exploration.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "499acb9a",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "ea9ad864",
   "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>release</th>\n",
       "      <th>static_ssid</th>\n",
       "      <th>strong</th>\n",
       "      <th>implicit</th>\n",
       "      <th>partial</th>\n",
       "      <th>code_size</th>\n",
       "      <th>compute_time</th>\n",
       "      <th>data_sent</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>82</td>\n",
       "      <td>171</td>\n",
       "      <td>342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>83</td>\n",
       "      <td>139</td>\n",
       "      <td>342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>83</td>\n",
       "      <td>134</td>\n",
       "      <td>206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>83</td>\n",
       "      <td>103</td>\n",
       "      <td>206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>88</td>\n",
       "      <td>171</td>\n",
       "      <td>383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>88</td>\n",
       "      <td>139</td>\n",
       "      <td>383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>87</td>\n",
       "      <td>135</td>\n",
       "      <td>247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>88</td>\n",
       "      <td>103</td>\n",
       "      <td>247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>81</td>\n",
       "      <td>171</td>\n",
       "      <td>302</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>82</td>\n",
       "      <td>139</td>\n",
       "      <td>302</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>82</td>\n",
       "      <td>135</td>\n",
       "      <td>166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>82</td>\n",
       "      <td>103</td>\n",
       "      <td>166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>86</td>\n",
       "      <td>170</td>\n",
       "      <td>343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>87</td>\n",
       "      <td>138</td>\n",
       "      <td>343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>87</td>\n",
       "      <td>135</td>\n",
       "      <td>207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>87</td>\n",
       "      <td>103</td>\n",
       "      <td>207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>132</td>\n",
       "      <td>148</td>\n",
       "      <td>342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>137</td>\n",
       "      <td>119</td>\n",
       "      <td>342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>132</td>\n",
       "      <td>117</td>\n",
       "      <td>206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>138</td>\n",
       "      <td>90</td>\n",
       "      <td>206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>139</td>\n",
       "      <td>148</td>\n",
       "      <td>383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>142</td>\n",
       "      <td>120</td>\n",
       "      <td>383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>139</td>\n",
       "      <td>116</td>\n",
       "      <td>247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>142</td>\n",
       "      <td>89</td>\n",
       "      <td>247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>131</td>\n",
       "      <td>145</td>\n",
       "      <td>302</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>136</td>\n",
       "      <td>120</td>\n",
       "      <td>302</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>132</td>\n",
       "      <td>115</td>\n",
       "      <td>166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>137</td>\n",
       "      <td>88</td>\n",
       "      <td>166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>138</td>\n",
       "      <td>148</td>\n",
       "      <td>343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>141</td>\n",
       "      <td>120</td>\n",
       "      <td>343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>138</td>\n",
       "      <td>116</td>\n",
       "      <td>207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>142</td>\n",
       "      <td>89</td>\n",
       "      <td>207</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    release  static_ssid  strong  implicit  partial  code_size  compute_time  \\\n",
       "0     False        False   False     False    False         82           171   \n",
       "1     False        False   False     False     True         83           139   \n",
       "2     False        False   False      True    False         83           134   \n",
       "3     False        False   False      True     True         83           103   \n",
       "4     False        False    True     False    False         88           171   \n",
       "5     False        False    True     False     True         88           139   \n",
       "6     False        False    True      True    False         87           135   \n",
       "7     False        False    True      True     True         88           103   \n",
       "8     False         True   False     False    False         81           171   \n",
       "9     False         True   False     False     True         82           139   \n",
       "10    False         True   False      True    False         82           135   \n",
       "11    False         True   False      True     True         82           103   \n",
       "12    False         True    True     False    False         86           170   \n",
       "13    False         True    True     False     True         87           138   \n",
       "14    False         True    True      True    False         87           135   \n",
       "15    False         True    True      True     True         87           103   \n",
       "16     True        False   False     False    False        132           148   \n",
       "17     True        False   False     False     True        137           119   \n",
       "18     True        False   False      True    False        132           117   \n",
       "19     True        False   False      True     True        138            90   \n",
       "20     True        False    True     False    False        139           148   \n",
       "21     True        False    True     False     True        142           120   \n",
       "22     True        False    True      True    False        139           116   \n",
       "23     True        False    True      True     True        142            89   \n",
       "24     True         True   False     False    False        131           145   \n",
       "25     True         True   False     False     True        136           120   \n",
       "26     True         True   False      True    False        132           115   \n",
       "27     True         True   False      True     True        137            88   \n",
       "28     True         True    True     False    False        138           148   \n",
       "29     True         True    True     False     True        141           120   \n",
       "30     True         True    True      True    False        138           116   \n",
       "31     True         True    True      True     True        142            89   \n",
       "\n",
       "    data_sent  \n",
       "0         342  \n",
       "1         342  \n",
       "2         206  \n",
       "3         206  \n",
       "4         383  \n",
       "5         383  \n",
       "6         247  \n",
       "7         247  \n",
       "8         302  \n",
       "9         302  \n",
       "10        166  \n",
       "11        166  \n",
       "12        343  \n",
       "13        343  \n",
       "14        207  \n",
       "15        207  \n",
       "16        342  \n",
       "17        342  \n",
       "18        206  \n",
       "19        206  \n",
       "20        383  \n",
       "21        383  \n",
       "22        247  \n",
       "23        247  \n",
       "24        302  \n",
       "25        302  \n",
       "26        166  \n",
       "27        166  \n",
       "28        343  \n",
       "29        343  \n",
       "30        207  \n",
       "31        207  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(\"./benchmark_data.csv\")\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "be9e5918",
   "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>release</th>\n",
       "      <th>static_ssid</th>\n",
       "      <th>strong</th>\n",
       "      <th>implicit</th>\n",
       "      <th>partial</th>\n",
       "      <th>code_size</th>\n",
       "      <th>compute_time</th>\n",
       "      <th>data_sent</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>82</td>\n",
       "      <td>171</td>\n",
       "      <td>342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>83</td>\n",
       "      <td>139</td>\n",
       "      <td>342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>83</td>\n",
       "      <td>134</td>\n",
       "      <td>206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>83</td>\n",
       "      <td>103</td>\n",
       "      <td>206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>88</td>\n",
       "      <td>171</td>\n",
       "      <td>383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>88</td>\n",
       "      <td>139</td>\n",
       "      <td>383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>87</td>\n",
       "      <td>135</td>\n",
       "      <td>247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>88</td>\n",
       "      <td>103</td>\n",
       "      <td>247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>81</td>\n",
       "      <td>171</td>\n",
       "      <td>302</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>82</td>\n",
       "      <td>139</td>\n",
       "      <td>302</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>82</td>\n",
       "      <td>135</td>\n",
       "      <td>166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>82</td>\n",
       "      <td>103</td>\n",
       "      <td>166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>86</td>\n",
       "      <td>170</td>\n",
       "      <td>343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>87</td>\n",
       "      <td>138</td>\n",
       "      <td>343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>87</td>\n",
       "      <td>135</td>\n",
       "      <td>207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>87</td>\n",
       "      <td>103</td>\n",
       "      <td>207</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    release  static_ssid  strong  implicit  partial  code_size  compute_time  \\\n",
       "0     False        False   False     False    False         82           171   \n",
       "1     False        False   False     False     True         83           139   \n",
       "2     False        False   False      True    False         83           134   \n",
       "3     False        False   False      True     True         83           103   \n",
       "4     False        False    True     False    False         88           171   \n",
       "5     False        False    True     False     True         88           139   \n",
       "6     False        False    True      True    False         87           135   \n",
       "7     False        False    True      True     True         88           103   \n",
       "8     False         True   False     False    False         81           171   \n",
       "9     False         True   False     False     True         82           139   \n",
       "10    False         True   False      True    False         82           135   \n",
       "11    False         True   False      True     True         82           103   \n",
       "12    False         True    True     False    False         86           170   \n",
       "13    False         True    True     False     True         87           138   \n",
       "14    False         True    True      True    False         87           135   \n",
       "15    False         True    True      True     True         87           103   \n",
       "\n",
       "    data_sent  \n",
       "0         342  \n",
       "1         342  \n",
       "2         206  \n",
       "3         206  \n",
       "4         383  \n",
       "5         383  \n",
       "6         247  \n",
       "7         247  \n",
       "8         302  \n",
       "9         302  \n",
       "10        166  \n",
       "11        166  \n",
       "12        343  \n",
       "13        343  \n",
       "14        207  \n",
       "15        207  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[~df.release]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "7f3d328f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(    release  static_ssid  strong  implicit  partial  code_size  compute_time  \\\n",
       " 16     True        False   False     False    False        132           148   \n",
       " 17     True        False   False     False     True        137           119   \n",
       " 18     True        False   False      True    False        132           117   \n",
       " 19     True        False   False      True     True        138            90   \n",
       " 20     True        False    True     False    False        139           148   \n",
       " 21     True        False    True     False     True        142           120   \n",
       " 22     True        False    True      True    False        139           116   \n",
       " 23     True        False    True      True     True        142            89   \n",
       " 24     True         True   False     False    False        131           145   \n",
       " 25     True         True   False     False     True        136           120   \n",
       " 26     True         True   False      True    False        132           115   \n",
       " 27     True         True   False      True     True        137            88   \n",
       " 28     True         True    True     False    False        138           148   \n",
       " 29     True         True    True     False     True        141           120   \n",
       " 30     True         True    True      True    False        138           116   \n",
       " 31     True         True    True      True     True        142            89   \n",
       " \n",
       "     data_sent  \n",
       " 16        342  \n",
       " 17        342  \n",
       " 18        206  \n",
       " 19        206  \n",
       " 20        383  \n",
       " 21        383  \n",
       " 22        247  \n",
       " 23        247  \n",
       " 24        302  \n",
       " 25        302  \n",
       " 26        166  \n",
       " 27        166  \n",
       " 28        343  \n",
       " 29        343  \n",
       " 30        207  \n",
       " 31        207  ,\n",
       "     release  static_ssid  strong  implicit  partial  code_size  compute_time  \\\n",
       " 0     False        False   False     False    False         82           171   \n",
       " 1     False        False   False     False     True         83           139   \n",
       " 2     False        False   False      True    False         83           134   \n",
       " 3     False        False   False      True     True         83           103   \n",
       " 4     False        False    True     False    False         88           171   \n",
       " 5     False        False    True     False     True         88           139   \n",
       " 6     False        False    True      True    False         87           135   \n",
       " 7     False        False    True      True     True         88           103   \n",
       " 8     False         True   False     False    False         81           171   \n",
       " 9     False         True   False     False     True         82           139   \n",
       " 10    False         True   False      True    False         82           135   \n",
       " 11    False         True   False      True     True         82           103   \n",
       " 12    False         True    True     False    False         86           170   \n",
       " 13    False         True    True     False     True         87           138   \n",
       " 14    False         True    True      True    False         87           135   \n",
       " 15    False         True    True      True     True         87           103   \n",
       " \n",
       "     data_sent  \n",
       " 0         342  \n",
       " 1         342  \n",
       " 2         206  \n",
       " 3         206  \n",
       " 4         383  \n",
       " 5         383  \n",
       " 6         247  \n",
       " 7         247  \n",
       " 8         302  \n",
       " 9         302  \n",
       " 10        166  \n",
       " 11        166  \n",
       " 12        343  \n",
       " 13        343  \n",
       " 14        207  \n",
       " 15        207  )"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rel_df = df[df.release]\n",
    "ser_df = df[~df.release]\n",
    "\n",
    "rel_df, ser_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "4d92d18a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "52.5"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# average code size difference between release mode and server mode\n",
    "code_sz_diffs = [r-s for (r, s) in zip(rel_df.code_size, ser_df.code_size)]\n",
    "sum(code_sz_diffs) / len(code_sz_diffs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "90245fe3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-18.8125"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# average compute time difference between release mode and server mode\n",
    "comp_t_diffs = [r-s for (r, s) in zip(rel_df.compute_time, ser_df.compute_time)]\n",
    "sum(comp_t_diffs) / len(comp_t_diffs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "13cdc1da",
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_diffs(df, feature):\n",
    "    df_a = df[df[feature]]\n",
    "    df_b = df[~df[feature]]\n",
    "    code_sz_diffs = [r-s for (r, s) in zip(df_a.code_size, df_b.code_size)]\n",
    "    sz_diff = sum(code_sz_diffs) / len(code_sz_diffs)\n",
    "    comp_t_diffs = [r-s for (r, s) in zip(df_a.compute_time, df_b.compute_time)]\n",
    "    ct_diff = sum(comp_t_diffs) / len(comp_t_diffs)\n",
    "    data_s_diffs = [r-s for (r, s) in zip(df_a.data_sent, df_b.data_sent)]\n",
    "    ds_diff = sum(data_s_diffs) / len(data_s_diffs)\n",
    "    return sz_diff, ct_diff, ds_diff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "5e255b41",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(52.5, -18.8125, 0.0)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "compute_diffs(df, \"release\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "15505359",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2.375, -29.5625, 0.0)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "compute_diffs(df, \"partial\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "fc6849fe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.375, -33.4375, -136.0)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "compute_diffs(df, \"implicit\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "49c04c30",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5.375, 0.1875, 41.0)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "compute_diffs(df, \"strong\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "d8d356e8",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-0.875, -0.4375, -40.0)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "compute_diffs(df, \"static_ssid\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "d1be24f9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# graph the change in code size and compute time by feature flag\n",
    "# adapted from https://matplotlib.org/stable/gallery/lines_bars_and_markers/barchart.html\n",
    "# accessed 14/04/2023\n",
    "\n",
    "# first setup the data\n",
    "variables = [\"release\", \"partial\", \"implicit\", \"strong\", \"static_ssid\"]\n",
    "measurables = [\"Size diff (Kib)\", \"Time diff (ms)\", \"Data Sent (bytes)\"]\n",
    "diffs = {\n",
    "    v: dict(zip(measurables, compute_diffs(df, v)))\n",
    "    for v in variables\n",
    "}\n",
    "data = {\n",
    "    t: tuple(diffs[v][t] for v in variables)\n",
    "    for t in measurables\n",
    "}\n",
    "\n",
    "# now setup the graph stuff\n",
    "y = np.arange(len(variables))\n",
    "width = 0.25\n",
    "multiplier = 0\n",
    "\n",
    "fig, ax = plt.subplots(layout=\"constrained\")\n",
    "\n",
    "for attribute, measurement in data.items():\n",
    "    offset = width * multiplier\n",
    "    rects = ax.barh(y + offset, measurement, width, label=attribute)\n",
    "    ax.bar_label(rects, padding=3)\n",
    "    multiplier += 1\n",
    "    \n",
    "# ax.set_ylabel('Mean Difference')\n",
    "ax.set_title('Effect of features of Code Size and Compute Time')\n",
    "ax.set_yticks(y + width, variables)\n",
    "ax.legend(loc='upper left')\n",
    "ax.set_facecolor(\"#f3f3f3\")\n",
    "ax.set_xlim(-155, 70)\n",
    "\n",
    "# plt.savefig(\"/home/tritoke/feature_effects.eps\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "de75f5ff",
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_perc_diffs(df, feature):\n",
    "    df_a = df[df[feature]]\n",
    "    df_b = df[~df[feature]]\n",
    "    code_sz_diffs = [100 * (r-s)/s for (r, s) in zip(df_a.code_size, df_b.code_size)]\n",
    "    sz_diff = sum(code_sz_diffs) / len(code_sz_diffs), min(code_sz_diffs), max(code_sz_diffs)\n",
    "    comp_t_diffs = [100 * (r-s)/s for (r, s) in zip(df_a.compute_time, df_b.compute_time)]\n",
    "    ct_diff = sum(comp_t_diffs) / len(comp_t_diffs), min(comp_t_diffs), max(comp_t_diffs)\n",
    "    data_s_diffs = [100 * (r-s)/s for (r, s) in zip(df_a.data_sent, df_b.data_sent)]\n",
    "    ds_diff = sum(data_s_diffs) / len(data_s_diffs), min(data_s_diffs), max(data_s_diffs)\n",
    "    return sz_diff, ct_diff, ds_diff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "fd6bb6e1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((-0.8584149556820144, -2.272727272727273, 0.0),\n",
       " (-0.354766621846865, -2.2222222222222223, 0.8403361344537815),\n",
       " (-14.43789459361864, -19.41747572815534, -10.443864229765014))"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "compute_perc_diffs(df, \"static_ssid\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "1c285048",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The PostScript backend does not support transparency; partially transparent artists will be rendered opaque.\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# graph the change in code size and compute time by feature flag\n",
    "# adapted from https://matplotlib.org/stable/gallery/lines_bars_and_markers/barchart.html\n",
    "# accessed 14/04/2023\n",
    "\n",
    "# first setup the data\n",
    "variables = [\"release\", \"partial\", \"implicit\", \"strong\", \"static_ssid\"]\n",
    "measurables = [\"Size diff (Kib)\", \"Time diff (ms)\", \"Data Sent (bytes)\"]\n",
    "diffs = {\n",
    "    v: dict(zip(measurables, compute_perc_diffs(df, v)))\n",
    "    for v in variables\n",
    "}\n",
    "data = {\n",
    "    t: tuple(diffs[v][t] for v in variables)\n",
    "    for t in measurables\n",
    "}\n",
    "\n",
    "# now setup the graph stuff\n",
    "y = np.arange(len(variables))\n",
    "width = 0.25\n",
    "multiplier = 0\n",
    "\n",
    "fig, ax = plt.subplots(layout=\"constrained\")\n",
    "\n",
    "for attribute, measurements in data.items():\n",
    "    offset = width * multiplier\n",
    "    m = [x[0] for x in measurements]\n",
    "    e = [abs(x[2] - x[1]) / 2 for x in measurements]\n",
    "    rects = ax.barh(y - offset, m, width, xerr=e, label=attribute)\n",
    "    ax.bar_label(rects, padding=3, fmt=\"{:.2f}%\")\n",
    "    multiplier += 1\n",
    "    \n",
    "ax.set_xlabel('Mean Percentage Difference')\n",
    "ax.set_title('Effect of features of Code Size and Compute Time')\n",
    "ax.set_yticks(y - width, variables)\n",
    "ax.legend(loc='upper right')\n",
    "ax.set_facecolor(\"#f3f3f3\")\n",
    "ax.set_xlim(-70, 90)\n",
    "\n",
    "# plt.savefig(\"/home/tritoke/feature_effects.eps\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "3eeaf4e7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "137.25"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.release].code_size.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "42c2f48f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "84.75"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[~df.release].code_size.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "721c55e9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(118.0, 136.8125)"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.release].compute_time.mean(), df[~df.release].compute_time.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "7880d10b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(274.5, 274.5)"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.release].data_sent.mean(), df[~df.release].data_sent.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "b8a65079",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "171"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.compute_time.max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "ab26093c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "166"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.data_sent.min()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "118d6443",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.927710843373494"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "486 / df.data_sent.min()"
   ]
  }
 ],
 "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": "3.10.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}