{
 "cells": [
  {
   "cell_type": "code",
   "id": "initial_id",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "import pickle\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "\n",
    "import config"
   ],
   "outputs": [],
   "execution_count": null
  },
  {
   "metadata": {},
   "cell_type": "code",
   "source": [
    "GROUP_LABELS = ['Golden', 'Synthetic: Backward', 'Synthetic: Forward', 'Synthetic: Backward → Forward',\n",
    "                'Synthetic: All']"
   ],
   "id": "3dedb547dfae89d5",
   "outputs": [],
   "execution_count": null
  },
  {
   "metadata": {},
   "cell_type": "code",
   "source": [
    "def load_data(s):\n",
    "    with open(config.OUTPUT_CHARTS_DIR / f\"{s}_data.pkl\", 'rb') as f:\n",
    "        groups = pickle.load(f)\n",
    "\n",
    "        data = []\n",
    "        for i, group in enumerate(groups):\n",
    "            for e in group:\n",
    "                data.append({\n",
    "                    'Subset': GROUP_LABELS[i],\n",
    "                    'value': e\n",
    "                })\n",
    "        return pd.DataFrame.from_records(data)"
   ],
   "id": "837c039c80b1093e",
   "outputs": [],
   "execution_count": null
  },
  {
   "metadata": {},
   "cell_type": "code",
   "source": [
    "name = 'deletions_norm'\n",
    "\n",
    "sns.set_theme(palette=\"colorblind\")\n",
    "\n",
    "ax = sns.displot(data=load_data(name), \n",
    "                 x='value', \n",
    "                 hue='Subset', \n",
    "                 kind='kde',\n",
    "                 aspect=1.5, \n",
    "                 common_norm=False,\n",
    "                 clip=(0, 10000))\n",
    "sns.move_legend(ax, \"center right\", bbox_to_anchor=(.70, .75))\n",
    "for line in ax.legend.get_lines():\n",
    "    line.set_linewidth(5.0)\n",
    "\n",
    "plt.title(\"Number of \\\"delete\\\" operations: distribution\")\n",
    "plt.xlabel(\"Number of \\\"delete\\\" operations\")\n",
    "\n",
    "plt.savefig(config.OUTPUT_CHARTS_DIR / f\"{name}_dist.png\",bbox_inches='tight')\n",
    "plt.savefig(config.OUTPUT_CHARTS_DIR / f\"{name}_dist.svg\",bbox_inches='tight')\n",
    "plt.show()"
   ],
   "id": "8c756559f9bd4ec",
   "outputs": [],
   "execution_count": null
  },
  {
   "metadata": {},
   "cell_type": "code",
   "source": [
    "name = 'insertions_norm'\n",
    "\n",
    "sns.set_theme(palette=\"colorblind\")\n",
    "\n",
    "ax = sns.displot(data=load_data(name), \n",
    "                 x='value', \n",
    "                 hue='Subset', \n",
    "                 kind='kde',\n",
    "                 aspect=1.5, \n",
    "                 common_norm=False,\n",
    "                 clip=(0, 10000))\n",
    "sns.move_legend(ax, \"center right\", bbox_to_anchor=(.70, .75))\n",
    "for line in ax.legend.get_lines():\n",
    "    line.set_linewidth(5.0)\n",
    "\n",
    "plt.title(\"Number of \\\"insert\\\" operations: distribution\")\n",
    "plt.xlabel(\"Number of \\\"insert\\\" operations\")\n",
    "\n",
    "plt.savefig(config.OUTPUT_CHARTS_DIR / f\"{name}_dist.png\",bbox_inches='tight')\n",
    "plt.savefig(config.OUTPUT_CHARTS_DIR / f\"{name}_dist.svg\",bbox_inches='tight')\n",
    "plt.show()"
   ],
   "id": "af1668213091e2c1",
   "outputs": [],
   "execution_count": null
  },
  {
   "metadata": {},
   "cell_type": "code",
   "source": [
    "import numpy as np\n",
    "\n",
    "def get_bins_for_report(intervals):\n",
    "    bins = []\n",
    "    for cur_interval in intervals:\n",
    "        bins += list(map(int, np.linspace(start=cur_interval[0], stop=cur_interval[1], num=cur_interval[2])))\n",
    "    print(\",\".join(list(map(str, bins))))\n",
    "    \n",
    "get_bins_for_report([\n",
    "    (0, 400, 30),\n",
    "    (401, 1200, 20)\n",
    "])"
   ],
   "id": "c2d245ada4ae54a",
   "outputs": [],
   "execution_count": null
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "## FUS logs",
   "id": "e293db4b140f920e"
  },
  {
   "metadata": {},
   "cell_type": "code",
   "source": [
    "def to_hist(data):\n",
    "    data = data.copy()\n",
    "    num_samples = sum(data['y'])\n",
    "    data['y'] = [e / num_samples / scale for e, scale in zip(data['y'], data['bin_size'])]\n",
    "    \n",
    "    return data"
   ],
   "id": "74f1d808e83aa528",
   "outputs": [],
   "execution_count": null
  },
  {
   "metadata": {},
   "cell_type": "code",
   "source": [
    "FUS_REPORT_TYPE = \"all_ide\"\n",
    "\n",
    "RAW_FUS_REPORT = f\"data/fus_raw_report_{FUS_REPORT_TYPE}.csv\"\n",
    "\n",
    "FUS_METRICS = [\"CommitMessageEditDistance\", \"CommitMessageLengthDiff\"]\n",
    "\n",
    "FUS_DATA = {}\n",
    "\n",
    "for metric in FUS_METRICS:\n",
    "    FUS_DATA[metric] = {\"x\": [], \"y\": [], \"bin_size\": []}\n",
    "    \n",
    "with open(RAW_FUS_REPORT, \"r\") as f:\n",
    "    current_metric = None\n",
    "    for line in f.readlines():\n",
    "        if line.isspace():\n",
    "            current_metric = None\n",
    "            continue\n",
    "        \n",
    "        skip_line = False\n",
    "        \n",
    "        for metric in FUS_METRICS:\n",
    "                if metric in line:\n",
    "                    current_metric = metric\n",
    "                    skip_line = True\n",
    "                    break\n",
    "                \n",
    "        \n",
    "        if not skip_line and current_metric is not None:\n",
    "            tokens = line.strip().split(\",\")\n",
    "            interval = tokens[0].replace(\"> \", \"\").split(\" - \")\n",
    "            x_left = float(interval[0])\n",
    "            x_right = float(interval[-1])\n",
    "            x = (x_left + x_right) / 2\n",
    "            y = float(tokens[-2])\n",
    "            bin_size = x_right - x_left + 1\n",
    "            FUS_DATA[current_metric][\"x\"].append(x)\n",
    "            FUS_DATA[current_metric][\"y\"].append(y)\n",
    "            FUS_DATA[current_metric][\"bin_size\"].append(bin_size)\n",
    "            \n",
    "            \n",
    "SKIP_FIRST = 2\n",
    "FUS_DATA_NO_HEAD = {}\n",
    "\n",
    "for metric in FUS_METRICS:\n",
    "    FUS_DATA_NO_HEAD[metric] = {\n",
    "        'x': FUS_DATA[metric]['x'][SKIP_FIRST:], \n",
    "        'y': FUS_DATA[metric]['y'][SKIP_FIRST:],\n",
    "        'bin_size': FUS_DATA[metric]['bin_size'][SKIP_FIRST:]}"
   ],
   "id": "ed09c1d2fec35d58",
   "outputs": [],
   "execution_count": null
  },
  {
   "metadata": {},
   "cell_type": "code",
   "source": [
    "name = 'editdist'\n",
    "\n",
    "sns.set_theme(palette=\"colorblind\")\n",
    "\n",
    "ax = sns.displot(data=load_data(name), \n",
    "                 x='value', \n",
    "                 hue='Subset', \n",
    "                 kind='kde',\n",
    "                 aspect=1.5, \n",
    "                 common_norm=False,\n",
    "                 clip=(0, 10000))\n",
    "\n",
    "\n",
    "sns.lineplot(data=to_hist(FUS_DATA_NO_HEAD[\"CommitMessageEditDistance\"]), x='x', y='y', style=True, dashes=[(2,2)], legend=False)\n",
    "\n",
    "sns.move_legend(ax, \"center right\", bbox_to_anchor=(.70, .75))\n",
    "for line in ax.legend.get_lines():\n",
    "    line.set_linewidth(5.0)\n",
    "\n",
    "plt.title(f\"Edit distance ({FUS_REPORT_TYPE}): distribution\")\n",
    "plt.xlabel(\"Edit disatnce\")\n",
    "\n",
    "plt.savefig(config.OUTPUT_CHARTS_DIR / f\"{name}_fus_{FUS_REPORT_TYPE}_skip_{SKIP_FIRST}_dist.png\",bbox_inches='tight')\n",
    "plt.savefig(config.OUTPUT_CHARTS_DIR / f\"{name}_fus_{FUS_REPORT_TYPE}_skip_{SKIP_FIRST}_dist.svg\",bbox_inches='tight')\n",
    "plt.show()"
   ],
   "id": "be2eb3c1499e196f",
   "outputs": [],
   "execution_count": null
  },
  {
   "metadata": {},
   "cell_type": "code",
   "source": [
    "name = 'lendiff'\n",
    "\n",
    "sns.set_theme(palette=\"colorblind\")\n",
    "\n",
    "ax = sns.displot(data=load_data(name), \n",
    "                 x='value', \n",
    "                 hue='Subset', \n",
    "                 kind='kde',\n",
    "                 aspect=1.5, \n",
    "                 common_norm=False,\n",
    "                 clip=(0, 10000))\n",
    "\n",
    "sns.lineplot(data=to_hist(FUS_DATA_NO_HEAD[\"CommitMessageLengthDiff\"]), x='x', y='y', style=True, dashes=[(2,2)], legend=False)\n",
    "\n",
    "sns.move_legend(ax, \"center right\", bbox_to_anchor=(.70, .75))\n",
    "for line in ax.legend.get_lines():\n",
    "    line.set_linewidth(5.0)\n",
    "\n",
    "\n",
    "plt.title(f\"Length difference ({FUS_REPORT_TYPE}): distribution\")\n",
    "plt.xlabel(\"Length difference\")\n",
    "\n",
    "plt.savefig(config.OUTPUT_CHARTS_DIR / f\"{name}_fus_{FUS_REPORT_TYPE}_skip_{SKIP_FIRST}_dist.png\",bbox_inches='tight')\n",
    "plt.savefig(config.OUTPUT_CHARTS_DIR / f\"{name}_fus_{FUS_REPORT_TYPE}_skip_{SKIP_FIRST}_dist.svg\",bbox_inches='tight')\n",
    "plt.show()"
   ],
   "id": "31221bbd8ce257db",
   "outputs": [],
   "execution_count": null
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}