assocRules.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The autoreload extension is already loaded. To reload it, use:\n",
      "  %reload_ext autoreload\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:git_processor:Using pre-existing git repository at C:\\dev\\gitAnalysis\\GIT_REPOS\\jadx\n",
      "INFO:git_processor:Using pre-existing cache at C:\\dev\\gitAnalysis\\GIT_REPOS_SHELVE\\jadx\n",
      "INFO:git_processor:Processing git repository located at C:\\dev\\gitAnalysis\\GIT_REPOS\\jadx\n"
     ]
    }
   ],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "import pandas as pd\n",
    "from git_analysis import AssocAnalyzer, ModelParams, HierarchyType\n",
    "from gen_callgraph_wrapper import GRAPH_DIR\n",
    "from graph import raw_graph_to_igraph\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import logging\n",
    "logging.basicConfig()\n",
    "logging.getLogger(\"git_processor\").setLevel(logging.INFO)\n",
    "logging.getLogger(\"java_change_detector\").setLevel(logging.ERROR)\n",
    "logging.getLogger(\"assoc_mining\").setLevel(logging.INFO)\n",
    "ana = AssocAnalyzer(\"https://github.com/skylot/jadx\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "params = {\n",
    "    \"jadx - Commits - methods\": ModelParams(\"commit\", HierarchyType.method, min_support=0.005, min_confidence=0),\n",
    "    \"jadx - Commits - type declarations\": ModelParams(\"commit\", HierarchyType.type_def, min_support=0.005, min_confidence=0),\n",
    "    \"jadx - Commits - packages\": ModelParams(\"commit\", HierarchyType.package, min_support=0.005, min_confidence=0),\n",
    "    \"jadx - Pull requests - methods\": ModelParams(\"pr\", HierarchyType.method, min_support=0.005, min_confidence=0),\n",
    "    \"jadx - Pull requests - type declarations\": ModelParams(\"pr\", HierarchyType.type_def, min_support=0.005, min_confidence=0),\n",
    "    \"jadx - Pull requests - packages\": ModelParams(\"pr\", HierarchyType.package, min_support=0.005, min_confidence=0)\n",
    "}\n",
    "models = {\n",
    "    s: ana.analyze_model(v)\n",
    "    for s, v in params.items()\n",
    "}\n",
    "for k, v in models.items():\n",
    "    g = ana.create_graph(v)\n",
    "    g.to_folder(GRAPH_DIR / k)\n",
    "\n",
    "# g = ana.create_graph(model)\n",
    "# model.params, model.describe()\n",
    "# g.to_folder(GRAPH_DIR / f\"jadx-git-{g.hash}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>#baskets</th>\n",
       "      <th>#items</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>basket type</th>\n",
       "      <th>item type</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">commit</th>\n",
       "      <th>method</th>\n",
       "      <td>1423</td>\n",
       "      <td>13359</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>type_def</th>\n",
       "      <td>1452</td>\n",
       "      <td>2888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>package</th>\n",
       "      <td>1455</td>\n",
       "      <td>192</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">pr</th>\n",
       "      <th>method</th>\n",
       "      <td>248</td>\n",
       "      <td>4343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>type_def</th>\n",
       "      <td>254</td>\n",
       "      <td>1072</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>package</th>\n",
       "      <td>254</td>\n",
       "      <td>106</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       #baskets  #items\n",
       "basket type item type                  \n",
       "commit      method         1423   13359\n",
       "            type_def       1452    2888\n",
       "            package        1455     192\n",
       "pr          method          248    4343\n",
       "            type_def        254    1072\n",
       "            package         254     106"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stat_df = pd.DataFrame.from_records([{\n",
    "    \"item type\": v.params.hierarchy_type.name,\n",
    "    \"basket type\": v.params.commit_or_pr,\n",
    "    \"#baskets\": len(v.one_hot),\n",
    "    \"#items\": len(v.one_hot.columns),\n",
    "} for v in models.values()])\n",
    "\n",
    "pd.pivot(stat_df, [\"basket type\", \"item type\"], [])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    4536.000000\n",
       "mean        0.018784\n",
       "std         0.013323\n",
       "min         0.007874\n",
       "25%         0.011811\n",
       "50%         0.015748\n",
       "75%         0.023622\n",
       "max         0.118110\n",
       "Name: support, dtype: float64"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "v.assoc_rules[\"support\"].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1728x720 with 12 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(nrows=len(models), ncols=2, figsize=(len(models) * 4,10), sharex='col')\n",
    "for i, (s, v) in enumerate(models.items()):\n",
    "    # sns.kdeplot(x=v.assoc_rules[\"support\"], ax=ax[i, 0])\n",
    "    sns.boxplot(x=v.assoc_rules[\"support\"], ax=ax[i, 0])\n",
    "    ax[i, 0].set_title(f\"Support - {s}\")\n",
    "\n",
    "    # sns.kdeplot(x=v.assoc_rules[\"confidence\"], ax=ax[i, 1])\n",
    "    sns.boxplot(x=v.assoc_rules[\"confidence\"], ax=ax[i, 1])\n",
    "    ax[i, 1].set_title(f\"Confidence - {s}\")\n",
    "fig.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "b08beb85a999dc244c14dbb949ad77b6cde1aba6b12954537daf387febb06a96"
  },
  "kernelspec": {
   "display_name": "Python 3.7.9 ('venv': venv)",
   "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": "3.7.9"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}