maddes8cht commited on
Commit
6ab6014
·
1 Parent(s): bd4cf9e

Add LICENSE, README, .gitignore, and dataset_card.yaml for preliminary financial dataset

Browse files
Files changed (4) hide show
  1. .gitignore +11 -0
  2. LICENSE +11 -0
  3. README.md +70 -0
  4. dataset_card.yaml +92 -0
.gitignore ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Exclude temporary and working files
2
+
3
+ \*.fetched.parquet \*.test.parquet \*.ufetch.parquet \*.utest.parquet *.fail*.parquet *.ufail*.parquet
4
+
5
+ # Include only valid parquet files
6
+
7
+ !\*.valid.parquet
8
+
9
+ # Other common ignores
10
+
11
+ \*.log \*.tmp \*.bak .DS_Store **pycache**/ \*.pyc .env logs/
LICENSE CHANGED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Custom License - Non-Commercial Research Use Only
2
+
3
+ This dataset contains historical financial time series data collected via the yfinance Python library, which retrieves public market data from Yahoo Finance.
4
+
5
+ These data remain the property of their respective original sources and are provided here strictly for non-commercial research and educational purposes.
6
+
7
+ Redistribution of this dataset may be limited by the terms of the original data providers.
8
+
9
+ NO WARRANTY IS PROVIDED. USE AT YOUR OWN RISK.
10
+
11
+ For any intended use beyond private research, please consult the terms of Yahoo Finance and other data providers.
README.md CHANGED
@@ -3,3 +3,73 @@ license: other
3
  license_name: license
4
  license_link: LICENSE
5
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  license_name: license
4
  license_link: LICENSE
5
  ---
6
+ # Preliminary Financial Time Series Dataset (Version 2)
7
+
8
+ ⚠️ Warning: This is a second preliminary dataset for development, testing, and feedback purposes. A stable, production-ready dataset will be released later. This time, all data in here will be available in future "stable" datasets (I guesss.. )
9
+
10
+ ## Overview
11
+
12
+ This dataset contains **parquet files** with time series data for various financial instruments, collected using an improved and more stable version of our data pipeline. It builds upon the first preliminary dataset, with a refined structure and more reliable data fetching processes. The data is sourced from publicly available sources like [Yahoo Finance](https://finance.yahoo.com) via the `yfinance` library.
13
+
14
+ Each financial instrument includes:
15
+ - **Daily candles**: Covering multiple years of historical data.
16
+ - **Hourly candles**: Covering at least 2 years.
17
+ - **Minute candles**: Covering a shorter, recent period with high granularity.
18
+
19
+ This multi-resolution format supports models that analyze both long-term trends and short-term patterns.
20
+
21
+ ## Purpose
22
+
23
+ This second preliminary dataset is designed to:
24
+ - Provide early access to reliable financial time series data.
25
+ - Enable testing and iteration of machine learning models for trading.
26
+ - Gather feedback to finalize a stable dataset format.
27
+ - Test the consistency of the data update pipeline over a few days.
28
+
29
+ ## Data Structure
30
+
31
+ The dataset is organized into a clear directory structure:
32
+ - **/data/{category}/{symbol}/{symbol}.days.valid.parquet**: Daily OHLC data.
33
+ - **/data/{category}/{symbol}/{symbol}.hours.valid.parquet**: Hourly OHLC data.
34
+ - **/data/{category}/{symbol}/{symbol}.minutes.valid.parquet**: Minute OHLC data.
35
+
36
+ Only files marked as `.valid.parquet` are included in this dataset to ensure data quality and consistency. Temporary files (e.g., `fetch`, `test`, `ufetch`, `utest`, `failXX`, `ufailXX`) are excluded via `.gitignore`.
37
+
38
+ ## Expected Changes
39
+
40
+ While the pipeline is more stable, this dataset remains preliminary. Potential changes include:
41
+ - Adjustments to file naming conventions.
42
+ - Reorganization into sharded folders (e.g., by year or month).
43
+ - Refinements to dataset split logic.
44
+
45
+ A stable, production-ready dataset will be released separately to ensure long-term consistency.
46
+
47
+ ## Goals
48
+ The Traders-Lab datasets aim to grow in two dimensions:
49
+ - **More stocks**: Additional symbols will be added over time, with rapid expansion expected soon.
50
+ - **More data**: Short-term datasets (hourly and minute candles) will grow as more data is accumulated. While continuity of current minute data is not guaranteed yet, future updates will ensure a continuous time history.
51
+
52
+ ## Non-Goals
53
+ The dataset is designed to be sufficiently up-to-date for training purposes, with data typically no more than a few days old. Real-time updates are not a goal.
54
+
55
+ ## License & Usage
56
+
57
+ This dataset is not licensed under a standard open data license. See the [`LICENSE`](./LICENSE) file for detailed usage permissions. It is intended **solely for research and educational purposes**. Redistribution may be restricted; please respect the terms of the original data providers, such as Yahoo Finance.
58
+
59
+ ## Accessing the Dataset
60
+
61
+ The dataset is hosted on Hugging Face under the [Traders-Lab organization](https://huggingface.co/Traders-Lab). To clone the dataset:
62
+
63
+ ```bash
64
+ # Ensure git-lfs is installed (https://git-lfs.com)
65
+ git lfs install
66
+ git clone https://huggingface.co/datasets/Traders-Lab/preliminary-v2
67
+ ```
68
+
69
+ ## Metadata
70
+
71
+ Dataset metadata is provided in the [`dataset_card.yml`](./dataset_card.yml) file, following Hugging Face's dataset card standards.
72
+
73
+ ## Feedback
74
+
75
+ We welcome feedback to improve the dataset! Please share your thoughts via the [Hugging Face Discussions](https://huggingface.co/datasets/Traders-Lab/preliminary-v2/discussions) or contact the Traders-Lab team.
dataset_card.yaml ADDED
@@ -0,0 +1,92 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ license: other
5
+ pretty_name: Preliminary Financial Time Series Dataset (Version 2)
6
+ task_categories:
7
+ - time-series-forecasting
8
+ tags:
9
+ - finance
10
+ - stock-market
11
+ - OHLC
12
+ - time-series
13
+ - trading
14
+ dataset_info:
15
+ features:
16
+ - name: symbol
17
+ dtype: string
18
+ - name: time
19
+ dtype: int64
20
+ - name: open
21
+ dtype: float64
22
+ - name: high
23
+ dtype: float64
24
+ - name: low
25
+ dtype: float64
26
+ - name: close
27
+ dtype: float64
28
+ - name: volume
29
+ dtype: int64
30
+ splits:
31
+ - name: daily
32
+ num_examples: null
33
+ - name: hourly
34
+ num_examples: null
35
+ - name: minute
36
+ num_examples: null
37
+ size_categories:
38
+ - n<1K
39
+ - 1K<n<10K
40
+ - 10K<n<100K
41
+ ---
42
+
43
+ # Preliminary Financial Time Series Dataset (Version 2)
44
+
45
+ ## Dataset Description
46
+
47
+ - **Repository**: [Traders-Lab/preliminary-v2](https://huggingface.co/datasets/Traders-Lab/preliminary-v2)
48
+ - **Paper**: None
49
+ - **Point of Contact**: [Traders-Lab Discussions](https://huggingface.co/datasets/Traders-Lab/preliminary-v2/discussions)
50
+ - **Size of Dataset**: Varies by symbol and interval (daily, hourly, minute)
51
+ - **Size of Downloaded Dataset**: Multiple parquet files, size depends on the number of symbols
52
+ - **Framework**: Data collected using `yfinance` from [Yahoo Finance](https://finance.yahoo.com)
53
+
54
+ This is a second preliminary dataset containing OHLC (Open, High, Low, Close) time series data for various financial instruments, intended for research and educational purposes. It includes daily, hourly, and minute-resolution data, organized by symbol and category.
55
+
56
+ ## Dataset Structure
57
+
58
+ The dataset is organized as follows:
59
+ - **/data/{category}/{symbol}/{symbol}.days.valid.parquet**: Daily OHLC data.
60
+ - **/data/{category}/{symbol}/{symbol}.hours.valid.parquet**: Hourly OHLC data.
61
+ - **/data/{category}/{symbol}/{symbol}.minutes.valid.parquet**: Minute OHLC data.
62
+
63
+ Only `.valid.parquet` files are included to ensure data quality. Temporary files are excluded.
64
+
65
+ ## Intended Uses
66
+
67
+ - **Primary Use**: Training and evaluating machine learning models for trading strategies.
68
+ - **Other Uses**: Time series analysis, financial research, and educational projects.
69
+
70
+ ## Limitations
71
+
72
+ - **Preliminary Nature**: The dataset is not yet stable and may undergo changes in structure or naming.
73
+ - **Data Freshness**: Data is typically a few days old, not real-time.
74
+ - **Coverage**: Not all symbols may have complete historical data, especially for minute-resolution data.
75
+
76
+ ## License
77
+
78
+ The dataset is not licensed under a standard open data license. See the [`LICENSE`](./LICENSE) file for details. Usage is restricted to research and educational purposes, and redistribution may be limited by the terms of the original data provider (Yahoo Finance).
79
+
80
+ ## How to Use
81
+
82
+ Clone the dataset using:
83
+ ```bash
84
+ git lfs install
85
+ git clone https://huggingface.co/datasets/Traders-Lab/preliminary-v2
86
+ ```
87
+
88
+ Load the parquet files using Python libraries like `pandas` or `duckdb`.
89
+
90
+ ## Feedback
91
+
92
+ Please provide feedback via the [Hugging Face Discussions](https://huggingface.co/datasets/Traders-Lab/preliminary-v2/discussions) to help improve the dataset.