results/plot_results_clef.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib.pyplot import xticks\n",
    "from matplotlib.pyplot import ylim\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>UNCERTAINTYclefw2vecMLPre@10</th>\n",
       "      <th>UNCERTAINTYclefw2vecRFre@10</th>\n",
       "      <th>UNCERTAINTYclefw2vecLRre@10</th>\n",
       "      <th>UNCERTAINTYclefw2vecSVMrbfre@10</th>\n",
       "      <th>UNCERTAINTYclefgloveMLPre@10</th>\n",
       "      <th>UNCERTAINTYclefgloveRFre@10</th>\n",
       "      <th>UNCERTAINTYclefgloveLRre@10</th>\n",
       "      <th>UNCERTAINTYclefgloveSVMrbfre@10</th>\n",
       "      <th>UNCERTAINTYclefBERTMLPre@10</th>\n",
       "      <th>UNCERTAINTYclefBERTRFre@10</th>\n",
       "      <th>...</th>\n",
       "      <th>RANDOMclefgloveRFre@10</th>\n",
       "      <th>RANDOMclefgloveLRre@10</th>\n",
       "      <th>RANDOMclefgloveSVMrbfre@10</th>\n",
       "      <th>RANDOMclefBERTMLPre@10</th>\n",
       "      <th>RANDOMclefBERTRFre@10</th>\n",
       "      <th>RANDOMclefBERTLRre@10</th>\n",
       "      <th>RANDOMclefBERTSVMrbfre@10</th>\n",
       "      <th>RANDOMclefBioBERTMLPre@10</th>\n",
       "      <th>RANDOMclefBioBERTRFre@10</th>\n",
       "      <th>RANDOMclefBioBERTLRre@10</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.097</td>\n",
       "      <td>0.138</td>\n",
       "      <td>0.079</td>\n",
       "      <td>0.013</td>\n",
       "      <td>0.036</td>\n",
       "      <td>0.104</td>\n",
       "      <td>0.085</td>\n",
       "      <td>0.006</td>\n",
       "      <td>0.256</td>\n",
       "      <td>0.286</td>\n",
       "      <td>...</td>\n",
       "      <td>0.074</td>\n",
       "      <td>0.068</td>\n",
       "      <td>0.008</td>\n",
       "      <td>0.032</td>\n",
       "      <td>0.078</td>\n",
       "      <td>0.119</td>\n",
       "      <td>0.018</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.059</td>\n",
       "      <td>0.138</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.140</td>\n",
       "      <td>0.163</td>\n",
       "      <td>0.105</td>\n",
       "      <td>0.064</td>\n",
       "      <td>0.085</td>\n",
       "      <td>0.128</td>\n",
       "      <td>0.128</td>\n",
       "      <td>0.031</td>\n",
       "      <td>0.294</td>\n",
       "      <td>0.348</td>\n",
       "      <td>...</td>\n",
       "      <td>0.082</td>\n",
       "      <td>0.088</td>\n",
       "      <td>0.027</td>\n",
       "      <td>0.033</td>\n",
       "      <td>0.083</td>\n",
       "      <td>0.133</td>\n",
       "      <td>0.030</td>\n",
       "      <td>0.061</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.145</td>\n",
       "      <td>0.181</td>\n",
       "      <td>0.120</td>\n",
       "      <td>0.121</td>\n",
       "      <td>0.107</td>\n",
       "      <td>0.144</td>\n",
       "      <td>0.147</td>\n",
       "      <td>0.124</td>\n",
       "      <td>0.328</td>\n",
       "      <td>0.385</td>\n",
       "      <td>...</td>\n",
       "      <td>0.098</td>\n",
       "      <td>0.092</td>\n",
       "      <td>0.058</td>\n",
       "      <td>0.035</td>\n",
       "      <td>0.086</td>\n",
       "      <td>0.138</td>\n",
       "      <td>0.087</td>\n",
       "      <td>0.061</td>\n",
       "      <td>0.068</td>\n",
       "      <td>0.150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.162</td>\n",
       "      <td>0.182</td>\n",
       "      <td>0.140</td>\n",
       "      <td>0.142</td>\n",
       "      <td>0.109</td>\n",
       "      <td>0.155</td>\n",
       "      <td>0.166</td>\n",
       "      <td>0.142</td>\n",
       "      <td>0.350</td>\n",
       "      <td>0.408</td>\n",
       "      <td>...</td>\n",
       "      <td>0.102</td>\n",
       "      <td>0.095</td>\n",
       "      <td>0.074</td>\n",
       "      <td>0.040</td>\n",
       "      <td>0.089</td>\n",
       "      <td>0.147</td>\n",
       "      <td>0.100</td>\n",
       "      <td>0.070</td>\n",
       "      <td>0.071</td>\n",
       "      <td>0.153</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.173</td>\n",
       "      <td>0.195</td>\n",
       "      <td>0.165</td>\n",
       "      <td>0.167</td>\n",
       "      <td>0.114</td>\n",
       "      <td>0.181</td>\n",
       "      <td>0.168</td>\n",
       "      <td>0.145</td>\n",
       "      <td>0.353</td>\n",
       "      <td>0.444</td>\n",
       "      <td>...</td>\n",
       "      <td>0.105</td>\n",
       "      <td>0.098</td>\n",
       "      <td>0.099</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.097</td>\n",
       "      <td>0.152</td>\n",
       "      <td>0.117</td>\n",
       "      <td>0.074</td>\n",
       "      <td>0.074</td>\n",
       "      <td>0.158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.175</td>\n",
       "      <td>0.197</td>\n",
       "      <td>0.176</td>\n",
       "      <td>0.181</td>\n",
       "      <td>0.120</td>\n",
       "      <td>0.196</td>\n",
       "      <td>0.182</td>\n",
       "      <td>0.151</td>\n",
       "      <td>0.372</td>\n",
       "      <td>0.503</td>\n",
       "      <td>...</td>\n",
       "      <td>0.109</td>\n",
       "      <td>0.107</td>\n",
       "      <td>0.118</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.102</td>\n",
       "      <td>0.154</td>\n",
       "      <td>0.128</td>\n",
       "      <td>0.076</td>\n",
       "      <td>0.077</td>\n",
       "      <td>0.161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.195</td>\n",
       "      <td>0.213</td>\n",
       "      <td>0.201</td>\n",
       "      <td>0.214</td>\n",
       "      <td>0.124</td>\n",
       "      <td>0.214</td>\n",
       "      <td>0.210</td>\n",
       "      <td>0.174</td>\n",
       "      <td>0.394</td>\n",
       "      <td>0.509</td>\n",
       "      <td>...</td>\n",
       "      <td>0.112</td>\n",
       "      <td>0.108</td>\n",
       "      <td>0.119</td>\n",
       "      <td>0.049</td>\n",
       "      <td>0.109</td>\n",
       "      <td>0.154</td>\n",
       "      <td>0.132</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.082</td>\n",
       "      <td>0.176</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.197</td>\n",
       "      <td>0.254</td>\n",
       "      <td>0.204</td>\n",
       "      <td>0.225</td>\n",
       "      <td>0.126</td>\n",
       "      <td>0.218</td>\n",
       "      <td>0.218</td>\n",
       "      <td>0.176</td>\n",
       "      <td>0.419</td>\n",
       "      <td>0.513</td>\n",
       "      <td>...</td>\n",
       "      <td>0.117</td>\n",
       "      <td>0.108</td>\n",
       "      <td>0.133</td>\n",
       "      <td>0.050</td>\n",
       "      <td>0.114</td>\n",
       "      <td>0.173</td>\n",
       "      <td>0.138</td>\n",
       "      <td>0.101</td>\n",
       "      <td>0.108</td>\n",
       "      <td>0.177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.200</td>\n",
       "      <td>0.259</td>\n",
       "      <td>0.205</td>\n",
       "      <td>0.228</td>\n",
       "      <td>0.129</td>\n",
       "      <td>0.223</td>\n",
       "      <td>0.251</td>\n",
       "      <td>0.224</td>\n",
       "      <td>0.419</td>\n",
       "      <td>0.546</td>\n",
       "      <td>...</td>\n",
       "      <td>0.125</td>\n",
       "      <td>0.108</td>\n",
       "      <td>0.133</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.116</td>\n",
       "      <td>0.175</td>\n",
       "      <td>0.148</td>\n",
       "      <td>0.120</td>\n",
       "      <td>0.108</td>\n",
       "      <td>0.185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.215</td>\n",
       "      <td>0.265</td>\n",
       "      <td>0.228</td>\n",
       "      <td>0.237</td>\n",
       "      <td>0.132</td>\n",
       "      <td>0.223</td>\n",
       "      <td>0.263</td>\n",
       "      <td>0.228</td>\n",
       "      <td>0.481</td>\n",
       "      <td>0.565</td>\n",
       "      <td>...</td>\n",
       "      <td>0.128</td>\n",
       "      <td>0.113</td>\n",
       "      <td>0.144</td>\n",
       "      <td>0.082</td>\n",
       "      <td>0.130</td>\n",
       "      <td>0.178</td>\n",
       "      <td>0.165</td>\n",
       "      <td>0.126</td>\n",
       "      <td>0.111</td>\n",
       "      <td>0.201</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   UNCERTAINTYclefw2vecMLPre@10  UNCERTAINTYclefw2vecRFre@10  \\\n",
       "0                         0.097                        0.138   \n",
       "1                         0.140                        0.163   \n",
       "2                         0.145                        0.181   \n",
       "3                         0.162                        0.182   \n",
       "4                         0.173                        0.195   \n",
       "5                         0.175                        0.197   \n",
       "6                         0.195                        0.213   \n",
       "7                         0.197                        0.254   \n",
       "8                         0.200                        0.259   \n",
       "9                         0.215                        0.265   \n",
       "\n",
       "   UNCERTAINTYclefw2vecLRre@10  UNCERTAINTYclefw2vecSVMrbfre@10  \\\n",
       "0                        0.079                            0.013   \n",
       "1                        0.105                            0.064   \n",
       "2                        0.120                            0.121   \n",
       "3                        0.140                            0.142   \n",
       "4                        0.165                            0.167   \n",
       "5                        0.176                            0.181   \n",
       "6                        0.201                            0.214   \n",
       "7                        0.204                            0.225   \n",
       "8                        0.205                            0.228   \n",
       "9                        0.228                            0.237   \n",
       "\n",
       "   UNCERTAINTYclefgloveMLPre@10  UNCERTAINTYclefgloveRFre@10  \\\n",
       "0                         0.036                        0.104   \n",
       "1                         0.085                        0.128   \n",
       "2                         0.107                        0.144   \n",
       "3                         0.109                        0.155   \n",
       "4                         0.114                        0.181   \n",
       "5                         0.120                        0.196   \n",
       "6                         0.124                        0.214   \n",
       "7                         0.126                        0.218   \n",
       "8                         0.129                        0.223   \n",
       "9                         0.132                        0.223   \n",
       "\n",
       "   UNCERTAINTYclefgloveLRre@10  UNCERTAINTYclefgloveSVMrbfre@10  \\\n",
       "0                        0.085                            0.006   \n",
       "1                        0.128                            0.031   \n",
       "2                        0.147                            0.124   \n",
       "3                        0.166                            0.142   \n",
       "4                        0.168                            0.145   \n",
       "5                        0.182                            0.151   \n",
       "6                        0.210                            0.174   \n",
       "7                        0.218                            0.176   \n",
       "8                        0.251                            0.224   \n",
       "9                        0.263                            0.228   \n",
       "\n",
       "   UNCERTAINTYclefBERTMLPre@10  UNCERTAINTYclefBERTRFre@10  \\\n",
       "0                        0.256                       0.286   \n",
       "1                        0.294                       0.348   \n",
       "2                        0.328                       0.385   \n",
       "3                        0.350                       0.408   \n",
       "4                        0.353                       0.444   \n",
       "5                        0.372                       0.503   \n",
       "6                        0.394                       0.509   \n",
       "7                        0.419                       0.513   \n",
       "8                        0.419                       0.546   \n",
       "9                        0.481                       0.565   \n",
       "\n",
       "             ...             RANDOMclefgloveRFre@10  RANDOMclefgloveLRre@10  \\\n",
       "0            ...                              0.074                   0.068   \n",
       "1            ...                              0.082                   0.088   \n",
       "2            ...                              0.098                   0.092   \n",
       "3            ...                              0.102                   0.095   \n",
       "4            ...                              0.105                   0.098   \n",
       "5            ...                              0.109                   0.107   \n",
       "6            ...                              0.112                   0.108   \n",
       "7            ...                              0.117                   0.108   \n",
       "8            ...                              0.125                   0.108   \n",
       "9            ...                              0.128                   0.113   \n",
       "\n",
       "   RANDOMclefgloveSVMrbfre@10  RANDOMclefBERTMLPre@10  RANDOMclefBERTRFre@10  \\\n",
       "0                       0.008                   0.032                  0.078   \n",
       "1                       0.027                   0.033                  0.083   \n",
       "2                       0.058                   0.035                  0.086   \n",
       "3                       0.074                   0.040                  0.089   \n",
       "4                       0.099                   0.044                  0.097   \n",
       "5                       0.118                   0.048                  0.102   \n",
       "6                       0.119                   0.049                  0.109   \n",
       "7                       0.133                   0.050                  0.114   \n",
       "8                       0.133                   0.057                  0.116   \n",
       "9                       0.144                   0.082                  0.130   \n",
       "\n",
       "   RANDOMclefBERTLRre@10  RANDOMclefBERTSVMrbfre@10  \\\n",
       "0                  0.119                      0.018   \n",
       "1                  0.133                      0.030   \n",
       "2                  0.138                      0.087   \n",
       "3                  0.147                      0.100   \n",
       "4                  0.152                      0.117   \n",
       "5                  0.154                      0.128   \n",
       "6                  0.154                      0.132   \n",
       "7                  0.173                      0.138   \n",
       "8                  0.175                      0.148   \n",
       "9                  0.178                      0.165   \n",
       "\n",
       "   RANDOMclefBioBERTMLPre@10  RANDOMclefBioBERTRFre@10  \\\n",
       "0                      0.057                     0.059   \n",
       "1                      0.061                     0.060   \n",
       "2                      0.061                     0.068   \n",
       "3                      0.070                     0.071   \n",
       "4                      0.074                     0.074   \n",
       "5                      0.076                     0.077   \n",
       "6                      0.080                     0.082   \n",
       "7                      0.101                     0.108   \n",
       "8                      0.120                     0.108   \n",
       "9                      0.126                     0.111   \n",
       "\n",
       "   RANDOMclefBioBERTLRre@10  \n",
       "0                     0.138  \n",
       "1                     0.148  \n",
       "2                     0.150  \n",
       "3                     0.153  \n",
       "4                     0.158  \n",
       "5                     0.161  \n",
       "6                     0.176  \n",
       "7                     0.177  \n",
       "8                     0.185  \n",
       "9                     0.201  \n",
       "\n",
       "[10 rows x 31 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('results/clef_results.csv', sep=';')\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "run_control": {
     "marked": false
    }
   },
   "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>UNCERTAINTYclefBERTMLPre@10</th>\n",
       "      <th>UNCERTAINTYclefBERTSVMrbfre@10</th>\n",
       "      <th>UNCERTAINTYclefBERTLRre@10</th>\n",
       "      <th>UNCERTAINTYclefBERTRFre@10</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.256</td>\n",
       "      <td>0.042</td>\n",
       "      <td>0.224</td>\n",
       "      <td>0.286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.294</td>\n",
       "      <td>0.214</td>\n",
       "      <td>0.350</td>\n",
       "      <td>0.348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.328</td>\n",
       "      <td>0.252</td>\n",
       "      <td>0.357</td>\n",
       "      <td>0.385</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.350</td>\n",
       "      <td>0.382</td>\n",
       "      <td>0.379</td>\n",
       "      <td>0.408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.353</td>\n",
       "      <td>0.427</td>\n",
       "      <td>0.421</td>\n",
       "      <td>0.444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.372</td>\n",
       "      <td>0.469</td>\n",
       "      <td>0.493</td>\n",
       "      <td>0.503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.394</td>\n",
       "      <td>0.485</td>\n",
       "      <td>0.521</td>\n",
       "      <td>0.509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.419</td>\n",
       "      <td>0.505</td>\n",
       "      <td>0.537</td>\n",
       "      <td>0.513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.419</td>\n",
       "      <td>0.547</td>\n",
       "      <td>0.543</td>\n",
       "      <td>0.546</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.481</td>\n",
       "      <td>0.560</td>\n",
       "      <td>0.561</td>\n",
       "      <td>0.565</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   UNCERTAINTYclefBERTMLPre@10  UNCERTAINTYclefBERTSVMrbfre@10  \\\n",
       "0                        0.256                           0.042   \n",
       "1                        0.294                           0.214   \n",
       "2                        0.328                           0.252   \n",
       "3                        0.350                           0.382   \n",
       "4                        0.353                           0.427   \n",
       "5                        0.372                           0.469   \n",
       "6                        0.394                           0.485   \n",
       "7                        0.419                           0.505   \n",
       "8                        0.419                           0.547   \n",
       "9                        0.481                           0.560   \n",
       "\n",
       "   UNCERTAINTYclefBERTLRre@10  UNCERTAINTYclefBERTRFre@10  \n",
       "0                       0.224                       0.286  \n",
       "1                       0.350                       0.348  \n",
       "2                       0.357                       0.385  \n",
       "3                       0.379                       0.408  \n",
       "4                       0.421                       0.444  \n",
       "5                       0.493                       0.503  \n",
       "6                       0.521                       0.509  \n",
       "7                       0.537                       0.513  \n",
       "8                       0.543                       0.546  \n",
       "9                       0.561                       0.565  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "al = 'UNCERTAINTY' # choose UNCERTAINTY or RANDOM active learning strategy \n",
    "dataset = 'clef' # choose clef or episte \n",
    "language_model = 'BERT' # choose glove, w2vec, BERT, BioBERT\n",
    "models = ['MLP', 'SVMrbf', 'LR', 'RF']\n",
    "metric = 're@10' # choose r@10, r@20, r@30, p@10, p@20, p@30 , MAP, lastrel, WSS\n",
    "\n",
    "COLUMNS = [al + dataset + language_model + m + metric for m in models ]\n",
    "\n",
    "\n",
    "# standard deviation \n",
    "df_stdevs = pd.read_csv('results/{}_results_stdevs.csv'.format(dataset), sep=';' )\n",
    "\n",
    "df[COLUMNS]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>UNCERTAINTYclefw2vecMLPre@10</th>\n",
       "      <th>UNCERTAINTYclefw2vecRFre@10</th>\n",
       "      <th>UNCERTAINTYclefw2vecLRre@10</th>\n",
       "      <th>UNCERTAINTYclefw2vecSVMrbfre@10</th>\n",
       "      <th>UNCERTAINTYclefgloveMLPre@10</th>\n",
       "      <th>UNCERTAINTYclefgloveRFre@10</th>\n",
       "      <th>UNCERTAINTYclefgloveLRre@10</th>\n",
       "      <th>UNCERTAINTYclefgloveSVMrbfre@10</th>\n",
       "      <th>UNCERTAINTYclefBERTMLPre@10</th>\n",
       "      <th>UNCERTAINTYclefBERTRFre@10</th>\n",
       "      <th>...</th>\n",
       "      <th>RANDOMclefgloveLRre@10</th>\n",
       "      <th>RANDOMclefgloveSVMrbfre@10</th>\n",
       "      <th>RANDOMclefBERTMLPre@10</th>\n",
       "      <th>RANDOMclefBERTRFre@10</th>\n",
       "      <th>RANDOMclefBERTLRre@10</th>\n",
       "      <th>RANDOMclefBERTSVMrbfre@10</th>\n",
       "      <th>RANDOMclefBioBERTMLPre@10</th>\n",
       "      <th>RANDOMclefBioBERTRFre@10</th>\n",
       "      <th>RANDOMclefBioBERTLRre@10</th>\n",
       "      <th>RANDOMclefBioBERTSVMrbfre@10</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.043</td>\n",
       "      <td>0.042</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.036</td>\n",
       "      <td>0.039</td>\n",
       "      <td>0.040</td>\n",
       "      <td>0.040</td>\n",
       "      <td>0.033</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.048</td>\n",
       "      <td>...</td>\n",
       "      <td>0.042</td>\n",
       "      <td>0.032</td>\n",
       "      <td>0.050</td>\n",
       "      <td>0.049</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.035</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.036</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.041</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.036</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.048</td>\n",
       "      <td>...</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.037</td>\n",
       "      <td>0.040</td>\n",
       "      <td>0.049</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.038</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.040</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.040</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.044</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.049</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.049</td>\n",
       "      <td>...</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.043</td>\n",
       "      <td>0.037</td>\n",
       "      <td>0.049</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.050</td>\n",
       "      <td>0.050</td>\n",
       "      <td>0.041</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.045</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.045</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.050</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.050</td>\n",
       "      <td>...</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.041</td>\n",
       "      <td>0.050</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.041</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.044</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.058</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.050</td>\n",
       "      <td>0.050</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.050</td>\n",
       "      <td>...</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.050</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.039</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.046</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.051</td>\n",
       "      <td>...</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.050</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.041</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.054</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.049</td>\n",
       "      <td>0.049</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.050</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.050</td>\n",
       "      <td>0.052</td>\n",
       "      <td>...</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.037</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.043</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.054</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.051</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.058</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.051</td>\n",
       "      <td>...</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.037</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.048</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.059</td>\n",
       "      <td>0.050</td>\n",
       "      <td>0.052</td>\n",
       "      <td>...</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.038</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.049</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.050</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.059</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.052</td>\n",
       "      <td>...</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.049</td>\n",
       "      <td>0.058</td>\n",
       "      <td>0.053</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   UNCERTAINTYclefw2vecMLPre@10  UNCERTAINTYclefw2vecRFre@10  \\\n",
       "0                         0.043                        0.042   \n",
       "1                         0.041                        0.045   \n",
       "2                         0.044                        0.048   \n",
       "3                         0.045                        0.046   \n",
       "4                         0.044                        0.048   \n",
       "5                         0.046                        0.048   \n",
       "6                         0.049                        0.049   \n",
       "7                         0.051                        0.053   \n",
       "8                         0.048                        0.054   \n",
       "9                         0.050                        0.052   \n",
       "\n",
       "   UNCERTAINTYclefw2vecLRre@10  UNCERTAINTYclefw2vecSVMrbfre@10  \\\n",
       "0                        0.044                            0.036   \n",
       "1                        0.046                            0.045   \n",
       "2                        0.051                            0.049   \n",
       "3                        0.047                            0.052   \n",
       "4                        0.058                            0.052   \n",
       "5                        0.054                            0.051   \n",
       "6                        0.055                            0.052   \n",
       "7                        0.058                            0.052   \n",
       "8                        0.054                            0.052   \n",
       "9                        0.054                            0.055   \n",
       "\n",
       "   UNCERTAINTYclefgloveMLPre@10  UNCERTAINTYclefgloveRFre@10  \\\n",
       "0                         0.039                        0.040   \n",
       "1                         0.051                        0.046   \n",
       "2                         0.051                        0.047   \n",
       "3                         0.050                        0.047   \n",
       "4                         0.050                        0.050   \n",
       "5                         0.051                        0.051   \n",
       "6                         0.050                        0.053   \n",
       "7                         0.051                        0.053   \n",
       "8                         0.051                        0.051   \n",
       "9                         0.051                        0.053   \n",
       "\n",
       "   UNCERTAINTYclefgloveLRre@10  UNCERTAINTYclefgloveSVMrbfre@10  \\\n",
       "0                        0.040                            0.033   \n",
       "1                        0.045                            0.036   \n",
       "2                        0.046                            0.051   \n",
       "3                        0.047                            0.053   \n",
       "4                        0.048                            0.052   \n",
       "5                        0.053                            0.052   \n",
       "6                        0.052                            0.053   \n",
       "7                        0.052                            0.053   \n",
       "8                        0.053                            0.059   \n",
       "9                        0.051                            0.059   \n",
       "\n",
       "   UNCERTAINTYclefBERTMLPre@10  UNCERTAINTYclefBERTRFre@10  \\\n",
       "0                        0.053                       0.048   \n",
       "1                        0.046                       0.048   \n",
       "2                        0.046                       0.049   \n",
       "3                        0.053                       0.050   \n",
       "4                        0.052                       0.050   \n",
       "5                        0.054                       0.051   \n",
       "6                        0.050                       0.052   \n",
       "7                        0.044                       0.051   \n",
       "8                        0.050                       0.052   \n",
       "9                        0.053                       0.052   \n",
       "\n",
       "               ...               RANDOMclefgloveLRre@10  \\\n",
       "0              ...                                0.042   \n",
       "1              ...                                0.045   \n",
       "2              ...                                0.045   \n",
       "3              ...                                0.044   \n",
       "4              ...                                0.044   \n",
       "5              ...                                0.045   \n",
       "6              ...                                0.046   \n",
       "7              ...                                0.046   \n",
       "8              ...                                0.046   \n",
       "9              ...                                0.046   \n",
       "\n",
       "   RANDOMclefgloveSVMrbfre@10  RANDOMclefBERTMLPre@10  RANDOMclefBERTRFre@10  \\\n",
       "0                       0.032                   0.050                  0.049   \n",
       "1                       0.037                   0.040                  0.049   \n",
       "2                       0.043                   0.037                  0.049   \n",
       "3                       0.045                   0.041                  0.050   \n",
       "4                       0.047                   0.044                  0.051   \n",
       "5                       0.047                   0.045                  0.050   \n",
       "6                       0.046                   0.037                  0.052   \n",
       "7                       0.047                   0.037                  0.053   \n",
       "8                       0.047                   0.038                  0.053   \n",
       "9                       0.048                   0.044                  0.054   \n",
       "\n",
       "   RANDOMclefBERTLRre@10  RANDOMclefBERTSVMrbfre@10  \\\n",
       "0                  0.053                      0.035   \n",
       "1                  0.053                      0.038   \n",
       "2                  0.051                      0.050   \n",
       "3                  0.055                      0.051   \n",
       "4                  0.052                      0.050   \n",
       "5                  0.052                      0.052   \n",
       "6                  0.052                      0.054   \n",
       "7                  0.053                      0.054   \n",
       "8                  0.053                      0.055   \n",
       "9                  0.053                      0.054   \n",
       "\n",
       "   RANDOMclefBioBERTMLPre@10  RANDOMclefBioBERTRFre@10  \\\n",
       "0                      0.054                     0.044   \n",
       "1                      0.054                     0.040   \n",
       "2                      0.050                     0.041   \n",
       "3                      0.048                     0.041   \n",
       "4                      0.044                     0.039   \n",
       "5                      0.051                     0.041   \n",
       "6                      0.046                     0.043   \n",
       "7                      0.045                     0.046   \n",
       "8                      0.046                     0.049   \n",
       "9                      0.047                     0.049   \n",
       "\n",
       "   RANDOMclefBioBERTLRre@10  RANDOMclefBioBERTSVMrbfre@10  \n",
       "0                     0.053                         0.036  \n",
       "1                     0.054                         0.040  \n",
       "2                     0.054                         0.045  \n",
       "3                     0.054                         0.049  \n",
       "4                     0.054                         0.053  \n",
       "5                     0.054                         0.054  \n",
       "6                     0.055                         0.054  \n",
       "7                     0.055                         0.055  \n",
       "8                     0.055                         0.055  \n",
       "9                     0.058                         0.053  \n",
       "\n",
       "[10 rows x 32 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_stdevs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/ipykernel_launcher.py:7: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  import sys\n"
     ]
    },
    {
     "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>BERTMLP</th>\n",
       "      <th>BERTSVMrbf</th>\n",
       "      <th>BERTLR</th>\n",
       "      <th>BERTRF</th>\n",
       "      <th>x</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.256</td>\n",
       "      <td>0.042</td>\n",
       "      <td>0.224</td>\n",
       "      <td>0.286</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.294</td>\n",
       "      <td>0.214</td>\n",
       "      <td>0.350</td>\n",
       "      <td>0.348</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.328</td>\n",
       "      <td>0.252</td>\n",
       "      <td>0.357</td>\n",
       "      <td>0.385</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.350</td>\n",
       "      <td>0.382</td>\n",
       "      <td>0.379</td>\n",
       "      <td>0.408</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.353</td>\n",
       "      <td>0.427</td>\n",
       "      <td>0.421</td>\n",
       "      <td>0.444</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.372</td>\n",
       "      <td>0.469</td>\n",
       "      <td>0.493</td>\n",
       "      <td>0.503</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.394</td>\n",
       "      <td>0.485</td>\n",
       "      <td>0.521</td>\n",
       "      <td>0.509</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.419</td>\n",
       "      <td>0.505</td>\n",
       "      <td>0.537</td>\n",
       "      <td>0.513</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.419</td>\n",
       "      <td>0.547</td>\n",
       "      <td>0.543</td>\n",
       "      <td>0.546</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.481</td>\n",
       "      <td>0.560</td>\n",
       "      <td>0.561</td>\n",
       "      <td>0.565</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  BERTMLP BERTSVMrbf BERTLR BERTRF   x\n",
       "0   0.256      0.042  0.224  0.286   1\n",
       "1   0.294      0.214  0.350  0.348   2\n",
       "2   0.328      0.252  0.357  0.385   3\n",
       "3   0.350      0.382  0.379  0.408   4\n",
       "4   0.353      0.427  0.421  0.444   5\n",
       "5   0.372      0.469  0.493  0.503   6\n",
       "6   0.394      0.485  0.521  0.509   7\n",
       "7   0.419      0.505  0.537  0.513   8\n",
       "8   0.419      0.547  0.543  0.546   9\n",
       "9   0.481      0.560  0.561  0.565  10"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_results = df[COLUMNS]\n",
    "df_stdevs_results = df_stdevs[COLUMNS]\n",
    "\n",
    "#change columns names \n",
    "df_results.columns = [[ language_model + m for m in models ]]\n",
    "\n",
    "df_results['x'] = range(1,11)\n",
    "\n",
    "df_results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.errorbar(\n",
    "    [i for i in range(10)],  # X\n",
    "    [x[0] for x in df_results['{}{}'.format(language_model, models[0])].values.tolist()], # Y\n",
    "    yerr=sum(df_stdevs_results[al + dataset + language_model + models[0] + metric])/10,     # Y-average-errors\n",
    "    fmt=\"o\", # format line like for plot()\n",
    "    linewidth=1,# width of plot line\n",
    "    markersize=8,\n",
    "    color = 'blue',\n",
    "    capsize=5,\n",
    "    ecolor='blue',\n",
    "    label='{}{}'.format(language_model,models[0]),\n",
    "    linestyle='dashed'\n",
    "    )\n",
    "\n",
    "plt.errorbar(\n",
    "    [i for i in range(10)],  # X\n",
    "    [x[0] for x in df_results['{}{}'.format(language_model, models[1])].values.tolist()], # Y\n",
    "    yerr=sum(df_stdevs_results[al + dataset + language_model + models[1] + metric])/10,     # Y-errors\n",
    "    fmt=\"o\", # format line like for plot()\n",
    "    linewidth=1,# width of plot line\n",
    "    markersize=8,\n",
    "    color = 'green',\n",
    "    capsize=5,\n",
    "    ecolor='green',\n",
    "    label='{}{}'.format(language_model,models[1]),\n",
    "    linestyle='dashed'\n",
    ")\n",
    "\n",
    "\n",
    "plt.errorbar(\n",
    "    [i for i in range(10)],  # X\n",
    "    [x[0] for x in df_results['{}{}'.format(language_model, models[2])].values.tolist()], # Y\n",
    "    yerr=sum(df_stdevs_results[al + dataset + language_model + models[2] + metric])/10,     # Y-errors\n",
    "    fmt=\"o\", # format line like for plot()\n",
    "    linewidth=1,# width of plot line\n",
    "    markersize=8,\n",
    "    color = 'orange',\n",
    "    capsize=5,\n",
    "    ecolor='orange',\n",
    "    label='{}{}'.format(language_model,models[2]),\n",
    "    linestyle='dashed'\n",
    ")\n",
    "\n",
    "plt.errorbar(\n",
    "    [i for i in range(10)],  # X\n",
    "    [x[0] for x in df_results['{}{}'.format(language_model, models[3])].values.tolist()], # Y\n",
    "    yerr=sum(df_stdevs_results[al + dataset + language_model + models[3] + metric])/10,     # Y-errors\n",
    "    fmt=\"o\", # format line like for plot()\n",
    "    linewidth=1,# width of plot line\n",
    "    markersize=8,\n",
    "    color = 'red',\n",
    "    capsize=5,\n",
    "    ecolor='red',\n",
    "    label='{}{}'.format(language_model,models[3]),\n",
    "    linestyle='dashed'\n",
    ")\n",
    "\n",
    "\n",
    "ylim(0,0.7)\n",
    "\n",
    "\n",
    "plt.ylabel(\"recall@10\")\n",
    "plt.xlabel(\"iterations\")\n",
    "plt.legend(loc='upper right')\n",
    "plt.title(' {} document representation {} Sampling'.format(language_model.capitalize(), al.capitalize()) )\n",
    "\n",
    "\n",
    "# save plot\n",
    "plt.savefig('{}_{}_{}.eps'.format(dataset, language_model, al ), format='eps')\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.0"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}