to the length of `obj` and n is equal to the number of properties specified.\n\n## Examples\n\nThis example illustrates some of the ways you can use `get` to return property values for the GPIB object `g`.\n\n```g = gpib('ni',0,1); out1 = get(g); out2 = get(g,{'PrimaryAddress','EOSCharCode'}); get(g,'EOIMode') ans = on```\n\n## Tips\n\nWhen specifying a property name, you can do so without regard to case, and you can make use of property name completion. For example, if `g` is a GPIB object, then these commands are all valid.\n\n```out = get(g,'EOSMode'); out = get(g,'eosmode'); out = get(g,'EOSM');```\n\n## Version History\n\nIntroduced before R2006a" ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.62179756,"math_prob":0.88480884,"size":1683,"snap":"2023-40-2023-50","text_gpt3_token_len":413,"char_repetition_ratio":0.18046457,"word_repetition_ratio":0.014492754,"special_character_ratio":0.23410577,"punctuation_ratio":0.112094395,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.95663226,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2023-09-22T05:23:16Z\",\"WARC-Record-ID\":\"<urn:uuid:76703372-03c8-4050-b24e-233c255ba836>\",\"Content-Length\":\"80349\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:d5501a64-3c71-4229-84a5-5decff65072b>\",\"WARC-Concurrent-To\":\"<urn:uuid:9ee69b5f-117b-40e6-a75a-76b34c54fc6c>\",\"WARC-IP-Address\":\"23.34.160.82\",\"WARC-Target-URI\":\"https://se.mathworks.com/help/instrument/get.html\",\"WARC-Payload-Digest\":\"sha1:TFWL5V55PQKGMOOU2AA7FHS3X3QRGBW4\",\"WARC-Block-Digest\":\"sha1:LU7I62WAD5VXTGWFR37W6H5IJFZJ4BKI\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2023/CC-MAIN-2023-40/CC-MAIN-2023-40_segments_1695233506329.15_warc_CC-MAIN-20230922034112-20230922064112-00187.warc.gz\"}"}
https://nforum.ncatlab.org/discussion/9836/	[ "# Start a new discussion\n\n## Not signed in\n\nWant to take part in these discussions? Sign in if you have an account, or apply for one below\n\n## Site Tag Cloud\n\nVanilla 1.1.10 is a product of Lussumo. More Information: Documentation, Community Support.\n\n• CommentRowNumber1.\n• CommentAuthorUrs\n• CommentTimeApr 19th 2019\n\nam starting something here, to be expanded…\n\n• CommentRowNumber2.\n• CommentAuthorUrs\n• CommentTimeJun 19th 2019\n\nam taking the liberty of adding pointer to\n\n• CommentRowNumber3.\n• CommentAuthorDavid_Corfield\n• CommentTimeJun 19th 2019\n\nWell, from a cursory look, there aren’t too many other references to choose. Cruikshank’s thesis has a relevant Chap. 7, perhaps the source for the entry included in cohomotopy\n\n• James Cruickshank, Twisted homotopy theory and the geometric equivariant 1-stem, Topology and its Applications Volume 129, Issue 3, 1 April 2003, Pages 251-271 (arXiv:10.1016/S0166-8641(02)00183-9)\n\nI think that’s just the stable version.\n\n• CommentRowNumber4.\n• CommentAuthorUrs\n• CommentTimeJun 19th 2019\n\nThanks. I have added the pointer.\n\n• CommentRowNumber5.\n• CommentAuthorUrs\n• CommentTimeSep 28th 2019\n• (edited Sep 28th 2019)\n\nadded statement of one form of what we like to call the “twisted Pontrjagin-Thom theorem”, currently it reads as follows:\n\nLet\n\n1. $X^n$ be a closed manifold of dimension $n$;\n\n2. $1 \\leq k \\in \\mathbb{N}$ a positive natural number.\n\nThen the scanning map constitutes a weak homotopy equivalence\n\n$\\underset{ \\color{blue} { \\phantom{a} \\atop \\text{ J-twisted Cohomotopy space}} }{ Maps_{{}_{/B O(n)}} \\Big( X^n \\;,\\; S^{ \\mathbf{n}_{def} + \\mathbf{k}_{\\mathrm{triv}} } \\!\\sslash\\! O(n) \\Big) } \\underoverset {\\simeq} { \\color{blue} \\text{scanning map} } {\\longleftarrow} \\underset{ \\mathclap{ \\color{blue} { \\phantom{a} \\atop { \\text{configuration space} \\atop \\text{of points} } } } }{ Conf \\big( X^n, S^k \\big) }$\n\nbetween\n\n1. the J-twisted (n+k)-Cohomotopy space of $X^n$, hence the space of sections of the $(n + k)$-spherical fibration over $X$ which is associated via the tangent bundle by the O(n)-action on $S^{n+k} = S(\\mathbb{R}^{n} \\times \\mathbb{R}^{k+1})$\n\n2. the configuration space of points on $X^n$ with labels in $S^k$.\n\n• CommentRowNumber6.\n• CommentAuthorUrs\n• CommentTimeMar 7th 2020\n\nAlso took the liberty of adding pointer to\n\n• CommentRowNumber7.\n• CommentAuthorUrs\n• CommentTimeMar 3rd 2021\n\nThe section titled “Twisted Pontrjagin-Thom theorem” really talked about the “May-Segal theorem” (i.e. the negative codimension version which gives configuration spaces of points, here).\n\nI have now instead added a brief pointer to twisted Pontrjagin theorem and gave the previous material a new header “Twisted May-Segal theorem”.\n\nI should really create an entry May-Segal theorem (as it’s somewhat buried at configuration space of points). But not right now." ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.8867082,"math_prob":0.8689214,"size":1723,"snap":"2022-27-2022-33","text_gpt3_token_len":442,"char_repetition_ratio":0.10645724,"word_repetition_ratio":0.02189781,"special_character_ratio":0.22925131,"punctuation_ratio":0.1260997,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9803414,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2022-08-19T11:04:27Z\",\"WARC-Record-ID\":\"<urn:uuid:0fde8e5b-975c-47a0-81b4-e6c578713be9>\",\"Content-Length\":\"60715\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:8f0d6704-633a-42b5-8bd7-9767c42de766>\",\"WARC-Concurrent-To\":\"<urn:uuid:2d0c70f1-ff29-4a9d-b3d4-acf1591e0317>\",\"WARC-IP-Address\":\"104.21.17.153\",\"WARC-Target-URI\":\"https://nforum.ncatlab.org/discussion/9836/\",\"WARC-Payload-Digest\":\"sha1:22J7BDMVBNOHYDM6MW22U3ELTP7UXE7Q\",\"WARC-Block-Digest\":\"sha1:LNZ4XLIMZSZUVTP5MXB2R7YY2NHANV4F\",\"WARC-Identified-Payload-Type\":\"application/xhtml+xml\",\"warc_filename\":\"/cc_download/warc_2022/CC-MAIN-2022-33/CC-MAIN-2022-33_segments_1659882573667.83_warc_CC-MAIN-20220819100644-20220819130644-00677.warc.gz\"}"}
http://currency7.com/tjs-to-lrd-exchange-rate-converter	[ "# Currency Converter · Tajikistani Somoni (TJS) to Liberian Dollar (LRD)\n\nThe currency calculator will convert exchange rate of Tajikistani somoni (TJS) to Liberian dollar (LRD).\n\n• Tajikistani somoni\nThe Tajikistani somoni (TJS) is the currency of Tajikistan. The currency code is TJS and currency symbol is SM. The Tajikistani somoni is subdivided into 100 diram (singular: diram). Frequently used Tajikistani somoni coins are in denominations of 1 somoni, 3 somoni, 5 somoni, 5 diram, 10 dirham, 20 dirham, 25 dirham, 50 dirham. Frequently used Tajikistani somoni banknotes are in denominations of 1 somoni, 3 somoni, 5 somoni, 10 somoni, 20 somoni, 50 somoni, 100 somoni, 200 somoni, 500 somoni, 1 dirham, 5 dirham, 20 dirham, 50 dirham.\n• Liberian dollar\nThe Liberian dollar (LRD) is the currency of Liberia. The currency code is LRD and currency symbol is \\$, or L\\$. The Liberian dollar is subdivided into 100 cents (singular: cent; symbol: ¢). Frequently used Liberian dollar coins are in denominations of L\\$1, 5 ¢, 10 ¢, 25 ¢, 50 ¢. Frequently used Liberian dollar banknotes are in denominations of L\\$5, L\\$10, L\\$20, L\\$50, L\\$100.\n• 1 TJS = 17.17 LRD\n• 5 TJS = 85.84 LRD\n• 10 TJS = 171.67 LRD\n• 20 TJS = 343.34 LRD\n• 25 TJS = 429.18 LRD\n• 50 TJS = 858.36 LRD\n• 100 TJS = 1,716.72 LRD\n• 200 TJS = 3,433.44 LRD\n• 250 TJS = 4,291.80 LRD\n• 500 TJS = 8,583.60 LRD\n• 1,000 TJS = 17,167.20 LRD\n• 2,000 TJS = 34,334.40 LRD\n• 2,500 TJS = 42,918.00 LRD\n• 5,000 TJS = 85,836.00 LRD\n• 10,000 TJS = 171,672.01 LRD\n• 10 LRD = 0.58 TJS\n• 50 LRD = 2.91 TJS\n• 100 LRD = 5.83 TJS\n• 250 LRD = 14.56 TJS\n• 500 LRD = 29.13 TJS\n• 1,000 LRD = 58.25 TJS\n• 2,000 LRD = 116.50 TJS\n• 2,500 LRD = 145.63 TJS\n• 5,000 LRD = 291.25 TJS\n• 10,000 LRD = 582.51 TJS\n• 20,000 LRD = 1,165.01 TJS\n• 50,000 LRD = 2,912.53 TJS\n• 100,000 LRD = 5,825.06 TJS\n• 250,000 LRD = 14,562.65 TJS\n• 500,000 LRD = 29,125.31 TJS\n\n## Popular TJS pairing\n\n` <a href=\"http://currency7.com/TJS-to-LRD-exchange-rate-converter?amount=300\">300 TJS in LRD</a> `" ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.5789073,"math_prob":0.99028623,"size":3095,"snap":"2023-40-2023-50","text_gpt3_token_len":1135,"char_repetition_ratio":0.27208024,"word_repetition_ratio":0.04638219,"special_character_ratio":0.3657512,"punctuation_ratio":0.17045455,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9707162,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2023-12-01T22:04:50Z\",\"WARC-Record-ID\":\"<urn:uuid:71d9b29d-72db-4eb1-8f61-b83214a072ad>\",\"Content-Length\":\"29442\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:278dd9c1-e0fc-493f-869e-b9295d0e843a>\",\"WARC-Concurrent-To\":\"<urn:uuid:39584974-e26d-45ca-9551-73cd9df52629>\",\"WARC-IP-Address\":\"70.35.206.41\",\"WARC-Target-URI\":\"http://currency7.com/tjs-to-lrd-exchange-rate-converter\",\"WARC-Payload-Digest\":\"sha1:PFEPMMMQ5JGN5Z7OHEQXPWY2XURRNUYE\",\"WARC-Block-Digest\":\"sha1:RTAFJJ7ZFSNGDRZOJZ6SZ2JDJVX5RJ24\",\"WARC-Identified-Payload-Type\":\"application/xhtml+xml\",\"warc_filename\":\"/cc_download/warc_2023/CC-MAIN-2023-50/CC-MAIN-2023-50_segments_1700679100308.37_warc_CC-MAIN-20231201215122-20231202005122-00111.warc.gz\"}"}
https://webapps.stackexchange.com/questions/160032/google-sheets-formulas-how-to-return-where-a-min-or-max-values-within-a-range-w	[ "# Google Sheets formulas: how to return where a MIN or MAX values within a range were found?\n\nI'm trying to find a way to get the index of a cell where the value was found; more specifically, I want to return the numerical value of the row where it was found, as I need the number to be used in some calculation.\n\nI tried using \"LOOKUP\" with the returned MIN or MAX value — but that for some reason returns the no match was found.\n\nAny suggestions? Thank you in advance.\n\n• Welcome to Web Applications Stack Exchange. Consider sharing a publicly editable sample spreadsheet with realistic-looking data, and show your hand-entered expected results there. Nov 5, 2021 at 10:45\n• The perceived need to find a row number often suggests that you in the end want to locate some value as well, rather than just finding the row number. See What is the XY problem? Nov 5, 2021 at 10:46\n• @doubleunary The value I want is the value of row where the MIN or MAX values are. The numerical content of the cells is already available through the simple =MIN() and =MAX() formulas.\n– TLSO\nNov 5, 2021 at 18:41\n\nUse `query()`, like this:\n\n``````=arrayformula(\nquery(\nsplit(\nflatten( row(A2:F) & \"→\" & A2:F ),\n\"→\", false, true\n),\n\"select Col1 where Col2 = \" & max(A2:F), 0\n)\n)\n``````\n• That worked! But I would like to know what each component of the formula does. Care to explain? Thank you.\n– TLSO\nNov 5, 2021 at 18:42\n• See query(), split() and flatten(). An `arrayformula()` wrapper lets you evaluate a formula over a range of cells rather than a single cell, returning multiple results automatically. It is required here primarily to make the `&` operator work over arrays. Nov 5, 2021 at 22:26\n• I assume it makes an array of two columns, one with numbers and one with the values and this syntax tells it somehow to return the number in column 1 when the adjacent cell in column 2 has the result for MAX or MIN? I'm not familiar with these formulas other than ARRAYFORMULA(), and I suppose Google's explanations could be formulated more clearly.\n– TLSO\nNov 5, 2021 at 22:51\n• That is correct. Nov 5, 2021 at 23:04" ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.97465146,"math_prob":0.88300323,"size":374,"snap":"2023-40-2023-50","text_gpt3_token_len":85,"char_repetition_ratio":0.14594595,"word_repetition_ratio":0.0,"special_character_ratio":0.22459893,"punctuation_ratio":0.0875,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.99360436,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2023-12-07T14:56:41Z\",\"WARC-Record-ID\":\"<urn:uuid:bc860a81-1706-4259-b334-4e857e6275e3>\",\"Content-Length\":\"169130\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:a7e2572a-105c-42bf-8878-dd54936febf6>\",\"WARC-Concurrent-To\":\"<urn:uuid:ef90f3a4-f331-48e4-89d4-40b92099ffcb>\",\"WARC-IP-Address\":\"104.18.43.226\",\"WARC-Target-URI\":\"https://webapps.stackexchange.com/questions/160032/google-sheets-formulas-how-to-return-where-a-min-or-max-values-within-a-range-w\",\"WARC-Payload-Digest\":\"sha1:WM5LGNUG7VHQGC766YB7JGP4BLKHFCME\",\"WARC-Block-Digest\":\"sha1:JWTEZZDILQ2SVU6RHF6IXLITMJ4TAINJ\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2023/CC-MAIN-2023-50/CC-MAIN-2023-50_segments_1700679100674.56_warc_CC-MAIN-20231207121942-20231207151942-00869.warc.gz\"}"}
https://openpunk.com/post/7	[ "## Making a Lua Bytecode parser in Python\n\n611 views lua python reverse-engineering\n\nby CPunch", null, "## [The repository for this project can be found here]\n\nSo recently I've been getting back into Lua, my first scripting language. I've already done a series about manipulating the LuaVM, (which you can read here) but this time I was interested in the LuaVM bytecode, specifically the Lua 5.1 bytecode. If you don't know what bytecode is or even how Lua works, here's a basic rundown:\n\n• LuaC is the Lua Compiler. Its job is to turn our human readable script into Lua Bytecode ready to be executed by the LVM (LuaVM)\n• This bytecode is everything the LVM needs to run!\n• Constants\n• Locals\n• Protos (The functions)\n• and even Debug information and line info\n\nNow I know what you're thinking \"Who cares! Why would I need to know this!\" Well, being able to parse bytecode can enable us to do many things! To name a few:\n\n• We could easily edit precompiled Lua Scripts embedded in a game. :eyes:\n• Help un-obfuscate scripts. :eyes:\n\nUnfortunately the Lua bytecode has no real standard and changes from version to version. Meaning, our Lua 5.1 Bytecode parser won't be able to read Lua 5.3 Bytecode. Another unfortunate thing is that there is NO official documentation for Lua Bytecode since there is no real standard. Luckily however some really cool guy who made ChunkSpy also wrote some super cool paper about the Lua 5.1 Bytecode!! You can read his amazing paper here! He talks about some really important stuff like how the instructions are encoded, and the Lua chunk header.\n\nTo start off I'm going to make a basic lua script and use luac to compile it.\n\n```print(\"hello world\")\n```\n\nI know, I know, I should be scripting for NASA. But the simplicity of this will help us tiptoe into the deepend of the Lua Bytecode later on.\n\nTo compile this script, save it as \"epic.lua\" and call luac like so:\n\n```luac -o epic.luac epic.lua\n```", null, "You won't see much but your script was just compiled into Lua bytecode! if you want you can even try to read the compiled script.", null, "Hmmm a lot of weird symbols. Those symbols before 'epic.luaA' is part of our chunk header, the ones after are our instructions. You can see our 'hello world' and 'print' is readable. Lua stores these as constants and are human readable,,, for the most part.\n\nAnyways lets actually talk about parsing this bytecode. All LVM Bytecode starts with a header. This just lets us know how to correctly parse the bytecode and what version it is. It's read as the following in this order:\n\n• First 4 bytes are hex 0x1b and 'Lua' - 4 Bytes\n• Lua revision - 1 Byte\n• Lua format - 1 Byte\n• Endian flag - 1 Byte\n• int size - 1 Byte\n• size_t size - 1 Byte\n• instruction size - 1 Byte\n• lua_Number size - 1 Byte\n• integral flag - 1 Byte\n\nor you can just reference the paper lol:", null, "```class LuaCompiler:\ndef __init__(self):\nself.chunks = []\nself.chunk = {}\nself.index = 0\n\ndef get_byte(self):\nb = self.bytecode[self.index]\nself.index = self.index + 1\nreturn b\n\ndef get_string(self, size):\ns = \"\".join(chr(x) for x in self.bytecode[self.index:self.index+size])\nself.index = self.index + size\nreturn s\n\ndef decode_bytecode(self, bytecode):\nself.bytecode = bytecode\n\nself.signature_byte = self.get_byte()\nself.signature = self.get_string(3)\nself.vm_version = self.get_byte()\nself.bytecode_format = self.get_byte()\nself.big_endian = (self.get_byte() == 0)\nself.int_size = self.get_byte()\nself.size_t = self.get_byte()\nself.instr_size = self.get_byte() # gets size of instructions\nself.l_number_size = self.get_byte() # size of lua_Number\nself.integral_flag = self.get_byte()\n\nprint(\"Lua VM version: \", hex(self.vm_version))\nprint(\"Big Endian: \", self.big_endian)\nprint(\"int_size: \", self.int_size)\nprint(\"size_t: \", self.size_t)\n```\n\nNow we're going to have to talk about Instructions. The Lua 5.1 VM has 38 different opcodes for 38 instructions. There are 3 main registers, A, B, and C. Not all instructions use all three, with 3 main different types of instructions, each with different ways to encode the registers.\n\n• iABC - This type of instruction uses all three registers, with each representing an unsigned integer.\n• iABx - This type of instruction uses A and B, both representing an unsigned integer as well.\n• iAsBx - This type of instruction uses A and B, however B can represent a negative number. However the B in this instruction is strange. instead of having 1 big represent the sign, the range is -131071 to 131071. It's encoded as a regular unsigned integer however, so to get the actual number, you subtract 131071.\n\nAll instructions start with the opcode [6 bits], and use the A register [8 bits] however are encoded differently per type:\n\n• iABC - B and C are both 9 bits\n• iABx and iAsBx - B is 18 bits\n\nYou can also reference this very helpful chart!", null, "In lua Bytecode, datatypes are stored as the following:\n\n• Integer - Usually 4 bytes long, integer.\n• Lua_TNumber - all lua numbers are stored as this, 8 Bytes long, double.\n• String\n• First byte is the size of the string\n• Characters\n• Boolean - 1 Byte, only (1 or 0)\n\nNow, let's write some code to read the datatypes before parsing the function chunk.\n\nWe'll need to be able to get the binary of bytes for the instructions, to do that I'll be using python's 'bin' function which turns any number into base 2, aka BInary. However we'll still need it to be 32 bits long, so any missing bits I'll just add a tracing 0.\n\nThat'll look like this:\n\n```# at [p]osition to k\ndef get_bits(num, p, k):\n# convert number into binary first\nbinary = bin(num)\n\n# remove first two characters\nbinary = binary[2:]\n\n# fill in missing bits\nfor i in range(32 - len(binary)):\nbinary = '0' + binary\n\nend = len(binary) - p + 1\nstart = len(binary) - k + 1\n\n# extract k bit sub-string\nkBitSubStr = binary[start : end]\n\n# convert extracted sub-string into decimal again\nreturn (int(kBitSubStr,2))\n```\n\nThis method lets us parse the binary of an instruction, and extract the specific bits we want, then convert them back into base 10. Pretty cool :)\n\nHowever, we still aren't done. We need to parse multiple bytes into a double, integer, etc. To do that, we'll start with what we wrote previously, but with a few more changes. I mainly used python's struct module to be able to parse these bytes.\n\nHere's what that looks like:\n\n```\nclass LuaCompiler:\ndef __init__(self):\nself.chunks = []\nself.chunk = {}\nself.index = 0\n\ndef get_byte(self):\nb = self.bytecode[self.index]\nself.index = self.index + 1\nreturn b\n\ndef get_int32(self):\ni = 0\nif (self.big_endian):\ni = int.from_bytes(self.bytecode[self.index:self.index+4], byteorder='big', signed=False)\nelse:\ni = int.from_bytes(self.bytecode[self.index:self.index+4], byteorder='little', signed=False)\nself.index = self.index + self.int_size\nreturn i\n\ndef get_int(self):\ni = 0\nif (self.big_endian):\ni = int.from_bytes(self.bytecode[self.index:self.index+self.int_size], byteorder='big', signed=False)\nelse:\ni = int.from_bytes(self.bytecode[self.index:self.index+self.int_size], byteorder='little', signed=False)\nself.index = self.index + self.int_size\nreturn i\n\ndef get_size_t(self):\ns = ''\nif (self.big_endian):\ns = int.from_bytes(self.bytecode[self.index:self.index+self.size_t], byteorder='big', signed=False)\nelse:\ns = int.from_bytes(self.bytecode[self.index:self.index+self.size_t], byteorder='little', signed=False)\nself.index = self.index + self.size_t\nreturn s\n\ndef get_double(self):\nif self.big_endian:\nf = struct.unpack('>d', bytearray(self.bytecode[self.index:self.index+8]))\nelse:\nf = struct.unpack('<d', bytearray(self.bytecode[self.index:self.index+8]))\nself.index = self.index + 8\nreturn f\n\ndef get_string(self, size):\nif (size == None):\nsize = self.get_size_t()\nif (size == 0):\nreturn None\n\ns = \"\".join(chr(x) for x in self.bytecode[self.index:self.index+size])\nself.index = self.index + size\nreturn s\n\ndef decode_bytecode(self, bytecode):\nself.bytecode = bytecode\n\nself.signature_byte = self.get_byte()\nself.signature = self.get_string(3)\nself.vm_version = self.get_byte()\nself.bytecode_format = self.get_byte()\nself.big_endian = (self.get_byte() == 0)\nself.int_size = self.get_byte()\nself.size_t = self.get_byte()\nself.instr_size = self.get_byte() # gets size of instructions\nself.l_number_size = self.get_byte() # size of lua_Number\nself.integral_flag = self.get_byte()\n\nprint(\"Lua VM version: \", hex(self.vm_version))\nprint(\"Big Endian: \", self.big_endian)\nprint(\"int_size: \", self.int_size)\nprint(\"size_t: \", self.size_t)\n```\n\nAlright now that we have that, we can decode our proto chunks.\n\nAfter the header is the first function chunk. This includes:\n\n• Name - dynamic size\n• First line - Integer\n• Last line - integer\n• Upvalues - 1 Byte\n• Arguments - 1 Byte\n• VArg - 1 Byte\n• Stack - 1 Byte\n• Instuction list\n• Number of instructions - Integer\n• Instruction list - Dynamic size\n• Constant list\n• Number of constants - Integer\n• Constants are stored as:\n• first byte is type of constant - 1 Byte\n• 4 main types of constants:\n• Type == 0: Nil - No data, ignore\n• Type == 1: Boolean - 1 Byte (1 or 0)\n• Type == 3: Lua_TNumber - 8 Bytes\n• Type == 4: String - Dynamic size, first byte is length of characters.\n• List of constants: - Dynamic size\n• Function prototypes\n• Number of protos - Integer\n• Functions chunks - Dynamic, it's big lol\n• Source lines\n• Number of lines - Integer\n• Lines - Integer, Dynamic size\n• Local List\n• Number of locals - Integer\n• Each local is stored as:\n• name - String\n• start line - Int\n• end line - Int\n• Upvalue list\n• Number of Upvalues - Integer\n• List of strings - strings, the names of the Upvalue\n\nHere's the code equivalent:\n\n``` def decode_chunk(self):\nchunk = {\n'INSTRUCTIONS': {},\n'CONSTANTS': {},\n'PROTOTYPES': {}\n}\n\nchunk['NAME'] = self.get_string(None)\nchunk['FIRST_LINE'] = self.get_int()\nchunk['LAST_LINE'] = self.get_int()\n\nchunk['UPVALUES'] = self.get_byte()\nchunk['ARGUMENTS'] = self.get_byte()\nchunk['VARG'] = self.get_byte()\nchunk['STACK'] = self.get_byte()\n\nif (not chunk['NAME'] == None):\nchunk['NAME'] = chunk['NAME'][1:-1]\n\n# parse instructions\nprint(\" DECODING INSTRUCTIONS\")\n\nnum = self.get_int()\nfor i in range(num):\ninstruction = {\n# opcode = opcode number;\n# type = [ABC, ABx, AsBx]\n# A, B, C, Bx, or sBx depending on type\n}\n\ndata = self.get_int32()\nopcode = get_bits(data, 1, 6)\ntp = lua_opcode_types[opcode]\n\ninstruction['OPCODE'] = opcode\ninstruction['TYPE'] = tp\ninstruction['A'] = get_bits(data, 7, 14)\n\nif instruction['TYPE'] == \"ABC\":\ninstruction['B'] = get_bits(data, 24, 32)\ninstruction['C'] = get_bits(data, 15, 23)\nelif instruction['TYPE'] == \"ABx\":\ninstruction['Bx'] = get_bits(data, 15, 32)\nelif instruction['TYPE'] == \"AsBx\":\ninstruction['sBx'] = get_bits(data, 15, 32) - 131071\n\nchunk['INSTRUCTIONS'][i] = instruction\n\nprint(lua_opcode_names[opcode], instruction)\n\n# get constants\nprint(\" DECODING CONSTANTS\")\n\nnum = self.get_int()\nfor i in range(num):\nconstant = {\n# type = constant type;\n# data = constant data;\n}\nconstant['TYPE'] = self.get_byte()\n\nif constant['TYPE'] == 1:\nconstant['DATA'] = (self.get_byte() != 0)\nelif constant['TYPE'] == 3:\nconstant['DATA'] = self.get_double()\nelif constant['TYPE'] == 4:\nconstant['DATA'] = self.get_string(None)[:-1]\n\nprint(constant)\n\nchunk['CONSTANTS'][i] = constant\n\n# parse protos\n\nprint(\" DECODING PROTOS\")\n\nnum = self.get_int()\nfor i in range(num):\nchunk['PROTOTYPES'][i] = self.decode_chunk()\n\n# debug stuff\nprint(\" DECODING DEBUG SYMBOLS\")\n\n# line numbers\nnum = self.get_int()\nfor i in range(num):\nself.get_int32()\n\n# locals\nnum = self.get_int()\nfor i in range(num):\nprint(self.get_string(None)[:-1]) # local name\nself.get_int32() # local start PC\nself.get_int32() # local end PC\n\n# upvalues\nnum = self.get_int()\nfor i in range(num):\nself.get_string(None) # upvalue name\n\nself.chunks.append(chunk)\n\nreturn chunk\n```\n\nSo, using this, let's go back to where we started. Let's try and parse our epic compile lua bytecode from the beginning.\n\n```with open('epic.luac', 'rb') as luac_file:\nbytecode = array.array('b', rawbytecode)\nself.decode_bytecode(bytecode)\nself.decode_chunk()\n\n```\n\nYour output should look something like:", null, "Congrats!!!! You just wasted a month of your life on a project!!! oh wait, I must be self-projecting again. Anyways, congrats! You have successfully parsed Lua 5.1 Bytecode! Thank you so much for reading this! More projects soon I promise!\n\nSep 15, 2019 by CPunch" ]	[ null, "https://openpunk.com/uploads/ZC0uzhDsm", null, "data:image/png;base64,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", null, "data:image/png;base64,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", null, "data:image/png;base64,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", null, "data:image/png;base64,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", null, "data:image/png;base64,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", null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.645591,"math_prob":0.9071366,"size":9755,"snap":"2020-34-2020-40","text_gpt3_token_len":2475,"char_repetition_ratio":0.19649267,"word_repetition_ratio":0.16630197,"special_character_ratio":0.28221425,"punctuation_ratio":0.20591286,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.9760682,"pos_list":[0,1,2,3,4,5,6,7,8,9,10,11,12],"im_url_duplicate_count":[null,2,null,null,null,null,null,null,null,null,null,null,null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2020-08-08T23:52:03Z\",\"WARC-Record-ID\":\"<urn:uuid:43b261af-3f33-437c-9f43-d841582593ae>\",\"Content-Length\":\"478579\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:e45d9916-a4e5-488a-a12e-b1333938638f>\",\"WARC-Concurrent-To\":\"<urn:uuid:9507983b-6d92-4178-a68d-a68deee4d10c>\",\"WARC-IP-Address\":\"104.28.26.35\",\"WARC-Target-URI\":\"https://openpunk.com/post/7\",\"WARC-Payload-Digest\":\"sha1:WD77LJZV6GZOOKOYSGIPWC37FAFWWGJQ\",\"WARC-Block-Digest\":\"sha1:W4A4MUDV56GY4SCFYV72K37FZ2OQPGKL\",\"WARC-Identified-Payload-Type\":\"text/html\",\"warc_filename\":\"/cc_download/warc_2020/CC-MAIN-2020-34/CC-MAIN-2020-34_segments_1596439738366.27_warc_CC-MAIN-20200808224308-20200809014308-00256.warc.gz\"}"}
https://inf.mit.bme.hu/en/cav14	[ "# CAV 2014 Measurements\n\nOn this page we provide all of our measurements comparing some state-of-the-art model checkers to our algorithm. The tools and algorithms included in the measurements are the following:\n\n• NuSMV2: traditional BDD-based fixed point computations,\n• NuXMV (NuSMV3): the newest generation of SAT-based model checking algorithms (so-called IC3), various abstraction and reduction techniques,\n• ITS-LTL: the newest variant of saturation based model checking algorithms based on different automaton representations and optimizations,\n• Hybrid MC: the prototype of our algorithm presented in the submitted paper currently implemented in C#.\n\nThe models are traditional asynchronous benchmark models [1,2]. In addition we used an industrial case-study of a safety logic in a nuclear power plant from :\n\n• Counter models a simple binary N-bit counter.\n• The Round Robin model (RR-N) is a resource sharing protocol and Slotted Ring (SR-N) is the model of a communication protocol with N participants.\n• DPhil is the well-known dining philosophers model.\n• The Flexible Manufacturing System (FMS-N) and Kanban models a production system. The parameter N refers to the complexity of the model checking problem.\n• PRISE example is a small, but important safety-critical industrial system: the safety function initiating an emergency procedure in a nuclear power plant. As verification requirement we used a complex fairness property stating that the initialization can be infinitely often activated under dangerous situations.\n\nThe following table shows the different models and expressions we evaluated in each tool, with the runtimes of the algorithms. To reproduce our results, one may download the following archives, in which we give our models, expressions and scripts, in some cases together with the tool itself. To run measurements in PetriDotNet, please refer to the README file in the Application folder. The other tools are run using the \"run-timeout.bat\" or \"run_timeout.sh\" script files. Be aware that NuSMV and NuXMV may require installation and modification of environment variables.\n\nWARNING: The following packages contain executables in order to provide a way to exactly reproduce results, use them only for this purpose. If you need any of the tools for other use, the newest versions are available from their official websites with the proper licensing (all of them are free for personal and academic use).\n\nAll measurements were done on a laptop with Intel® Core™ i3-350M CPU (3M Cache, 2.26 GHz), 8 GB 1333 MHz DDR3 RAM, and a Samsung 840 EVO SSD. ITS and Hybrid MC were run on Microsoft Windows 8, while runtimes of NuSMV and NuXMV were measured on Ubuntu 13.10.\n\nMemory consumption is not measured, as Hybrid MC is run in a managed environment (.Net), and the garbage collection mechanism could make the measurements not reproducible (memory consumption highly depends on the schedule of the GC). Experiments showed that Hybrid MC uses only a small amount of memory (despite it is run in a managed environment), and ITS-LTL is also efficient in terms of memory usage. NuSMV and NuXMV both has problems with memory, since they need to transform the input model to a low-level formalism, while Hybrid-MC and ITS both work on high-level models. Comparison to these tools is justified by the fact that they are well-known tools with stable algorithms, and highly poopular on the market.\n\nDue to the limited precision of measurements, runtimes less than 0.01 are just signed \"<0.01\". Timeout was set to 1200s, cases where algorithms did not finish within this time limit are denoted by \">1200.00\". When an algorithm ran out of memory, the table shows \"memory\" instead of the runtime. The sign \"---\" marks cases where the conversion of benchmark models to the tool's format was not done due to technical reasons (such as format incompatibilities, the size of the resulting model, or simply the fact that model checking of smaller models with the corresponding tool already failed to complete).\n\nSCC detection statistics show some information about the performance of our hybrid approach. FP denotes the number of incremental symbolic fixed point computations actually started. R shows how many such computations were prevented by observing recurrent states, while A shows the same for explicit runs on abstractions.\n\nModel Expression Result Runtime (s) SCC detection statistics4\nNuSMV21 NuXMV1 ITS-LTL2 Hybrid MC3 FP R A\nCounter-10 !G(bit9=0) true 0.01\n(0.02)\n0.01 (0.02) --- <0.01\n(<0.01)\n0 2 23\nCounter-15 !G(bit14=0) true 0.21\n(0.23)\n0.02\n(0.05)\n--- <0.01\n(<0.01)\n0 2 38\nCounter-20 !G(bit19=0) true 15.30\n(15.36)\n0.02\n(0.07)\n--- <0.01\n(0.01)\n0 2 53\nCounter-50 !G(bit49=0) true >1200.00 0.04\n(0.70)\n--- 0.01\n(0.02)\n0 2 143\nDPhil-50 F(HasLeft1=1&HasRight1=1) false >1200.00 2.15\n(108.58)\n0.20 0.01\n(0.03)\n1 2 5\nF(HasLeft50=1&HasRight50=1) false >1200.00 2.31\n(108.24)\n0.20 <0.01\n(0.02)\n1 1 3\nGF(HasRight1=1) -> GF(HasLeft1=1&HasRight1=1) false >1200.00 176.72\n(281.71)\n0.20 0.01\n(0.03)\n8 14 4\nGF(HasRight50=1) -> GF(HasLeft50=1&HasRight50=1) false >1200.00 178.87\n(284.28)\n0.22 0.13\n(0.15)\n7 12 6\nDPhil-100 F(HasLeft1=1&HasRight1=1) false >1200.00 >1200.00 0.64 0.01\n(0.08)\n1 2 5\nF(HasLeft100=1&HasRight100=1) false >1200.00 >1200.00 0.63 0.01\n(0.07)\n1 1 3\nGF(HasRight1=1) -> GF(HasLeft1=1&HasRight1=1) false >1200.00 >1200.00 0.64 0.01\n(0.08)\n8 14 4\nGF(HasRight100=1) -> GF(HasLeft100=1&HasRight100=1) false >1200.00 >1200.00 0.64 0.51\n(0.58)\n7 12 6\nDPhil-200 F(HasLeft1=1&HasRight1=1) false memory memory 2.30 0.02\n(0.31)\n1 2 5\nF(HasLeft200=1&HasRight200=1) false memory memory 2.16 0.01\n(0.29)\n1 1 3\nGF(HasRight1=1) -> GF(HasLeft1=1&HasRight1=1) false memory memory 2.45 0.04\n(0.32)\n8 14 4\nGF(HasRight200=1) -> GF(HasLeft200=1&HasRight200=1) false memory memory 2.19 2.18\n(2.44)\n7 12 6\nDPhil-300 F(HasLeft1=1&HasRight1=1) false --- --- 5.19 0.04\n(0.67)\n1 2 5\nF(HasLeft300=1&HasRight300=1) false --- --- 5.03 0.02\n(0.64)\n1 1 3\nGF(HasRight1=1) -> GF(HasLeft1=1&HasRight1=1) false --- --- 5.14 0.06\n(0.69)\n8 14 4\nGF(HasRight300=1) -> GF(HasLeft300=1&HasRight300=1) false --- --- 5.13 5.25\n(5.89)\n7 12 6\nFMS-5 F(P1s=P2s=P3s=5) false >1200.00 0.03\n(0.64)\n0.13 <0.01\n(0.01)\n1 28 17\n!G(P1wP2>0&P2=P3=5) true >1200.00 0.02\n(0.63)\n0.02 <0.01\n(<0.01)\n0 0 0\nFMS-10 F(P1s=P2s=P3s=10) false >1200.00 0.02\n(1.38)\n0.53 0.01\n(0.01)\n1 53 37\n!G(P1wP2>0&P2=P3=10) true >1200.00 0.02\n(1.36)\n0.03 <0.01\n(<0.01)\n0 0 0\nFMS-20 F(P1s=P2s=P3s=20) false >1200.00 0.03\n(1.54)\n3.70 0.02\n(0.02)\n1 103 77\n!G(P1wP2>0&P2=P3=20) true >1200.00 0.03\n(1.56)\n0.03 <0.01\n(<0.01)\n0 0 0\nKaban-10 !G(Pout1>0\|Pout2>0\|Pout3>0\|Pout4>0) true 0.02\n(0.47)\n0.02\n(0.50)\n0.02 <0.01\n(<0.01)\n0 0 0\n!G(Pback2=9\|Pback3=9) true 0.07\n(0.52)\n0.02\n(0.49)\n0.02 <0.01\n(<0.01)\n0 0 0\nKanban-20 !G(Pout1>0\|Pout2>0\|Pout3>0\|Pout4>0) true 0.71\n(1.26)\n0.02\n(0.60)\n0.02 <0.01\n(<0.01)\n0 0 0\n!G(Pback2=19\|Pback3=19) true 9.51\n(10.07)\n0.02\n(0.62)\n0.03 <0.01\n(<0.01)\n0 0 0\nKanban-30 !G(Pout1>0\|Pout2>0\|Pout3>0\|Pout4>0) true 0.76\n(1.34)\n0.03\n(0.64)\n0.03 <0.01\n(<0.01)\n0 0 0\n!G(Pback2=29\|Pback3=29) true 13.64\n(14.22)\n0.03\n(0.65)\n0.02 <0.01\n(<0.01)\n0 0 0\nRR-10 !G(True) false >1200.00 0.08\n(1.68)\n0.08 0.03\n(0.03)\n1 110 186\nRR-50 !G(True) false >1200.00 4.47\n(813.12)\n0.64 1.28\n(1.30)\n1 2550 2986\nRR-100 !G(True) false memory memory 2.13 8.86\n(8.96)\n1 10100 10986\nSR-5 !G(C1=0) true 32.03\n(32.36)\n0.02\n(0.38)\n0.03 <0.01\n(<0.01)\n0 0 0\n!G(C5 = 0) true 14.45\n(14.77)\n0.04\n(0.40)\n0.03 <0.01\n(<0.01)\n0 0 0\n!G(H1=0\|G1=0); false 39.43\n(39.77)\n0.03\n(0.40)\n0.06 0.02\n(0.02)\n1 4 95\nGF(H1=1) -> GF(H1=1&G1=1) false >1200.00 >1200.00 0.12 0.06\n(0.06)\n1 66 333\nSR-10 !G(C1=0) true >1200.00 0.10\n(2.01)\n0.03 <0.01\n(<0.01)\n0 0 0\n!G(C10 = 0) true >1200.00 0.10\n(1.98)\n0.03 <0.01\n(<0.01)\n0 0 0\n!G(H1=0\|G1=0); false >1200.00 0.11\n(1.99)\n0.22 0.07\n(0.07)\n1 7 433\nGF(H1=1) -> GF(H1=1&G1=1) false >1200.00 >1200.00 1.08 0.37\n(0.37)\n1 111 3444\nSR-50 !G(C1=0) true >1200.00 6.11\n(595.26)\n0.19 <0.01\n(0.03)\n0 0 0\n!G(C50 = 0) true >1200.00 5.87\n(595.61)\n0.19 <0.01\n(0.04)\n0 0 0\n!G(H1=0\|G1=0); false >1200.00 6.10\n(597.61)\n18.33 2.68\n(2.71)\n1 27 19423\nGF(H1=1) -> GF(H1=1&G1=1) false >1200.00 >1200.00 807.58 32.97\n(33.00)\n1 351 368021\nSR-100 !G(C1=0) true memory memory 0.58 <0.01\n(0.14)\n0 0 0\n!G(C100 = 0) true memory memory 0.58 <0.01\n(0.14)\n0 0 0\n!G(H1=0\|G1=0); false memory memory 168.73 16.25\n(16.39)\n1 52 119473\nGF(H1=1) -> GF(H1=1&G1=1) false memory memory >1200.00 251.91\n(252.06)\n1 651 2812346\nPRISE-S G(P15OUTPUT_2!=true U (P15OUTPUT_2!=true& P18outPUT_1=true)) false >1200.00 >1200.00 --- 4.98\n(5.13)\n1 69435 13180\n!GF(P15OUTPUT_2=true\|P15OUTPUT_2=false) false >1200.00 >1200.00 --- 4.19\n(4.21)\n1 64925 0\n!GF(P15OUTPUT_2=true) -> GF(P18outPUT_1=true) false >1200.00 >1200.00 --- 5.05\n(5.08)\n1 74436 13283\nGF(P15OUTPUT_2=true) -> GF(P18outPUT_1=true) false >1200.00 >1200.00 --- 7.55\n(7.58)\n1 152486 18\nPRISE-M G(P15OUTPUT_2!=true U (P15OUTPUT_2!=true& P18outPUT_1=true)) false --- --- --- 12.51\n(12.73)\n1 179680 32968\n!GF(P15OUTPUT_2=true\|P15OUTPUT_2=false) false --- --- --- 10.64\n(10.67)\n1 163437 0\n!GF(P15OUTPUT_2=true) -> GF(P18outPUT_1=true) false --- --- --- 12.64\n(12.67)\n1 192437 33347\nGF(P15OUTPUT_2=true) -> GF(P18outPUT_1=true) false --- --- --- 20.12\n(20.14)\n1 389210 36\n\n1 Runtime of the algorithm, total runtime with loading and conversion time shown in parentheses.\n\n2 Runtime results provided by the tool itself.\n\n3 Runtime of state space exploration and model checking, total runtime with initialization and the building of the property automaton is shown in parentheses.\n\n4 Order of applying these techniques is the following: (non-empty state and transition set) → (presence of accepting states) → presence of recurrent states → explicit search in the abstraction → incremental symbolic fixed-point computation. This way, the shown metrics mark cases where the technique was actually applied and broke the chain.\n\n Model Checking Contest: http://mcc.lip6.fr/models.php" ]	[ null ]	{"ft_lang_label":"__label__en","ft_lang_prob":0.65663946,"math_prob":0.98001164,"size":11252,"snap":"2020-45-2020-50","text_gpt3_token_len":4295,"char_repetition_ratio":0.1857219,"word_repetition_ratio":0.06409459,"special_character_ratio":0.4749378,"punctuation_ratio":0.17558722,"nsfw_num_words":0,"has_unicode_error":false,"math_prob_llama3":0.97208166,"pos_list":[0],"im_url_duplicate_count":[null],"WARC_HEADER":"{\"WARC-Type\":\"response\",\"WARC-Date\":\"2020-10-24T17:26:56Z\",\"WARC-Record-ID\":\"<urn:uuid:a765a2b8-5884-4e98-a695-e56b29fba68e>\",\"Content-Length\":\"52773\",\"Content-Type\":\"application/http; msgtype=response\",\"WARC-Warcinfo-ID\":\"<urn:uuid:7feccc69-e7a1-4f76-9f2c-f06623d83505>\",\"WARC-Concurrent-To\":\"<urn:uuid:08837676-f9f3-48d8-9f53-690e962054ae>\",\"WARC-IP-Address\":\"152.66.252.223\",\"WARC-Target-URI\":\"https://inf.mit.bme.hu/en/cav14\",\"WARC-Payload-Digest\":\"sha1:BQEUGNPADT74QYTWIP3DPPSNOFWVGBI2\",\"WARC-Block-Digest\":\"sha1:NJNR6DNHSLIJHWZEG7XUORRCLFL7T5TQ\",\"WARC-Identified-Payload-Type\":\"application/xhtml+xml\",\"warc_filename\":\"/cc_download/warc_2020/CC-MAIN-2020-45/CC-MAIN-2020-45_segments_1603107884322.44_warc_CC-MAIN-20201024164841-20201024194841-00400.warc.gz\"}"}