File size: 6,131 Bytes
77ba698 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 |
{
"cells": [
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"{'prob': {'file': 'prob.json'}}"
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import json\n",
"import numpy as np\n",
"import plotly.graph_objects as go\n",
"RED_FULL=\"rgba(255, 0, 0, 1)\"\n",
"\n",
"# Define the function 1 - (1 - x^8)^14\n",
"def func1(x):\n",
" return 1 - np.power(1 - np.power(x, 8), 14)\n",
"\n",
"# Define the function 1 - (1 - x^20)^450\n",
"def func2(x):\n",
" return 1 - np.power(1 - np.power(x, 20), 450)\n",
"\n",
"# Generate x values from 0 to 1\n",
"x = np.linspace(0, 1, 1000)\n",
"\n",
"# Calculate y values for each function\n",
"y1 = func1(x)\n",
"y2 = func2(x)\n",
"\n",
"# Create traces\n",
"trace1 = go.Scatter(x=x, y=y1, mode='lines', name='FineWeb: 1-(1-s^8)^14')\n",
"trace2 = go.Scatter(x=x, y=y2, mode='lines', name='RefinedWeb: 1-(1-s^20)^450')\n",
"vertical_line = go.Scatter(x=[0.75, 0.75], y=[0, 1], mode='lines', line=dict(color='red', dash='dash'), name='Threshold')\n",
"\n",
"# Define layout\n",
"layout = {\n",
" 'title': {\n",
" 'text': 'MinHash parameters',\n",
" },\n",
" 'xaxis': {\n",
" 'title': {\n",
" 'text': 'Document similarity (s)',\n",
" },\n",
" },\n",
" 'yaxis': {\n",
" 'title': {\n",
" 'text': 'Matched as dups probability',\n",
" },\n",
" },\n",
"}\n",
"\n",
"\n",
"def normalize_run_name(run_name):\n",
" return run_name.replace(\"/\", \"_\")\n",
"\n",
"\n",
"def save_for_plot(dir_name, df, views, xlabel=\"Dataset\", ylabel=\"Matched as dups probability\", plot_name=\"plot name\", custom_layout={}, ranges={}, x_column=None, default_metric=None):\n",
" import os\n",
" files = {}\n",
" os.makedirs(f\"data/plots/{dir_name}\", exist_ok=True)\n",
" for view in views:\n",
" data = {}\n",
" for run_name in df[\"runname\"].unique():\n",
" run_name_only=df[df[\"runname\"]==run_name]\n",
" data[run_name] = {\n",
" \"x\": run_name_only[x_column].tolist() if x_column else [run_name],\n",
" \"y\": run_name_only[view].tolist(),\n",
" \"label\": run_name,\n",
" }\n",
" file_name = f\"{normalize_run_name(view)}.json\"\n",
" files[view] = {\"file\": f\"{file_name}\"}\n",
" with open(f\"data/plots/{dir_name}/{file_name}\", \"w\") as f:\n",
" json.dump({\n",
" \"data\": data,\n",
" \"layout\": {\n",
" \"title\": {\n",
" \"text\": plot_name,\n",
" },\n",
" \"xaxis\": {\n",
" \"title\": {\n",
" \"text\": xlabel,\n",
" },\n",
" },\n",
" \"yaxis\": {\n",
" # \"range\": ranges.get(view, None),\n",
" \"title\": {\n",
" \"text\": ylabel,\n",
" },\n",
" },\n",
" \"shapes\": [\n",
" {\n",
" \"type\": \"line\",\n",
" \"x0\": 0.75,\n",
" \"y0\": 0.0,\n",
" \"x1\": 0.75,\n",
" \"y1\": 1.2,\n",
" \"xref\": \"x\",\n",
" \"yref\": \"y\",\n",
" \"line\": {\n",
" \"color\": RED_FULL,\n",
" \"width\": 1,\n",
" \"dash\": \"dashdot\"\n",
" },\n",
" \"showarrow\": False\n",
" }\n",
" ],\n",
" **custom_layout,\n",
" },\n",
" }, f)\n",
" with open(f\"data/plots/{dir_name}/index.json\", \"w\") as f:\n",
" json.dump({\n",
" \"files\": files,\n",
" \"settings\": {\n",
" \"defaultMetric\": default_metric,\n",
" \"slider\": None,\n",
" \"autoSetXRange\": False,\n",
" }\n",
" }, f)\n",
" return files\n",
"\n",
"import pandas as pd\n",
"df = pd.DataFrame({\n",
" \"runname\": [\"FineWeb: 1-(1-s^8)^14\"]*len(x) + [\"RefinedWeb: 1-(1-s^20)^450\"]*len(x),\n",
" \"similarity\": x.tolist()+x.tolist(),\n",
" \"prob\": y1.tolist()+y2.tolist(),\n",
" \"view\": [\"normal\"]*2*len(x)\n",
"})\n",
"\n",
"custom_layout = {\n",
" \"legend\": {\n",
" \"orientation\": \"v\",\n",
" \"xanchor\": \"left\",\n",
" \"yanchor\": \"top\",\n",
" \"x\": 0,\n",
" \"y\": 1,\n",
" },\n",
"}\n",
"\n",
"save_for_plot(\"minhash_params\", df, [\"prob\"], xlabel=\"Document similarity (s)\", plot_name=\"MinHash parameters\", custom_layout=custom_layout, ranges={}, x_column=\"similarity\", default_metric=\"prob\")"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "datatrove",
"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": "3.12.2"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
|