{ "cells": [ { "cell_type": "code", "execution_count": 4, "id": "84aba5f9-77d5-4638-9121-f85a21c0f6c4", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Loading data from experiment_results/exp_20250319_141119_all_results.json...\n", "Loaded 16 valid results\n", "Analysis results will be saved to analysis_output_exp_20250319_141119\n" ] }, { "ename": "TypeError", "evalue": "Object of type int64 is not JSON serializable", "output_type": "error", "traceback": [ "\u001b[31m---------------------------------------------------------------------------\u001b[39m", "\u001b[31mTypeError\u001b[39m Traceback (most recent call last)", "\u001b[36mCell\u001b[39m\u001b[36m \u001b[39m\u001b[32mIn[4]\u001b[39m\u001b[32m, line 468\u001b[39m\n\u001b[32m 464\u001b[39m \u001b[38;5;28mprint\u001b[39m(\u001b[33mf\u001b[39m\u001b[33m\"\u001b[39m\u001b[33mAnalysis complete. Results saved to \u001b[39m\u001b[38;5;132;01m{\u001b[39;00moutput_dir\u001b[38;5;132;01m}\u001b[39;00m\u001b[33m\"\u001b[39m)\n\u001b[32m 465\u001b[39m \u001b[38;5;28;01mreturn\u001b[39;00m output_dir\n\u001b[32m--> \u001b[39m\u001b[32m468\u001b[39m \u001b[43mperform_full_analysis\u001b[49m\u001b[43m(\u001b[49m\u001b[33;43m\"\u001b[39;49m\u001b[33;43mexperiment_results/exp_20250319_141119_all_results.json\u001b[39;49m\u001b[33;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n", "\u001b[36mCell\u001b[39m\u001b[36m \u001b[39m\u001b[32mIn[4]\u001b[39m\u001b[32m, line 426\u001b[39m, in \u001b[36mperform_full_analysis\u001b[39m\u001b[34m(input_file)\u001b[39m\n\u001b[32m 408\u001b[39m basic_stats = {\n\u001b[32m 409\u001b[39m \u001b[33m'\u001b[39m\u001b[33mtotal_simulations\u001b[39m\u001b[33m'\u001b[39m: \u001b[38;5;28mlen\u001b[39m(df),\n\u001b[32m 410\u001b[39m \u001b[33m'\u001b[39m\u001b[33mparameter_ranges\u001b[39m\u001b[33m'\u001b[39m: {\n\u001b[32m (...)\u001b[39m\u001b[32m 422\u001b[39m }\n\u001b[32m 423\u001b[39m }\n\u001b[32m 425\u001b[39m \u001b[38;5;28;01mwith\u001b[39;00m \u001b[38;5;28mopen\u001b[39m(os.path.join(output_dir, \u001b[33m'\u001b[39m\u001b[33mbasic_stats.json\u001b[39m\u001b[33m'\u001b[39m), \u001b[33m'\u001b[39m\u001b[33mw\u001b[39m\u001b[33m'\u001b[39m) \u001b[38;5;28;01mas\u001b[39;00m f:\n\u001b[32m--> \u001b[39m\u001b[32m426\u001b[39m \u001b[43mjson\u001b[49m\u001b[43m.\u001b[49m\u001b[43mdump\u001b[49m\u001b[43m(\u001b[49m\u001b[43mbasic_stats\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mf\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mindent\u001b[49m\u001b[43m=\u001b[49m\u001b[32;43m2\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[32m 428\u001b[39m \u001b[38;5;66;03m# Create visualizations\u001b[39;00m\n\u001b[32m 429\u001b[39m \u001b[38;5;28mprint\u001b[39m(\u001b[33m\"\u001b[39m\u001b[33mGenerating visualizations...\u001b[39m\u001b[33m\"\u001b[39m)\n", "\u001b[36mFile \u001b[39m\u001b[32m/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/json/__init__.py:179\u001b[39m, in \u001b[36mdump\u001b[39m\u001b[34m(obj, fp, skipkeys, ensure_ascii, check_circular, allow_nan, cls, indent, separators, default, sort_keys, **kw)\u001b[39m\n\u001b[32m 173\u001b[39m iterable = \u001b[38;5;28mcls\u001b[39m(skipkeys=skipkeys, ensure_ascii=ensure_ascii,\n\u001b[32m 174\u001b[39m check_circular=check_circular, allow_nan=allow_nan, indent=indent,\n\u001b[32m 175\u001b[39m separators=separators,\n\u001b[32m 176\u001b[39m default=default, sort_keys=sort_keys, **kw).iterencode(obj)\n\u001b[32m 177\u001b[39m \u001b[38;5;66;03m# could accelerate with writelines in some versions of Python, at\u001b[39;00m\n\u001b[32m 178\u001b[39m \u001b[38;5;66;03m# a debuggability cost\u001b[39;00m\n\u001b[32m--> \u001b[39m\u001b[32m179\u001b[39m \u001b[43m\u001b[49m\u001b[38;5;28;43;01mfor\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mchunk\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;129;43;01min\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43miterable\u001b[49m\u001b[43m:\u001b[49m\n\u001b[32m 180\u001b[39m \u001b[43m \u001b[49m\u001b[43mfp\u001b[49m\u001b[43m.\u001b[49m\u001b[43mwrite\u001b[49m\u001b[43m(\u001b[49m\u001b[43mchunk\u001b[49m\u001b[43m)\u001b[49m\n", "\u001b[36mFile \u001b[39m\u001b[32m/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/json/encoder.py:432\u001b[39m, in \u001b[36m_make_iterencode.._iterencode\u001b[39m\u001b[34m(o, _current_indent_level)\u001b[39m\n\u001b[32m 430\u001b[39m \u001b[38;5;28;01myield from\u001b[39;00m _iterencode_list(o, _current_indent_level)\n\u001b[32m 431\u001b[39m \u001b[38;5;28;01melif\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(o, \u001b[38;5;28mdict\u001b[39m):\n\u001b[32m--> \u001b[39m\u001b[32m432\u001b[39m \u001b[38;5;28;01myield from\u001b[39;00m _iterencode_dict(o, _current_indent_level)\n\u001b[32m 433\u001b[39m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[32m 434\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m markers \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n", "\u001b[36mFile \u001b[39m\u001b[32m/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/json/encoder.py:406\u001b[39m, in \u001b[36m_make_iterencode.._iterencode_dict\u001b[39m\u001b[34m(dct, _current_indent_level)\u001b[39m\n\u001b[32m 404\u001b[39m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[32m 405\u001b[39m chunks = _iterencode(value, _current_indent_level)\n\u001b[32m--> \u001b[39m\u001b[32m406\u001b[39m \u001b[38;5;28;01myield from\u001b[39;00m chunks\n\u001b[32m 407\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m newline_indent \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[32m 408\u001b[39m _current_indent_level -= \u001b[32m1\u001b[39m\n", "\u001b[36mFile \u001b[39m\u001b[32m/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/json/encoder.py:406\u001b[39m, in \u001b[36m_make_iterencode.._iterencode_dict\u001b[39m\u001b[34m(dct, _current_indent_level)\u001b[39m\n\u001b[32m 404\u001b[39m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[32m 405\u001b[39m chunks = _iterencode(value, _current_indent_level)\n\u001b[32m--> \u001b[39m\u001b[32m406\u001b[39m \u001b[38;5;28;01myield from\u001b[39;00m chunks\n\u001b[32m 407\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m newline_indent \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[32m 408\u001b[39m _current_indent_level -= \u001b[32m1\u001b[39m\n", "\u001b[36mFile \u001b[39m\u001b[32m/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/json/encoder.py:326\u001b[39m, in \u001b[36m_make_iterencode.._iterencode_list\u001b[39m\u001b[34m(lst, _current_indent_level)\u001b[39m\n\u001b[32m 324\u001b[39m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[32m 325\u001b[39m chunks = _iterencode(value, _current_indent_level)\n\u001b[32m--> \u001b[39m\u001b[32m326\u001b[39m \u001b[38;5;28;01myield from\u001b[39;00m chunks\n\u001b[32m 327\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m newline_indent \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[32m 328\u001b[39m _current_indent_level -= \u001b[32m1\u001b[39m\n", "\u001b[36mFile \u001b[39m\u001b[32m/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/json/encoder.py:439\u001b[39m, in \u001b[36m_make_iterencode.._iterencode\u001b[39m\u001b[34m(o, _current_indent_level)\u001b[39m\n\u001b[32m 437\u001b[39m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\u001b[33m\"\u001b[39m\u001b[33mCircular reference detected\u001b[39m\u001b[33m\"\u001b[39m)\n\u001b[32m 438\u001b[39m markers[markerid] = o\n\u001b[32m--> \u001b[39m\u001b[32m439\u001b[39m o = \u001b[43m_default\u001b[49m\u001b[43m(\u001b[49m\u001b[43mo\u001b[49m\u001b[43m)\u001b[49m\n\u001b[32m 440\u001b[39m \u001b[38;5;28;01myield from\u001b[39;00m _iterencode(o, _current_indent_level)\n\u001b[32m 441\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m markers \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n", "\u001b[36mFile \u001b[39m\u001b[32m/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/json/encoder.py:180\u001b[39m, in \u001b[36mJSONEncoder.default\u001b[39m\u001b[34m(self, o)\u001b[39m\n\u001b[32m 161\u001b[39m \u001b[38;5;28;01mdef\u001b[39;00m\u001b[38;5;250m \u001b[39m\u001b[34mdefault\u001b[39m(\u001b[38;5;28mself\u001b[39m, o):\n\u001b[32m 162\u001b[39m \u001b[38;5;250m \u001b[39m\u001b[33;03m\"\"\"Implement this method in a subclass such that it returns\u001b[39;00m\n\u001b[32m 163\u001b[39m \u001b[33;03m a serializable object for ``o``, or calls the base implementation\u001b[39;00m\n\u001b[32m 164\u001b[39m \u001b[33;03m (to raise a ``TypeError``).\u001b[39;00m\n\u001b[32m (...)\u001b[39m\u001b[32m 178\u001b[39m \n\u001b[32m 179\u001b[39m \u001b[33;03m \"\"\"\u001b[39;00m\n\u001b[32m--> \u001b[39m\u001b[32m180\u001b[39m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mTypeError\u001b[39;00m(\u001b[33mf\u001b[39m\u001b[33m'\u001b[39m\u001b[33mObject of type \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mo.\u001b[34m__class__\u001b[39m.\u001b[34m__name__\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[33m \u001b[39m\u001b[33m'\u001b[39m\n\u001b[32m 181\u001b[39m \u001b[33mf\u001b[39m\u001b[33m'\u001b[39m\u001b[33mis not JSON serializable\u001b[39m\u001b[33m'\u001b[39m)\n", "\u001b[31mTypeError\u001b[39m: Object of type int64 is not JSON serializable" ] } ], "source": [ "import os\n", "import json\n", "import argparse\n", "import numpy as np\n", "import pandas as pd\n", "import matplotlib.pyplot as plt\n", "import seaborn as sns\n", "from scipy import stats\n", "from pathlib import Path\n", "\n", "def load_experiment_data(filename):\n", " \"\"\"\n", " Load the experiment results from a JSON file.\n", " \n", " Args:\n", " filename: Path to the results JSON file\n", " \n", " Returns:\n", " Pandas DataFrame with the experiment results\n", " \"\"\"\n", " print(f\"Loading data from {filename}...\")\n", " \n", " with open(filename, 'r') as f:\n", " data = json.load(f)\n", " \n", " # Check if this is the all_results file directly\n", " if isinstance(data, list) and len(data) > 0 and 'ibfs_config' in data[0]:\n", " results = data\n", " # Check if this is the experiment_id_analysis.json file\n", " elif 'summary' in data:\n", " # We need to find and load the corresponding all_results file\n", " experiment_id = data.get('experiment_id')\n", " base_dir = os.path.dirname(filename)\n", " all_results_file = os.path.join(base_dir, f\"{experiment_id}_all_results.json\")\n", " \n", " if os.path.exists(all_results_file):\n", " print(f\"Found associated all_results file: {all_results_file}\")\n", " with open(all_results_file, 'r') as f:\n", " results = json.load(f)\n", " else:\n", " # Need to reconstruct from individual simulation files\n", " print(f\"No all_results file found. Reconstructing from individual simulation files...\")\n", " results = []\n", " for file in os.listdir(base_dir):\n", " if file.startswith(experiment_id) and file.endswith(\".json\") and \"_sim_\" in file:\n", " with open(os.path.join(base_dir, file), 'r') as f:\n", " try:\n", " sim_data = json.load(f)\n", " results.append(sim_data)\n", " except json.JSONDecodeError:\n", " print(f\"Error loading {file} - skipping\")\n", " else:\n", " raise ValueError(\"Unrecognized file format. Please provide an experiment results file.\")\n", " \n", " # Filter out any results with errors\n", " valid_results = [r for r in results if 'error' not in r]\n", " \n", " # Convert to DataFrame for easier analysis\n", " rows = []\n", " for result in valid_results:\n", " # Extract the core configuration\n", " row = {\n", " 'query': result.get('query', ''),\n", " 'similarity_score': result.get('similarity_score', 0),\n", " 'diversity_level': result.get('ibfs_config', {}).get('diversity_level', ''),\n", " 'branching_factor': result.get('ibfs_config', {}).get('branching_factor', 0),\n", " 'max_depth': result.get('ibfs_config', {}).get('max_depth', 0),\n", " 'epsilon': result.get('user_config', {}).get('epsilon', 0),\n", " 'strategy_path': result.get('strategy_path', []),\n", " 'path_length': len(result.get('strategy_path', [])),\n", " 'experiment_id': result.get('experiment_id', '')\n", " }\n", " rows.append(row)\n", " \n", " df = pd.DataFrame(rows)\n", " print(f\"Loaded {len(df)} valid results\")\n", " return df\n", "\n", "def create_output_directory(experiment_id):\n", " \"\"\"Create an output directory for analysis artifacts.\"\"\"\n", " output_dir = f\"analysis_output_{experiment_id}\"\n", " os.makedirs(output_dir, exist_ok=True)\n", " return output_dir\n", "\n", "def plot_similarity_by_parameter(df, parameter, output_dir):\n", " \"\"\"\n", " Create a box plot showing similarity scores grouped by a parameter.\n", " \n", " Args:\n", " df: DataFrame with experiment results\n", " parameter: The parameter to group by (e.g., 'diversity_level')\n", " output_dir: Directory to save the plot\n", " \"\"\"\n", " plt.figure(figsize=(10, 6))\n", " sns.boxplot(x=parameter, y='similarity_score', data=df)\n", " plt.title(f'Similarity Scores by {parameter}')\n", " plt.ylabel('Similarity Score')\n", " plt.xlabel(parameter.replace('_', ' ').title())\n", " plt.grid(axis='y', linestyle='--', alpha=0.7)\n", " \n", " # Save the figure\n", " filename = os.path.join(output_dir, f'similarity_by_{parameter}.png')\n", " plt.savefig(filename, dpi=300, bbox_inches='tight')\n", " plt.close()\n", " print(f\"Saved plot to {filename}\")\n", "\n", "def plot_parameter_interaction(df, param1, param2, output_dir):\n", " \"\"\"\n", " Create a heatmap showing the interaction between two parameters.\n", " \n", " Args:\n", " df: DataFrame with experiment results\n", " param1: First parameter (e.g., 'branching_factor')\n", " param2: Second parameter (e.g., 'max_depth')\n", " output_dir: Directory to save the plot\n", " \"\"\"\n", " # Group by both parameters and calculate mean similarity score\n", " grouped = df.groupby([param1, param2])['similarity_score'].mean().reset_index()\n", " pivot_table = grouped.pivot(index=param1, columns=param2, values='similarity_score')\n", " \n", " plt.figure(figsize=(10, 8))\n", " sns.heatmap(pivot_table, annot=True, cmap='viridis', fmt='.3f', cbar_kws={'label': 'Avg Similarity Score'})\n", " plt.title(f'Interaction between {param1} and {param2}')\n", " plt.xlabel(param2.replace('_', ' ').title())\n", " plt.ylabel(param1.replace('_', ' ').title())\n", " \n", " # Save the figure\n", " filename = os.path.join(output_dir, f'interaction_{param1}_{param2}.png')\n", " plt.savefig(filename, dpi=300, bbox_inches='tight')\n", " plt.close()\n", " print(f\"Saved plot to {filename}\")\n", "\n", "def run_statistical_tests(df, output_dir):\n", " \"\"\"\n", " Run statistical tests to determine significant effects.\n", " \n", " Args:\n", " df: DataFrame with experiment results\n", " output_dir: Directory to save the results\n", " \"\"\"\n", " test_results = {}\n", " \n", " # Test for effect of diversity level\n", " if 'diversity_level' in df.columns and len(df['diversity_level'].unique()) > 1:\n", " groups = [df[df['diversity_level'] == level]['similarity_score'].values \n", " for level in df['diversity_level'].unique()]\n", " \n", " # ANOVA test if we have more than 2 groups\n", " if len(groups) > 2:\n", " f_stat, p_value = stats.f_oneway(*groups)\n", " test_results['diversity_level_anova'] = {\n", " 'test': 'ANOVA',\n", " 'F_statistic': float(f_stat),\n", " 'p_value': float(p_value),\n", " 'significant': p_value < 0.05\n", " }\n", " # T-test if we have exactly 2 groups\n", " elif len(groups) == 2:\n", " t_stat, p_value = stats.ttest_ind(groups[0], groups[1], equal_var=False)\n", " test_results['diversity_level_ttest'] = {\n", " 'test': 't-test',\n", " 'groups': list(df['diversity_level'].unique()),\n", " 't_statistic': float(t_stat),\n", " 'p_value': float(p_value),\n", " 'significant': p_value < 0.05\n", " }\n", " \n", " # Test for effect of branching factor\n", " if 'branching_factor' in df.columns and len(df['branching_factor'].unique()) > 1:\n", " groups = [df[df['branching_factor'] == b]['similarity_score'].values \n", " for b in sorted(df['branching_factor'].unique())]\n", " \n", " if len(groups) > 2:\n", " f_stat, p_value = stats.f_oneway(*groups)\n", " test_results['branching_factor_anova'] = {\n", " 'test': 'ANOVA',\n", " 'F_statistic': float(f_stat),\n", " 'p_value': float(p_value),\n", " 'significant': p_value < 0.05\n", " }\n", " elif len(groups) == 2:\n", " t_stat, p_value = stats.ttest_ind(groups[0], groups[1], equal_var=False)\n", " test_results['branching_factor_ttest'] = {\n", " 'test': 't-test',\n", " 'groups': list(sorted(df['branching_factor'].unique())),\n", " 't_statistic': float(t_stat),\n", " 'p_value': float(p_value),\n", " 'significant': p_value < 0.05\n", " }\n", " \n", " # Test for effect of max depth\n", " if 'max_depth' in df.columns and len(df['max_depth'].unique()) > 1:\n", " groups = [df[df['max_depth'] == d]['similarity_score'].values \n", " for d in sorted(df['max_depth'].unique())]\n", " \n", " if len(groups) > 2:\n", " f_stat, p_value = stats.f_oneway(*groups)\n", " test_results['max_depth_anova'] = {\n", " 'test': 'ANOVA',\n", " 'F_statistic': float(f_stat),\n", " 'p_value': float(p_value),\n", " 'significant': p_value < 0.05\n", " }\n", " elif len(groups) == 2:\n", " t_stat, p_value = stats.ttest_ind(groups[0], groups[1], equal_var=False)\n", " test_results['max_depth_ttest'] = {\n", " 'test': 't-test',\n", " 'groups': list(sorted(df['max_depth'].unique())),\n", " 't_statistic': float(t_stat),\n", " 'p_value': float(p_value),\n", " 'significant': p_value < 0.05\n", " }\n", " \n", " # Test for effect of epsilon\n", " if 'epsilon' in df.columns and len(df['epsilon'].unique()) > 1:\n", " groups = [df[df['epsilon'] == e]['similarity_score'].values \n", " for e in sorted(df['epsilon'].unique())]\n", " \n", " if len(groups) > 2:\n", " f_stat, p_value = stats.f_oneway(*groups)\n", " test_results['epsilon_anova'] = {\n", " 'test': 'ANOVA',\n", " 'F_statistic': float(f_stat),\n", " 'p_value': float(p_value),\n", " 'significant': p_value < 0.05\n", " }\n", " elif len(groups) == 2:\n", " t_stat, p_value = stats.ttest_ind(groups[0], groups[1], equal_var=False)\n", " test_results['epsilon_ttest'] = {\n", " 'test': 't-test',\n", " 'groups': list(sorted(df['epsilon'].unique())),\n", " 't_statistic': float(t_stat),\n", " 'p_value': float(p_value),\n", " 'significant': p_value < 0.05\n", " }\n", " \n", " # Save the test results\n", " with open(os.path.join(output_dir, 'statistical_tests.json'), 'w') as f:\n", " json.dump(test_results, f, indent=2)\n", " \n", " # Also save a human-readable summary\n", " with open(os.path.join(output_dir, 'statistical_tests_summary.txt'), 'w') as f:\n", " f.write(\"Statistical Test Results Summary\\n\")\n", " f.write(\"===============================\\n\\n\")\n", " \n", " for test_name, result in test_results.items():\n", " f.write(f\"Test: {test_name}\\n\")\n", " f.write(f\"Type: {result['test']}\\n\")\n", " if 'groups' in result:\n", " f.write(f\"Groups: {result['groups']}\\n\")\n", " \n", " if result['test'] == 'ANOVA':\n", " f.write(f\"F-statistic: {result['F_statistic']:.4f}\\n\")\n", " elif result['test'] == 't-test':\n", " f.write(f\"t-statistic: {result['t_statistic']:.4f}\\n\")\n", " \n", " f.write(f\"p-value: {result['p_value']:.4f}\\n\")\n", " f.write(f\"Significant: {result['significant']}\\n\\n\")\n", " \n", " print(f\"Saved statistical test results to {output_dir}\")\n", " return test_results\n", "\n", "def plot_final_configs(df, output_dir):\n", " \"\"\"\n", " Create visualizations of the best-performing configurations.\n", " \n", " Args:\n", " df: DataFrame with experiment results\n", " output_dir: Directory to save the plots\n", " \"\"\"\n", " # Group by configuration and calculate mean similarity score\n", " grouped = df.groupby(['diversity_level', 'branching_factor', 'max_depth', 'epsilon'])['similarity_score'].agg(['mean', 'std', 'count']).reset_index()\n", " \n", " # Sort by mean similarity score (descending)\n", " grouped = grouped.sort_values('mean', ascending=False)\n", " \n", " # Plot top configurations\n", " top_n = min(10, len(grouped))\n", " top_configs = grouped.head(top_n)\n", " \n", " plt.figure(figsize=(12, 8))\n", " \n", " # Create labels for x-axis\n", " config_labels = [f\"D:{row['diversity_level']}, B:{row['branching_factor']}, M:{row['max_depth']}, E:{row['epsilon']}\" \n", " for _, row in top_configs.iterrows()]\n", " \n", " # Plot with error bars\n", " plt.bar(range(top_n), top_configs['mean'], yerr=top_configs['std'], capsize=5, color='skyblue')\n", " plt.xticks(range(top_n), config_labels, rotation=45, ha='right')\n", " plt.ylabel('Mean Similarity Score')\n", " plt.title(f'Top {top_n} Configurations by Mean Similarity Score')\n", " plt.grid(axis='y', linestyle='--', alpha=0.7)\n", " plt.tight_layout()\n", " \n", " # Save the figure\n", " filename = os.path.join(output_dir, 'top_configurations.png')\n", " plt.savefig(filename, dpi=300, bbox_inches='tight')\n", " plt.close()\n", " print(f\"Saved plot to {filename}\")\n", "\n", "def plot_query_performance(df, output_dir):\n", " \"\"\"\n", " Create visualizations of performance across different queries.\n", " \n", " Args:\n", " df: DataFrame with experiment results\n", " output_dir: Directory to save the plots\n", " \"\"\"\n", " # Check if we have multiple queries\n", " if len(df['query'].unique()) <= 1:\n", " print(\"Only one query found - skipping query performance analysis\")\n", " return\n", " \n", " # For each query, find the best-performing configuration\n", " query_results = []\n", " \n", " for query in df['query'].unique():\n", " query_df = df[df['query'] == query]\n", " \n", " # Group by configuration\n", " grouped = query_df.groupby(['diversity_level', 'branching_factor', 'max_depth', 'epsilon'])['similarity_score'].mean().reset_index()\n", " \n", " # Get the best configuration\n", " best_config = grouped.loc[grouped['similarity_score'].idxmax()]\n", " \n", " query_results.append({\n", " 'query': query,\n", " 'best_config': {\n", " 'diversity_level': best_config['diversity_level'],\n", " 'branching_factor': best_config['branching_factor'],\n", " 'max_depth': best_config['max_depth'],\n", " 'epsilon': best_config['epsilon']\n", " },\n", " 'best_score': best_config['similarity_score'],\n", " 'avg_score': query_df['similarity_score'].mean(),\n", " 'min_score': query_df['similarity_score'].min(),\n", " 'max_score': query_df['similarity_score'].max()\n", " })\n", " \n", " # Create a plot of best scores by query\n", " query_df = pd.DataFrame(query_results)\n", " \n", " plt.figure(figsize=(10, 6))\n", " plt.bar(range(len(query_df)), query_df['best_score'], color='darkgreen')\n", " plt.xticks(range(len(query_df)), [q[:30] + '...' if len(q) > 30 else q for q in query_df['query']], rotation=45, ha='right')\n", " plt.ylabel('Best Similarity Score')\n", " plt.title('Best Performance by Query')\n", " plt.grid(axis='y', linestyle='--', alpha=0.7)\n", " plt.tight_layout()\n", " \n", " # Save the figure\n", " filename = os.path.join(output_dir, 'best_scores_by_query.png')\n", " plt.savefig(filename, dpi=300, bbox_inches='tight')\n", " plt.close()\n", " print(f\"Saved plot to {filename}\")\n", " \n", " # Save the query analysis results\n", " with open(os.path.join(output_dir, 'query_analysis.json'), 'w') as f:\n", " json.dump(query_results, f, indent=2)\n", "\n", "def analyze_path_length_effect(df, output_dir):\n", " \"\"\"\n", " Analyze the effect of actual path length on performance.\n", " \n", " Args:\n", " df: DataFrame with experiment results\n", " output_dir: Directory to save the plots\n", " \"\"\"\n", " if 'path_length' not in df.columns:\n", " print(\"Path length information not available - skipping path length analysis\")\n", " return\n", " \n", " # Group by path length\n", " grouped = df.groupby('path_length')['similarity_score'].agg(['mean', 'std', 'count']).reset_index()\n", " \n", " plt.figure(figsize=(10, 6))\n", " plt.bar(grouped['path_length'], grouped['mean'], yerr=grouped['std'], capsize=5, color='purple')\n", " plt.xticks(grouped['path_length'])\n", " plt.xlabel('Path Length')\n", " plt.ylabel('Mean Similarity Score')\n", " plt.title('Effect of Strategy Path Length on Performance')\n", " plt.grid(axis='y', linestyle='--', alpha=0.7)\n", " \n", " # Save the figure\n", " filename = os.path.join(output_dir, 'path_length_effect.png')\n", " plt.savefig(filename, dpi=300, bbox_inches='tight')\n", " plt.close()\n", " print(f\"Saved plot to {filename}\")\n", "\n", "def perform_full_analysis(input_file):\n", " \"\"\"\n", " Perform a full analysis of the experiment results.\n", " \n", " Args:\n", " input_file: Path to the experiment results file\n", " \"\"\"\n", " # Load the data\n", " df = load_experiment_data(input_file)\n", " \n", " # Get the experiment ID\n", " experiment_id = df['experiment_id'].iloc[0] if 'experiment_id' in df.columns and not df.empty else \"unknown_experiment\"\n", " \n", " # Create output directory\n", " output_dir = create_output_directory(experiment_id)\n", " print(f\"Analysis results will be saved to {output_dir}\")\n", " \n", " # Save some basic statistics\n", " basic_stats = {\n", " 'total_simulations': len(df),\n", " 'parameter_ranges': {\n", " 'diversity_levels': list(df['diversity_level'].unique()) if 'diversity_level' in df.columns else [],\n", " 'branching_factors': list(sorted(df['branching_factor'].unique())) if 'branching_factor' in df.columns else [],\n", " 'max_depths': list(sorted(df['max_depth'].unique())) if 'max_depth' in df.columns else [],\n", " 'epsilon_values': list(sorted(df['epsilon'].unique())) if 'epsilon' in df.columns else []\n", " },\n", " 'similarity_stats': {\n", " 'mean': float(df['similarity_score'].mean()),\n", " 'median': float(df['similarity_score'].median()),\n", " 'min': float(df['similarity_score'].min()),\n", " 'max': float(df['similarity_score'].max()),\n", " 'std': float(df['similarity_score'].std())\n", " }\n", " }\n", " \n", " with open(os.path.join(output_dir, 'basic_stats.json'), 'w') as f:\n", " json.dump(basic_stats, f, indent=2)\n", " \n", " # Create visualizations\n", " print(\"Generating visualizations...\")\n", " \n", " # Individual parameter effects\n", " for param in ['diversity_level', 'branching_factor', 'max_depth', 'epsilon']:\n", " if param in df.columns and len(df[param].unique()) > 1:\n", " plot_similarity_by_parameter(df, param, output_dir)\n", " \n", " # Parameter interactions\n", " interaction_pairs = [\n", " ('diversity_level', 'branching_factor'),\n", " ('diversity_level', 'max_depth'),\n", " ('branching_factor', 'max_depth'),\n", " ('epsilon', 'diversity_level')\n", " ]\n", " \n", " for param1, param2 in interaction_pairs:\n", " if param1 in df.columns and param2 in df.columns and len(df[param1].unique()) > 1 and len(df[param2].unique()) > 1:\n", " plot_parameter_interaction(df, param1, param2, output_dir)\n", " \n", " # Run statistical tests\n", " print(\"Running statistical tests...\")\n", " test_results = run_statistical_tests(df, output_dir)\n", " \n", " # Plot top configurations\n", " print(\"Analyzing top configurations...\")\n", " plot_final_configs(df, output_dir)\n", " \n", " # Query-specific analysis\n", " print(\"Analyzing performance by query...\")\n", " plot_query_performance(df, output_dir)\n", " \n", " # Path length analysis\n", " print(\"Analyzing path length effects...\")\n", " analyze_path_length_effect(df, output_dir)\n", " \n", " print(f\"Analysis complete. Results saved to {output_dir}\")\n", " return output_dir\n", "\n", "\n", "perform_full_analysis(\"experiment_results/exp_20250319_141119_all_results.json\")\n" ] }, { "cell_type": "code", "execution_count": 2, "id": "0b51f8cf-6f33-46dc-9209-3c08dc64944e", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Collecting seaborn\n", " Obtaining dependency information for seaborn from https://files.pythonhosted.org/packages/83/11/00d3c3dfc25ad54e731d91449895a79e4bf2384dc3ac01809010ba88f6d5/seaborn-0.13.2-py3-none-any.whl.metadata\n", " Using cached seaborn-0.13.2-py3-none-any.whl.metadata (5.4 kB)\n", "Collecting scipy\n", " Obtaining dependency information for scipy from https://files.pythonhosted.org/packages/a4/98/e5c964526c929ef1f795d4c343b2ff98634ad2051bd2bbadfef9e772e413/scipy-1.15.2-cp312-cp312-macosx_14_0_arm64.whl.metadata\n", " Using cached scipy-1.15.2-cp312-cp312-macosx_14_0_arm64.whl.metadata (61 kB)\n", "Requirement already satisfied: numpy!=1.24.0,>=1.20 in ./.venv/lib/python3.12/site-packages (from seaborn) (2.2.3)\n", "Requirement already satisfied: pandas>=1.2 in ./.venv/lib/python3.12/site-packages (from seaborn) (2.2.3)\n", "Requirement already satisfied: matplotlib!=3.6.1,>=3.4 in ./.venv/lib/python3.12/site-packages (from seaborn) (3.10.1)\n", "Requirement already satisfied: contourpy>=1.0.1 in ./.venv/lib/python3.12/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (1.3.1)\n", "Requirement already satisfied: cycler>=0.10 in ./.venv/lib/python3.12/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (0.12.1)\n", "Requirement already satisfied: fonttools>=4.22.0 in ./.venv/lib/python3.12/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (4.56.0)\n", "Requirement already satisfied: kiwisolver>=1.3.1 in ./.venv/lib/python3.12/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (1.4.8)\n", "Requirement already satisfied: packaging>=20.0 in ./.venv/lib/python3.12/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (24.2)\n", "Requirement already satisfied: pillow>=8 in ./.venv/lib/python3.12/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (11.1.0)\n", "Requirement already satisfied: pyparsing>=2.3.1 in ./.venv/lib/python3.12/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (3.2.1)\n", "Requirement already satisfied: python-dateutil>=2.7 in ./.venv/lib/python3.12/site-packages (from matplotlib!=3.6.1,>=3.4->seaborn) (2.9.0.post0)\n", "Requirement already satisfied: pytz>=2020.1 in ./.venv/lib/python3.12/site-packages (from pandas>=1.2->seaborn) (2025.1)\n", "Requirement already satisfied: tzdata>=2022.7 in ./.venv/lib/python3.12/site-packages (from pandas>=1.2->seaborn) (2025.1)\n", "Requirement already satisfied: six>=1.5 in ./.venv/lib/python3.12/site-packages (from python-dateutil>=2.7->matplotlib!=3.6.1,>=3.4->seaborn) (1.17.0)\n", "Using cached seaborn-0.13.2-py3-none-any.whl (294 kB)\n", "Using cached scipy-1.15.2-cp312-cp312-macosx_14_0_arm64.whl (22.4 MB)\n", "Installing collected packages: scipy, seaborn\n", "Successfully installed scipy-1.15.2 seaborn-0.13.2\n", "\n", "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m A new release of pip is available: \u001b[0m\u001b[31;49m23.2.1\u001b[0m\u001b[39;49m -> \u001b[0m\u001b[32;49m25.0.1\u001b[0m\n", "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m To update, run: \u001b[0m\u001b[32;49mpip install --upgrade pip\u001b[0m\n" ] } ], "source": [ "!pip install seaborn scipy " ] }, { "cell_type": "code", "execution_count": 7, "id": "f7c32e9c-fbc7-415e-b484-526d079dd9b2", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Loading data from experiment_results/exp_20250319_141119_all_results.json...\n", "Loaded 16 valid results\n" ] }, { "ename": "TypeError", "evalue": "Object of type int64 is not JSON serializable", "output_type": "error", "traceback": [ "\u001b[31m---------------------------------------------------------------------------\u001b[39m", "\u001b[31mTypeError\u001b[39m Traceback (most recent call last)", "\u001b[36mCell\u001b[39m\u001b[36m \u001b[39m\u001b[32mIn[7]\u001b[39m\u001b[32m, line 522\u001b[39m\n\u001b[32m 517\u001b[39m \u001b[38;5;28;01mreturn\u001b[39;00m output_dir\n\u001b[32m 519\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m \u001b[34m__name__\u001b[39m == \u001b[33m\"\u001b[39m\u001b[33m__main__\u001b[39m\u001b[33m\"\u001b[39m:\n\u001b[32m--> \u001b[39m\u001b[32m522\u001b[39m \u001b[43mperform_full_analysis\u001b[49m\u001b[43m(\u001b[49m\u001b[33;43m\"\u001b[39;49m\u001b[33;43mexperiment_results/exp_20250319_141119_all_results.json\u001b[39;49m\u001b[33;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n", "\u001b[36mCell\u001b[39m\u001b[36m \u001b[39m\u001b[32mIn[7]\u001b[39m\u001b[32m, line 489\u001b[39m, in \u001b[36mperform_full_analysis\u001b[39m\u001b[34m(input_file)\u001b[39m\n\u001b[32m 471\u001b[39m basic_stats = {\n\u001b[32m 472\u001b[39m \u001b[33m'\u001b[39m\u001b[33mtotal_simulations\u001b[39m\u001b[33m'\u001b[39m: \u001b[38;5;28mlen\u001b[39m(df),\n\u001b[32m 473\u001b[39m \u001b[33m'\u001b[39m\u001b[33mparameter_ranges\u001b[39m\u001b[33m'\u001b[39m: {\n\u001b[32m (...)\u001b[39m\u001b[32m 485\u001b[39m }\n\u001b[32m 486\u001b[39m }\n\u001b[32m 488\u001b[39m \u001b[38;5;28;01mwith\u001b[39;00m \u001b[38;5;28mopen\u001b[39m(os.path.join(output_dir, \u001b[33m'\u001b[39m\u001b[33mbasic_stats.json\u001b[39m\u001b[33m'\u001b[39m), \u001b[33m'\u001b[39m\u001b[33mw\u001b[39m\u001b[33m'\u001b[39m) \u001b[38;5;28;01mas\u001b[39;00m f:\n\u001b[32m--> \u001b[39m\u001b[32m489\u001b[39m \u001b[43mjson\u001b[49m\u001b[43m.\u001b[49m\u001b[43mdump\u001b[49m\u001b[43m(\u001b[49m\u001b[43mbasic_stats\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mf\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mindent\u001b[49m\u001b[43m=\u001b[49m\u001b[32;43m2\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[32m 491\u001b[39m \u001b[38;5;66;03m# Print basic statistics\u001b[39;00m\n\u001b[32m 492\u001b[39m \u001b[38;5;28mprint\u001b[39m(\u001b[33m\"\u001b[39m\u001b[38;5;130;01m\\n\u001b[39;00m\u001b[33m============= EXPERIMENT ANALYSIS =============\u001b[39m\u001b[33m\"\u001b[39m)\n", "\u001b[36mFile \u001b[39m\u001b[32m/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/json/__init__.py:179\u001b[39m, in \u001b[36mdump\u001b[39m\u001b[34m(obj, fp, skipkeys, ensure_ascii, check_circular, allow_nan, cls, indent, separators, default, sort_keys, **kw)\u001b[39m\n\u001b[32m 173\u001b[39m iterable = \u001b[38;5;28mcls\u001b[39m(skipkeys=skipkeys, ensure_ascii=ensure_ascii,\n\u001b[32m 174\u001b[39m check_circular=check_circular, allow_nan=allow_nan, indent=indent,\n\u001b[32m 175\u001b[39m separators=separators,\n\u001b[32m 176\u001b[39m default=default, sort_keys=sort_keys, **kw).iterencode(obj)\n\u001b[32m 177\u001b[39m \u001b[38;5;66;03m# could accelerate with writelines in some versions of Python, at\u001b[39;00m\n\u001b[32m 178\u001b[39m \u001b[38;5;66;03m# a debuggability cost\u001b[39;00m\n\u001b[32m--> \u001b[39m\u001b[32m179\u001b[39m \u001b[43m\u001b[49m\u001b[38;5;28;43;01mfor\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mchunk\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;129;43;01min\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43miterable\u001b[49m\u001b[43m:\u001b[49m\n\u001b[32m 180\u001b[39m \u001b[43m \u001b[49m\u001b[43mfp\u001b[49m\u001b[43m.\u001b[49m\u001b[43mwrite\u001b[49m\u001b[43m(\u001b[49m\u001b[43mchunk\u001b[49m\u001b[43m)\u001b[49m\n", "\u001b[36mFile \u001b[39m\u001b[32m/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/json/encoder.py:432\u001b[39m, in \u001b[36m_make_iterencode.._iterencode\u001b[39m\u001b[34m(o, _current_indent_level)\u001b[39m\n\u001b[32m 430\u001b[39m \u001b[38;5;28;01myield from\u001b[39;00m _iterencode_list(o, _current_indent_level)\n\u001b[32m 431\u001b[39m \u001b[38;5;28;01melif\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(o, \u001b[38;5;28mdict\u001b[39m):\n\u001b[32m--> \u001b[39m\u001b[32m432\u001b[39m \u001b[38;5;28;01myield from\u001b[39;00m _iterencode_dict(o, _current_indent_level)\n\u001b[32m 433\u001b[39m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[32m 434\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m markers \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n", "\u001b[36mFile \u001b[39m\u001b[32m/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/json/encoder.py:406\u001b[39m, in \u001b[36m_make_iterencode.._iterencode_dict\u001b[39m\u001b[34m(dct, _current_indent_level)\u001b[39m\n\u001b[32m 404\u001b[39m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[32m 405\u001b[39m chunks = _iterencode(value, _current_indent_level)\n\u001b[32m--> \u001b[39m\u001b[32m406\u001b[39m \u001b[38;5;28;01myield from\u001b[39;00m chunks\n\u001b[32m 407\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m newline_indent \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[32m 408\u001b[39m _current_indent_level -= \u001b[32m1\u001b[39m\n", "\u001b[36mFile \u001b[39m\u001b[32m/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/json/encoder.py:406\u001b[39m, in \u001b[36m_make_iterencode.._iterencode_dict\u001b[39m\u001b[34m(dct, _current_indent_level)\u001b[39m\n\u001b[32m 404\u001b[39m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[32m 405\u001b[39m chunks = _iterencode(value, _current_indent_level)\n\u001b[32m--> \u001b[39m\u001b[32m406\u001b[39m \u001b[38;5;28;01myield from\u001b[39;00m chunks\n\u001b[32m 407\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m newline_indent \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[32m 408\u001b[39m _current_indent_level -= \u001b[32m1\u001b[39m\n", "\u001b[36mFile \u001b[39m\u001b[32m/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/json/encoder.py:326\u001b[39m, in \u001b[36m_make_iterencode.._iterencode_list\u001b[39m\u001b[34m(lst, _current_indent_level)\u001b[39m\n\u001b[32m 324\u001b[39m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[32m 325\u001b[39m chunks = _iterencode(value, _current_indent_level)\n\u001b[32m--> \u001b[39m\u001b[32m326\u001b[39m \u001b[38;5;28;01myield from\u001b[39;00m chunks\n\u001b[32m 327\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m newline_indent \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[32m 328\u001b[39m _current_indent_level -= \u001b[32m1\u001b[39m\n", "\u001b[36mFile \u001b[39m\u001b[32m/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/json/encoder.py:439\u001b[39m, in \u001b[36m_make_iterencode.._iterencode\u001b[39m\u001b[34m(o, _current_indent_level)\u001b[39m\n\u001b[32m 437\u001b[39m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\u001b[33m\"\u001b[39m\u001b[33mCircular reference detected\u001b[39m\u001b[33m\"\u001b[39m)\n\u001b[32m 438\u001b[39m markers[markerid] = o\n\u001b[32m--> \u001b[39m\u001b[32m439\u001b[39m o = \u001b[43m_default\u001b[49m\u001b[43m(\u001b[49m\u001b[43mo\u001b[49m\u001b[43m)\u001b[49m\n\u001b[32m 440\u001b[39m \u001b[38;5;28;01myield from\u001b[39;00m _iterencode(o, _current_indent_level)\n\u001b[32m 441\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m markers \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n", "\u001b[36mFile \u001b[39m\u001b[32m/Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/json/encoder.py:180\u001b[39m, in \u001b[36mJSONEncoder.default\u001b[39m\u001b[34m(self, o)\u001b[39m\n\u001b[32m 161\u001b[39m \u001b[38;5;28;01mdef\u001b[39;00m\u001b[38;5;250m \u001b[39m\u001b[34mdefault\u001b[39m(\u001b[38;5;28mself\u001b[39m, o):\n\u001b[32m 162\u001b[39m \u001b[38;5;250m \u001b[39m\u001b[33;03m\"\"\"Implement this method in a subclass such that it returns\u001b[39;00m\n\u001b[32m 163\u001b[39m \u001b[33;03m a serializable object for ``o``, or calls the base implementation\u001b[39;00m\n\u001b[32m 164\u001b[39m \u001b[33;03m (to raise a ``TypeError``).\u001b[39;00m\n\u001b[32m (...)\u001b[39m\u001b[32m 178\u001b[39m \n\u001b[32m 179\u001b[39m \u001b[33;03m \"\"\"\u001b[39;00m\n\u001b[32m--> \u001b[39m\u001b[32m180\u001b[39m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mTypeError\u001b[39;00m(\u001b[33mf\u001b[39m\u001b[33m'\u001b[39m\u001b[33mObject of type \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mo.\u001b[34m__class__\u001b[39m.\u001b[34m__name__\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[33m \u001b[39m\u001b[33m'\u001b[39m\n\u001b[32m 181\u001b[39m \u001b[33mf\u001b[39m\u001b[33m'\u001b[39m\u001b[33mis not JSON serializable\u001b[39m\u001b[33m'\u001b[39m)\n", "\u001b[31mTypeError\u001b[39m: Object of type int64 is not JSON serializable" ] } ], "source": [ "import os\n", "import json\n", "import argparse\n", "import numpy as np\n", "import pandas as pd\n", "import matplotlib.pyplot as plt\n", "import seaborn as sns\n", "from scipy import stats\n", "\n", "def load_experiment_data(filename):\n", " \"\"\"Load experiment results from a JSON file.\"\"\"\n", " print(f\"Loading data from {filename}...\")\n", " \n", " with open(filename, 'r') as f:\n", " data = json.load(f)\n", " \n", " # Handle all_results file format\n", " if isinstance(data, list):\n", " results = data\n", " # Handle analysis file format\n", " elif 'summary' in data:\n", " experiment_id = data.get('experiment_id')\n", " all_results_file = os.path.join(os.path.dirname(filename), f\"{experiment_id}_all_results.json\")\n", " \n", " if os.path.exists(all_results_file):\n", " with open(all_results_file, 'r') as f:\n", " results = json.load(f)\n", " else:\n", " print(\"Error: Could not find all_results file\")\n", " return None\n", " else:\n", " print(\"Error: Unrecognized file format\")\n", " return None\n", " \n", " # Create DataFrame\n", " rows = []\n", " for result in results:\n", " # Skip results with errors\n", " if 'error' in result:\n", " continue\n", " \n", " row = {\n", " 'query': result.get('query', ''),\n", " 'similarity_score': result.get('similarity_score', 0),\n", " 'diversity_level': result.get('ibfs_config', {}).get('diversity_level', ''),\n", " 'branching_factor': result.get('ibfs_config', {}).get('branching_factor', 0),\n", " 'max_depth': result.get('ibfs_config', {}).get('max_depth', 0),\n", " 'epsilon': result.get('user_config', {}).get('epsilon', 0),\n", " 'experiment_id': result.get('experiment_id', '')\n", " }\n", " \n", " # Add strategy path information if available\n", " if 'strategy_path' in result:\n", " row['path_length'] = len(result.get('strategy_path', []))\n", " \n", " rows.append(row)\n", " \n", " df = pd.DataFrame(rows)\n", " print(f\"Loaded {len(df)} valid results\")\n", " return df\n", "\n", "def create_parameter_plots(df, output_dir):\n", " \"\"\"Create plots showing the effects of different parameters.\"\"\"\n", " os.makedirs(output_dir, exist_ok=True)\n", " \n", " # Set seaborn style for better visualization\n", " sns.set(style=\"whitegrid\")\n", " \n", " # Plot 1: Similarity scores by diversity level\n", " if 'diversity_level' in df.columns and len(df['diversity_level'].unique()) > 1:\n", " plt.figure(figsize=(10, 6))\n", " ax = sns.boxplot(x='diversity_level', y='similarity_score', data=df, palette=\"Set3\")\n", " ax.set_title('Similarity Scores by Diversity Level', fontsize=16)\n", " ax.set_xlabel('Diversity Level', fontsize=14)\n", " ax.set_ylabel('Similarity Score', fontsize=14)\n", " \n", " # Add mean values as text above each box\n", " for i, level in enumerate(sorted(df['diversity_level'].unique())):\n", " mean_val = df[df['diversity_level'] == level]['similarity_score'].mean()\n", " ax.text(i, df['similarity_score'].max() + 0.01, f'Mean: {mean_val:.3f}', \n", " horizontalalignment='center', fontsize=12)\n", " \n", " plt.tight_layout()\n", " plt.savefig(os.path.join(output_dir, 'diversity_effect.png'), dpi=300, bbox_inches='tight')\n", " plt.show()\n", " \n", " # Plot 2: Similarity scores by branching factor\n", " if 'branching_factor' in df.columns and len(df['branching_factor'].unique()) > 1:\n", " plt.figure(figsize=(10, 6))\n", " ax = sns.boxplot(x='branching_factor', y='similarity_score', data=df, palette=\"Blues\")\n", " ax.set_title('Similarity Scores by Branching Factor', fontsize=16)\n", " ax.set_xlabel('Branching Factor', fontsize=14)\n", " ax.set_ylabel('Similarity Score', fontsize=14)\n", " \n", " # Add mean values as text above each box\n", " for i, factor in enumerate(sorted(df['branching_factor'].unique())):\n", " mean_val = df[df['branching_factor'] == factor]['similarity_score'].mean()\n", " ax.text(i, df['similarity_score'].max() + 0.01, f'Mean: {mean_val:.3f}', \n", " horizontalalignment='center', fontsize=12)\n", " \n", " plt.tight_layout()\n", " plt.savefig(os.path.join(output_dir, 'branching_effect.png'), dpi=300, bbox_inches='tight')\n", " plt.show()\n", " \n", " # Plot 3: Similarity scores by max depth\n", " if 'max_depth' in df.columns and len(df['max_depth'].unique()) > 1:\n", " plt.figure(figsize=(10, 6))\n", " ax = sns.boxplot(x='max_depth', y='similarity_score', data=df, palette=\"Greens\")\n", " ax.set_title('Similarity Scores by Max Depth', fontsize=16)\n", " ax.set_xlabel('Max Depth', fontsize=14)\n", " ax.set_ylabel('Similarity Score', fontsize=14)\n", " \n", " # Add mean values as text above each box\n", " for i, depth in enumerate(sorted(df['max_depth'].unique())):\n", " mean_val = df[df['max_depth'] == depth]['similarity_score'].mean()\n", " ax.text(i, df['similarity_score'].max() + 0.01, f'Mean: {mean_val:.3f}', \n", " horizontalalignment='center', fontsize=12)\n", " \n", " plt.tight_layout()\n", " plt.savefig(os.path.join(output_dir, 'depth_effect.png'), dpi=300, bbox_inches='tight')\n", " plt.show()\n", " \n", " # Plot 4: Similarity scores by epsilon\n", " if 'epsilon' in df.columns and len(df['epsilon'].unique()) > 1:\n", " plt.figure(figsize=(10, 6))\n", " ax = sns.boxplot(x='epsilon', y='similarity_score', data=df, palette=\"Oranges\")\n", " ax.set_title('Similarity Scores by Epsilon (Noise Parameter)', fontsize=16)\n", " ax.set_xlabel('Epsilon', fontsize=14)\n", " ax.set_ylabel('Similarity Score', fontsize=14)\n", " \n", " # Add mean values as text above each box\n", " for i, eps in enumerate(sorted(df['epsilon'].unique())):\n", " mean_val = df[df['epsilon'] == eps]['similarity_score'].mean()\n", " ax.text(i, df['similarity_score'].max() + 0.01, f'Mean: {mean_val:.3f}', \n", " horizontalalignment='center', fontsize=12)\n", " \n", " plt.tight_layout()\n", " plt.savefig(os.path.join(output_dir, 'epsilon_effect.png'), dpi=300, bbox_inches='tight')\n", " plt.show()\n", " \n", " # Plot 5: Heatmap of diversity level and branching factor\n", " if ('diversity_level' in df.columns and 'branching_factor' in df.columns and \n", " len(df['diversity_level'].unique()) > 1 and len(df['branching_factor'].unique()) > 1):\n", " \n", " pivot_table = df.pivot_table(\n", " values='similarity_score', \n", " index='diversity_level', \n", " columns='branching_factor',\n", " aggfunc='mean'\n", " )\n", " \n", " plt.figure(figsize=(10, 8))\n", " ax = sns.heatmap(pivot_table, annot=True, cmap=\"YlGnBu\", fmt=\".3f\", linewidths=.5)\n", " ax.set_title('Interaction: Diversity Level × Branching Factor', fontsize=16)\n", " ax.set_xlabel('Branching Factor', fontsize=14)\n", " ax.set_ylabel('Diversity Level', fontsize=14)\n", " \n", " plt.tight_layout()\n", " plt.savefig(os.path.join(output_dir, 'diversity_branching_interaction.png'), dpi=300, bbox_inches='tight')\n", " plt.show()\n", " \n", " # Plot 6: Heatmap of diversity level and max depth\n", " if ('diversity_level' in df.columns and 'max_depth' in df.columns and \n", " len(df['diversity_level'].unique()) > 1 and len(df['max_depth'].unique()) > 1):\n", " \n", " pivot_table = df.pivot_table(\n", " values='similarity_score', \n", " index='diversity_level', \n", " columns='max_depth',\n", " aggfunc='mean'\n", " )\n", " \n", " plt.figure(figsize=(10, 8))\n", " ax = sns.heatmap(pivot_table, annot=True, cmap=\"YlOrRd\", fmt=\".3f\", linewidths=.5)\n", " ax.set_title('Interaction: Diversity Level × Max Depth', fontsize=16)\n", " ax.set_xlabel('Max Depth', fontsize=14)\n", " ax.set_ylabel('Diversity Level', fontsize=14)\n", " \n", " plt.tight_layout()\n", " plt.savefig(os.path.join(output_dir, 'diversity_depth_interaction.png'), dpi=300, bbox_inches='tight')\n", " plt.show()\n", " \n", " # Plot 7: Heatmap of branching factor and max depth\n", " if ('branching_factor' in df.columns and 'max_depth' in df.columns and \n", " len(df['branching_factor'].unique()) > 1 and len(df['max_depth'].unique()) > 1):\n", " \n", " pivot_table = df.pivot_table(\n", " values='similarity_score', \n", " index='branching_factor', \n", " columns='max_depth',\n", " aggfunc='mean'\n", " )\n", " \n", " plt.figure(figsize=(10, 8))\n", " ax = sns.heatmap(pivot_table, annot=True, cmap=\"PuBuGn\", fmt=\".3f\", linewidths=.5)\n", " ax.set_title('Interaction: Branching Factor × Max Depth', fontsize=16)\n", " ax.set_xlabel('Max Depth', fontsize=14)\n", " ax.set_ylabel('Branching Factor', fontsize=14)\n", " \n", " plt.tight_layout()\n", " plt.savefig(os.path.join(output_dir, 'branching_depth_interaction.png'), dpi=300, bbox_inches='tight')\n", " plt.show()\n", "\n", "def plot_top_configurations(df, output_dir):\n", " \"\"\"Create plots showing top configurations.\"\"\"\n", " # Group by all key parameters\n", " grouped = df.groupby(['diversity_level', 'branching_factor', 'max_depth'])['similarity_score'].agg(\n", " ['mean', 'std', 'count']).reset_index()\n", " \n", " # Sort by mean similarity score (descending)\n", " grouped = grouped.sort_values('mean', ascending=False)\n", " \n", " # Plot top 10 configurations (or all if less than 10)\n", " top_n = min(10, len(grouped))\n", " top_configs = grouped.head(top_n)\n", " \n", " plt.figure(figsize=(12, 8))\n", " \n", " # Create labels for x-axis\n", " config_labels = [f\"D:{row['diversity_level']}, B:{row['branching_factor']}, M:{row['max_depth']}\" \n", " for _, row in top_configs.iterrows()]\n", " \n", " # Plot with error bars\n", " ax = plt.bar(range(top_n), top_configs['mean'], yerr=top_configs['std'], \n", " capsize=5, color='skyblue', alpha=0.7)\n", " plt.xticks(range(top_n), config_labels, rotation=45, ha='right', fontsize=12)\n", " plt.ylabel('Mean Similarity Score', fontsize=14)\n", " plt.title('Top Configurations by Mean Similarity Score', fontsize=16)\n", " plt.ylim(0, 1.0) # Assuming similarity scores are between 0 and 1\n", " plt.grid(axis='y', linestyle='--', alpha=0.7)\n", " \n", " # Add value labels on top of bars\n", " for i, value in enumerate(top_configs['mean']):\n", " plt.text(i, value + 0.02, f\"{value:.3f}\", ha='center', fontsize=11)\n", " \n", " plt.tight_layout()\n", " plt.savefig(os.path.join(output_dir, 'top_configurations.png'), dpi=300, bbox_inches='tight')\n", " plt.show()\n", " \n", " # Output the top configurations as text\n", " with open(os.path.join(output_dir, 'top_configurations.txt'), 'w') as f:\n", " f.write(\"Top Configurations by Mean Similarity Score\\n\")\n", " f.write(\"=========================================\\n\\n\")\n", " \n", " for i, (_, row) in enumerate(top_configs.iterrows(), 1):\n", " f.write(f\"{i}. Diversity: {row['diversity_level']}, Branching: {row['branching_factor']}, \")\n", " f.write(f\"Max Depth: {row['max_depth']}, Score: {row['mean']:.4f}\")\n", " f.write(f\" (std: {row['std']:.3f}, n: {row['count']})\\n\")\n", " \n", " # Print top configurations\n", " print(\"\\nTop configurations by mean similarity score:\")\n", " for i, (_, row) in enumerate(top_configs.iterrows(), 1):\n", " print(f\"{i}. Diversity: {row['diversity_level']}, Branching: {row['branching_factor']}, \"\n", " f\"Max Depth: {row['max_depth']}, Score: {row['mean']:.4f}\")\n", "\n", "def run_statistical_tests(df, output_dir):\n", " \"\"\"Run statistical tests to check for significant effects.\"\"\"\n", " results = {}\n", " \n", " # Function to run and format test results\n", " def run_test_for_param(param):\n", " param_values = sorted(df[param].unique())\n", " \n", " if len(param_values) == 2:\n", " # T-test for two groups\n", " group1 = df[df[param] == param_values[0]]['similarity_score']\n", " group2 = df[df[param] == param_values[1]]['similarity_score']\n", " \n", " t_stat, p_val = stats.ttest_ind(group1, group2, equal_var=False)\n", " \n", " # Calculate effect size (Cohen's d)\n", " mean1, mean2 = group1.mean(), group2.mean()\n", " sd1, sd2 = group1.std(), group2.std()\n", " n1, n2 = len(group1), len(group2)\n", " \n", " # Pooled standard deviation\n", " sd_pooled = np.sqrt(((n1 - 1) * sd1**2 + (n2 - 1) * sd2**2) / (n1 + n2 - 2))\n", " \n", " # Cohen's d\n", " d = abs(mean1 - mean2) / sd_pooled\n", " \n", " return {\n", " 'test': 't-test',\n", " 'parameter': param,\n", " 'groups': param_values,\n", " 'statistic': float(t_stat),\n", " 'p_value': float(p_val),\n", " 'significant': p_val < 0.05,\n", " 'effect_size': float(d),\n", " 'effect_size_type': \"Cohen's d\",\n", " 'mean_values': {str(param_values[0]): float(mean1), str(param_values[1]): float(mean2)}\n", " }\n", " \n", " elif len(param_values) > 2:\n", " # ANOVA for more than two groups\n", " groups = [df[df[param] == val]['similarity_score'] for val in param_values]\n", " f_stat, p_val = stats.f_oneway(*groups)\n", " \n", " # Calculate mean for each group\n", " mean_values = {str(val): float(df[df[param] == val]['similarity_score'].mean()) for val in param_values}\n", " \n", " # Calculate effect size (Eta squared)\n", " ss_between = sum(len(group) * ((group.mean() - df['similarity_score'].mean()) ** 2) for group in groups)\n", " ss_total = sum((df['similarity_score'] - df['similarity_score'].mean()) ** 2)\n", " eta_squared = ss_between / ss_total if ss_total > 0 else 0\n", " \n", " return {\n", " 'test': 'ANOVA',\n", " 'parameter': param,\n", " 'groups': list(map(str, param_values)),\n", " 'statistic': float(f_stat),\n", " 'p_value': float(p_val),\n", " 'significant': p_val < 0.05,\n", " 'effect_size': float(eta_squared),\n", " 'effect_size_type': \"Eta squared\",\n", " 'mean_values': mean_values\n", " }\n", " \n", " # Test all available parameters\n", " for param in ['diversity_level', 'branching_factor', 'max_depth', 'epsilon']:\n", " if param in df.columns and len(df[param].unique()) > 1:\n", " results[f'{param}_test'] = run_test_for_param(param)\n", " \n", " # Write results to a detailed text file\n", " with open(os.path.join(output_dir, 'statistical_results.txt'), 'w') as f:\n", " f.write(\"Statistical Test Results\\n\")\n", " f.write(\"=======================\\n\\n\")\n", " \n", " for test_name, result in results.items():\n", " f.write(f\"Test: {test_name}\\n\")\n", " f.write(f\"Parameter: {result['parameter']}\\n\")\n", " f.write(f\"Type: {result['test']}\\n\")\n", " f.write(f\"Groups: {result['groups']}\\n\")\n", " \n", " if result['test'] == 'ANOVA':\n", " f.write(f\"F-statistic: {result['statistic']:.4f}\\n\")\n", " elif result['test'] == 't-test':\n", " f.write(f\"t-statistic: {result['statistic']:.4f}\\n\")\n", " \n", " f.write(f\"p-value: {result['p_value']:.4f}\\n\")\n", " f.write(f\"Significant: {result['significant']}\\n\")\n", " f.write(f\"Effect size ({result['effect_size_type']}): {result['effect_size']:.4f}\\n\")\n", " f.write(\"Mean values by group:\\n\")\n", " \n", " for group, mean in result['mean_values'].items():\n", " f.write(f\" - {group}: {mean:.4f}\\n\")\n", " \n", " f.write(\"\\n\")\n", " \n", " # Save results as JSON for potential further analysis\n", " with open(os.path.join(output_dir, 'statistical_results.json'), 'w') as f:\n", " json.dump(results, f, indent=2)\n", " \n", " # Print significant findings\n", " significant_results = [r for r in results.values() if r['significant']]\n", " \n", " if significant_results:\n", " print(\"\\nSignificant effects found:\")\n", " for result in significant_results:\n", " effect_size_interp = \"\"\n", " if result['effect_size_type'] == \"Cohen's d\":\n", " if result['effect_size'] < 0.2:\n", " effect_size_interp = \"negligible effect\"\n", " elif result['effect_size'] < 0.5:\n", " effect_size_interp = \"small effect\"\n", " elif result['effect_size'] < 0.8:\n", " effect_size_interp = \"medium effect\"\n", " else:\n", " effect_size_interp = \"large effect\"\n", " elif result['effect_size_type'] == \"Eta squared\":\n", " if result['effect_size'] < 0.01:\n", " effect_size_interp = \"negligible effect\"\n", " elif result['effect_size'] < 0.06:\n", " effect_size_interp = \"small effect\"\n", " elif result['effect_size'] < 0.14:\n", " effect_size_interp = \"medium effect\"\n", " else:\n", " effect_size_interp = \"large effect\"\n", " \n", " print(f\"- {result['parameter']}: p={result['p_value']:.4f}, {result['effect_size_type']}={result['effect_size']:.3f} ({effect_size_interp})\")\n", " # Print mean differences to understand direction\n", " means = result['mean_values']\n", " print(f\" Mean values: {', '.join([f'{k}={v:.3f}' for k, v in means.items()])}\")\n", " else:\n", " print(\"\\nNo significant effects found in the parameters tested.\")\n", " \n", " return results\n", "\n", "def analyze_query_performance(df, output_dir):\n", " \"\"\"Analyze performance across different queries.\"\"\"\n", " if 'query' not in df.columns or len(df['query'].unique()) <= 1:\n", " print(\"Query analysis skipped - insufficient query variation\")\n", " return\n", " \n", " # Get performance by query\n", " query_performance = df.groupby('query')['similarity_score'].agg(['mean', 'std', 'min', 'max', 'count']).reset_index()\n", " query_performance = query_performance.sort_values('mean', ascending=False)\n", " \n", " # Plot query performance\n", " plt.figure(figsize=(12, 6))\n", " \n", " # Truncate long query text for display\n", " query_labels = [q[:30] + '...' if len(q) > 30 else q for q in query_performance['query']]\n", " \n", " # Plot with error bars\n", " plt.bar(range(len(query_performance)), query_performance['mean'], \n", " yerr=query_performance['std'], capsize=5, color='lightgreen', alpha=0.7)\n", " plt.xticks(range(len(query_performance)), query_labels, rotation=45, ha='right')\n", " plt.ylabel('Mean Similarity Score')\n", " plt.title('Performance by Query')\n", " plt.ylim(0, 1.0)\n", " plt.grid(axis='y', linestyle='--', alpha=0.7)\n", " \n", " # Add value labels on top of bars\n", " for i, value in enumerate(query_performance['mean']):\n", " plt.text(i, value + 0.02, f\"{value:.3f}\", ha='center')\n", " \n", " plt.tight_layout()\n", " plt.savefig(os.path.join(output_dir, 'query_performance.png'), dpi=300, bbox_inches='tight')\n", " plt.show()\n", " \n", " # For each query, find best configuration\n", " query_best_configs = []\n", " \n", " for query in df['query'].unique():\n", " query_df = df[df['query'] == query]\n", " \n", " # Group by configuration\n", " grouped = query_df.groupby(['diversity_level', 'branching_factor', 'max_depth'])['similarity_score'].mean().reset_index()\n", " \n", " # Get the best configuration\n", " best_config = grouped.loc[grouped['similarity_score'].idxmax()]\n", " \n", " query_best_configs.append({\n", " 'query': query,\n", " 'best_config': {\n", " 'diversity_level': best_config['diversity_level'],\n", " 'branching_factor': int(best_config['branching_factor']),\n", " 'max_depth': int(best_config['max_depth']),\n", " },\n", " 'best_score': float(best_config['similarity_score']),\n", " 'mean_score': float(query_df['similarity_score'].mean()),\n", " })\n", " \n", " # Save query analysis to file\n", " with open(os.path.join(output_dir, 'query_analysis.json'), 'w') as f:\n", " json.dump(query_best_configs, f, indent=2)\n", " \n", " # Print best configurations for each query\n", " print(\"\\nBest configuration for each query:\")\n", " for result in query_best_configs:\n", " query_display = result['query'] if len(result['query']) < 50 else result['query'][:47] + \"...\"\n", " config = result['best_config']\n", " print(f\"- {query_display}\")\n", " print(f\" Best config: D:{config['diversity_level']}, B:{config['branching_factor']}, M:{config['max_depth']}\")\n", " print(f\" Score: {result['best_score']:.4f} (mean across all configs: {result['mean_score']:.4f})\")\n", "\n", "def perform_full_analysis(input_file):\n", " \"\"\"Perform a comprehensive analysis of the experiment results.\"\"\"\n", " # Load the data\n", " df = load_experiment_data(input_file)\n", " if df is None:\n", " return\n", " \n", " # Create output directory\n", " experiment_id = df['experiment_id'].iloc[0] if not df.empty and 'experiment_id' in df.columns else \"unknown\"\n", " output_dir = f\"analysis_{experiment_id}\"\n", " os.makedirs(output_dir, exist_ok=True)\n", " \n", " # Save basic statistics\n", " basic_stats = {\n", " 'total_simulations': len(df),\n", " 'parameter_ranges': {\n", " 'diversity_levels': list(df['diversity_level'].unique()) if 'diversity_level' in df.columns else [],\n", " 'branching_factors': list(sorted(df['branching_factor'].unique())) if 'branching_factor' in df.columns else [],\n", " 'max_depths': list(sorted(df['max_depth'].unique())) if 'max_depth' in df.columns else [],\n", " 'epsilon_values': list(sorted(df['epsilon'].unique())) if 'epsilon' in df.columns else []\n", " },\n", " 'similarity_stats': {\n", " 'mean': float(df['similarity_score'].mean()),\n", " 'median': float(df['similarity_score'].median()),\n", " 'min': float(df['similarity_score'].min()),\n", " 'max': float(df['similarity_score'].max()),\n", " 'std': float(df['similarity_score'].std())\n", " }\n", " }\n", " \n", " with open(os.path.join(output_dir, 'basic_stats.json'), 'w') as f:\n", " json.dump(basic_stats, f, indent=2)\n", " \n", " # Print basic statistics\n", " print(\"\\n============= EXPERIMENT ANALYSIS =============\")\n", " print(f\"Total simulations: {basic_stats['total_simulations']}\")\n", " print(f\"Overall similarity score: mean={basic_stats['similarity_stats']['mean']:.4f}, \"\n", " f\"median={basic_stats['similarity_stats']['median']:.4f}, \"\n", " f\"min={basic_stats['similarity_stats']['min']:.4f}, \"\n", " f\"max={basic_stats['similarity_stats']['max']:.4f}, \"\n", " f\"std={basic_stats['similarity_stats']['std']:.4f}\")\n", " \n", " # Create parameter plots\n", " print(\"\\nGenerating parameter effect plots...\")\n", " create_parameter_plots(df, output_dir)\n", " \n", " # Plot top configurations\n", " print(\"\\nAnalyzing top configurations...\")\n", " plot_top_configurations(df, output_dir)\n", " \n", " # Run statistical tests\n", " print(\"\\nRunning statistical tests...\")\n", " run_statistical_tests(df, output_dir)\n", " \n", " # Analyze query performance\n", " print(\"\\nAnalyzing query performance...\")\n", " analyze_query_performance(df, output_dir)\n", " \n", " print(f\"\\nAnalysis complete. Results saved to {output_dir}\")\n", " return output_dir\n", "\n", "if __name__ == \"__main__\":\n", "\n", " \n", " perform_full_analysis(\"experiment_results/exp_20250319_141119_all_results.json\")" ] }, { "cell_type": "code", "execution_count": 16, "id": "435a0d02-9262-4467-878f-191ffe679521", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Loading data from experiment_results/exp_20250319_160627_all_results.json...\n", "Loaded 80 valid results\n" ] } ], "source": [ "fn = \"experiment_results/exp_20250319_160627_all_results.json\"\n", "df = load_experiment_data(fn)" ] }, { "cell_type": "code", "execution_count": 17, "id": "2a140217-f62d-4cce-8786-7b5d44cf4aee", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
querysimilarity_scorediversity_levelbranching_factormax_depthepsilonexperiment_idpath_length
0What are the environmental impacts of electric...0.6low210.2exp_20250319_1606271
1What are the environmental impacts of electric...0.5low210.2exp_20250319_1606271
2What are the environmental impacts of electric...0.5low210.2exp_20250319_1606271
3What are the environmental impacts of electric...0.6low210.2exp_20250319_1606271
4What are the environmental impacts of electric...0.3low210.2exp_20250319_1606271
...........................
75What are the environmental impacts of electric...0.3medium420.2exp_20250319_1606272
76What are the environmental impacts of electric...0.4medium420.2exp_20250319_1606272
77What are the environmental impacts of electric...0.6medium420.2exp_20250319_1606272
78What are the environmental impacts of electric...0.5medium420.2exp_20250319_1606272
79What are the environmental impacts of electric...0.3medium420.2exp_20250319_1606272
\n", "

80 rows × 8 columns

\n", "
" ], "text/plain": [ " query similarity_score \\\n", "0 What are the environmental impacts of electric... 0.6 \n", "1 What are the environmental impacts of electric... 0.5 \n", "2 What are the environmental impacts of electric... 0.5 \n", "3 What are the environmental impacts of electric... 0.6 \n", "4 What are the environmental impacts of electric... 0.3 \n", ".. ... ... \n", "75 What are the environmental impacts of electric... 0.3 \n", "76 What are the environmental impacts of electric... 0.4 \n", "77 What are the environmental impacts of electric... 0.6 \n", "78 What are the environmental impacts of electric... 0.5 \n", "79 What are the environmental impacts of electric... 0.3 \n", "\n", " diversity_level branching_factor max_depth epsilon experiment_id \\\n", "0 low 2 1 0.2 exp_20250319_160627 \n", "1 low 2 1 0.2 exp_20250319_160627 \n", "2 low 2 1 0.2 exp_20250319_160627 \n", "3 low 2 1 0.2 exp_20250319_160627 \n", "4 low 2 1 0.2 exp_20250319_160627 \n", ".. ... ... ... ... ... \n", "75 medium 4 2 0.2 exp_20250319_160627 \n", "76 medium 4 2 0.2 exp_20250319_160627 \n", "77 medium 4 2 0.2 exp_20250319_160627 \n", "78 medium 4 2 0.2 exp_20250319_160627 \n", "79 medium 4 2 0.2 exp_20250319_160627 \n", "\n", " path_length \n", "0 1 \n", "1 1 \n", "2 1 \n", "3 1 \n", "4 1 \n", ".. ... \n", "75 2 \n", "76 2 \n", "77 2 \n", "78 2 \n", "79 2 \n", "\n", "[80 rows x 8 columns]" ] }, "execution_count": 17, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df" ] }, { "cell_type": "code", "execution_count": 22, "id": "64b8ccc4-3db0-4cfa-a64b-1bef41e63601", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 22, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAA/AAAAPdCAYAAAAtbYPmAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjEsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvc2/+5QAAAAlwSFlzAAAPYQAAD2EBqD+naQAAcxVJREFUeJzt3QmcVWX9P/DvsMuuoqBEkksJKaAgiEtoorikaGlEJkRG/UzUJC3JBLfEFbEkSRP3BXdbDE2E3EgUXNBcUlNIWV1A0UDh/l/P+b9mmpEBh2GYmTPzfr9e5zX3nHvOuc+9c+99zuc+z3lOUaFQKAQAAABQqzWo6QIAAAAAn0+ABwAAgBwQ4AEAACAHBHgAAADIAQEeAAAAckCABwAAgBwQ4AEAACAHBHgAAADIAQEeAAAAckCAh1KuvfbaaNu2bbU81ve///04/PDDIw/22Wef7LVZm+nTp0dRUVG8//77URtV5LXu3LlzjB8/PqrTPffcE9tvv300bNgwfvrTn1brYwNUJ/Vr+dSvG4f6lbpMgIeN7I033sgq32eeeaamixLz58+P7373u/HlL385GjRooFIr5cknn4wf/ehH1fqYP/7xj+PII4+MefPmxTnnnFOn3mvleffdd+OEE06Ir3zlK7HJJpvEF7/4xTjxxBNj6dKlNV00IIdq03ee+nXt1K/Vq1AoxEEHHZSVN/2QQd0jwEM9smLFithiiy3iV7/6VXTv3r1Gy7Jq1apYvXp11BbpdWnevHm1Pd6HH34YixYtigEDBsTWW28drVq1itrkk08+qfJ9vv3229l08cUXx/PPP5+1Ok2ZMiWOPfbYKn8sgOqkfl079evGr19LS70dUnin7hLgqbYuYqnlLf0ivemmm0b79u3jqquuiuXLl8ewYcOyL9fU1emvf/1rmQooHdh/6UtfylrrUqvdZZddVnL/f//73/jqV79a5lfd1157LdvXpEmTKlSuFCBSK2CqWI444oh455131ljn3nvvjV133TWaNWsW2267bZx11lnx6aefltyfviSvuOKK7NfOVM60zh133FFyfyp/sssuu2TrpteitBRmttpqq9h8883j+OOP36hf7KkbW3oNhwwZEm3atKny/T/22GPRrVu37LXafffds5D22e6Tf/zjH6Nr167RtGnTmDt3bvbL/P777x/t2rXLytSvX7+YPXt2mf2m1+0Pf/hD9j9K/6sddtgh209pL7zwQnzjG9+I1q1bZ++BvffeO3s/VPS1/mwXv4o8ZppPy9Pz3XfffeO6666rUFfH1CWy+IDi61//erZNWpbef4MHD46OHTtmj7nzzjvHLbfcUmbbdFB24YUXZp+X9Bqm9++vf/3rdb7X0jZnn312fOELX8i26dGjRxacP9uyMHny5Oz1T8/npptuiqq20047xZ133hmHHnpobLfddtlzT2X/05/+VOYzBVSc+lX9qn79n/pavxZLPQQuueSSCn9OyakCVIN+/foVWrVqVTjnnHMKr7zySva3YcOGhYMOOqhw5ZVXZsuOO+64wuabb15Yvnx5ts3KlSsLo0ePLjz55JOF119/vXDjjTcWmjdvXpg8eXLJfp9++ulCkyZNCvfcc0/h008/Ley+++6FI444okJl+sc//lFo0KBB4YILLii8/PLLhcsuu6zQtm3bQps2bUrWefjhhwutW7cuXHvttYXXXnut8MADDxQ6d+5cOPPMM0vWSR+jVO6rrroq28+vfvWr7Ln985//zO6fOXNmts6DDz5YmD9/fuGdd97Jlg8dOjTb9//93/8VXnzxxcKf/vSn7Pml12NtUnlatGixzim9ThX9n5x00kkVXveaa65Z6/3Tpk3LnmOXLl2y1+i5554rfOMb38heq/R/TNL2jRs3Luyxxx6Fxx57rPDSSy9l/+upU6cWbrjhhuw1SK/ZscceW2jfvn1h2bJlZV7jL3zhC4Wbb7658K9//atw4oknFlq2bFnyWv7nP/8pbLbZZoVvfvOb2fsl/R8mTZqUPUZFX+ttttmmcOmll1b4MdN7Mj2fU045JXucW265pdCxY8dsu/fee2+dr+eKFSuyMqZ177zzzux9kZal53HRRRdl7+v0fvvNb36TvZeeeOKJkm1//vOfFzbddNPsPfnqq68WHnnkkey9t6732rhx47Lnn8qYypr2kcqePnfJv//972y79P9K5UnP7e233y637AceeOA6339du3YtrI9U9nbt2q3XNsD/qF/Vr+rX/6nP9Wv6n6f3SfrMFr/Od9999zq3IZ8EeKpFqqD22muvkvl0MJC+jI455piSZekLMX3ZzJgxY637Of744wvf+ta3yiy78MILswAwYsSIwlZbbVVYsmRJhco0ePDgwsEHH1xm2aBBg8ocYOy3336F8847r8w6qTJMj1MslTlVXKX16dMnO2Aq/eWdKo3SUqWXKrX0WhQ76qijsjKszUcffZRVduuaSlfM1X2Aceutt5YsS5XbJptsUnJAmLZP6zzzzDPrfKxVq1ZlB6PpIKBY2i4duBX78MMPs2V//etfs/lRo0YVvvSlL5UczHxWRV7r8g4w1vWYv/jFLwo77bRTmcc5/fTTK3SAkaR10rrptVuXQw45pPCzn/0su53+t02bNi05oPistb3Xtt5668Kvf/3rMst22223wk9+8pMy240fP/5zy50Ogtb1/nvjjTcKFbV48eLCF7/4xcIvf/nLCm8DlKV+Vb+qX8uqr/Xrj370o+xHmmICfN3VqKZ7AFB/pK5fxdKooKmbVerCVCx1+0vSeUvFJkyYkHUDSl3BPv7441i5cmXWPam0n/3sZ9kgHZdffnnWRTDttyJefPHFrPtWaX379i3T9enZZ5/Nuq0Vd6Eq7nqYuhd+9NFHJed0pe0+u5+KDHSSuiim16JY6n42Z86cta6fuhCmrl21VenXYbPNNsu6ZabXuViTJk3KvA+ShQsXZucMpi5u6X+fXt/02qb/eWmlt2vRokXWla/4vZJe69Slr3HjxlX2Wn/eY7788sux2267lVm/d+/esSHScz/vvPPitttui7feeit7v6fzKovfZ+m1TPP77bdfhfe5bNmy7LzzPffcs8zyNJ/e36X16tXrc/eXuh9WhVSuQw45JOvueeaZZ1bJPqG+Ur+uSf2qfq1P9Ws65eChhx6Kp59+utL7ID8EeKrNZ7/80zlBpZcVD7hRPPDKrbfeGqecckp2Lk+quNI5TRdddFE88cQTZfaTvvBfeeWVrPL417/+FQceeGCVDoSSzsn75je/ucZ96TymjfGarGvgmUceeSQ7F3Bdfv/738fRRx8dtVE6QPrswCpDhw7Nzk1L5w5us8022Tlk6f+dKteKvlZpv1X9Wld2mw2R3t/pdUjnCqaD73RQk85rLX4tKvI8N0R6vM+T3n/pfbg26X+Yzpdclw8++CD7nKbP9N13373OA0Pg86lf16R+Vb/Wp/o1hfc0LsFnL9X4rW99K/sBJv2IQ90hwFNrpV/m99hjj/jJT35Ssuyzg6YkP/jBD7Iv4zQgz/Dhw6N///7RpUuXz91/WuezByv/+Mc/ysynwXXSL8Gf96t82i4NXFN6Pg10UvyrePGvvxsq/YL7eS0PxS0tNSE97zToS/Lee+9lB36f979I/+ff/e53cfDBB2fz6ZIvS5YsWa/HTb/kpwFu0qA51RUGU+vHfffdV2ZZGjBoQ6TXYuDAgfG9730vm08HM+k1TK3USRrQJx1kTJ06NX74wx+usX1577XUqpFG4U37TgPolH6syrRopIGHUmvd2nze659aLNLIwOlAMrUYVMWBOrB+1K9rUr+WT/2aj/r1tNNOW6Pc6bN76aWXZgPHUrcI8NRa6cv0+uuvj/vvvz8b/fOGG27IvsCLRwIt7gI4Y8aMeO6556JTp07xl7/8Jft1PFV0xV+2a5OuP526OaWRU9OXenqc0t37ktGjR2cjr6ZKM11PNF3bNXWLSqO/nnvuuSXr3X777Vnlv9dee2Wji86cOTOuvvrq7L4tt9wyqxTSvtMopSmwVHaE2qro4ld8gJJaPxYvXpzNp9equBLbEGkk1tTFMh3knH766dnIt4cffvjn/p/T/za9fincnXrqqev9S/iIESPit7/9bXznO9+JUaNGZa9veg+kCjQdCGysa8yOGzcufvGLX2QHt+l1TCMBJ5W9fEt6LdIIy48//ng2mnTaf+oCWfy/Se+d9Hg///nPs/9Zev+m/2H6RT6VYW3vtfSajhkzJhv5PXWRveaaa7LyVmYk3A3p4pf+vwcccEDWhfPGG2/M5tNUfJmh0l0wgY1H/bom9Wv51K/5qF87dOiQTZ+VPl+lP9fUDS4jR62VvsBT17pBgwZFnz59sm5gpVsLXnrppeyLM/26nA4uknQ7/bp8xhlnfO7+02VY0qV2UpeqdM3WBx54IDtXrLTUUvjnP/85uy+dj5W2Sb9mpm5MpaVugKlLYvqlOh0UpUuTFFcKjRo1it/85jdZ17v0S206mKlJqeUiTbNmzYqbb745u1386/yGOv/88+Okk06Knj17xoIFC7LLg33egV46EEutCak15phjjskO/FJFuT7SQU3qPpYOmtKv4Onx0/92Y7YWpAoxHQzcdddd2f89XeooHVQlqXW5MtL7L70O6X2XLlGTKuPPHqCl93Y6LzUd/KbWl/T5KD5vcG3vtfSajhw5Mtsu/SKfDkCKL9FTndLli1KrXDo3Mh0op/Mki6fUMgRUD/XrxqF+rRrqV1i3ojSS3eesA6xD+jU4ncf7eb+E51mq7L7//e9nE2uXBmOaOHGiMApQBdSvFFO/wv/oQg9QSalFKrUcpRaKdM5bGiQndTcEACpP/Qprpws9dVYazbNly5blTulSIrCh0qjMqRtd6s55zjnnZF3oii+J5v0H1FW+39jY1K+wdrrQU2el63yubTTPdA3VNFExafCYNDjLZ68RzNp5/wF1le+3qqN+XX/ef9R3AjwAAADkgC70AAAAkAP1LsCnDgfpWpg6HgBA1VG/AsDGV+8C/AcffBBt2rTJ/gIAVUP9CgAbX70L8AAAAJBHAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOVArAvyECROic+fO0axZs+jTp0/MnDlzneu///77cfzxx8dWW20VTZs2jS9/+ctx3333VVt5AQAAoLo1iho2efLkGDlyZEycODEL7+PHj48BAwbEyy+/HFtuueUa669cuTL233//7L477rgjOnbsGG+++Wa0bdu2RsoPAAAA1aGoUCgUogal0L7bbrvF5Zdfns2vXr06OnXqFCeccEKcdtppa6yfgv5FF10UL730UjRu3Hi9H2/ZsmXRpk2bWLp0abRu3bpKngMA1HfqVwCo413oU2v6rFmzon///v8rUIMG2fyMGTPK3eaPf/xj9O3bN+tC3759+9hpp53ivPPOi1WrVpW7/ooVK7KDitITALBh1K8AUM8C/JIlS7LgnYJ4aWl+wYIF5W7z+uuvZ13n03bpvPczzjgjLrnkkjj33HPLXX/s2LFZi0DxlFr3AYANo34FgHo6iN36SF3s0/nvV155ZfTs2TMGDRoUp59+eta1vjyjRo3KuvMVT/Pmzav2MgNAXaN+BYB6Nohdu3btomHDhrFw4cIyy9N8hw4dyt0mjTyfzn1P2xXr0qVL1mKfuuQ3adKkzPpplPo0AQBVR/0KAPWsBT6F7dSKPnXq1DIt7Gk+nedenj333DNeffXVbL1ir7zyShbsPxveAQAAoK6o8S706RJyV111VVx33XXx4osvxnHHHRfLly+PYcOGZfcPGTIk66ZXLN3/7rvvxkknnZQF97/85S/ZIHZpUDsAAACoq2r8OvDpHPbFixfH6NGjs27wPXr0iClTppQMbDd37txsZPpiaZCc+++/P04++eTo1q1bdh34FOZ/8Ytf1OCzAAAAgDp+Hfjq5jq1AFD11K8AUA+60AMAAACfT4AHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAca1XQBAAAA6rNCoRDLly8vmW/RokUUFRXVaJmonQR4AACAGpTC+8CBA0vm77333mjZsmWNlonaSRd6AAAAyAEBHgAAAHJAgAcAAIAcEOABAAAgBwR4AAAAyAGj0AOwXlzqBgCgZgjwAKwXl7oBAKgZutADAABADgjwAAAAkAMCPAAAAOSAAA8AAAA5IMADAABADgjwAAAAkAMCPAAAAOSAAA8AAAA5IMADAABADgjwAAAAkAMCPAAAAOSAAA8AAAA5IMADAABADgjwAAAAkAMCPAAAAOSAAA8AAAA5IMADAABADgjwAAAAkAMCPAAAAOSAAA8AAAA50KimCwBQ1Xqeen1NF6FOK/p0ZbQpNb/PGbdGoVGTGixR3TbroiE1XQQAoJbQAg8AAAA5IMADAABADgjwAAAAkAMCPAAAAOSAQezIlUKhEMuXLy+Zb9GiRRQVFdVomQAAAKqDAE+upPA+cODAkvl77703WrZsWaNlAgAAqA660AMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQA7UiwE+YMCE6d+4czZo1iz59+sTMmTPXuu61114bRUVFZaa0HQAAANRlNR7gJ0+eHCNHjowxY8bE7Nmzo3v37jFgwIBYtGjRWrdp3bp1zJ8/v2R68803q7XMAAB1SaFQiA8//LBkSvMA1D6NaroA48aNi+HDh8ewYcOy+YkTJ8Zf/vKXmDRpUpx22mnlbpNa3Tt06FCh/a9YsSKbii1btqyKSg4A9Zf6tW5Zvnx5DBw4sGT+3nvvjZYtW9ZomQCoZS3wK1eujFmzZkX//v3/V6AGDbL5GTNmrHW79MvwNttsE506dcoqmxdeeGGt644dOzbatGlTMqVtAIANo34FgHoW4JcsWRKrVq2K9u3bl1me5hcsWFDuNl/5yley1vn0y/CNN94Yq1evjj322CP+85//lLv+qFGjYunSpSXTvHnzNspzAYD6RP0KAPWwC/366tu3bzYVS+G9S5cu8fvf/z7OOeecNdZv2rRpNgFQNQoNG8fSboPLzFP/qF8BoJ4F+Hbt2kXDhg1j4cKFZZan+Yqe4964cePYZZdd4tVXX91IpQSgjKKiKDRqUtOlAACod2q0C32TJk2iZ8+eMXXq1JJlqUt8mi/dyr4uqQv+nDlzYqutttqIJQUAAIB63oU+XUJu6NCh0atXr+jdu3eMHz8+Gwm1eFT6IUOGRMeOHbPBcpKzzz47dt9999h+++3j/fffj4suuii7jNwPf/jDGn4mAAAAUIcD/KBBg2Lx4sUxevTobOC6Hj16xJQpU0oGtps7d242Mn2x9957L7vsXFp30003zVrwH3/88ejatWsNPgsAAACo4wE+GTFiRDaVZ/r06WXmL7300mwCAACqR89Tr6/pItRpRZ+ujDal5vc541bjzWxksy4aEnlUo+fAAwAAABUjwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAONKrpAtQ1PU+9vqaLUKcVfboy2pSa3+eMW6PQqEkNlqjum3XRkJouAlDPqVs3PvVr9VK3ApWlBR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMiBRjVdAAAAalahYeNY2m1wmXmg+vgMkqsW+AkTJkTnzp2jWbNm0adPn5g5c2aFtrv11lujqKgoDj/88I1eRgCAOquoKAqNmpRMaR6oRj6D5CXAT548OUaOHBljxoyJ2bNnR/fu3WPAgAGxaNGidW73xhtvxCmnnBJ77713tZUVAAAA6m2AHzduXAwfPjyGDRsWXbt2jYkTJ0bz5s1j0qRJa91m1apVcfTRR8dZZ50V2267bbWWFwAAAOpdgF+5cmXMmjUr+vfv/78CNWiQzc+YMWOt25199tmx5ZZbxrHHHvu5j7FixYpYtmxZmQkA2DDqVwCoZwF+yZIlWWt6+/btyyxP8wsWLCh3m0cffTSuvvrquOqqqyr0GGPHjo02bdqUTJ06daqSsgNAfaZ+BYB62IV+fXzwwQdxzDHHZOG9Xbt2Fdpm1KhRsXTp0pJp3rx5G72cAFDXqV8BoJ5dRi6F8IYNG8bChQvLLE/zHTp0WGP91157LRu87tBDDy1Ztnr16uxvo0aN4uWXX47tttuuzDZNmzbNJgCg6qhfASBHLfCPPPJIfO9734u+ffvGW2+9lS274YYbsi7uFdWkSZPo2bNnTJ06tUwgT/Npv5+14447xpw5c+KZZ54pmQ477LDYd999s9u67wEAAFBXVSrA33nnndml3jbZZJN4+umns4FsktSF7rzzzluvfaVLyKUu8dddd128+OKLcdxxx8Xy5cuzUemTIUOGZN30knSd+J122qnM1LZt22jVqlV2O/0gAAAAAHVRpQL8ueeem13uLQXvxo0blyzfc889s2u5r49BgwbFxRdfHKNHj44ePXpkLelTpkwpGdhu7ty5MX/+/MoUEwAAAOr3OfDpXPOvfe1rayxPo9C+//77672/ESNGZFN5pk+fvs5tr7322vV+PAAAAKgXLfBpgLlXX311jeXp/Pdtt922KsoFAAAAbGgL/PDhw+Okk06KSZMmRVFRUbz99tsxY8aMOOWUU+KMM86ozC6hQgoNG8fSboPLzAMAANQHlQrwp512WjZa/H777RcfffRR1p0+XUomBfgTTjih6ksJxYqKotDIYIUAAED9s94BftWqVfHYY4/F8ccfH6eeemrWlf7DDz+Mrl27RsuWLTdOKQEAAKCeW+8A37BhwzjggAOyS76lS7il4A4AAADUwkHs0jXXX3/99aovDQAAAFC114FP57v/+c9/zq7RvmzZsjITAAAAUAsGsTv44IOzv4cddlg2Cn2xQqGQzafz5AEAAIAaDvDTpk2rwiIAAAAAGyXA9+vXrzKbAQAAANUZ4JP3338/rr766mw0+uSrX/1q/OAHP4g2bdpUdpcAAABAVQ5i99RTT8V2220Xl156abz77rvZNG7cuGzZ7NmzK7NLAAAAoKpb4E8++eRsALurrroqGjX6/7v49NNP44c//GH89Kc/jYcffrgyuwUAAACqMsCnFvjS4T3bUaNG8fOf/zx69epVmV0CAAAAVd2FvnXr1jF37tw1ls+bNy9atWpVmV0CAAAAVR3gBw0aFMcee2xMnjw5C+1puvXWW7Mu9IMHD67MLgEAAICq7kJ/8cUXR1FRUQwZMiQ79z1p3LhxHHfccXH++edXZpcAAABAVQf4Jk2axGWXXRZjx46N1157LVuWRqBv3rx5ZXYHAAAAbIwAv3Tp0li1alVsttlmsfPOO5csT5eTS4PZpXPkAQAAgBo+B/473/lOds77Z912223ZfQAAAEAtCPBPPPFE7Lvvvmss32effbL7AAAAgFoQ4FesWFEyeF1pn3zySXz88cdVUS4AAABgQwN8796948orr1xj+cSJE6Nnz56V2SUAAABQ1YPYnXvuudG/f/949tlnY7/99suWTZ06NZ588sl44IEHKrNLAAAAoKpb4Pfcc8+YMWNGdOrUKRu47k9/+lNsv/328dxzz8Xee+9dmV0CAAAAVd0Cn/To0SNuuummym4OAAAAbOwW+NmzZ8ecOXNK5u+99944/PDD45e//GWsXLmyMrsEAAAAqjrA//jHP45XXnklu/3666/HoEGDonnz5nH77bfHz3/+88rsEgAAAKjqAJ/Ce+pCn6TQ3q9fv7j55pvj2muvjTvvvLMyuwQAAACqOsAXCoVYvXp1dvvBBx+Mgw8+OLudBrVbsmRJZXYJAAAAVHWA79WrV3YpuRtuuCH+/ve/xyGHHJIt//e//x3t27evzC4BAACAqg7w48ePzwayGzFiRJx++unZJeSSO+64I/bYY4/K7BIAAACo6svIdevWrcwo9MUuuuiiaNiwYcn8LbfcEocddli0aNGiMg8DAAAAbEgL/No0a9YsGjduXGa0+oULF1blQwAAAEC9VKUBvrzB7gAAAIBaHuABAACAqiHAAwAAQA4I8AAAAJADAjwAAADU9wC/zTbblBmVHgAAAKjGAD906NB4+OGHP3e9559/Pjp16lSZhwAAAAA2NMAvXbo0+vfvHzvssEOcd9558dZbb1VmNwAAAMDGDPD33HNPFtqPO+64mDx5cnTu3DkOOuiguOOOO+KTTz6pzC4BAACAjXEO/BZbbBEjR46MZ599Np544onYfvvt45hjjomtt946Tj755PjXv/5V2V0DAAAAVT2I3fz58+Nvf/tbNjVs2DAOPvjgmDNnTnTt2jUuvfTSDd09AAAAUNkAn7rJ33nnnfGNb3wjG2n+9ttvj5/+9Kfx9ttvx3XXXRcPPvhg3HbbbXH22WdXfYkBAACgHmpUmY222mqrWL16dQwePDhmzpwZPXr0WGOdfffdN9q2bVsVZQQAAIB6r1IBPnWNP+qoo6JZs2ZrXSeF93//+98bUjYAAABgQ7rQT5s2rdzR5pcvXx4/+MEPKrNLAAAAoKoDfDrP/eOPP15jeVp2/fXXV2aXAAAAQFV1oV+2bFkUCoVs+uCDD8p0oV+1alXcd999seWWW67PLgEAAICqDvDpvPaioqJs+vKXv7zG/Wn5WWedtT67BAAAAKo6wKdz31Pr+9e//vXsMnKbbbZZyX1NmjTJLim39dZbr88uAQAAgKoO8P369cv+ptHlv/jFL2Yt7gAAAEAtCvDPPfdc7LTTTtGgQYNYunRpzJkzZ63rduvWrarKBwAAAKxPgO/Ro0csWLAgG6Qu3U6t76k7/Wel5WlAOwAAAKAGAnzqNr/FFluU3AYAAABq4XXg0wB1qXX9k08+yUaaX716dbasvGl9TZgwITp37pxdlq5Pnz4xc+bMta571113Ra9evbIR8Vu0aJH1BrjhhhvW+zEBAACgTgb4Yo0bN85GoK8qkydPjpEjR8aYMWNi9uzZ0b179xgwYEAsWrSo3PXTyPenn356zJgxIzsvf9iwYdl0//33V1mZAAAAIPcBPjn88MPjnnvuqZICjBs3LoYPH56F8K5du8bEiROjefPmMWnSpHLX32effeKII46ILl26xHbbbRcnnXRSNmjeo48+Wu76K1asiGXLlpWZAIANo34FgFp+GbliO+ywQ5x99tnx2GOPRc+ePbOu7KWdeOKJFdrPypUrY9asWTFq1KiSZWmU+/79+2ct7J8nDaL30EMPxcsvvxwXXHBBueuMHTs26/IPAFQd9SsA5CTAX3311dk56Cl8p6m0dJ58RQP8kiVLshHr27dvX2Z5mn/ppZfWul26jF3Hjh2zX/8bNmwYv/vd72L//fcvd93040Dqol8stRB06tSpQuUDAMqnfgWAnAT4mh6FvlWrVvHMM8/Ehx9+GFOnTs0OILbddtuse/1nNW3aNJsAgKqjfgWAnAT4qtKuXbusBX3hwoVllqf5Dh06rHW71M1+++23z26nUehffPHFrCtfeQEeAAAA6nWA/89//hN//OMfY+7cudm57J8dmK4imjRpkp1Dn1rR08B4Sbo8XZofMWJEhcuStknd6QEAAKCuqlSATwH7sMMOy7qtp3PVd9ppp3jjjTeyQeV23XXX9dpX6v4+dOjQ7NruvXv3jvHjx8fy5cuzUemTIUOGZOe7pxb2JP1N66YR6FNov++++7LrwF9xxRWVeSoAAABQdwN8GrjmlFNOyUafTeejp+vCb7nllnH00UfHgQceuF77GjRoUCxevDhGjx4dCxYsyLrET5kypWRgu9TCn7rMF0vh/ic/+UnWA2CTTTaJHXfcMW688cZsPwAAAFBXVSrAp3POb7nllv+/g0aN4uOPP46WLVtml5YbOHBgHHfcceu1v9Rdfm1d5qdPn15m/txzz80mAAAAqE/+17S9HtJ134vPe99qq63itddeK3NpOAAAAKAWtMDvvvvu8eijj0aXLl3i4IMPjp/97GcxZ86cuOuuu7L7AAAAgFoQ4NMo8+ka7Ek6Dz7dnjx5cuywww4VHoEeAAAA2MgBPo0+X7o7/cSJEyuzGwAAAGBjngMPAAAA1NIW+E033TSKiooqtO677767IWUCAAAAKhvgx48fX9FVAQAAgJoK8EOHDq3qxwYAAACqOsAvW7YsWrduXXJ7XYrXAwAAAGrgHPj58+fHlltuGW3bti33fPhCoZAtX7VqVRUVDwAAAFivAP/QQw/FZpttlt2eNm2aVw8AAABqY4Dv169fubcBAACAWhTgP+u///1vPPfcc7Fo0aJYvXp1mfsOO+ywqigbAAAAsCEBfsqUKTFkyJBYsmTJGvc5Bx4AAACqXoPKbHTCCSfEUUcdlQ1ql1rfS0/COwAAANSSAL9w4cIYOXJktG/fvupLBAAAAFRNgD/yyCNj+vTpldkUAAAAqK5z4C+//PKsC/0jjzwSO++8czRu3LjM/SeeeGJldgsAAABUZYC/5ZZb4oEHHohmzZplLfFp4Lpi6bYADwAAALUgwJ9++ulx1llnxWmnnRYNGlSqFz4AAACwHiqVvleuXBmDBg0S3gEAAKCaVCqBDx06NCZPnlz1pQEAAACqrgt9utb7hRdeGPfff39069ZtjUHsxo0bV5ndAgAAAFUZ4OfMmRO77LJLdvv5558vc1/pAe0AAACAGgzw06ZNq6KHBwAAACrCKHQAAABQl1rgv/nNb8a1114brVu3zm6vy1133VUVZQMAAADWN8C3adOm5Pz2dBsAAACohQH+mmuuKbn9u9/9LlavXh0tWrTI5t9444245557okuXLjFgwICNU1IAAACoxyp1DvzAgQPjhhtuyG6///77sfvuu8cll1wShx9+eFxxxRVVXUYAAACo9yoV4GfPnh177713dvuOO+6I9u3bx5tvvhnXX399/OY3v6nqMgIAAEC9V6kA/9FHH0WrVq2y2w888EA2qF2DBg2ylvgU5AEAAIBaEOC333777Jz3efPmxf333x8HHHBAtnzRokXZKPUAAABALQjwo0ePjlNOOSU6d+4cffr0ib59+5a0xu+yyy5VXEQAAACgwqPQl3bkkUfGXnvtFfPnz4/u3buXLN9vv/3iiCOOqMryAQAAAJUN8EmHDh2yqbTevXtXRZkAAACAquhCDwAAAFQvAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAdqRYCfMGFCdO7cOZo1axZ9+vSJmTNnrnXdq666Kvbee+/YdNNNs6l///7rXB8AAADqghoP8JMnT46RI0fGmDFjYvbs2dG9e/cYMGBALFq0qNz1p0+fHoMHD45p06bFjBkzolOnTnHAAQfEW2+9Ve1lBwAAgHoT4MeNGxfDhw+PYcOGRdeuXWPixInRvHnzmDRpUrnr33TTTfGTn/wkevToETvuuGP84Q9/iNWrV8fUqVOrvewAAABQXRrV5IOvXLkyZs2aFaNGjSpZ1qBBg6xbfGpdr4iPPvooPvnkk9hss83KvX/FihXZVGzZsmVVUHIAqN/UrwBQz1rglyxZEqtWrYr27duXWZ7mFyxYUKF9/OIXv4itt946C/3lGTt2bLRp06ZkSl3uAYANo34FgHrYhX5DnH/++XHrrbfG3XffnQ2AV57Uur906dKSad68edVeTgCoa9SvAFDPutC3a9cuGjZsGAsXLiyzPM136NBhndtefPHFWYB/8MEHo1u3bmtdr2nTptkEAFQd9SsA1LMW+CZNmkTPnj3LDEBXPCBd375917rdhRdeGOecc05MmTIlevXqVU2lBQAAgHraAp+kS8gNHTo0C+K9e/eO8ePHx/Lly7NR6ZMhQ4ZEx44ds3PtkgsuuCBGjx4dN998c3bt+OJz5Vu2bJlNAAAAUBfVeIAfNGhQLF68OAvlKYyny8OllvXige3mzp2bjUxf7IorrshGrz/yyCPL7CddR/7MM8+s9vIDAABAvQjwyYgRI7KpPNOnTy8z/8Ybb1RTqQAAAKD2yPUo9AAAAFBfCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAO1IoAP2HChOjcuXM0a9Ys+vTpEzNnzlzrui+88EJ861vfytYvKiqK8ePHV2tZAQAAoF4G+MmTJ8fIkSNjzJgxMXv27OjevXsMGDAgFi1aVO76H330UWy77bZx/vnnR4cOHaq9vAAAAFAvA/y4ceNi+PDhMWzYsOjatWtMnDgxmjdvHpMmTSp3/d122y0uuuii+M53vhNNmzb93P2vWLEili1bVmYCADaM+hUA6lmAX7lyZcyaNSv69+//vwI1aJDNz5gxo0oeY+zYsdGmTZuSqVOnTlWyXwCoz9SvAFDPAvySJUti1apV0b59+zLL0/yCBQuq5DFGjRoVS5cuLZnmzZtXJfsFgPpM/QoA1a9R1HGpm31FutoDABWnfgWAetYC365du2jYsGEsXLiwzPI0b4A6AAAAqCUBvkmTJtGzZ8+YOnVqybLVq1dn83379q3JogEAAECtUuNd6NMl5IYOHRq9evWK3r17Z9d1X758eTYqfTJkyJDo2LFjNlhO8cB3//znP0tuv/XWW/HMM89Ey5YtY/vtt6/R5wIAAAB1NsAPGjQoFi9eHKNHj84GruvRo0dMmTKlZGC7uXPnZiPTF3v77bdjl112KZm/+OKLs6lfv34xffr0GnkOAAAAUOcDfDJixIhsKs9nQ3nnzp2jUChUU8kAAACgdqjRc+ABAACAihHgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByoFYE+AkTJkTnzp2jWbNm0adPn5g5c+Y617/99ttjxx13zNbfeeed47777qu2sgIAAEC9DPCTJ0+OkSNHxpgxY2L27NnRvXv3GDBgQCxatKjc9R9//PEYPHhwHHvssfH000/H4Ycfnk3PP/98tZcdAAAA6k2AHzduXAwfPjyGDRsWXbt2jYkTJ0bz5s1j0qRJ5a5/2WWXxYEHHhinnnpqdOnSJc4555zYdddd4/LLL6/2sgMAAEB1aRQ1aOXKlTFr1qwYNWpUybIGDRpE//79Y8aMGeVuk5anFvvSUov9PffcU+76K1asyKZiS5cuzf4uW7YsNoZVKz7eKPuFmrKxPisbk88hdcnG/gy2atUqioqK1nu76qxffaapa9StULc/h60qWbfW+gC/ZMmSWLVqVbRv377M8jT/0ksvlbvNggULyl0/LS/P2LFj46yzzlpjeadOnTao7FBftPnt/9V0EaBe29ifwRS8W7duvd7bqV+h8tStULc/h0srWbfW+gBfHVLrfukW+9WrV8e7774bm2+++Ub7VYSN/2tZOkCcN2/eRvtgAOvmc1h3pFaCylC/1j0+11CzfAbrjlaVrFtrfYBv165dNGzYMBYuXFhmeZrv0KFDuduk5euzftOmTbOptLZt225w2al56YvNlxvULJ/D+kv9Wnf5XEPN8hmk1g5i16RJk+jZs2dMnTq1zC/4ab5v377lbpOWl14/+dvf/rbW9QEAAKAuqPEu9Kn73dChQ6NXr17Ru3fvGD9+fCxfvjwblT4ZMmRIdOzYMTvXLjnppJOiX79+cckll8QhhxwSt956azz11FNx5ZVX1vAzAQAAgDoc4AcNGhSLFy+O0aNHZwPR9ejRI6ZMmVIyUN3cuXOzkemL7bHHHnHzzTfHr371q/jlL38ZO+ywQzYC/U477VSDz4LqlLpsjhkzZo2um0D18TmEusfnGmqWzyAVUVQoFAoVWhMAAACon+fAAwAAABUjwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPBQyrXXXhtt27atlsf6/ve/H4cffnjkQefOnWP69Om14nWrjH322Sd++tOfrnOdoqKiuOeee6I6XXnlldGpU6do0KBBjB8/vlofG6A6qV/Lp37dONSv1GUCPGxkb7zxRlZ5PfPMMzVdlLjrrrti//33jy222CJat24dffv2jfvvv7+mi1UrzJ8/Pw466KBqe7xly5bFiBEj4he/+EW89dZb8aMf/WiD95kOAtN77f3334/a+lk49thj40tf+lJssskmsd1228WYMWNi5cqVNV00IIfUr/mgfq1eK1asiB49etSazwZVT4CHeuThhx/ODjDuu+++mDVrVuy7775x6KGHxtNPP13tZaltoa1Dhw7RtGnTanu8uXPnxieffBKHHHJIbLXVVtG8efOoLQqFQnz66adVvt+XXnopVq9eHb///e/jhRdeiEsvvTQmTpwYv/zlL6v8sQCqk/p17dSvG79+Le3nP/95bL311hv1MahZAjzV1sXqhBNOyLpZbbrpptG+ffu46qqrYvny5TFs2LBo1apVbL/99vHXv/61ZJtVq1aVaa37yle+EpdddlnJ/f/973/jq1/9aplfVl977bVsX5MmTapQuVLXtC9+8YvZl/sRRxwR77zzzhrr3HvvvbHrrrtGs2bNYtttt42zzjqrzJdv+oXziiuuyH5dTuVM69xxxx0l96fyJ7vssku2bnotSrv44ouzCmbzzTeP448/Pqt0NpbUjSx9se+2226xww47xHnnnZf9/dOf/lQl+09d5NL+0ms1YMCAmDdvXsl9Z555ZvaL8B/+8IfsNUnrJFOmTIm99tor6yKYXoNvfOMb2f/xsy0sqXUjHRCl/1X37t1jxowZZR77sccey17bdH96j6XHf++990ruT8ExPffNNtssO5hI5VlbF7+KPmZ6D6cuesXvn3HjxlWoq2N63+28887Z7fR+SY+VHjM974EDB2afj5YtW2b/pwcffHCNX9ZTq0J63HRAlD43V199dbZ9KmuSnn/aZ+pGWrzNiSeeGFtuuWX2uqfX+8knn1yjZSF9/nr27Jnt99FHH42qduCBB8Y111wTBxxwQPa8DzvssDjllFOy1xmoHPWr+lX9+j/1tX4tlh7ngQceyN771GEFqAb9+vUrtGrVqnDOOecUXnnllexvw4YNCwcddFDhyiuvzJYdd9xxhc0337ywfPnybJuVK1cWRo8eXXjyyScLr7/+euHGG28sNG/evDB58uSS/T799NOFJk2aFO65557Cp59+Wth9990LRxxxRIXK9I9//KPQoEGDwgUXXFB4+eWXC5dddlmhbdu2hTZt2pSs8/DDDxdat25duPbaawuvvfZa4YEHHih07ty5cOaZZ5askz5GqdxXXXVVtp9f/epX2XP75z//md0/c+bMbJ0HH3ywMH/+/MI777yTLR86dGi27//7v/8rvPjii4U//elP2fNLr8fapPK0aNFinVN6nSpq1apVhU6dOhV++9vfrnO9bbbZpjBt2rS13n/NNdcUGjduXOjVq1fh8ccfLzz11FOF3r17F/bYY4+SdcaMGZOV78ADDyzMnj278Oyzz2bL77jjjsKdd95Z+Ne//pX9Pw899NDCzjvvnJUt+fe//529fjvuuGPhz3/+c/YaH3nkkVmZPvnkk2ydtF3Tpk2z99AzzzxTeP7557PntHjx4pL3X3qt0/8tvdeuu+66QlFRUfb/LJYe4+67767wYz766KPZ++eiiy7K7p8wYUJhs802K/P+WZuPPvooez+kx0jvj/S+SO/fVPaJEycW5syZk5UzvZeaNWtWePPNN0u2/fa3v539z+66667sPZn2c+utt2bbp9cx7TOVJ+3z/fffz7Y58cQTC1tvvXXhvvvuK7zwwgvZe2/TTTcteS+m/23arlu3btlr8uqrr5bc91ldu3Zd5/sv/X/Xx+mnn17o2bPnem0D/I/6Vf2qfv2f+ly/LliwoNCxY8fsc138Oqf/H3WPAE+1SF/we+21V8l8+jJMX0bHHHNMybL0hZi+bGbMmLHW/Rx//PGFb33rW2WWXXjhhYV27doVRowYUdhqq60KS5YsqVCZBg8eXDj44IPLLBs0aFCZCmK//fYrnHfeeWXWueGGG7LHKZbKnA4SSuvTp09W2SVr+xJNX/KpwkqvRbGjjjoqK8O6KqZUEa9rWrZsWaGi0sFVqmgWLly4wQcY6Tmmg7Zi6aApLXviiSdKDjDSQciiRYvW+VjpoCBtlyrZ0q/fH/7wh5J1UiWZlqXHKP5f7rnnnhV+/yW77bZb4Re/+MU6DzDW9Zjp/3TIIYeU2efRRx9doQOMJL0f0v7SY63LV7/61ZIDwHTgkLb529/+Vu66xQcK7733XsmyDz/8MHvdb7rpppJl6eA9HXCkz07p7dKB+ud544031vn++89//lOoqLR+OvBb10E1sG7qV/Wr+rWs+li/rl69Ogv46Qe8RICv2xrVdA8A6o9u3bqV3G7YsGHWnau4m1OSujUlixYtKlk2YcKErLteOp/p448/zs7rSt3ESvvZz36Wdc26/PLLs65Dab8V8eKLL2bdskpLg86kLmfFnn322azr2K9//esyXQ9T98KPPvqo5LyqtN1n91ORgUNSF8X0WhRLXf3mzJmz1vVTF8LUpasq3HzzzVl3xdSFMXX92lCNGjXKuqQV23HHHbPubul17t27d7Zsm222yQb4Ke1f//pXjB49Op544olYsmRJ1hUvSf/znXbaqdz3T3qdit8r6XHSa33UUUets3ylty/eR+n32udt89nHfPnll9d4/6Tn+ec//zkq68MPP8y6Hv7lL3/JBv1JXUnT+z69Fkl6nun90q9fvwrvM3UbTN1G99xzz5JljRs3zsqa/jel9erV63P3l/6HVSENLJS61Kf/2/Dhw6tkn1BfqV/XpH5Vv9an+vW3v/1tfPDBBzFq1KhK74P8cA481SZ9qZWWzgkqvSzNJ8UVzK233pqdH5vO00vn86Qv13Q+32cHZ0lf+K+88kr2xZsqq6qUvvBTJZweu3hKBwDpcYrPMavq16T4+ZfnkUceyc7dWtd00003fe7jptf2hz/8Ydx2223Rv3//qC4tWrRYY1ka5Ofdd9/NzndLBxlpSj77f17XeyUdeFX1a/15j7kxpPf73XffnZ07mf7X6f2WDsKLX4uKPM+q/v+Ud1C8rvdfRUYafvvtt7PzCffYY4/sUj/AhlG/rkn9qn6tT/XrQw89lI0jkM6xTz/4FP8YlX44GDp0aJU+F2qeFnhqrfTLfDrA/8lPflKyrPTgK8V+8IMfZF/C6UAkteSlCrNLly6fu/+0TnFlVuwf//hHmfk0uE76JfjzfpVP2w0ZMqTMfBpUJ2nSpElJy8KGSl/En9fyUNzSsja33HJL9pqlg4w0QmtVSb9mP/XUUyWtAel1S5dbWdf/Ig1qlNZLBxd77713tqwyg7ukX/KnTp2aHQxWlzToU+mBapLPzlfmPZ8GxilueUgHuGnwnGLpfZ4OcP7+97+Xe2BY3nstXaotLU/7Lv51P7UYpLJ+3rV7y5NGWF7XQFCfdxCUWt5TeE+D+aQB7dI1eoHqpX5dk/q1fOrXfNSvv/nNb+Lcc88t80N5Gmxw8uTJ0adPn/UuC7WbAE+tlUZbvf7667PrqKZRVW+44YbsS7F41NniLoDpF8fnnnsuGzU0dY06+uijswq++Mt2bdKooanbUxqpM41Mmh6ndPe+JHU9S6O2ppF0jzzyyCxspG5/zz//fJkvyttvvz2r/NPoo+kX+pkzZ2Yjlyap+1z60k37/sIXvpC1LLRp06ZSr8mGdvFL3frSL7FptOH0hb5gwYKS/Va2TKV/TU8jIadKJP36m67Buvvuu5cccJQnjeaaumSmVtjUhS51ZTvttNPW+7FTl7FU+aaD0f/7v//L/vfTpk3Luv21a9cuNob0XL/2ta9lI+OmVo7063fqYlrcklDZ93wamTftL+3njDPOKNMi0blz5+z/lw4Q0+ucRu598803s1ayb3/729kBRNoudTM8+OCDs/9r+tX+uOOOi1NPPTUbITi9ly+88MKsi2o6KF9fG9LFL4X3NJJx2kf63C1evLjkvjRyMVA91K9rUr+WT/2aj/o1PXZpqWzFPzKkzwZ1i6YPaq0f//jH8c1vfjMGDRqUVYbp1+TSrQXpmtLpS/N3v/tddnCRpNvpPK/0xfx5UuWXfplOlW36ok7dCH/1q1+VWSf9epm+rNN96fyztE26dvVnv2TTL9PpF/f0S3U6KEq/wnft2jW7L1W2qTJI175O1+VMBzM1JVXk6Zf8dDmdVKEXTyeddNIG7zudr5guv/Ld7343O3BLlUf65Xdd0gFbet3SNXPT+Xgnn3xyXHTRRev92F/+8pez/1E6+EsHNOkcyXTuYXrtN5b0HNM1zNMBRnr/pAPIVP4N6fqZ9pUOulLLWDrISO+/1EpVWrqkUjrYTZ+FdK5gahVLl4tKOnbsmL0X00FaailKB3nJ+eefH9/61rfimGOOyfb36quvZgfU6bGq09/+9rfssVNrTjqgKP0eBKqP+rXqqV+rjvoV1q0ojWT3OesA65B+kU3nVR1++OFRV6VfptO1VT97jV3KSpV9OvBN59cBsGHUrxRTv8L/6EIPUEmpe+j++++fDU6Tuvddd911WSsVAFB56ldYO13oqbPSaJ1rG8kzjUIKGyqdi5kOMNL5gam7X+rKmUYf/rzRZCsykjFAbaV+ZWNTv8La6UJPnZUGzErX+CxPGmwkTVTM+PHjsy6MqasfFZMGv1nbaLLp/LlWrVpVe5kAqoL6teqoX9ef+pX6ToAHAACAHKh3XejT7xXLli3L/gIAVUP9CgAbX70L8B988EF2Pc70FwCoGupXANj46l2ABwAAgDwS4AEAACAHBHgAAADIAQEeAAAAckCABwAAgBwQ4AEAACAHBHgAAADIAQEeAAAAckCABwAAgBwQ4AEAACAHBHgAAADIAQEeAAAAckCABwAAgBwQ4AEAACAHBHgAAADIAQEeAAAAckCABwAAgBwQ4AEAACAHGtV0AQAAAOqzQqEQy5cvL5lv0aJFFBUV1WiZqJ0EeAAAgBqUwvvAgQNL5u+9995o2bJljZaJ2kkXegAAAMgBAR4AAAByQIAHAACAHKgVAX7ChAnRuXPnaNasWfTp0ydmzpy5zvXff//9OP7442OrrbaKpk2bxpe//OW47777qq28AAAAUO8GsZs8eXKMHDkyJk6cmIX38ePHx4ABA+Lll1+OLbfcco31V65cGfvvv3923x133BEdO3aMN998M9q2bVsj5QcAAIB6EeDHjRsXw4cPj2HDhmXzKcj/5S9/iUmTJsVpp522xvpp+bvvvhuPP/54NG7cOFuWWu8BAACgLqvRLvSpNX3WrFnRv3///xWoQYNsfsaMGeVu88c//jH69u2bdaFv37597LTTTnHeeefFqlWryl1/xYoVsWzZsjITALBh1K8AUM8C/JIlS7LgnYJ4aWl+wYIF5W7z+uuvZ13n03bpvPczzjgjLrnkkjj33HPLXX/s2LHRpk2bkqlTp04b5bkAQH2ifgWAejqI3fpYvXp1dv77lVdeGT179oxBgwbF6aefnnW9L8+oUaNi6dKlJdO8efOqvcwAUNeoXwGgnp0D365du2jYsGEsXLiwzPI036FDh3K3SSPPp3Pf03bFunTpkrXYpy75TZo0KbN+GqU+TQBA1VG/AkA9a4FPYTu1ok+dOrVMC3uaT+e5l2fPPfeMV199NVuv2CuvvJIF+8+GdwAAAKgrarwLfbqE3FVXXRXXXXddvPjii3HcccfF8uXLS0alHzJkSNZNr1i6P41Cf9JJJ2XBPY1YnwaxS4PaAQAAQF1V45eRS+ewL168OEaPHp11g+/Ro0dMmTKlZGC7uXPnZiPTF0uD5Nx///1x8sknR7du3bLrwKcw/4tf/KIGnwUAAABsXEWFQqEQ9Ui6zE0aLTcNuNO6deuaLg4A1AnqV4DK+/DDD2PgwIEl8/fee2+0bNmyRstE7VTjXegBAACAzyfAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA40qukCAABQswqFQixfvrxkvkWLFlFUVFSjZQJgTQI8AEA9l8L7wIEDS+bvvffeaNmyZY2WCYA16UIPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOSDAAwAAQA4I8AAAAJADAjwAAADkgAAPAAAAOVArAvyECROic+fO0axZs+jTp0/MnDlzretee+21UVRUVGZK2wEAAEBdVuMBfvLkyTFy5MgYM2ZMzJ49O7p37x4DBgyIRYsWrXWb1q1bx/z580umN998s1rLDAAAAPUuwI8bNy6GDx8ew4YNi65du8bEiROjefPmMWnSpLVuk1rdO3ToUDK1b99+reuuWLEili1bVmYCADaM+hUAql+jqEErV66MWbNmxahRo0qWNWjQIPr37x8zZsxY63YffvhhbLPNNrF69erYdddd47zzzouvfvWr5a47duzYOOusszZK+QGgvlK/Qv3S89Tra7oIdVrRpyujTan5fc64NQqNmtRgieq+WRcNiTyq0Rb4JUuWxKpVq9ZoQU/zCxYsKHebr3zlK1nr/L333hs33nhjFuL32GOP+M9//lPu+unHgaVLl5ZM8+bN2yjPBQDqE/UrANSzFvjK6Nu3bzYVS+G9S5cu8fvf/z7OOeecNdZv2rRpNgEAVUf9CgD1rAW+Xbt20bBhw1i4cGGZ5Wk+ndteEY0bN45ddtklXn311Y1USgAAAKjnAb5JkybRs2fPmDp1asmy1CU+zZduZV+X1AV/zpw5sdVWW23EkgIAAEA970KfLiE3dOjQ6NWrV/Tu3TvGjx8fy5cvz0alT4YMGRIdO3bMBstJzj777Nh9991j++23j/fffz8uuuii7DJyP/zhD2v4mQAAAEAdDvCDBg2KxYsXx+jRo7OB63r06BFTpkwpGdhu7ty52cj0xd57773ssnNp3U033TRrwX/88cezS9ABAABAXVXjAT4ZMWJENpVn+vTpZeYvvfTSbAIAAID6pEbPgQcAAAAqRoAHAACAHBDgAQAAIAcEeAAAAMiBWjGIHVRUoVDILjNYrEWLFlFUVFSjZQIAAKgOAjy5ksL7wIEDS+bvvffeaNmyZY2WCQAAoDoI8ABArdbz1Otrugh1XtGnK6NNqfl9zrg1Co2a1GCJ6rZZFw2p6SIAOeUceAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMiBRjVdAADypVAoxPLly0vmW7RoEUVFRTVaJgCA+kCAB2C9pPA+cODAkvl77703WrZsWaNlAgCoD3ShBwAAgBwQ4AEAACAHdKEHAKjnCg0bx9Jug8vMA1D7CPAAAPVdUVEUGjWp6VIA8Dl0oQcAAIAcEOABAAAgBwR4AAAAyAEBHgAAAHJAgAcAAIAcEOABAAAgB2pFgJ8wYUJ07tw5mjVrFn369ImZM2dWaLtbb701ioqK4vDDD9/oZQQAAIB6HeAnT54cI0eOjDFjxsTs2bOje/fuMWDAgFi0aNE6t3vjjTfilFNOib333rvaygoAAAD1NsCPGzcuhg8fHsOGDYuuXbvGxIkTo3nz5jFp0qS1brNq1ao4+uij46yzzoptt922WssLAAAA9S7Ar1y5MmbNmhX9+/f/X4EaNMjmZ8yYsdbtzj777Nhyyy3j2GOP/dzHWLFiRSxbtqzMBABsGPUrAFS/RlGDlixZkrWmt2/fvszyNP/SSy+Vu82jjz4aV199dTzzzDMVeoyxY8dmLfXVpeep11fbY9VHRZ+ujDal5vc549YoNGpSgyWq+2ZdNKSmiwDUQtVdvwIAtaAL/fr44IMP4phjjomrrroq2rVrV6FtRo0aFUuXLi2Z5s2bt9HLCQB1nfoVAOpZC3wK4Q0bNoyFCxeWWZ7mO3TosMb6r732WjZ43aGHHlqybPXq1dnfRo0axcsvvxzbbbddmW2aNm2aTQBA1VG/AlSdQsPGsbTb4DLzUKUt8I888kh873vfi759+8Zbb72VLbvhhhuyLu4V1aRJk+jZs2dMnTq1TCBP82m/n7XjjjvGnDlzsu7zxdNhhx0W++67b3a7U6dOlX06AAAANaOoKDsttHhK81BlAf7OO+/MLvW2ySabxNNPP50NZJOkLnTnnXfeeu0rXUIudYm/7rrr4sUXX4zjjjsuli9fno1KnwwZMiTrppek68TvtNNOZaa2bdtGq1atstvpBwEAAACoiyoV4M8999zscm8peDdu/L/uHXvuuWd2Lff1MWjQoLj44otj9OjR0aNHj6wlfcqUKSUD282dOzfmz59fmWICAABA/T4HPp1r/rWvfW2N5W3atIn3339/vfc3YsSIbCrP9OnT17nttddeu96PBwAAAPWiBT4NMPfqq6+usTyd/77ttttWRbkAAACADQ3ww4cPj5NOOimeeOKJKCoqirfffjtuuummOOWUU7Jz2AEAAIBa0IX+tNNOy0aL32+//eKjjz7KutOnS8mkAH/CCSdUcREBAACA9Q7wq1atisceeyyOP/74OPXUU7Ou9B9++GF07do1WrZsuXFKCQAAAPXcegf4hg0bxgEHHJBd8i1dwi0FdwAAAKAWngOfrrn++uuvV31pAAAAgKq9Dnw63/3Pf/5zdo32ZcuWlZkAAACAWjCI3cEHH5z9Peyww7JR6IsVCoVsPp0nDwAAANRwgJ82bVoVFgEAAADYKAG+X79+ldkMAAAAqM4An7z//vtx9dVXZ6PRJ1/96lfjBz/4QbRp06ayuwQAAACqchC7p556Krbbbru49NJL4913382mcePGZctmz55dmV0CAAAAVd0Cf/LJJ2cD2F111VXRqNH/38Wnn34aP/zhD+OnP/1pPPzww5XZLQAAAFCVAT61wJcO79mOGjWKn//859GrV6/K7BIAAACo6i70rVu3jrlz566xfN68edGqVavK7BIAAACo6gA/aNCgOPbYY2Py5MlZaE/TrbfemnWhHzx4cGV2CQAAAFR1F/qLL744ioqKYsiQIdm570njxo3juOOOi/PPP78yuwQAAACqOsA3adIkLrvsshg7dmy89tpr2bI0An3z5s0rszsAAABgYwT4pUuXxqpVq2KzzTaLnXfeuWR5upxcGswunSMPAAAA1PA58N/5zneyc94/67bbbsvuAwAAAGpBC/wTTzwR48aNW2P5PvvsE6effnpVlAug0nqeen1NF6FOK/p0ZbQpNb/PGbdGoVGTGixR3TbroiE1XQQAIM8t8CtWrCgZvK60Tz75JD7++OOqKBcAAACwoQG+d+/eceWVV66xfOLEidGzZ8/K7BIAAACo6i705557bvTv3z+effbZ2G+//bJlU6dOjSeffDIeeOCByuwSAAAAqOoW+D333DNmzJgRnTp1ygau+9Of/hTbb799PPfcc7H33ntXZpcAAABAVbfAJz169IibbrqpspsDAAAAG7sFfvbs2TFnzpyS+XvvvTcOP/zw+OUvfxkrV66szC4BAACAqg7wP/7xj+OVV17Jbr/++usxaNCgaN68edx+++3x85//vDK7BAAAAKo6wKfwnrrQJym09+vXL26++ea49tpr484776zMLgEAAICqDvCFQiFWr16d3X7wwQfj4IMPzm6nQe2WLFlSmV0CAAAAVR3ge/XqlV1K7oYbboi///3vccghh2TL//3vf0f79u0rs0sAAACgqgP8+PHjs4HsRowYEaeffnp2CbnkjjvuiD322KMyuwQAAACq+jJy3bp1KzMKfbGLLrooGjZsWDJ/yy23xGGHHRYtWrSozMMAAAAAG9ICvzbNmjWLxo0blxmtfuHChVX5EAAAAFAvVWmAL2+wOwAAAKCWB3gAAACgagjwAAAAkAMCPAAAAOSAAA8AAAD1PcBvs802ZUalBwAAAKoxwA8dOjQefvjhz13v+eefj06dOlXmIQAAAIANDfBLly6N/v37xw477BDnnXdevPXWW5XZDQAAALAxA/w999yThfbjjjsuJk+eHJ07d46DDjoo7rjjjvjkk0+qvpQAAABQz1X6HPgtttgiRo4cGc8++2w88cQTsf3228cxxxwTW2+9dZx88snxr3/9q2pLCgAAAPXYBg9iN3/+/Pjb3/6WTQ0bNoyDDz445syZE127do1LL720akoJAAAA9VylAnzqJn/nnXfGN77xjWyk+dtvvz1++tOfxttvvx3XXXddPPjgg3HbbbfF2WefXfUlBgAAgHqoUWU22mqrrWL16tUxePDgmDlzZvTo0WONdfbdd99o27ZtVZQRAAAA6r1KBfjUNf6oo46KZs2arXWdFN7//e9/b0jZAAAAgA3pQj9t2rRyR5tfvnx5/OAHP6jMLgEAAICqDvDpPPePP/54jeVp2fXXX1+ZXUKFFBo2jqXdBpdMaR4AAKA+WK8u9MuWLYtCoZBNH3zwQZku9KtWrYr77rsvttxyy41RTvj/ioqi0KhJTZcCAACgdgf4dF57UVFRNn35y19e4/60/KyzzqrK8gEAAADrG+DTue+p9f3rX/96dhm5zTbbrOS+Jk2aZJeU23rrrTdGOQEAAKBeW68A369fv+xvGl3+i1/8YtbiDgAAANSiAP/cc8/FTjvtFA0aNIilS5fGnDlz1rput27dqqp8AAAAwPoE+B49esSCBQuyQerS7dT6nrrTf1Zanga0AwAAAGogwKdu81tssUXJbQAAAKAWXgc+DVCXWtc/+eSTbKT51atXZ8vKm9bXhAkTonPnztll6fr06RMzZ85c67p33XVX9OrVKxsRv0WLFllvgBtuuGG9HxMAAADqZIAv1rhx42wE+qoyefLkGDlyZIwZMyZmz54d3bt3jwEDBsSiRYvKXT+NfH/66afHjBkzsvPyhw0blk33339/lZUJAAAAch/gk8MPPzzuueeeKinAuHHjYvjw4VkI79q1a0ycODGaN28ekyZNKnf9ffbZJ4444ojo0qVLbLfddnHSSSdlg+Y9+uij5a6/YsWKWLZsWZkJANgw6lcAqOWXkSu2ww47xNlnnx2PPfZY9OzZM+vKXtqJJ55Yof2sXLkyZs2aFaNGjSpZlka579+/f9bC/nnSIHoPPfRQvPzyy3HBBReUu87YsWOzLv8AQNVRvwJATgL81VdfnZ2DnsJ3mkpL58lXNMAvWbIkG7G+ffv2ZZan+Zdeemmt26XL2HXs2DH79b9hw4bxu9/9Lvbff/9y100/DqQu+sVSC0GnTp0qVD4AoHzqVwDISYCv6VHoW7VqFc8880x8+OGHMXXq1OwAYtttt826139W06ZNswmAqlFo2DiWdhtcZp76R/0KADkJ8FWlXbt2WQv6woULyyxP8x06dFjrdqmb/fbbb5/dTqPQv/jii1lXvvICPABVrKgoCo2a1HQpAADqnUoH+P/85z/xxz/+MebOnZudy/7ZgekqokmTJtk59KkVPQ2Ml6TL06X5ESNGVLgsaZvUnR4AAADqqkoF+BSwDzvssKzbejpXfaeddoo33ngjG1Ru1113Xa99pe7vQ4cOza7t3rt37xg/fnwsX748G5U+GTJkSHa+e2phT9LftG4agT6F9vvuuy+7DvwVV1xRmacCAAAAdTfAp4FrTjnllGz02XQ+erou/JZbbhlHH310HHjggeu1r0GDBsXixYtj9OjRsWDBgqxL/JQpU0oGtkst/KnLfLEU7n/yk59kPQA22WST2HHHHePGG2/M9gMAAAB1VaUCfDrn/JZbbvn/O2jUKD7++ONo2bJldmm5gQMHxnHHHbde+0vd5dfWZX769Oll5s8999xsAgAAgPrkf03b6yFd9734vPetttoqXnvttTKXhgMAAABqQQv87rvvHo8++mh06dIlDj744PjZz34Wc+bMibvuuiu7DwAAAKgFAT6NMp+uwZ6k8+DT7cmTJ8cOO+xQ4RHoAQAAgI0c4NPo86W700+cOLEyuwEAAAA25jnwAAAAQC1tgd90002jqKioQuu+++67G1ImAAAAoLIBfvz48RVdFQAAAKipAD906NCqfmwAAACgqgP8smXLonXr1iW316V4PQAAAKAGzoGfP39+bLnlltG2bdtyz4cvFArZ8lWrVlVR8QAAAID1CvAPPfRQbLbZZtntadOmefUAAACgNgb4fv36lXsbAAAAqEUB/rP++9//xnPPPReLFi2K1atXl7nvsMMOq4qyAQAAABsS4KdMmRJDhgyJJUuWrHGfc+ABAACg6jWozEYnnHBCHHXUUdmgdqn1vfQkvAMAAEAtCfALFy6MkSNHRvv27au+RAAAAEDVBPgjjzwypk+fXplNAQAAgOo6B/7yyy/PutA/8sgjsfPOO0fjxo3L3H/iiSdWZrcAAABAVQb4W265JR544IFo1qxZ1hKfBq4rlm4L8AAAAFALAvzpp58eZ511Vpx22mnRoEGleuEDAAAA66FS6XvlypUxaNAg4R0AAACqSaUS+NChQ2Py5MlVXxoAAACg6rrQp2u9X3jhhXH//fdHt27d1hjEbty4cZXZLQAAAFCVAX7OnDmxyy67ZLeff/75MveVHtAOAAAAqMEAP23atCp6eAAAAKAijEIHAAAAdakF/pvf/GZce+210bp16+z2utx1111VUTYAAABgfQN8mzZtSs5vT7cBAACAWhjgr7nmmpLbv/vd72L16tXRokWLbP6NN96Ie+65J7p06RIDBgzYOCUFAACAeqxS58APHDgwbrjhhuz2+++/H7vvvntccsklcfjhh8cVV1xR1WUEAACAeq9SAX727Nmx9957Z7fvuOOOaN++fbz55ptx/fXXx29+85uqLiMAAADUe5UK8B999FG0atUqu/3AAw9kg9o1aNAga4lPQR4AAACoBQF+++23z855nzdvXtx///1xwAEHZMsXLVqUjVIPAAAA1IIAP3r06DjllFOic+fO0adPn+jbt29Ja/wuu+xSxUUEAAAAKjwKfWlHHnlk7LXXXjF//vzo3r17yfL99tsvjjjiiKosHwAAAFDZAJ906NAhm0rr3bt3VZQJAAAAqIou9AAAAED1EuABAAAgBwR4AAAAyAEBHgAAAHJAgAcAAIAcEOABAAAgBwR4AAAAyAEBHgAAAHJAgAcAAIAcEOABAAAgBwR4AAAAyAEBHgAAAHJAgAcAAIAcEOABAAAgBwR4AAAAyAEBHgAAAHJAgAcAAIAcEOABAAAgBwR4AAAAyAEBHgAAAHKgVgT4CRMmROfOnaNZs2bRp0+fmDlz5lrXveqqq2LvvfeOTTfdNJv69++/zvUBAACgLqjxAD958uQYOXJkjBkzJmbPnh3du3ePAQMGxKJFi8pdf/r06TF48OCYNm1azJgxIzp16hQHHHBAvPXWW9VedgAAAKg3AX7cuHExfPjwGDZsWHTt2jUmTpwYzZs3j0mTJpW7/k033RQ/+clPokePHrHjjjvGH/7wh1i9enVMnTq12ssOAAAA1aVR1KCVK1fGrFmzYtSoUSXLGjRokHWLT63rFfHRRx/FJ598Eptttlm5969YsSKbii1btqwKSg4A9Zv6FQDqWQv8kiVLYtWqVdG+ffsyy9P8ggULKrSPX/ziF7H11ltnob88Y8eOjTZt2pRMqcs9ALBh1K8AUA+70G+I888/P2699da4++67swHwypNa95cuXVoyzZs3r9rLCQB1jfoVAOpZF/p27dpFw4YNY+HChWWWp/kOHTqsc9uLL744C/APPvhgdOvWba3rNW3aNJsAgKqjfgWAetYC36RJk+jZs2eZAeiKB6Tr27fvWre78MIL45xzzokpU6ZEr169qqm0AAAAUE9b4JN0CbmhQ4dmQbx3794xfvz4WL58eTYqfTJkyJDo2LFjdq5dcsEFF8To0aPj5ptvzq4dX3yufMuWLbMJAAAA6qIaD/CDBg2KxYsXZ6E8hfF0ebjUsl48sN3cuXOzkemLXXHFFdno9UceeWSZ/aTryJ955pnVXn4AAACoFwE+GTFiRDaVZ/r06WXm33jjjWoqFQAAANQeuR6FHgAAAOoLAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMiBWhHgJ0yYEJ07d45mzZpFnz59YubMmWtd94UXXohvfetb2fpFRUUxfvz4ai0rAAAA1MsAP3ny5Bg5cmSMGTMmZs+eHd27d48BAwbEokWLyl3/o48+im233TbOP//86NChQ7WXFwAAAOplgB83blwMHz48hg0bFl27do2JEydG8+bNY9KkSeWuv9tuu8VFF10U3/nOd6Jp06afu/8VK1bEsmXLykwAwIZRvwJAPQvwK1eujFmzZkX//v3/V6AGDbL5GTNmVMljjB07Ntq0aVMyderUqUr2CwD1mfoVAOpZgF+yZEmsWrUq2rdvX2Z5ml+wYEGVPMaoUaNi6dKlJdO8efOqZL8AUJ+pXwGg+jWKOi51s69IV3sAoOLUrwBQz1rg27VrFw0bNoyFCxeWWZ7mDVAHAAAAtSTAN2nSJHr27BlTp04tWbZ69epsvm/fvjVZNAAAAKhVarwLfbqE3NChQ6NXr17Ru3fv7Lruy5cvz0alT4YMGRIdO3bMBsspHvjun//8Z8ntt956K5555plo2bJlbL/99jX6XAAAAKDOBvhBgwbF4sWLY/To0dnAdT169IgpU6aUDGw3d+7cbGT6Ym+//XbssssuJfMXX3xxNvXr1y+mT59eI88BAAAA6nyAT0aMGJFN5flsKO/cuXMUCoVqKhkAAADUDjV6DjwAAABQMQI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAO1IoAP2HChOjcuXM0a9Ys+vTpEzNnzlzn+rfffnvsuOOO2fo777xz3HfffdVWVgAAAKiXAX7y5MkxcuTIGDNmTMyePTu6d+8eAwYMiEWLFpW7/uOPPx6DBw+OY489Np5++uk4/PDDs+n555+v9rIDAABAvQnw48aNi+HDh8ewYcOia9euMXHixGjevHlMmjSp3PUvu+yyOPDAA+PUU0+NLl26xDnnnBO77rprXH755dVedgAAAKgujaIGrVy5MmbNmhWjRo0qWdagQYPo379/zJgxo9xt0vLUYl9aarG/5557yl1/xYoV2VRs6dKl2d9ly5bFxrBqxccbZb9QUzbWZ2Vj8jmkLtnYn8FWrVpFUVHRem9XnfWrzzR1jboV6vbnsFUl69ZaH+CXLFkSq1ativbt25dZnuZfeumlcrdZsGBBueun5eUZO3ZsnHXWWWss79Sp0waVHeqLNr/9v5ouAtRrG/szmIJ369at13s79StUnroV6vbncGkl69ZaH+CrQ2rdL91iv3r16nj33Xdj880332i/irDxfy1LB4jz5s3baB8MYN18DuuO1EpQGerXusfnGmqWz2Dd0aqSdWutD/Dt2rWLhg0bxsKFC8ssT/MdOnQod5u0fH3Wb9q0aTaV1rZt2w0uOzUvfbH5coOa5XNYf6lf6y6fa6hZPoPU2kHsmjRpEj179oypU6eW+QU/zfft27fcbdLy0usnf/vb39a6PgAAANQFNd6FPnW/Gzp0aPTq1St69+4d48ePj+XLl2ej0idDhgyJjh07ZufaJSeddFL069cvLrnkkjjkkEPi1ltvjaeeeiquvPLKGn4mAAAAUIcD/KBBg2Lx4sUxevTobCC6Hj16xJQpU0oGqps7d242Mn2xPfbYI26++eb41a9+Fb/85S9jhx12yEag32mnnWrwWVCdUpfNMWPGrNF1E6g+PodQ9/hcQ83yGaQiigqFQqFCawIAAAD18xx4AAAAoGIEeAAAAMgBAR4AAAByQICnVtlnn33ipz/9aU0XA1jPz2rnzp2zq4gAtY+6FfJB3UouRqEHIP+efPLJaNGiRU0XAwDqDHUr5RHgAdhgW2yxRU0XAQDqFHUr5dGFnlrrvffeiyFDhsSmm24azZs3j4MOOij+9a9/Zfelqx+mL7U77rijZP0ePXrEVlttVTL/6KOPZtfR/Oijj2qk/FBT3e9OOOGErAte+uy0b98+rrrqqli+fHkMGzYsWrVqFdtvv3389a9/Ldnm+eefzz5fLVu2zNY/5phjYsmSJSX3p23TZzHdnz5jl1xyyRqPW7qb3xtvvBFFRUXxzDPPlNz//vvvZ8umT5+ezae/af7++++PXXbZJTbZZJP4+te/HosWLcrK1qVLl2jdunV897vf9RmGKqRuhfWnbqU2EeCptb7//e/HU089FX/84x9jxowZ2YHFwQcfHJ988kn25fS1r32t5AsrHZC8+OKL8fHHH8dLL72ULfv73/8eu+22W3aAAvXJddddF+3atYuZM2dmBxzHHXdcHHXUUbHHHnvE7Nmz44ADDsgOJFLlnSr/VLmnij593qZMmRILFy6Mb3/72yX7O/XUU7PP07333hsPPPBA9rlL+6kKZ555Zlx++eXx+OOPx7x587LHTQcrN998c/zlL3/JHu+3v/1tlTwWoG6FylK3UmsUoBbp169f4aSTTiq88sorhfT2fOyxx0ruW7JkSWGTTTYp3Hbbbdn8b37zm8JXv/rV7PY999xT6NOnT2HgwIGFK664IlvWv3//wi9/+csaeiZQc5+hvfbaq2T+008/LbRo0aJwzDHHlCybP39+9vmaMWNG4ZxzzikccMABZfYxb9687P6XX3658MEHHxSaNGlS8rlL3nnnneyzmD6rxbbZZpvCpZdemt3+97//nW3/9NNPl9z/3nvvZcumTZuWzae/af7BBx8sWWfs2LHZstdee61k2Y9//OPCgAEDqvAVgvpH3QobRt1KbaIFnlop/eLfqFGj6NOnT8myzTffPL7yla9k9yX9+vWLf/7zn7F48eLsF8zUvSlN6RfM1JKQfnVM81DfdOvWreR2w4YNs8/OzjvvXLIsdeVLUpe6Z599NqZNm5Z14Suedtxxx+z+1157LZtWrlxZ5rO42WabZZ/Fqi5rKldq1dt2223LLEvlBDacuhUqT91KbWEQO3IrfWmmL7t0gJGmX//619GhQ4e44IILslE704FG6tYE9U3jxo3LzKdusaWXpflk9erV8eGHH8ahhx6afW4+K52T9+qrr6734zdo8P9/G05dc4ulz+PnlfWz5SxelsoJVA91K5RP3UptoQWeWikNsvHpp5/GE088UbLsnXfeiZdffjm6du1a8uWz9957Z+cOvfDCC7HXXntlvziuWLEifv/730evXr1cegM+x6677pp9ftJAOWkAntJT+vxst912WcVf+rOYzot95ZVXPnfU3Pnz55csKz3oDlAz1K1QPdStbEwCPLXSDjvsEAMHDozhw4dnI96mrkjf+973omPHjtnyYqkb3y233JKNkpu6J6VfJ9MAPDfddFPWDRBYt+OPPz7efffdGDx4cNa6lrr1pdFr06i6q1atyj5Xxx57bDbYzkMPPZSNqpsGwSpuCShPGvV29913j/PPPz/rlpta8X71q19V6/MC1qRuheqhbmVjEuCpta655pro2bNnfOMb34i+fftmXYbuu+++Mt2A0oFE+iIsfT5euv3ZZUD5tt5663jssceyz0waQTd1n02XyWnbtm3JgcRFF12Utcil7oD9+/fPWuTSZ3NdJk2alLX0pfXS/s4999xqekbAuqhbYeNTt7IxFaWR7DbqIwAAAAAbTAs8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI8AAAA5IAADwAAADkgwAMAAEAOCPAAAACQAwI81CH77LNP/PSnP81ud+7cOcaPH1+j5bn22mujbdu21facq8v3v//9OPzww6v1MQGoGerW6qFuhYoR4KGOevLJJ+NHP/pRjZZh0KBB8corr5TMn3nmmdGjR48aLRMAVJa6FahpjWq6AMDGscUWW2zU/RcKhVi1alU0arT2r5FNNtkkmwCgLlC3AjVNCzzk1PLly2PIkCHRsmXL2GqrreKSSy4pc3/pbn7f/e53s1/sS/vkk0+iXbt2cf3112fzq1evjrFjx8aXvvSl7MCge/fucccdd5SsP3369CgqKoq//vWv0bNnz2jatGk8+uij8eyzz8a+++4brVq1itatW2f3PfXUU2t080u3zzrrrGz9tJ80pWU/+MEP4hvf+MYaZdtyyy3j6quvXu/XZcWKFXHKKadEx44do0WLFtGnT5+s7MmyZcuy55aeQ2l33313Vv6PPvoom583b158+9vfzsq+2WabxcCBA+ONN95Y77IAkC/q1vKpW6H20AIPOXXqqafG3//+97j33nuzCvmXv/xlzJ49u9xudEcffXQcddRR8eGHH2YHJcn999+fVapHHHFENp8OMG688caYOHFi7LDDDvHwww/H9773vay1oV+/fiX7Ou200+Liiy+ObbfdNjbddNP42te+FrvssktcccUV0bBhw3jmmWeicePGa5QhHeQ8//zzMWXKlHjwwQezZW3atIkvf/nL2T7mz5+fHSwlf/7zn7OyffbAqCJGjBgR//znP+PWW2+NrbfeOjuAOPDAA2POnDnZ80oHNDfffHMcdNBBJdvcdNNN2Xl3zZs3zw5wBgwYEH379o1HHnkkawU599xzs30899xz0aRJk/UuEwD5oG4tn7oVapECkDsffPBBoUmTJoXbbrutZNk777xT2GSTTQonnXRSNr/NNtsULr300uz2J598UmjXrl3h+uuvL1l/8ODBhUGDBmW3//vf/xaaN29eePzxx8s8zrHHHputl0ybNq2QvjLuueeeMuu0atWqcO2115ZbzmuuuabQpk2bkvkxY8YUunfvvsZ6Xbt2LVxwwQUl84ceemjh+9//foVei379+pU85zfffLPQsGHDwltvvVVmnf32268watSo7Pbdd99daNmyZWH58uXZ/NKlSwvNmjUr/PWvf83mb7jhhsJXvvKVwurVq0u2X7FiRfba3n///dn80KFDCwMHDqxQ+QDIB3Xr/6hbofbShR5y6LXXXouVK1dmXdiKpe5oX/nKV8pdP/3SnbqtpV/Di7sIptaF1HqQvPrqq9mv8vvvv3/WilA8pS6A6bFK69WrV5n5kSNHxg9/+MPo379/nH/++WusXxFp+2uuuSa7vXDhwqwbXur+t75SS0A6dzC1PJR+Hqk1pbhcBx98cNaK8cc//jGbv/POO7Puian8SeqGmF6P1O2vePv02v73v/+t1HMDIB/UreVTt0Ltogs91BPpgCJ111u0aFH87W9/y85XS13XktT9L/nLX/6Snd9WWjofr7R07ltpafTbdB5g2jYdHIwZMybrYlfcfbAi0vmGqfvgjBkz4vHHH8/OFdx7773X+zmm55G6Gs6aNSv7W1px98bUTe/II4/Muvp95zvfyf6m7oTFAwalfaRzDYsPyKpz8CIA8kXdqm6F6ibAQw5tt9122S/dTzzxRHzxi1/Mlr333nvZZWVKn1NX2h577BGdOnWKyZMnZwcD6by94vPpunbtmh1MzJ07d63br0v6VT5NJ598cgwePDj7xb+8g4xUwadf8T9r8803z86TS9ulA41hw4ZFZaTzBdP+04HUug5S0gFXahF54YUX4qGHHsrOwyu26667Zq9ROvcxtR4AUD+oW8unboXaRYCHHEq/eB977LHZYDupgk4V4umnnx4NGqz7rJj0a34aSCcdjEybNq1keerSlkaXTQcJacTcvfbaK5YuXRqPPfZYVtEOHTq03P19/PHHWRnSr+7pl/3//Oc/2TVyv/Wtb5W7fhq999///nc2GM8XvvCF7HGLWyFSV780CE46SFjb432edKCTDiBSq0MaOTgddCxevDimTp0a3bp1i0MOOSRbLw3s06FDh2zdVO7S3SXTsosuuigbHffss8/Oyvnmm2/GXXfdFT//+c+zeQDqHnVr+dStULs4Bx5yKlWE6ZfwQw89NDvHLB0YpO5p65Iq0DSKbOrKt+eee5a575xzzokzzjgjGzG3S5cuWRfA1HUvVcJrk7rSvfPOO1mlnir4dC5gGoE2XdKmPOngI+03XRondZm75ZZbSu5LzyGNlJtGqU0j3FZWamlI5fnZz36WnbeYWh/SgU9xa0qSLrOTWjPSOXnF5yoWS6PlplGC0/rf/OY3s9ciHdCl8/S0GgDUberW8qlbofYoSiPZ1XQhANL5cengJx0kpModANgw6laoe3ShB2pU6la4ZMmSrFte27Zt47DDDqvpIgFArqlboe4S4IEalQb3SV0J0/lv1157bcmItcX3pUGA1iZ1WSzdfQ8AULdCXaYLPVBrffrpp/HGG2+s9f40cE/pgxIAYN3UrZBvAjwAAADkgFHoAQAAIAcEeAAAAMgBAR4AAAByQIAHAACAHBDgAQAAIAcEeAAAAMgBAR4AAACi9vt/fXLH3YpwsNIAAAAASUVORK5CYII=", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "sns.catplot(data=df, x='diversity_level', y='similarity_score', \n", " col='branching_factor', \n", " row = \"max_depth\",\n", " kind='bar') " ] }, { "cell_type": "code", "execution_count": 23, "id": "1fefb07c-d5a4-472c-845f-fb4909895b7e", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Loading data from experiment_results/exp_20250319_160627_all_results.json...\n", "Loaded 80 valid results\n" ] } ], "source": [ "def load_experiment_data(filename):\n", " \"\"\"Load experiment results from a JSON file.\"\"\"\n", " print(f\"Loading data from {filename}...\")\n", " \n", " with open(filename, 'r') as f:\n", " data = json.load(f)\n", " \n", " # Handle all_results file format\n", " if isinstance(data, list):\n", " results = data\n", " # Handle analysis file format\n", " elif 'summary' in data:\n", " experiment_id = data.get('experiment_id')\n", " all_results_file = os.path.join(os.path.dirname(filename), f\"{experiment_id}_all_results.json\")\n", " \n", " if os.path.exists(all_results_file):\n", " with open(all_results_file, 'r') as f:\n", " results = json.load(f)\n", " else:\n", " print(\"Error: Could not find all_results file\")\n", " return None\n", " else:\n", " print(\"Error: Unrecognized file format\")\n", " return None\n", " \n", " # Create DataFrame\n", " rows = []\n", " for result in results:\n", " # Skip results with errors\n", " if 'error' in result:\n", " continue\n", " \n", " row = {\n", " 'query': result.get('query', ''),\n", " 'similarity_score': result.get('similarity_score', 0),\n", " 'diversity_level': result.get('ibfs_config', {}).get('diversity_level', ''),\n", " 'branching_factor': result.get('ibfs_config', {}).get('branching_factor', 0),\n", " 'max_depth': result.get('ibfs_config', {}).get('max_depth', 0),\n", " 'epsilon': result.get('user_config', {}).get('epsilon', 0),\n", " 'experiment_id': result.get('experiment_id', ''), \n", " 'pref_answer':result.get('user_preferred_answer', ''), \n", " 'final_answer':result.get('final_answer', '')\n", " }\n", " \n", " # Add strategy path information if available\n", " if 'strategy_path' in result:\n", " row['path_length'] = len(result.get('strategy_path', []))\n", " \n", " rows.append(row)\n", " \n", " df = pd.DataFrame(rows)\n", " print(f\"Loaded {len(df)} valid results\")\n", " return df\n", "\n", "fn = \"experiment_results/exp_20250319_160627_all_results.json\"\n", "df = load_experiment_data(fn)" ] }, { "cell_type": "code", "execution_count": 28, "id": "e352fcd2-b49b-4a30-8d3c-dd1e14bfcb67", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "[nltk_data] Downloading package punkt to /Users/jashkina/nltk_data...\n", "[nltk_data] Package punkt is already up-to-date!\n", "/Users/jashkina/Documents/LocResearch/ibfs/ibfs_project/.venv/lib/python3.12/site-packages/nltk/translate/bleu_score.py:577: UserWarning: \n", "The hypothesis contains 0 counts of 4-gram overlaps.\n", "Therefore the BLEU score evaluates to 0, independently of\n", "how many N-gram overlaps of lower order it contains.\n", "Consider using lower n-gram order or use SmoothingFunction()\n", " warnings.warn(_msg)\n", "/Users/jashkina/Documents/LocResearch/ibfs/ibfs_project/.venv/lib/python3.12/site-packages/nltk/translate/bleu_score.py:577: UserWarning: \n", "The hypothesis contains 0 counts of 3-gram overlaps.\n", "Therefore the BLEU score evaluates to 0, independently of\n", "how many N-gram overlaps of lower order it contains.\n", "Consider using lower n-gram order or use SmoothingFunction()\n", " warnings.warn(_msg)\n" ] } ], "source": [ "import pandas as pd\n", "import numpy as np\n", "import json\n", "import os\n", "import nltk\n", "from sklearn.feature_extraction.text import TfidfVectorizer\n", "from sklearn.metrics.pairwise import cosine_similarity\n", "from nltk.translate.bleu_score import sentence_bleu\n", "from rouge import Rouge\n", "from rapidfuzz.distance import Levenshtein\n", "\n", "# Ensure NLTK data is available\n", "nltk.download('punkt')\n", "\n", "# Load experiment data function\n", "def load_experiment_data(filename):\n", " \"\"\"Load experiment results from a JSON file.\"\"\"\n", " \n", " with open(filename, 'r') as f:\n", " data = json.load(f)\n", " \n", " # Handle all_results file format\n", " if isinstance(data, list):\n", " results = data\n", " # Handle analysis file format\n", " elif 'summary' in data:\n", " experiment_id = data.get('experiment_id')\n", " all_results_file = os.path.join(os.path.dirname(filename), f\"{experiment_id}_all_results.json\")\n", " \n", " if os.path.exists(all_results_file):\n", " with open(all_results_file, 'r') as f:\n", " results = json.load(f)\n", " else:\n", " return None\n", " else:\n", " return None\n", " \n", " # Create DataFrame\n", " rows = []\n", " for result in results:\n", " if 'error' in result:\n", " continue\n", " \n", " row = {\n", " 'query': result.get('query', ''),\n", " 'similarity_score': result.get('similarity_score', 0),\n", " 'diversity_level': result.get('ibfs_config', {}).get('diversity_level', ''),\n", " 'branching_factor': result.get('ibfs_config', {}).get('branching_factor', 0),\n", " 'max_depth': result.get('ibfs_config', {}).get('max_depth', 0),\n", " 'epsilon': result.get('user_config', {}).get('epsilon', 0),\n", " 'experiment_id': result.get('experiment_id', ''), \n", " 'pref_answer': result.get('user_preferred_answer', ''), \n", " 'final_answer': result.get('final_answer', '')\n", " }\n", " \n", " if 'strategy_path' in result:\n", " row['path_length'] = len(result.get('strategy_path', []))\n", " \n", " rows.append(row)\n", " \n", " df = pd.DataFrame(rows)\n", " return df\n", "\n", "# Load dataset\n", "fn = \"experiment_results/exp_20250319_160627_all_results.json\"\n", "df = load_experiment_data(fn)\n", "\n", "# Ensure pref_answer and final_answer exist\n", "df = df.dropna(subset=['pref_answer', 'final_answer'])\n", "\n", "# Cosine similarity using TF-IDF\n", "vectorizer = TfidfVectorizer()\n", "tfidf_matrix = vectorizer.fit_transform(df['pref_answer'].tolist() + df['final_answer'].tolist())\n", "tfidf_pref = tfidf_matrix[:len(df)]\n", "tfidf_final = tfidf_matrix[len(df):]\n", "df['cosine_similarity'] = [cosine_similarity(tfidf_pref[i], tfidf_final[i])[0, 0] for i in range(len(df))]\n", "\n", "# Levenshtein Distance (normalized)\n", "df['levenshtein_distance'] = df.apply(lambda row: Levenshtein.distance(row['pref_answer'], row['final_answer']), axis=1)\n", "df['levenshtein_ratio'] = df.apply(lambda row: Levenshtein.normalized_similarity(row['pref_answer'], row['final_answer']), axis=1)\n", "\n", "# Jaccard Similarity\n", "def jaccard_similarity(str1, str2):\n", " set1, set2 = set(str1.split()), set(str2.split())\n", " intersection = len(set1 & set2)\n", " union = len(set1 | set2)\n", " return intersection / union if union != 0 else 0\n", "\n", "df['jaccard_similarity'] = df.apply(lambda row: jaccard_similarity(row['pref_answer'], row['final_answer']), axis=1)\n", "\n", "# BLEU Score\n", "df['bleu_score'] = df.apply(lambda row: sentence_bleu(\n", " [nltk.word_tokenize(row['pref_answer'])], \n", " nltk.word_tokenize(row['final_answer']),\n", " weights=(0.5, 0.5, 0, 0)), axis=1) # Bi-gram BLEU score\n", "\n", "# ROUGE Score\n", "rouge = Rouge()\n", "df['rouge_score'] = df.apply(lambda row: rouge.get_scores(row['final_answer'], row['pref_answer'])[0]['rouge-l']['f'], axis=1)\n" ] }, { "cell_type": "code", "execution_count": 30, "id": "fcce9a85-ad0b-4865-a3fa-d8a95b7d2c53", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import matplotlib.pyplot as plt\n", "\n", "# Define similarity metrics\n", "metrics = ['cosine_similarity', 'levenshtein_ratio', 'jaccard_similarity', 'bleu_score']\n", "\n", "# Define parameters to analyze\n", "parameters = ['diversity_level', 'branching_factor', 'max_depth']\n", "\n", "# Convert categorical columns to strings for proper grouping\n", "df['diversity_level'] = df['diversity_level'].astype(str)\n", "df['branching_factor'] = df['branching_factor'].astype(str)\n", "df['max_depth'] = df['max_depth'].astype(str)\n", "\n", "# Create bar charts for each metric across parameters\n", "for metric in metrics:\n", " for param in parameters:\n", " plt.figure(figsize=(8, 5))\n", " df.groupby(param)[metric].mean().plot(kind='bar', title=f'{metric} by {param}')\n", " plt.ylabel(metric)\n", " plt.xlabel(param)\n", " plt.xticks(rotation=45)\n", " plt.grid(axis='y', linestyle='--', alpha=0.7)\n", " plt.show()\n" ] }, { "cell_type": "code", "execution_count": 33, "id": "75efaa29-f478-437a-9a25-32b52261351f", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 33, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "sns.heatmap(df.corr(numeric_only=True), annot=True)" ] }, { "cell_type": "code", "execution_count": null, "id": "e73f649d-7b83-4932-b91e-24d283696342", "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.12.7" } }, "nbformat": 4, "nbformat_minor": 5 }