From dd2bf52934121ac3204c002d167271b804c74dce Mon Sep 17 00:00:00 2001
From: rbusche <rbusche@inserve.de>
Date: Sat, 20 Jun 2020 17:48:54 +0200
Subject: [PATCH 1/3] Add plot for datset sizes

---
 notebooks/data/data_set_and_label_sizes.csv |  25 ++
 notebooks/dataset_plot.ipynb                | 347 ++++++++++++++++++++
 2 files changed, 372 insertions(+)
 create mode 100644 notebooks/data/data_set_and_label_sizes.csv
 create mode 100644 notebooks/dataset_plot.ipynb

diff --git a/notebooks/data/data_set_and_label_sizes.csv b/notebooks/data/data_set_and_label_sizes.csv
new file mode 100644
index 0000000..482e4ed
--- /dev/null
+++ b/notebooks/data/data_set_and_label_sizes.csv
@@ -0,0 +1,25 @@
+,class,task,dataset,size
+0,positive,relevance_scoring,train,126.0
+1,positive,relevance_scoring,test,38.0
+2,positive,relevance_scoring,validation,
+3,positive,key_date_extraction,train,82.0
+4,positive,key_date_extraction,test,27.0
+5,positive,key_date_extraction,validation,
+6,positive,key_count_extraction,train,257.0
+7,positive,key_count_extraction,test,98.0
+8,positive,key_count_extraction,validation,
+9,positive,relevance_scoring_CNN,train,67.0
+10,positive,relevance_scoring_CNN,test,42.0
+11,positive,relevance_scoring_CNN,validation,55.0
+12,negative,relevance_scoring,train,2301.0
+13,negative,relevance_scoring,test,771.0
+14,negative,relevance_scoring,validation,
+15,negative,key_date_extraction,train,161.0
+16,negative,key_date_extraction,test,54.0
+17,negative,key_date_extraction,validation,
+18,negative,key_count_extraction,train,2629.0
+19,negative,key_count_extraction,test,865.0
+20,negative,key_count_extraction,validation,
+21,negative,relevance_scoring_CNN,train,1874.0
+22,negative,relevance_scoring_CNN,test,606.0
+23,negative,relevance_scoring_CNN,validation,592.0
diff --git a/notebooks/dataset_plot.ipynb b/notebooks/dataset_plot.ipynb
new file mode 100644
index 0000000..5b1a084
--- /dev/null
+++ b/notebooks/dataset_plot.ipynb
@@ -0,0 +1,347 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 30,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "import seaborn as sns\n",
+    "import matplotlib.pyplot as plt"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 24,
+   "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>class</th>\n",
+       "      <th>task</th>\n",
+       "      <th>dataset</th>\n",
+       "      <th>size</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>relevance scoring</td>\n",
+       "      <td>train</td>\n",
+       "      <td>126.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>relevance scoring</td>\n",
+       "      <td>test</td>\n",
+       "      <td>38.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>relevance scoring</td>\n",
+       "      <td>validation</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>key date extraction</td>\n",
+       "      <td>train</td>\n",
+       "      <td>82.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>key date extraction</td>\n",
+       "      <td>test</td>\n",
+       "      <td>27.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>key date extraction</td>\n",
+       "      <td>validation</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>key count extraction</td>\n",
+       "      <td>train</td>\n",
+       "      <td>257.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>key count extraction</td>\n",
+       "      <td>test</td>\n",
+       "      <td>98.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>key count extraction</td>\n",
+       "      <td>validation</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>relevance scoring CNN</td>\n",
+       "      <td>train</td>\n",
+       "      <td>67.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>10</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>relevance scoring CNN</td>\n",
+       "      <td>test</td>\n",
+       "      <td>42.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>11</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>relevance scoring CNN</td>\n",
+       "      <td>validation</td>\n",
+       "      <td>55.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>relevance scoring</td>\n",
+       "      <td>train</td>\n",
+       "      <td>2301.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>13</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>relevance scoring</td>\n",
+       "      <td>test</td>\n",
+       "      <td>771.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>14</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>relevance scoring</td>\n",
+       "      <td>validation</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>15</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>key date extraction</td>\n",
+       "      <td>train</td>\n",
+       "      <td>161.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>16</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>key date extraction</td>\n",
+       "      <td>test</td>\n",
+       "      <td>54.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>17</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>key date extraction</td>\n",
+       "      <td>validation</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>18</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>key count extraction</td>\n",
+       "      <td>train</td>\n",
+       "      <td>2629.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>19</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>key count extraction</td>\n",
+       "      <td>test</td>\n",
+       "      <td>865.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>20</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>key count extraction</td>\n",
+       "      <td>validation</td>\n",
+       "      <td>NaN</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>21</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>relevance scoring CNN</td>\n",
+       "      <td>train</td>\n",
+       "      <td>1874.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>22</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>relevance scoring CNN</td>\n",
+       "      <td>test</td>\n",
+       "      <td>606.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>23</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>relevance scoring CNN</td>\n",
+       "      <td>validation</td>\n",
+       "      <td>592.0</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "       class                   task     dataset    size\n",
+       "0   positive      relevance scoring       train   126.0\n",
+       "1   positive      relevance scoring        test    38.0\n",
+       "2   positive      relevance scoring  validation     NaN\n",
+       "3   positive    key date extraction       train    82.0\n",
+       "4   positive    key date extraction        test    27.0\n",
+       "5   positive    key date extraction  validation     NaN\n",
+       "6   positive   key count extraction       train   257.0\n",
+       "7   positive   key count extraction        test    98.0\n",
+       "8   positive   key count extraction  validation     NaN\n",
+       "9   positive  relevance scoring CNN       train    67.0\n",
+       "10  positive  relevance scoring CNN        test    42.0\n",
+       "11  positive  relevance scoring CNN  validation    55.0\n",
+       "12  negative      relevance scoring       train  2301.0\n",
+       "13  negative      relevance scoring        test   771.0\n",
+       "14  negative      relevance scoring  validation     NaN\n",
+       "15  negative    key date extraction       train   161.0\n",
+       "16  negative    key date extraction        test    54.0\n",
+       "17  negative    key date extraction  validation     NaN\n",
+       "18  negative   key count extraction       train  2629.0\n",
+       "19  negative   key count extraction        test   865.0\n",
+       "20  negative   key count extraction  validation     NaN\n",
+       "21  negative  relevance scoring CNN       train  1874.0\n",
+       "22  negative  relevance scoring CNN        test   606.0\n",
+       "23  negative  relevance scoring CNN  validation   592.0"
+      ]
+     },
+     "execution_count": 24,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "data = pd.read_csv(\"data/data_set_and_label_sizes.csv\").drop(columns=\"Unnamed: 0\")\n",
+    "data = data.assign(task=lambda df: df[\"task\"].str.replace(\"_\", \" \"))\n",
+    "data"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 62,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<seaborn.axisgrid.FacetGrid at 0x7fe88760ff50>"
+      ]
+     },
+     "execution_count": 62,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 720x720 with 4 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.rcParams[\"axes.labelsize\"] = \"xx-large\"\n",
+    "plt.rcParams[\"xtick.labelsize\"] = \"x-large\"\n",
+    "plt.rcParams[\"ytick.labelsize\"] = \"x-large\"\n",
+    "\n",
+    "\n",
+    "g = sns.catplot(col=\"task\", x=\"class\", y=\"size\", hue=\"dataset\", kind=\"bar\", col_wrap=2, ci=None, data=data, legend=False)\n",
+    "#plt.setp(g._legend.get_texts(), fontsize=\"x-large\")\n",
+    "plt.legend(bbox_to_anchor=(1.5, 1), loc=\"center right\", borderaxespad=0., fontsize=\"x-large\")\n",
+    "g"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 63,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "g.savefig(\"data_set_and_label_sizes.png\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 47,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<matplotlib.legend.Legend at 0x7fe888aab990>"
+      ]
+     },
+     "execution_count": 47,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "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": "3.7.6"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}

From 8b864a795057a71e056ea19b3e09e311c747896d Mon Sep 17 00:00:00 2001
From: rbusche <rbusche@inserve.de>
Date: Sun, 21 Jun 2020 15:49:05 +0200
Subject: [PATCH 2/3] Adapt plot

---
 notebooks/dataset_plot.ipynb | 335 ++++++++++++++++++++++++++++++++---
 1 file changed, 306 insertions(+), 29 deletions(-)

diff --git a/notebooks/dataset_plot.ipynb b/notebooks/dataset_plot.ipynb
index 5b1a084..391a371 100644
--- a/notebooks/dataset_plot.ipynb
+++ b/notebooks/dataset_plot.ipynb
@@ -2,18 +2,19 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 30,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [],
    "source": [
     "import pandas as pd\n",
     "import seaborn as sns\n",
-    "import matplotlib.pyplot as plt"
+    "import matplotlib.pyplot as plt\n",
+    "import numpy as np"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 24,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [
     {
@@ -244,7 +245,7 @@
        "23  negative  relevance scoring CNN  validation   592.0"
       ]
      },
-     "execution_count": 24,
+     "execution_count": 2,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -257,69 +258,345 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 62,
+   "execution_count": 3,
    "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>class</th>\n",
+       "      <th>task</th>\n",
+       "      <th>dataset</th>\n",
+       "      <th>size</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>key date extraction</td>\n",
+       "      <td>train</td>\n",
+       "      <td>82.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>key date extraction</td>\n",
+       "      <td>test</td>\n",
+       "      <td>27.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>key count extraction</td>\n",
+       "      <td>train</td>\n",
+       "      <td>257.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>key count extraction</td>\n",
+       "      <td>test</td>\n",
+       "      <td>98.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>15</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>key date extraction</td>\n",
+       "      <td>train</td>\n",
+       "      <td>161.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>16</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>key date extraction</td>\n",
+       "      <td>test</td>\n",
+       "      <td>54.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>18</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>key count extraction</td>\n",
+       "      <td>train</td>\n",
+       "      <td>2629.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>19</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>key count extraction</td>\n",
+       "      <td>test</td>\n",
+       "      <td>865.0</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
       "text/plain": [
-       "<seaborn.axisgrid.FacetGrid at 0x7fe88760ff50>"
+       "       class                  task dataset    size\n",
+       "3   positive   key date extraction   train    82.0\n",
+       "4   positive   key date extraction    test    27.0\n",
+       "6   positive  key count extraction   train   257.0\n",
+       "7   positive  key count extraction    test    98.0\n",
+       "15  negative   key date extraction   train   161.0\n",
+       "16  negative   key date extraction    test    54.0\n",
+       "18  negative  key count extraction   train  2629.0\n",
+       "19  negative  key count extraction    test   865.0"
       ]
      },
-     "execution_count": 62,
+     "execution_count": 3,
      "metadata": {},
      "output_type": "execute_result"
-    },
+    }
+   ],
+   "source": [
+    "key_extraction = data.query(\"task.str.startswith('key')\").dropna()\n",
+    "key_extraction"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [
     {
      "data": {
-      "image/png": "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\n",
+      "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>class</th>\n",
+       "      <th>task</th>\n",
+       "      <th>dataset</th>\n",
+       "      <th>size</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>relevance scoring</td>\n",
+       "      <td>train</td>\n",
+       "      <td>126.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>relevance scoring</td>\n",
+       "      <td>test</td>\n",
+       "      <td>38.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>relevance scoring CNN</td>\n",
+       "      <td>train</td>\n",
+       "      <td>67.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>10</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>relevance scoring CNN</td>\n",
+       "      <td>test</td>\n",
+       "      <td>42.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>11</th>\n",
+       "      <td>positive</td>\n",
+       "      <td>relevance scoring CNN</td>\n",
+       "      <td>validation</td>\n",
+       "      <td>55.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>relevance scoring</td>\n",
+       "      <td>train</td>\n",
+       "      <td>2301.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>13</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>relevance scoring</td>\n",
+       "      <td>test</td>\n",
+       "      <td>771.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>21</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>relevance scoring CNN</td>\n",
+       "      <td>train</td>\n",
+       "      <td>1874.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>22</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>relevance scoring CNN</td>\n",
+       "      <td>test</td>\n",
+       "      <td>606.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>23</th>\n",
+       "      <td>negative</td>\n",
+       "      <td>relevance scoring CNN</td>\n",
+       "      <td>validation</td>\n",
+       "      <td>592.0</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
       "text/plain": [
-       "<Figure size 720x720 with 4 Axes>"
+       "       class                   task     dataset    size\n",
+       "0   positive      relevance scoring       train   126.0\n",
+       "1   positive      relevance scoring        test    38.0\n",
+       "9   positive  relevance scoring CNN       train    67.0\n",
+       "10  positive  relevance scoring CNN        test    42.0\n",
+       "11  positive  relevance scoring CNN  validation    55.0\n",
+       "12  negative      relevance scoring       train  2301.0\n",
+       "13  negative      relevance scoring        test   771.0\n",
+       "21  negative  relevance scoring CNN       train  1874.0\n",
+       "22  negative  relevance scoring CNN        test   606.0\n",
+       "23  negative  relevance scoring CNN  validation   592.0"
       ]
      },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
+     "execution_count": 4,
+     "metadata": {},
+     "output_type": "execute_result"
     }
    ],
    "source": [
-    "plt.rcParams[\"axes.labelsize\"] = \"xx-large\"\n",
-    "plt.rcParams[\"xtick.labelsize\"] = \"x-large\"\n",
-    "plt.rcParams[\"ytick.labelsize\"] = \"x-large\"\n",
-    "\n",
-    "\n",
-    "g = sns.catplot(col=\"task\", x=\"class\", y=\"size\", hue=\"dataset\", kind=\"bar\", col_wrap=2, ci=None, data=data, legend=False)\n",
-    "#plt.setp(g._legend.get_texts(), fontsize=\"x-large\")\n",
-    "plt.legend(bbox_to_anchor=(1.5, 1), loc=\"center right\", borderaxespad=0., fontsize=\"x-large\")\n",
-    "g"
+    "relevance_scoring = data.query(\"task.str.startswith('relevance scoring')\").dropna()\n",
+    "relevance_scoring"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 63,
+   "execution_count": 5,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<seaborn.axisgrid.FacetGrid at 0x7f84c87e0e90>"
+      ]
+     },
+     "execution_count": 5,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 720x360 with 2 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
    "source": [
-    "g.savefig(\"data_set_and_label_sizes.png\")"
+    "def show_values_on_bars(axs):\n",
+    "    def _show_on_single_plot(ax):        \n",
+    "        for p in ax.patches:\n",
+    "            _x = p.get_x() + p.get_width() / 2\n",
+    "            _y = p.get_y() + p.get_height() + 15\n",
+    "            if pd.notna(p.get_height()):\n",
+    "                value = str(int(p.get_height()))\n",
+    "                ax.text(_x, _y, value, ha=\"center\") \n",
+    "\n",
+    "    if isinstance(axs, np.ndarray):\n",
+    "        for idx, ax in np.ndenumerate(axs):\n",
+    "            _show_on_single_plot(ax)\n",
+    "    else:\n",
+    "        _show_on_single_plot(axs)\n",
+    "\n",
+    "def dataset_plot(data):\n",
+    "    plt.rcParams[\"axes.labelsize\"] = \"xx-large\"\n",
+    "    plt.rcParams[\"xtick.labelsize\"] = \"x-large\"\n",
+    "    plt.rcParams[\"ytick.labelsize\"] = \"x-large\"\n",
+    "    \n",
+    "    with sns.axes_style(\"whitegrid\"):\n",
+    "        g = sns.catplot(col=\"task\", x=\"class\", y=\"size\", hue=\"dataset\", kind=\"bar\", col_wrap=2, ci=None, data=data, legend=False)\n",
+    "        plt.legend(bbox_to_anchor=(1.5, 1), loc=\"center right\", borderaxespad=0., fontsize=\"x-large\")\n",
+    "        show_values_on_bars(g.axes)\n",
+    "        return g\n",
+    "\n",
+    "plot = dataset_plot(key_extraction)\n",
+    "plot.savefig(\"data_set_and_label_sizes_key_extraction.png\")\n",
+    "plot"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 47,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "<matplotlib.legend.Legend at 0x7fe888aab990>"
+       "<seaborn.axisgrid.FacetGrid at 0x7f84c780f950>"
       ]
      },
-     "execution_count": 47,
+     "execution_count": 6,
      "metadata": {},
      "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 720x360 with 2 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
     }
    ],
+   "source": [
+    "plot = dataset_plot(relevance_scoring)\n",
+    "plot.savefig(\"data_set_and_label_sizes_relevance_scoring.png\")\n",
+    "plot"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
    "source": []
   }
  ],

From 4d975ea84b2016f0a33afffc1ff1092677443063 Mon Sep 17 00:00:00 2001
From: rbusche <rbusche@inserve.de>
Date: Tue, 23 Jun 2020 22:00:55 +0200
Subject: [PATCH 3/3] Save as pdf

---
 notebooks/dataset_plot.ipynb | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/notebooks/dataset_plot.ipynb b/notebooks/dataset_plot.ipynb
index 391a371..b22a1dd 100644
--- a/notebooks/dataset_plot.ipynb
+++ b/notebooks/dataset_plot.ipynb
@@ -510,7 +510,7 @@
     {
      "data": {
       "text/plain": [
-       "<seaborn.axisgrid.FacetGrid at 0x7f84c87e0e90>"
+       "<seaborn.axisgrid.FacetGrid at 0x7f8ea0abefd0>"
       ]
      },
      "execution_count": 5,
@@ -556,7 +556,7 @@
     "        return g\n",
     "\n",
     "plot = dataset_plot(key_extraction)\n",
-    "plot.savefig(\"data_set_and_label_sizes_key_extraction.png\")\n",
+    "plot.savefig(\"data_set_and_label_sizes_key_extraction.pdf\")\n",
     "plot"
    ]
   },
@@ -568,7 +568,7 @@
     {
      "data": {
       "text/plain": [
-       "<seaborn.axisgrid.FacetGrid at 0x7f84c780f950>"
+       "<seaborn.axisgrid.FacetGrid at 0x7f8e5d850650>"
       ]
      },
      "execution_count": 6,
@@ -588,7 +588,7 @@
    ],
    "source": [
     "plot = dataset_plot(relevance_scoring)\n",
-    "plot.savefig(\"data_set_and_label_sizes_relevance_scoring.png\")\n",
+    "plot.savefig(\"data_set_and_label_sizes_relevance_scoring.pdf\")\n",
     "plot"
    ]
   },