result_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Read CSV result file generated via bayesian learning run in Colab."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_result = pd.read_csv('tuning_results_3.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have 1000 networks sampled from the Bayesian posterior distribution.<br>\n",
    "N = Minimum number of networks (percentage) we want to agree on the classification of a certain image<br>\n",
    "P = Probabilty of each of those networks (who agree) predicting that image as that label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>N</th>\n",
       "      <th>P</th>\n",
       "      <th>Total Images</th>\n",
       "      <th>Skipped Images</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>Coverage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.95</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>339</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0.90</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>337</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.005900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>0.85</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>311</td>\n",
       "      <td>0.964286</td>\n",
       "      <td>0.082596</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>0.80</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>285</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.159292</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73</th>\n",
       "      <td>0.75</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>257</td>\n",
       "      <td>0.939024</td>\n",
       "      <td>0.241888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>0.70</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>177</td>\n",
       "      <td>0.925926</td>\n",
       "      <td>0.477876</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>0.65</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>124</td>\n",
       "      <td>0.897674</td>\n",
       "      <td>0.634218</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>0.60</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>70</td>\n",
       "      <td>0.884758</td>\n",
       "      <td>0.793510</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>137</th>\n",
       "      <td>0.55</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>31</td>\n",
       "      <td>0.847403</td>\n",
       "      <td>0.908555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>0.50</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>2</td>\n",
       "      <td>0.818991</td>\n",
       "      <td>0.994100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>0.45</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>0</td>\n",
       "      <td>0.817109</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>185</th>\n",
       "      <td>0.40</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>0</td>\n",
       "      <td>0.817109</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>201</th>\n",
       "      <td>0.35</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>0</td>\n",
       "      <td>0.817109</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>217</th>\n",
       "      <td>0.30</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>0</td>\n",
       "      <td>0.817109</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>233</th>\n",
       "      <td>0.25</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>0</td>\n",
       "      <td>0.817109</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>249</th>\n",
       "      <td>0.20</td>\n",
       "      <td>0.5</td>\n",
       "      <td>339</td>\n",
       "      <td>0</td>\n",
       "      <td>0.817109</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        N    P  Total Images  Skipped Images  Accuracy  Coverage\n",
       "9    0.95  0.5           339             339       NaN  0.000000\n",
       "25   0.90  0.5           339             337  0.500000  0.005900\n",
       "41   0.85  0.5           339             311  0.964286  0.082596\n",
       "57   0.80  0.5           339             285  0.944444  0.159292\n",
       "73   0.75  0.5           339             257  0.939024  0.241888\n",
       "89   0.70  0.5           339             177  0.925926  0.477876\n",
       "105  0.65  0.5           339             124  0.897674  0.634218\n",
       "121  0.60  0.5           339              70  0.884758  0.793510\n",
       "137  0.55  0.5           339              31  0.847403  0.908555\n",
       "153  0.50  0.5           339               2  0.818991  0.994100\n",
       "169  0.45  0.5           339               0  0.817109  1.000000\n",
       "185  0.40  0.5           339               0  0.817109  1.000000\n",
       "201  0.35  0.5           339               0  0.817109  1.000000\n",
       "217  0.30  0.5           339               0  0.817109  1.000000\n",
       "233  0.25  0.5           339               0  0.817109  1.000000\n",
       "249  0.20  0.5           339               0  0.817109  1.000000"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_result[df_result.P == 0.5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>N</th>\n",
       "      <th>P</th>\n",
       "      <th>Total Images</th>\n",
       "      <th>Skipped Images</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>Coverage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>144</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.95</td>\n",
       "      <td>339</td>\n",
       "      <td>85</td>\n",
       "      <td>0.877953</td>\n",
       "      <td>0.749263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.90</td>\n",
       "      <td>339</td>\n",
       "      <td>71</td>\n",
       "      <td>0.876866</td>\n",
       "      <td>0.790560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.85</td>\n",
       "      <td>339</td>\n",
       "      <td>55</td>\n",
       "      <td>0.862676</td>\n",
       "      <td>0.837758</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>147</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.80</td>\n",
       "      <td>339</td>\n",
       "      <td>44</td>\n",
       "      <td>0.850847</td>\n",
       "      <td>0.870206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.75</td>\n",
       "      <td>339</td>\n",
       "      <td>35</td>\n",
       "      <td>0.845395</td>\n",
       "      <td>0.896755</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.70</td>\n",
       "      <td>339</td>\n",
       "      <td>31</td>\n",
       "      <td>0.844156</td>\n",
       "      <td>0.908555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>150</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.65</td>\n",
       "      <td>339</td>\n",
       "      <td>25</td>\n",
       "      <td>0.837580</td>\n",
       "      <td>0.926254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>151</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.60</td>\n",
       "      <td>339</td>\n",
       "      <td>16</td>\n",
       "      <td>0.826625</td>\n",
       "      <td>0.952802</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>152</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.55</td>\n",
       "      <td>339</td>\n",
       "      <td>10</td>\n",
       "      <td>0.817629</td>\n",
       "      <td>0.970501</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.50</td>\n",
       "      <td>339</td>\n",
       "      <td>2</td>\n",
       "      <td>0.818991</td>\n",
       "      <td>0.994100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>154</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.45</td>\n",
       "      <td>339</td>\n",
       "      <td>0</td>\n",
       "      <td>0.817109</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>155</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.40</td>\n",
       "      <td>339</td>\n",
       "      <td>0</td>\n",
       "      <td>0.811209</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>156</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.35</td>\n",
       "      <td>339</td>\n",
       "      <td>0</td>\n",
       "      <td>0.808260</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>157</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.30</td>\n",
       "      <td>339</td>\n",
       "      <td>0</td>\n",
       "      <td>0.808260</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.25</td>\n",
       "      <td>339</td>\n",
       "      <td>0</td>\n",
       "      <td>0.808260</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.20</td>\n",
       "      <td>339</td>\n",
       "      <td>0</td>\n",
       "      <td>0.808260</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       N     P  Total Images  Skipped Images  Accuracy  Coverage\n",
       "144  0.5  0.95           339              85  0.877953  0.749263\n",
       "145  0.5  0.90           339              71  0.876866  0.790560\n",
       "146  0.5  0.85           339              55  0.862676  0.837758\n",
       "147  0.5  0.80           339              44  0.850847  0.870206\n",
       "148  0.5  0.75           339              35  0.845395  0.896755\n",
       "149  0.5  0.70           339              31  0.844156  0.908555\n",
       "150  0.5  0.65           339              25  0.837580  0.926254\n",
       "151  0.5  0.60           339              16  0.826625  0.952802\n",
       "152  0.5  0.55           339              10  0.817629  0.970501\n",
       "153  0.5  0.50           339               2  0.818991  0.994100\n",
       "154  0.5  0.45           339               0  0.817109  1.000000\n",
       "155  0.5  0.40           339               0  0.811209  1.000000\n",
       "156  0.5  0.35           339               0  0.808260  1.000000\n",
       "157  0.5  0.30           339               0  0.808260  1.000000\n",
       "158  0.5  0.25           339               0  0.808260  1.000000\n",
       "159  0.5  0.20           339               0  0.808260  1.000000"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_result[df_result.N == 0.5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create arrays with the values of the Accuracy and Coverage columns above for visualization purpose"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "ns = df_result[df_result.P == 0.5]['N'].values\n",
    "accs2 = df_result[df_result.P == 0.5]['Accuracy'].values\n",
    "covs2 = df_result[df_result.P == 0.5]['Coverage'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "ps = df_result[df_result.N == 0.5]['P'].values\n",
    "accs = df_result[df_result.N == 0.5]['Accuracy'].values\n",
    "covs =  df_result[df_result.N == 0.5]['Coverage'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(15, 10))\n",
    "\n",
    "ax = fig.subplots(nrows=2, ncols=2)\n",
    "\n",
    "ax[0][0].plot(ps, accs)\n",
    "ax[0][0].set_xlabel('P')\n",
    "ax[0][0].set_ylabel('Accuracy')\n",
    "ax[0][0].set_title('Accuracy against P at N = 0.5', y=-0.2)\n",
    "\n",
    "ax[0][1].plot(ps, covs)\n",
    "ax[0][1].set_xlabel('P')\n",
    "ax[0][1].set_ylabel('Coverage')\n",
    "ax[0][1].set_title('Coverage against P at N = 0.5', y=-0.2)\n",
    "\n",
    "ax[1][0].plot(ns[2:], accs2[2:])\n",
    "ax[1][0].set_xlabel('N')\n",
    "ax[1][0].set_ylabel('Accuracy')\n",
    "ax[1][0].set_title('Accuracy against N at P = 0.5', y=-0.2)\n",
    "\n",
    "ax[1][1].plot(ns[2:], covs2[2:])\n",
    "ax[1][1].set_xlabel('N')\n",
    "ax[1][1].set_ylabel('Coverage')\n",
    "ax[1][1].set_title('Coverage against N at P = 0.5', y=-0.2)\n",
    "\n",
    "plt.subplots_adjust(top = 0.99, bottom=0.03, hspace=0.4, wspace=0.3)\n",
    "plt.tight_layout()\n",
    "plt.savefig('graph.png')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "At constant P, higher values of N lead to higher accuarcy on the images that are predicted, but a lot of images are skipped for prediction (lower coverage).<br>\n",
    "At constant N, higher values of P lead to higher accuarcy on the images that are predicted, but a lot of images are skipped for prediction (lower coverage)."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}