02.basic-benchmarks-seaborn.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "26df0e34",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.12.1'"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "sns.set_theme(style=\"white\")\n",
    "import matplotlib.ticker as ticker\n",
    "sns.__version__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "f10f00c9",
   "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>s</th>\n",
       "      <th>h:m:s</th>\n",
       "      <th>max_rss</th>\n",
       "      <th>max_vms</th>\n",
       "      <th>max_uss</th>\n",
       "      <th>max_pss</th>\n",
       "      <th>io_in</th>\n",
       "      <th>io_out</th>\n",
       "      <th>mean_load</th>\n",
       "      <th>cpu_time</th>\n",
       "      <th>n_query</th>\n",
       "      <th>n_against</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1347.0233</td>\n",
       "      <td>0:22:27</td>\n",
       "      <td>12910.21</td>\n",
       "      <td>15135.63</td>\n",
       "      <td>13073.72</td>\n",
       "      <td>13073.72</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.14</td>\n",
       "      <td>2762.7</td>\n",
       "      <td>37225.36</td>\n",
       "      <td>100</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           s    h:m:s   max_rss   max_vms   max_uss   max_pss  io_in  io_out  \\\n",
       "0  1347.0233  0:22:27  12910.21  15135.63  13073.72  13073.72   0.01    0.14   \n",
       "\n",
       "   mean_load  cpu_time  n_query  n_against  \n",
       "0     2762.7  37225.36      100      10000  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def load_benchmark_and_annotate(filename, n_query, n_against):\n",
    "    df = pd.read_csv(filename, sep='\\t')\n",
    "    df['n_query'] = int(n_query)\n",
    "    df['n_against'] = int(n_against)\n",
    "    \n",
    "    return df\n",
    "\n",
    "load_benchmark_and_annotate('benchmarks/a_100_vs_a.txt', 100, 10000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "723d1087",
   "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>s</th>\n",
       "      <th>h:m:s</th>\n",
       "      <th>max_rss</th>\n",
       "      <th>max_vms</th>\n",
       "      <th>max_uss</th>\n",
       "      <th>max_pss</th>\n",
       "      <th>io_in</th>\n",
       "      <th>io_out</th>\n",
       "      <th>mean_load</th>\n",
       "      <th>cpu_time</th>\n",
       "      <th>n_query</th>\n",
       "      <th>n_against</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1347.0233</td>\n",
       "      <td>0:22:27</td>\n",
       "      <td>12910.21</td>\n",
       "      <td>15135.63</td>\n",
       "      <td>13073.72</td>\n",
       "      <td>13073.72</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.14</td>\n",
       "      <td>2762.70</td>\n",
       "      <td>37225.36</td>\n",
       "      <td>100</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1291.4875</td>\n",
       "      <td>0:21:31</td>\n",
       "      <td>14541.96</td>\n",
       "      <td>18260.40</td>\n",
       "      <td>14601.04</td>\n",
       "      <td>14601.05</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.24</td>\n",
       "      <td>3044.57</td>\n",
       "      <td>39323.17</td>\n",
       "      <td>200</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1454.4474</td>\n",
       "      <td>0:24:14</td>\n",
       "      <td>16348.15</td>\n",
       "      <td>18954.26</td>\n",
       "      <td>16429.44</td>\n",
       "      <td>16429.44</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.31</td>\n",
       "      <td>2828.24</td>\n",
       "      <td>41136.09</td>\n",
       "      <td>300</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1551.7362</td>\n",
       "      <td>0:25:51</td>\n",
       "      <td>14438.77</td>\n",
       "      <td>16864.46</td>\n",
       "      <td>14666.07</td>\n",
       "      <td>14666.08</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.40</td>\n",
       "      <td>2751.87</td>\n",
       "      <td>42702.26</td>\n",
       "      <td>400</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1603.6622</td>\n",
       "      <td>0:26:43</td>\n",
       "      <td>18893.50</td>\n",
       "      <td>21000.83</td>\n",
       "      <td>18968.33</td>\n",
       "      <td>18968.33</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.48</td>\n",
       "      <td>2802.76</td>\n",
       "      <td>44947.46</td>\n",
       "      <td>500</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1696.6006</td>\n",
       "      <td>0:28:16</td>\n",
       "      <td>17783.73</td>\n",
       "      <td>19868.45</td>\n",
       "      <td>17888.35</td>\n",
       "      <td>17888.35</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.59</td>\n",
       "      <td>2755.67</td>\n",
       "      <td>46753.81</td>\n",
       "      <td>600</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1610.4366</td>\n",
       "      <td>0:26:50</td>\n",
       "      <td>16855.73</td>\n",
       "      <td>19042.42</td>\n",
       "      <td>16989.04</td>\n",
       "      <td>16989.04</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.70</td>\n",
       "      <td>3032.36</td>\n",
       "      <td>48835.22</td>\n",
       "      <td>700</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1770.2948</td>\n",
       "      <td>0:29:30</td>\n",
       "      <td>16316.45</td>\n",
       "      <td>19000.11</td>\n",
       "      <td>16440.92</td>\n",
       "      <td>16440.92</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.85</td>\n",
       "      <td>2863.60</td>\n",
       "      <td>50694.71</td>\n",
       "      <td>800</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1806.3763</td>\n",
       "      <td>0:30:06</td>\n",
       "      <td>16779.61</td>\n",
       "      <td>19569.78</td>\n",
       "      <td>16286.82</td>\n",
       "      <td>16286.82</td>\n",
       "      <td>95967.58</td>\n",
       "      <td>0.97</td>\n",
       "      <td>2889.70</td>\n",
       "      <td>52198.87</td>\n",
       "      <td>900</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           s    h:m:s   max_rss   max_vms   max_uss   max_pss     io_in  \\\n",
       "0  1347.0233  0:22:27  12910.21  15135.63  13073.72  13073.72      0.01   \n",
       "0  1291.4875  0:21:31  14541.96  18260.40  14601.04  14601.05      0.01   \n",
       "0  1454.4474  0:24:14  16348.15  18954.26  16429.44  16429.44      0.02   \n",
       "0  1551.7362  0:25:51  14438.77  16864.46  14666.07  14666.08      0.02   \n",
       "0  1603.6622  0:26:43  18893.50  21000.83  18968.33  18968.33      0.03   \n",
       "0  1696.6006  0:28:16  17783.73  19868.45  17888.35  17888.35      0.03   \n",
       "0  1610.4366  0:26:50  16855.73  19042.42  16989.04  16989.04      0.04   \n",
       "0  1770.2948  0:29:30  16316.45  19000.11  16440.92  16440.92      0.04   \n",
       "0  1806.3763  0:30:06  16779.61  19569.78  16286.82  16286.82  95967.58   \n",
       "\n",
       "   io_out  mean_load  cpu_time  n_query  n_against  \n",
       "0    0.14    2762.70  37225.36      100      10000  \n",
       "0    0.24    3044.57  39323.17      200      10000  \n",
       "0    0.31    2828.24  41136.09      300      10000  \n",
       "0    0.40    2751.87  42702.26      400      10000  \n",
       "0    0.48    2802.76  44947.46      500      10000  \n",
       "0    0.59    2755.67  46753.81      600      10000  \n",
       "0    0.70    3032.36  48835.22      700      10000  \n",
       "0    0.85    2863.60  50694.71      800      10000  \n",
       "0    0.97    2889.70  52198.87      900      10000  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "query_line = [ load_benchmark_and_annotate(f\"benchmarks/a_{x}_vs_a.txt\", x, 10000) for x in range(100, 1000, 100) ]\n",
    "#query_line.append(load_benchmark_and_annotate('benchmarks/a_vs_a.txt', 1000, 10000))\n",
    "query_line_df = pd.concat(query_line)\n",
    "query_line_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e726b596",
   "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>s</th>\n",
       "      <th>h:m:s</th>\n",
       "      <th>max_rss</th>\n",
       "      <th>max_vms</th>\n",
       "      <th>max_uss</th>\n",
       "      <th>max_pss</th>\n",
       "      <th>io_in</th>\n",
       "      <th>io_out</th>\n",
       "      <th>mean_load</th>\n",
       "      <th>cpu_time</th>\n",
       "      <th>n_query</th>\n",
       "      <th>n_against</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>154.1568</td>\n",
       "      <td>0:02:34</td>\n",
       "      <td>8026.29</td>\n",
       "      <td>8707.47</td>\n",
       "      <td>8042.18</td>\n",
       "      <td>8042.25</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.11</td>\n",
       "      <td>2112.52</td>\n",
       "      <td>3256.75</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>269.0769</td>\n",
       "      <td>0:04:29</td>\n",
       "      <td>8977.02</td>\n",
       "      <td>9920.71</td>\n",
       "      <td>8961.81</td>\n",
       "      <td>8961.88</td>\n",
       "      <td>0.13</td>\n",
       "      <td>0.25</td>\n",
       "      <td>2450.90</td>\n",
       "      <td>6594.63</td>\n",
       "      <td>1000</td>\n",
       "      <td>2000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>354.2737</td>\n",
       "      <td>0:05:54</td>\n",
       "      <td>11786.54</td>\n",
       "      <td>13038.87</td>\n",
       "      <td>11863.02</td>\n",
       "      <td>11863.11</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.36</td>\n",
       "      <td>2943.38</td>\n",
       "      <td>10428.58</td>\n",
       "      <td>1000</td>\n",
       "      <td>3000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>478.0915</td>\n",
       "      <td>0:07:58</td>\n",
       "      <td>13964.38</td>\n",
       "      <td>15485.38</td>\n",
       "      <td>14074.45</td>\n",
       "      <td>14074.53</td>\n",
       "      <td>0.26</td>\n",
       "      <td>0.48</td>\n",
       "      <td>3003.24</td>\n",
       "      <td>14358.91</td>\n",
       "      <td>1000</td>\n",
       "      <td>4000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>704.6193</td>\n",
       "      <td>0:11:44</td>\n",
       "      <td>15294.20</td>\n",
       "      <td>18001.37</td>\n",
       "      <td>15028.22</td>\n",
       "      <td>15028.30</td>\n",
       "      <td>0.33</td>\n",
       "      <td>0.64</td>\n",
       "      <td>2722.32</td>\n",
       "      <td>19182.34</td>\n",
       "      <td>1000</td>\n",
       "      <td>5000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1073.2904</td>\n",
       "      <td>0:17:53</td>\n",
       "      <td>12150.98</td>\n",
       "      <td>14048.74</td>\n",
       "      <td>12201.99</td>\n",
       "      <td>12201.99</td>\n",
       "      <td>0.39</td>\n",
       "      <td>0.73</td>\n",
       "      <td>2635.62</td>\n",
       "      <td>28288.41</td>\n",
       "      <td>1000</td>\n",
       "      <td>6000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1269.1121</td>\n",
       "      <td>0:21:09</td>\n",
       "      <td>15887.25</td>\n",
       "      <td>17586.37</td>\n",
       "      <td>15507.43</td>\n",
       "      <td>15507.43</td>\n",
       "      <td>0.46</td>\n",
       "      <td>0.79</td>\n",
       "      <td>2923.52</td>\n",
       "      <td>37104.07</td>\n",
       "      <td>1000</td>\n",
       "      <td>7000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1419.2742</td>\n",
       "      <td>0:23:39</td>\n",
       "      <td>16026.34</td>\n",
       "      <td>18511.11</td>\n",
       "      <td>16068.09</td>\n",
       "      <td>16068.09</td>\n",
       "      <td>0.52</td>\n",
       "      <td>0.89</td>\n",
       "      <td>3018.64</td>\n",
       "      <td>42844.40</td>\n",
       "      <td>1000</td>\n",
       "      <td>8000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1576.3652</td>\n",
       "      <td>0:26:16</td>\n",
       "      <td>16715.54</td>\n",
       "      <td>19374.40</td>\n",
       "      <td>16759.83</td>\n",
       "      <td>16759.83</td>\n",
       "      <td>7.45</td>\n",
       "      <td>1.00</td>\n",
       "      <td>3052.70</td>\n",
       "      <td>48122.82</td>\n",
       "      <td>1000</td>\n",
       "      <td>9000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           s    h:m:s   max_rss   max_vms   max_uss   max_pss  io_in  io_out  \\\n",
       "0   154.1568  0:02:34   8026.29   8707.47   8042.18   8042.25   0.07    0.11   \n",
       "0   269.0769  0:04:29   8977.02   9920.71   8961.81   8961.88   0.13    0.25   \n",
       "0   354.2737  0:05:54  11786.54  13038.87  11863.02  11863.11   0.20    0.36   \n",
       "0   478.0915  0:07:58  13964.38  15485.38  14074.45  14074.53   0.26    0.48   \n",
       "0   704.6193  0:11:44  15294.20  18001.37  15028.22  15028.30   0.33    0.64   \n",
       "0  1073.2904  0:17:53  12150.98  14048.74  12201.99  12201.99   0.39    0.73   \n",
       "0  1269.1121  0:21:09  15887.25  17586.37  15507.43  15507.43   0.46    0.79   \n",
       "0  1419.2742  0:23:39  16026.34  18511.11  16068.09  16068.09   0.52    0.89   \n",
       "0  1576.3652  0:26:16  16715.54  19374.40  16759.83  16759.83   7.45    1.00   \n",
       "\n",
       "   mean_load  cpu_time  n_query  n_against  \n",
       "0    2112.52   3256.75     1000       1000  \n",
       "0    2450.90   6594.63     1000       2000  \n",
       "0    2943.38  10428.58     1000       3000  \n",
       "0    3003.24  14358.91     1000       4000  \n",
       "0    2722.32  19182.34     1000       5000  \n",
       "0    2635.62  28288.41     1000       6000  \n",
       "0    2923.52  37104.07     1000       7000  \n",
       "0    3018.64  42844.40     1000       8000  \n",
       "0    3052.70  48122.82     1000       9000  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "against_line = [ load_benchmark_and_annotate(f\"benchmarks/a_vs_a_{x}.txt\", 1000, x) for x in range(1000, 10000, 1000) ]\n",
    "#against_line.append(load_benchmark_and_annotate('benchmarks/a_vs_a.txt', 1000, 10000))\n",
    "against_line_df = pd.concat(against_line)\n",
    "against_line_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "9dbb2183",
   "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>s</th>\n",
       "      <th>h:m:s</th>\n",
       "      <th>max_rss</th>\n",
       "      <th>max_vms</th>\n",
       "      <th>max_uss</th>\n",
       "      <th>max_pss</th>\n",
       "      <th>io_in</th>\n",
       "      <th>io_out</th>\n",
       "      <th>mean_load</th>\n",
       "      <th>cpu_time</th>\n",
       "      <th>n_query</th>\n",
       "      <th>n_against</th>\n",
       "      <th>bench_type</th>\n",
       "      <th>varied_n</th>\n",
       "      <th>min</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1347.0233</td>\n",
       "      <td>0:22:27</td>\n",
       "      <td>12910.21</td>\n",
       "      <td>15135.63</td>\n",
       "      <td>13073.72</td>\n",
       "      <td>13073.72</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.14</td>\n",
       "      <td>2762.70</td>\n",
       "      <td>37225.36</td>\n",
       "      <td>100</td>\n",
       "      <td>10000</td>\n",
       "      <td>vary_n_queries</td>\n",
       "      <td>100</td>\n",
       "      <td>22.450388</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1291.4875</td>\n",
       "      <td>0:21:31</td>\n",
       "      <td>14541.96</td>\n",
       "      <td>18260.40</td>\n",
       "      <td>14601.04</td>\n",
       "      <td>14601.05</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.24</td>\n",
       "      <td>3044.57</td>\n",
       "      <td>39323.17</td>\n",
       "      <td>200</td>\n",
       "      <td>10000</td>\n",
       "      <td>vary_n_queries</td>\n",
       "      <td>200</td>\n",
       "      <td>21.524792</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1454.4474</td>\n",
       "      <td>0:24:14</td>\n",
       "      <td>16348.15</td>\n",
       "      <td>18954.26</td>\n",
       "      <td>16429.44</td>\n",
       "      <td>16429.44</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.31</td>\n",
       "      <td>2828.24</td>\n",
       "      <td>41136.09</td>\n",
       "      <td>300</td>\n",
       "      <td>10000</td>\n",
       "      <td>vary_n_queries</td>\n",
       "      <td>300</td>\n",
       "      <td>24.240790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1551.7362</td>\n",
       "      <td>0:25:51</td>\n",
       "      <td>14438.77</td>\n",
       "      <td>16864.46</td>\n",
       "      <td>14666.07</td>\n",
       "      <td>14666.08</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.40</td>\n",
       "      <td>2751.87</td>\n",
       "      <td>42702.26</td>\n",
       "      <td>400</td>\n",
       "      <td>10000</td>\n",
       "      <td>vary_n_queries</td>\n",
       "      <td>400</td>\n",
       "      <td>25.862270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1603.6622</td>\n",
       "      <td>0:26:43</td>\n",
       "      <td>18893.50</td>\n",
       "      <td>21000.83</td>\n",
       "      <td>18968.33</td>\n",
       "      <td>18968.33</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.48</td>\n",
       "      <td>2802.76</td>\n",
       "      <td>44947.46</td>\n",
       "      <td>500</td>\n",
       "      <td>10000</td>\n",
       "      <td>vary_n_queries</td>\n",
       "      <td>500</td>\n",
       "      <td>26.727703</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           s    h:m:s   max_rss   max_vms   max_uss   max_pss  io_in  io_out  \\\n",
       "0  1347.0233  0:22:27  12910.21  15135.63  13073.72  13073.72   0.01    0.14   \n",
       "0  1291.4875  0:21:31  14541.96  18260.40  14601.04  14601.05   0.01    0.24   \n",
       "0  1454.4474  0:24:14  16348.15  18954.26  16429.44  16429.44   0.02    0.31   \n",
       "0  1551.7362  0:25:51  14438.77  16864.46  14666.07  14666.08   0.02    0.40   \n",
       "0  1603.6622  0:26:43  18893.50  21000.83  18968.33  18968.33   0.03    0.48   \n",
       "\n",
       "   mean_load  cpu_time  n_query  n_against      bench_type  varied_n  \\\n",
       "0    2762.70  37225.36      100      10000  vary_n_queries       100   \n",
       "0    3044.57  39323.17      200      10000  vary_n_queries       200   \n",
       "0    2828.24  41136.09      300      10000  vary_n_queries       300   \n",
       "0    2751.87  42702.26      400      10000  vary_n_queries       400   \n",
       "0    2802.76  44947.46      500      10000  vary_n_queries       500   \n",
       "\n",
       "         min  \n",
       "0  22.450388  \n",
       "0  21.524792  \n",
       "0  24.240790  \n",
       "0  25.862270  \n",
       "0  26.727703  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# for seaborn, we need these to be in a single dataframe\n",
    "query_line_df['bench_type'] = \"vary_n_queries\"\n",
    "query_line_df[\"varied_n\"] = query_line_df['n_query']\n",
    "against_line_df['bench_type'] = 'vary_n_subjects'\n",
    "against_line_df['varied_n'] = against_line_df['n_against']\n",
    "\n",
    "benchD = pd.concat([query_line_df, against_line_df])\n",
    "benchD['min'] = benchD['s']/60\n",
    "benchD.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "cf9eb64f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# seaborn version -- no real point to this bc we need to access\n",
    "# the individual axes anyway, since we need diff labels. but here's the SAME PLOT\n",
    "g = sns.relplot(data=benchD, x='varied_n', y='min', col=\"bench_type\", \\\n",
    "                markers=True,marker=\"o\",\\\n",
    "                kind='line', facet_kws=dict(sharex=False))\n",
    "g.set(ylim=(0, 35))\n",
    "\n",
    "titles = [\"varying number of queries (10k subjects)\", 'varying number of subjects (1000 queries)']\n",
    "x_labels = ['num queries', 'num subject datasets']\n",
    "y_label = 'walltime (m)'\n",
    "tick_dist=[200, 2000]\n",
    "\n",
    "for n, ax in enumerate(g.axes.flat):\n",
    "    ax.set_title(titles[n])\n",
    "    ax.set_xlabel(x_labels[n])\n",
    "    ax.set_ylabel(y_label)\n",
    "    ax.xaxis.set_major_locator(ticker.MultipleLocator(tick_dist[n]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "d8560759",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/t/miniforge3/envs/py38/lib/python3.8/site-packages/seaborn/regression.py:582: UserWarning: sharex is deprecated from the `lmplot` function signature. Please update your code to pass it using `facet_kws`.\n",
      "  warnings.warn(msg, UserWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# linear regression version\n",
    "g = sns.lmplot(data=benchD, x='varied_n', y='min', col=\"bench_type\",\n",
    "                markers=\"o\",sharex=False, scatter_kws={\"s\": 30})              \n",
    "g.set(ylim=(0, 35))\n",
    "\n",
    "titles = [\"varying number of queries (10k subjects)\", 'varying number of subjects (1000 queries)']\n",
    "x_labels = ['Number of Queries', 'Number of Subject Datasets']\n",
    "y_label = 'walltime (m)'\n",
    "tick_dist=[200, 2000]\n",
    "\n",
    "for n, ax in enumerate(g.axes.flat):\n",
    "    ax.set_title(titles[n], size=16, pad=25)\n",
    "    ax.set_xlabel(x_labels[n], size=14, labelpad=10)\n",
    "    ax.set_ylabel(y_label, size=14, labelpad=10)\n",
    "    ax.xaxis.set_major_locator(ticker.MultipleLocator(tick_dist[n]))\n",
    "\n",
    "# a little more space between subplots\n",
    "plt.subplots_adjust(wspace=0.15)\n",
    "plt.tight_layout()\n",
    "plt.savefig('subsample_benchmarks.svg')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "a03072e4",
   "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>s</th>\n",
       "      <th>h:m:s</th>\n",
       "      <th>max_rss</th>\n",
       "      <th>max_vms</th>\n",
       "      <th>max_uss</th>\n",
       "      <th>max_pss</th>\n",
       "      <th>io_in</th>\n",
       "      <th>io_out</th>\n",
       "      <th>mean_load</th>\n",
       "      <th>cpu_time</th>\n",
       "      <th>ident</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1394.1251</td>\n",
       "      <td>0:23:14</td>\n",
       "      <td>16106.77</td>\n",
       "      <td>18410.19</td>\n",
       "      <td>16167.91</td>\n",
       "      <td>16167.94</td>\n",
       "      <td>95974.41</td>\n",
       "      <td>1.14</td>\n",
       "      <td>2914.11</td>\n",
       "      <td>40626.79</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1394.1750</td>\n",
       "      <td>0:23:14</td>\n",
       "      <td>17553.49</td>\n",
       "      <td>20057.43</td>\n",
       "      <td>17685.95</td>\n",
       "      <td>17685.99</td>\n",
       "      <td>94003.57</td>\n",
       "      <td>1.16</td>\n",
       "      <td>2771.80</td>\n",
       "      <td>38643.81</td>\n",
       "      <td>b</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1357.7237</td>\n",
       "      <td>0:22:37</td>\n",
       "      <td>18257.87</td>\n",
       "      <td>20041.38</td>\n",
       "      <td>18277.77</td>\n",
       "      <td>18277.80</td>\n",
       "      <td>91280.15</td>\n",
       "      <td>1.01</td>\n",
       "      <td>2794.07</td>\n",
       "      <td>37936.37</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1505.0091</td>\n",
       "      <td>0:25:05</td>\n",
       "      <td>16107.48</td>\n",
       "      <td>19187.17</td>\n",
       "      <td>16137.28</td>\n",
       "      <td>16137.32</td>\n",
       "      <td>93738.74</td>\n",
       "      <td>1.02</td>\n",
       "      <td>2605.24</td>\n",
       "      <td>39209.39</td>\n",
       "      <td>d</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1596.4828</td>\n",
       "      <td>0:26:36</td>\n",
       "      <td>14046.44</td>\n",
       "      <td>15868.38</td>\n",
       "      <td>14061.43</td>\n",
       "      <td>14061.46</td>\n",
       "      <td>91948.11</td>\n",
       "      <td>0.96</td>\n",
       "      <td>2367.13</td>\n",
       "      <td>37791.57</td>\n",
       "      <td>e</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           s    h:m:s   max_rss   max_vms   max_uss   max_pss     io_in  \\\n",
       "0  1394.1251  0:23:14  16106.77  18410.19  16167.91  16167.94  95974.41   \n",
       "0  1394.1750  0:23:14  17553.49  20057.43  17685.95  17685.99  94003.57   \n",
       "0  1357.7237  0:22:37  18257.87  20041.38  18277.77  18277.80  91280.15   \n",
       "0  1505.0091  0:25:05  16107.48  19187.17  16137.28  16137.32  93738.74   \n",
       "0  1596.4828  0:26:36  14046.44  15868.38  14061.43  14061.46  91948.11   \n",
       "\n",
       "   io_out  mean_load  cpu_time ident  \n",
       "0    1.14    2914.11  40626.79     a  \n",
       "0    1.16    2771.80  38643.81     b  \n",
       "0    1.01    2794.07  37936.37     c  \n",
       "0    1.02    2605.24  39209.39     d  \n",
       "0    0.96    2367.13  37791.57     e  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def load_benchmark_replicate(filename, ident):\n",
    "    df = pd.read_csv(filename, sep='\\t')\n",
    "    df['ident'] = ident\n",
    "    \n",
    "    return df\n",
    "\n",
    "replicates_df = [ load_benchmark_replicate(f\"benchmarks/a_vs_{x}.txt\", x) for x in ('a','b','c','d','e')]\n",
    "#query_line.append(load_benchmark_and_annotate('benchmarks/a_vs_a.txt', 1000, 10000))\n",
    "replicates_df = pd.concat(replicates_df)\n",
    "replicates_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "92f2cdc7",
   "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>ident</th>\n",
       "      <th>benchmark</th>\n",
       "      <th>val</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>a</td>\n",
       "      <td>mins</td>\n",
       "      <td>23.235418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>b</td>\n",
       "      <td>mins</td>\n",
       "      <td>23.236250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>c</td>\n",
       "      <td>mins</td>\n",
       "      <td>22.628728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>d</td>\n",
       "      <td>mins</td>\n",
       "      <td>25.083485</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>e</td>\n",
       "      <td>mins</td>\n",
       "      <td>26.608047</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  ident benchmark        val\n",
       "0     a      mins  23.235418\n",
       "1     b      mins  23.236250\n",
       "2     c      mins  22.628728\n",
       "3     d      mins  25.083485\n",
       "4     e      mins  26.608047"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# for seaborn, melt dataframe\n",
    "replicates_df['mins'] = replicates_df['s']/60\n",
    "replicates_df['max_rss_gb'] = replicates_df['max_rss']/1000\n",
    "replicates_df['io_in_gb'] = replicates_df['io_in']/1000\n",
    "\n",
    "val_vars = ['mins', 'max_rss_gb','io_in_gb'] # 'max_vms', 'max_uss','max_pss', 'io_in','io_out','mean_load','cpu_time']\n",
    "melted_reps = pd.melt(replicates_df, id_vars = 'ident', value_vars=val_vars, var_name='benchmark',value_name='val')\n",
    "melted_reps.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "08ee075e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x288 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1, 3, figsize=(14, 4), sharey=False)\n",
    "fig.suptitle('performance benchmarks', size=20, y=1.06)\n",
    "\n",
    "xlabels=['time (min)', 'max RSS (Gb)', 'IO in (Gb)']\n",
    "ylabels=['min', 'Gb', 'Gb']\n",
    "\n",
    "for n,val in enumerate(val_vars):\n",
    "    sns.boxplot(ax=axes[n], data=replicates_df, y=val_vars[n],\n",
    "                medianprops={\"color\": \"coral\"},showcaps=False,\n",
    "                boxprops={\"facecolor\": (.4, .6, .8, .5)})#, notch=True)\n",
    "    sns.swarmplot(ax=axes[n], data=replicates_df, y=val_vars[n], color='black')\n",
    "    axes[n].yaxis.set_major_locator(ticker.MultipleLocator(1))\n",
    "    axes[n].set_title(xlabels[n], size=14)\n",
    "\n",
    "    #axes[n].set_xlabel(xlabels[n], size=14)\n",
    "    axes[n].set_ylabel(xlabels[n], size=14)\n",
    "plt.subplots_adjust(wspace=0.4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "0f99fa57",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x216 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1, 3, figsize=(6, 3), sharey=False)\n",
    "fig.suptitle('performance benchmarks', size=20, y=1.06)\n",
    "\n",
    "xlabels=['time (min)', 'max RSS (Gb)', 'IO in (Gb)']\n",
    "ylabels=['min', 'Gb', 'Gb']\n",
    "\n",
    "for n,val in enumerate(val_vars):\n",
    "    sns.boxplot(ax=axes[n], data=replicates_df, y=val_vars[n],\n",
    "                medianprops={\"color\": \"coral\"},showcaps=False,\n",
    "                boxprops={\"facecolor\": (.4, .6, .8, .5)})#, notch=True)\n",
    "    sns.swarmplot(ax=axes[n], data=replicates_df, y=val_vars[n], color='black')\n",
    "    axes[n].yaxis.set_major_locator(ticker.MultipleLocator(2))\n",
    "    #axes[n].set_title(xlabels[n], size=14)\n",
    "\n",
    "    #axes[n].set_xlabel(xlabels[n], size=14)\n",
    "    axes[n].set_ylabel(xlabels[n], size=14)\n",
    "plt.subplots_adjust(wspace=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "93be9183",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAADeCAYAAADSHRzIAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAAA6+0lEQVR4nO3deVxN+f/A8dctIhoMsoYRysgufDMp+1ISNciSYYYJjeyyNLJnNyZrviZfJlvTaCYjjGkYNLbGGPuWrEVkixQ5vz/8OuMqXKlu6f18PDzc+znnfs773E69z/mcz/l8NIqiKAghhBCAgb4DEEIIkXtIUhBCCKGSpCCEEEIlSUEIIYRKkoIQQgiVJAUhhBAqSQr5WEBAAC1atKB27do0b96cpKQkfYf0XmnVqhXW1tb6DkMnP/74I5aWlqxevVrfobyVcePGYWlpyalTp/QdynujgL4DEPqxZ88e5s+fT5kyZejbty+FChXC2NhY32EJIfRMkkI+dfLkSQC8vLzo1q2bnqMRQuQW0nyUT6WkpADw4Ycf6jkSIURuIkkhF0hrF719+zZjxozB2tqaJk2aMGTIEM6dO5du/ZSUFFasWIGDgwN16tTBxsaGUaNGceXKFa31/P39sbS05M8//6Rbt27Url2b9u3bY2lpyeLFiwHw9PTE0tKSH3/8Uf3c1q1bcXNzo379+jRo0AA3Nzd++eUXrbqvXr2KpaUlixYtYvr06dSvX5+mTZsSHh6utk8fPHiQgIAAWrVqRd26denSpQt79uwB4IcffqBjx47Uq1cPJycntm3blm4/z549y5gxY7C3t6d27do0bNgQNzc3tm/fnuF+XrhwgQULFqj3SRwdHVm/fn26elNTUwkMDKRz587Ur18fe3t7xowZk+77UxSF9evX07VrV+rWrUvjxo0ZNGiQepWlq7Nnz/LZZ59Rr149PvnkEyZNmsTt27fTrZeYmMi8efNo06aNep/H19c33brZtb9p+xwYGEj79u2pXbs2rVu3ZunSpTx9+lRdJ+1nv3TpUnbs2KF+P61atSIwMBCAqKgoevXqRf369WnVqhX+/v5adQAkJCQwe/Zs9TioV68ejo6OLF++XGvdtOMpPDycL774gjp16tCyZcsM4wc4ffo0jRs3pnHjxhw/flwtX7t2LS4uLjRo0ICGDRvSq1cvwsPDM6wjP5Pmo1xk4MCBxMfH4+rqSlxcHL/++isHDx7k+++/p2bNmgA8efKEgQMHsn//furWrUufPn24ffs24eHh7N27l7Vr12JhYaFV7+jRozE3N8fd3Z2HDx/SqVMnDh48yMGDB3FwcMDc3JyPP/4YgNmzZ/Pdd99hampKp06dANi1axcjR47k5MmTjBkzRqvuTZs2AdCzZ0+io6OpX78+f/75JwAzZ84kLi4OR0dHHj16xE8//cTgwYNxc3MjJCQEBwcH/vOf/xAaGsqIESOoXLkytWrVAuCff/7B3d0dIyMj2rVrR8mSJbl06RK//fYbXl5eLF++nJYtW2rFMmbMGK5fv067du0oUKAAP//8M5MnT8bQ0JDu3bsD8OzZMzw8PNizZw/Vq1fn008/5c6dO2zdupX9+/fzww8/ULZsWQC8vb356aefqFGjBm5ubiQlJREeHo6bmxsrVqzAxsbmjT/Tx48f06dPH8zMzOjduzfHjh1j48aNHDhwgJCQEExMTAB48OABvXr14uzZs9jY2NCuXTuuXr3Kpk2b2LNnDxs2bKBMmTLZur8AK1euJCkpiY4dO2JnZ8eOHTtYtGgR9+7dY/z48Vrb37FjB0uXLqVDhw5YW1vz888/M2vWLK5du8bGjRuxs7OjZ8+ebNu2jcWLF1OsWDE+++wzdX+7d+9ObGwsrVq1ok2bNiQkJPDrr7+ycOFC7t27h7e3t9b2pk+fTpkyZXB3d+fq1atUqlQp3fcdExPDF198oSbC2rVrA887VcyfPx8rKyvc3Nx48uQJ27ZtY/jw4SQnJ9OlS5c3/izzDUXonbe3t2JhYaG0bNlSuX37tlq+bds2xcLCQunTp49atnLlSsXCwkKZM2eOVh3//POPYmVlpbi6uqpl3377rWJhYaG4uLgoqampWuunLfv111/VskOHDikWFhZKly5dtOK4ffu20qlTJ8XCwkI5ePCgoiiKcuXKFcXCwkKxtLRUTp06pVV3SEiIYmFhoTRs2FC5du2aWj5//nzFwsJC+fjjj7U+8+OPP6bbp88//1ypVauWcv78ea26f/nlF8XCwkIZOXJkun15+fuLiopSLCwslG7duqllwcHBioWFheLl5aUkJyer5WFhYYqFhYUybdo0RVEUZevWrep2njx5oq53+fJlpUmTJkrz5s21Pp+Rli1bKhYWFsrQoUO1vv958+YpFhYWyoIFC9SyyZMnKxYWFsr333+vVcfOnTvVeLNzf9N+ZvXr11cuXLigrhcXF6fUrVtXsba2Vvch7Wf/8vGzZ88etfzF/Uhb/9NPP1XLVqxYoVhYWCibNm3S2t/r168rtWvXVj755BO1LC02Ozs75dGjR1rrp/3unDx5UomLi1Natmyp1K9fXzl06JDWek2aNFHatGmj9bOMjY1Vateurbi4uCjiX9J8lIsMHjyYkiVLqu/bt29Po0aNOHjwIDdu3ACeN7sUK1aMESNGaH22Tp06dOjQgWPHjqVrcmrbti0GBm/+Uac1IY0dO1YrjpIlSzJq1CgAQkJCtD5TpUoV9SrmZe3ataNChQrq+4YNGwJgY2Oj9Zm6desCcO3aNbWsX79+zJ07l2rVqmnV2bRpU4AMm19cXV214m7YsCHFihXTqjetGWzChAkYGRmp5Y6OjgwaNEiN8YcffgBg4sSJFCjw7wV1pUqVcHNz48aNG0RGRma43y/SaDSMGTNG6/sfOnQoJiYmhIWFAfD06VNCQ0OpUaMGvXv31vp869atadiwIb/++iuJiYnZtr9pOnbsiLm5ufq+bNmyWFlZcf/+fe7du6e1bsWKFWnTpo3W9gGKFCmCm5ubWm5mZkbp0qW14rK1tWXKlCnpztDLly9PpUqVSEhI4GV2dnav7CF39+5d+vfvz+3bt1m2bFm6rsCKopCQkKDV5FSuXDnCw8NZt25dhnXmV9J8lIs0btw4XVndunWJiori9OnTmJiYcPHiRUxNTVm2bFm6dW/dugXAqVOnqFGjhlpuZmam0/ZPnz6NgYEBjRo1Srcsrez06dNa5a+ru3Llylrv036hX/5MoUKFgH9vfgM0b94cgPj4eE6fPs3ly5e5ePEiUVFRwPN28pdVrVo1XZmJiYnWH9PTp09ToUIFrSYTeP7H+8VEe+LECQoVKkRQUFC6Oi9evAg8/55btGiRbvmLypQpk66Zw8jIiJo1a3L48GEePHhAXFwcjx49IjU1FX9//3R1JCcnk5qaypkzZ7R+Nlm5v2k++uijdGUlSpQA4NGjR1odE6pUqaK1XpEiRYDnf2wNDQ21lhUqVIj79++r72vVqkWtWrV4+PAhR48e5dKlS8TExHDs2DEuXbqU4c/3dcfa2LFjuXnzJhUqVFBPMl7Uo0cPAgIC1PtwdnZ22NvbU6dOnVfWmV9JUshFXv7FBShdujTwvA027Zc9Pj5evVGckZfP6AoXLqzT9hMTEylUqJDWGWWaDz74AGNj43QPuKX9Qc/Iq87qMqr/ZdevX2f69OlERESgKAoGBgZ89NFHNGrU6JU3ejOqV6PRoLwwZcj9+/fV7/R1Hjx4wNOnT9/qe87Iq7ZVtGhR4Pkf2rQ/ltHR0W+1vazc39fVmUZ5aeqVd/n5Jicns2DBAjZu3KgeU2XLlqVx48Z8+OGHxMfHp/vM6461hIQE7O3t2b17N/7+/unuR4wcOZIqVaqwYcMG/vnnH44ePYq/vz9Vq1bF19dXp/tD+YUkhVzk8ePH6X7RHjx4ADzvOpp2JmZtbZ3hGey7Klq0KElJSdy/f59ixYppLUtOTubx48c50oVVURQ8PDw4f/48Hh4etGnThho1alC4cGFu3bpFcHBwpusuUqQIDx8+zHDZo0eP1O+4SJEiFC1alF27dmV6W4DW2fGLbt68iUajoVixYmqCcHZ2Zs6cOe+0vZfpur85bdasWaxbt4727dvTu3dvLC0t1SuSjh07ZpgUXmfKlCk4OTnRqVMn/ve//+Hk5KR2WoDnyfLTTz/l008/5fbt20RGRvLrr7+yY8cOBg8eTEREhFZTXH4m9xRykWPHjqUrO3LkCAUKFMDKyooPPviAChUqcP78eR4/fpxu3dDQUPz9/bl69Wqmtp/Wzp/WRPOiqKgoFEWhevXqmar7bZw5c4azZ8/Stm1bRowYQZ06ddSrnQsXLgDpz1p1ZWFhwfXr1zP8o9OlSxfat28PgKWlJXFxcRmut2vXLhYuXJiuKS0j169fT3f/4+7du8TExGBubo6xsTFVq1bFyMiIEydOZLhfq1evZunSpdy5c0fX3VTpur85bcuWLZQqVYpFixbRtGlTNSE8fvyY69evA2/3M7aysqJQoUL4+PiQmpqq/g9w584d/P392bx5MwClSpXCycmJb7/9FhcXF5KSkt66m/H7TJJCLuLv76/VHrxt2zYOHDhA69at1V+arl27cvfuXebNm8ezZ8/Udc+fP8/UqVMJDAxU131bLi4uACxYsEDrRl9CQoJ6Buvs7Jyput9GWvPDyzcb7969q8bxcp93XXXu3BlFUZg3b55Wu3V4eDiXLl1SmxG6du2KoihMmzZN617HzZs38fX1JSAgQD3Df53U1FSWLFmivlcUhQULFpCUlISrqyvwvFnEwcGB8+fPq/380xw4cIA5c+YQEhJC8eLFs21/c1qhQoVITk7WupJKTU1lxowZ6gnPkydP3rpee3t72rZty4kTJ1i7di3w/Ap4zZo1LFy4kLt372qtn5aAXuwQkd9J81EucvHiRbp06UKLFi24ceMGO3fupGzZsowbN05d58svv1SfR4iKiqJJkybcv3+fbdu2kZSUxLx589S+72+rcePG9O/fX33QKe05gN9//534+HgGDhyY4c3wrPbRRx9Rt25dDh06RK9evWjYsCF37txh586dpKSkYGxsnKmzZoBPP/2UHTt2EBoaypkzZ2jatCk3btxgx44dmJmZqTdfXVxciIiIYPv27Zw5c4bmzZvz9OlTwsPDuXv3LqNGjcqwn/zLSpUqRWhoKOfOnaN27dr89ddf/P3331hbW9O3b191PW9vb44cOcLs2bP57bffqFu3rhpXgQIFmDlzpk49yDK7vznNycmJ7777DldXV9q0acPTp0/Zu3cvFy9epGTJkiQkJHD37t10z2boYuLEiezbt49FixbRvn17ypcvj5eXF9OnT6dTp060bduWwoULc+jQIY4dO4azs7NWj6v8Tq4UcpH58+dTq1YtQkJCiIqKokuXLgQHB2udxRQuXJg1a9YwdOhQkpOTWbduHbt376Zhw4asWbNGfeAss8aNG8fcuXOpWLEiYWFhhIeHU7VqVfz9/Rk9evS77qJODAwMWLp0KS4uLly9epW1a9dy+PBh7OzsCAkJ4ZNPPiEmJobLly+/dd2GhoYsW7aM4cOH8/jxY4KCgti/fz9OTk6sW7dOPRvXaDR8++23TJw4EWNjY4KDgwkPD6d69eosWbKEL7/8UqftmZqaEhgYyJMnT/j++++JjY1l4MCB/Pe//6VgwYLqeiVLlmTTpk18/vnn3LhxQ93nVq1asWnTJrUrbnbtb04bMWIEQ4cOxcDAgHXr1rFz504qVqzIqlWrGDRoEAC7d+/OVN3ly5dnyJAhPHr0iClTpgDg7u7OwoULMTMzY+vWrQQFBZGSksL48eOZOXNmlu3X+0CjZLZxVmSZcePGsXnzZkJDQ9Uni4UQQh/kSkEIIYRKkoIQQgiVJAUhhBAquacghBBCJVcKQgghVJIUhBBCqCQpCCGEUElSEEIIoZKkIIQQQiVJQQghhEqSghBCCJUkBSGEECpJCkIIIVSSFIQQQqgkKQghhFBJUhBCCKGSpCCEEEIlSUEIIYRKkoIQQghVjiaFxYsX4+joiKOjI3PmzAFg7969dO7cmU6dOjF27FhSUlJyMiSRC2R0XERGRuLk5ES7du1YuHChniMUOS0gIID27dvj5OTEsmXLADhx4gSurq507twZDw8P7t+/r+co31NKDtm3b5/So0cPJTk5WUlJSVH69u2r7NixQ7Gzs1POnz+vKIqiDB06VNm0aVNOhSRygYyOi7CwMMXe3l65fPmy8uTJE+Xzzz9Xdu3ape9QRQ7Zt2+f0qlTJ+XBgwfK06dPFQ8PD2X79u1Kz5491ePAz89PWbBggZ4jfT/l2JWCqakp48aNw8jIiIIFC1KtWjWuX79OamoqiYmJpKamkpycTKFChXIqJJELZHRcxMTEUKVKFSpVqkSBAgVwcnJi27Zt+g5V5JCTJ09ia2uLiYkJhoaGNG/enJ07d/Ls2TMePnwIQFJSEoULF9ZzpO+nAjm1oRo1aqivY2Ji2Lp1Kxs2bKBixYq4u7tjYmKCmZkZHTp00Km+x48fc/z4cUxNTTE0NMyusPOE1NRU4uPjqV27dp77RcnouHB3d8fU1FQtL1OmDDdu3NCpPjku/pVXjwsrKytmzpyJh4cHxsbGREREoCgK48aNo3///sycORNjY2M2bdqkc51yXPzrTcdFjiWFNOfOncPDwwNvb2+KFi3KvHnz2LJlC2ZmZvj5+eHn54evr+8b6zl+/Di9e/fOgYjzjqCgIKytrfUdRqa8eFwUKFCAixcvai3XaDQ61SPHRXp57biwsbHBxcUFd3d3SpQogY2NDQcOHGDixIn873//o27dugQGBuLt7U1AQIBOdcpxkd6rjoscTQpRUVF4eXkxYcIEHB0dCQ8Px8LCgsqVKwPQvXt3hg8frlNdaWeSQUFBlCtXLrtCzhPi4uLo3bu31tl1XvLycXHw4EFu3bqlLr958yZlypTRqS45Lv6VV4+LxMRE2rZtS//+/QEIDAykUKFCFCpUiLp16wLQo0cPFi1apHOdclz8603HRY4lhdjYWDw9PVm4cCE2NjYAWFhYMHv2bG7dukXp0qX57bffqFOnjk71pV0ClitXDjMzs2yLOy/Ji5fFGR0X9erV4+LFi1y6dAkzMzO2bNmCq6urTvXJcZFeXjsurl69ire3NyEhISQlJREcHMy0adMYOnQo0dHRmJubv9XfCpDjIiOvOi5yLCmsWrWK5ORkZs2apZa5ubkxbNgw+vbti6GhIVWqVGHq1Kk5FZLIBV51XMyaNYuhQ4eSnJyMvb29zveaRN5Xs2ZN2rVrR+fOnUlNTaVfv340atQIPz8/hg8fjqIolCpVipkzZ+o71PdSjiUFHx8ffHx8MlzWtWvXnApD5DKvOy5+/vnnHI5G5Baenp54enpqldnb22Nvb6+niPIPeaJZCCGESpKCEEIIlSQFIYQQKkkKQgghVJIUhBBCqCQpCCGEUElSEEIIoZKkIIQQQiVJQQghhEqSghBCCJUkBSGEECpJCkIIIVSSFIQQQqgkKQghhFBJUhBCCKGSpCCEEEIlSUEIIYRKkoIQQgiVJAUhhBAqSQpCCCFUkhSEEEKoJCkIIYRQSVIQQgihkqQghBBCJUlBCCGESpKCEEIIlSQFIYQQKkkKQgghVJIUhBBCqCQpCCGEUElSEEIIoZKk8A6io6Oxt7enYMGC2NvbEx0dre+QhBDinRTQdwB5Wf/+/fnjjz8A+OOPP+jfvz+7d+/Wc1Q568GDBxw/fpyEhAQ0Gg2mpqZ8/PHHmJiY6Ds0kccFBAQQEhKCkZERDg4ODB48mOjoaHx9fbl37x6mpqYsWLCA4sWL6zvU94okhXcQGRn52vfvq6dPn7J161bWrVvHsWPHMDQ0pFixYiiKwr179wBo0KABPXr0wMHBAQMDuSAVbycyMpKwsDBCQkIwNjbG09OT7du3s2DBAiZOnIidnR3z5s0jICCAMWPG6Dvc94r8tr6DZs2avfb9+ygyMhJnZ2e2bNmCs7Mz4eHhHD16lL1797Jv3z6OHTtGaGgonTp1YuPGjXTs2JF9+/bpO2yRx5w8eRJbW1tMTEwwNDSkefPmbN68mSJFimBnZwfAoEGD6N27t54j1U1eamqWpPAOAgMDqVSpEgYGBtjZ2REYGKjvkLJdSEgIS5cuJSAggJ49e1K5cmU0Go26XKPRUL16dXr06MHatWtZsmQJwcHBeoxY5EVWVlbs3buXu3fvkpycTEREBL///julS5fG29sbJycnfH19KVKkiL5D1UlaU/PTp0/VpubcSpLCOzA3N6d3796MHTuW3bt3Y25uru+Qst38+fOpUqWKzutXr16db775JvsCEu8lGxsbXFxccHd3Z8CAATRq1AiAgwcP0qdPH8LCwqhUqRKzZs3Sc6S6yUtNzXJPQbyTBw8e8PPPPxMdHU3BggWpVq0aTk5OFC5cWN+hiTwsMTGRtm3bqmfUgYGBWFhYoNFoqFOnDgCdOnXCy8tLn2HqrFmzZmqnlLT3uZVcKYhM+/PPP2nRogWrV68mPj6eq1evsmzZMtq2bcvp06f1HZ7Iw65evYqnpydPnz7lwYMHBAcHM3nyZBISEtRjKyIiAisrKz1Hqpu81NQsVwoi02bOnEn37t0ZO3asel8hNTWVadOm4ePjww8//KDnCEVeVbNmTdq1a0fnzp1JTU2lX79+NGrUiCVLluDj40NSUhLlypVjzpw5+g5VJ2lNzQB+fn56jub1JCmITLt06RLffvut1o1mQ0ND+vbtS9euXfUYmf5ER0fTv39/IiMjadasGYGBgfniXlN28PT0xNPTU6usXr16crKRzaT5SGSavb09W7ZsSVe+fft2mjZtqoeI9C8v9TIRIiNypSDeytixY9XXqampLFmyhF27dlGnTh0MDAw4e/YsR44cybdXCnmpl4kQGZGkIN6KoaGh+rp48eLqH//k5GQAKlWqRKVKlUhNTdVLfPqWl3qZCJERSQrireT2m2T6FhgYSIsWLbh27Rq2tra5upeJEBnJ0aSwePFiwsPDgeft0WPHjuXIkSP4+fnx8OFDLC0tmTVrFkZGRjkZlnhLz549IyIigjZt2gAwefJk9UoBoGHDhnTr1k3n+hITE3Fzc2P58uWYmZmxd+9e5syZw7Nnz6hVqxbTp0/PM8dEXuplIkRGcuxGc2RkJHv37mXz5s2EhoZy4sQJNm/ezNChQ5k6dSq//PILgPQsyOUePHhA9+7dmTBhAteuXQPgp59+4v79+zx9+pTY2FimTJnC2bNndarv6NGj9OzZk5iYGLVs4sSJLFy4kC1btvD48WN++umn7NgVIUQGdL5SeNchkk1NTRk3bpx6xletWjWuXbtG/fr1qVmzJgA+Pj75ti06r1iyZAkajYadO3dSrFgxtXzcuHFUqlQJAHd3dwIDA3U6U960aRO+vr7pbmAnJiaSmppKcnIyhQoVyvodEdkmJiaGPXv2cOzYMW7fvo2BgQGlS5emdu3a2NvbY2Zmpu8QxWu8Nilk5RDJNWrUUF/HxMSwdetWPv/8c4oUKYKnpyeXL1/G2tqacePGZdGuiezw66+/Mm3aNK2E8LIBAwbg6+urU30zZsxIVzZ58mTc3d0xMTHBzMyMDh06ZDpekXMOHTrEkiVLiIqKok6dOlSvXh0rKyuePXvGnTt32Lx5M7NmzaJJkyZ4eHjQpEkTfYcsMvDKpBAZGcmMGTOoWLEizs7OzJkzh0qVKqkPKimKwoULF4iKimLjxo34+/szadIkPvnkk9du8Ny5c3h4eODt7U1cXBx79+5l48aNVKhQgYkTJxIQEMDQoUOzdi9Flrl582a6h7G6deumdcVYo0YNEhISMlV/fHw88+bNY8uWLZiZmeHn54efn5/OSUbox+jRo7lx4wY9e/Zk8eLFr2xBePjwIdu2bWP+/PmYmZkxf/78HI5UvMkrk0LaEMmvGhEzbYjktGGSz58/z+LFi1+bFKKiovDy8mLChAk4OjoSHBxMvXr11GaHjh078v3337/jLons9OGHHxIfH0+5cuXUsgkTJmitExcXR+nSpTNV/+HDh7GwsKBy5coAdO/eneHDh2c6XpEznJ2dad68+RvXK1q0KK6urri6uua7WQrzile292T1EMmxsbF4enoyb948HB0dAbC1teXEiRPExsYC8Pvvv+eZAa7yqyZNmrxxfoSNGzdia2ubqfotLCz4559/uHXrFgC//fabOiqmyL10SQgvs7e3z4ZIxLvS+Ubz/fv3OXv2LE+fPkVRFK1lNjY2b/z8qlWrSE5O1hr/3M3NjalTpzJo0CCSk5P5+OOP8fb2fovwRU4bOHAgbm5uaDQavLy8KFWqlLrs/v37LF68mIiICH788cdM1V+tWjWGDRtG3759MTQ0pEqVKkydOjWrwhc55IcffmDDhg3qkOrm5uZ88cUXajdmkXvplBQ2b97MlClTePz4cbplGo2GU6dOvbEOHx8ffHx8MlzWokULXcIQuYClpSXLli1j4sSJBAcH89FHH1GiRAkePHjAxYsXKVOmDMuXL1ebBHUVERGhvu7atWu+HSbjfbBgwQI2bNhA37598fT05NmzZ/zzzz+MHTsWLy8v+vXrp+8QxWvolBS++eYbunfvjpeXl85dUMX76z//+Q/btm1j3759/P333yQkJFC8eHGGDRuGnZ1dnnnQTGSP4OBgZs+eTcuWLdWy1q1b8/HHHzN9+nRJCrmcTknh/v379O3bVxKCUBUsWJAWLVrIVZ7IUPny5dOVValShZSUFD1EI96GTk80t2rVih07dmR3LCIP6NWr11uN/Ll7927c3NyyMSKRGzx79kz9N3jwYHx9fTl//ry6/MqVK8yYMYNBgwbpMUqhC52uFEqWLMnChQv55ZdfqFy5MgULFtRanldmPxLvbtq0aUyfPp0pU6bQunVrPvnkE6pVq8aHH35Iamoqd+/e5cyZMxw+fJjw8HAqVqzI9OnT9R22yGa1atXSmmxJURScnJwoVKgQGo2Gx48fo9FouHDhAp9//rkeIxVvolNSSExMpFOnTtkdi8gDqlWrRmBgIH/99RcbNmxg9OjR3L17V2udkiVLYmtry9y5c2nUqJF+AhU5as2aNfoOQWQRnZKCjPYoXtawYUMaNmwIwLVr17TGuHnxwbb8Jjo6mqCgIK5du0ZkZGS+mY5Thqx4f7wyKSxatIgvv/wSY2NjFi1a9MoK0vqri/yrYsWKVKxYUd9hvNHKlSuJjo7O1m0EBQVx5coVAP744w9atGihDqWdXczNzRk4cGC2bkMXx44dY/ny5cyZM4eiRYvSsGFDkpKS1OXW1tasXbtWjxEKXbwyKRw+fJj+/ftjbGzM4cOHX1nBi+2IQuRm0dHRnD13gVLl3u4Zirdx9f+HE3/x/e0H2dfj5nbclWyr+238888/9OnTBwcHB1JSUihatCiKouDn50fZsmWJi4vj66+/JiwsDCcnJ32HK17jlUnhxYwu2V28L0qVq4RTv1HZVn9kZCQnjuxX39eq1yRbtxe2OncMKLd06VJ69uzJ+PHj1TKNRkOjRo3UBxnPnDlDSEiIJIVcTudhLnbt2kV0dHS6fsYajQYPD48sD0yIvGjYpAUsmjqS0/8cpmZda4ZNWqDvkHLEkSNHGDFihFbZy8PhODs7ExISkpNhiUzQKSmMHz+esLAwqlWrRuHChbWWSVIQACkpKZw7d46SJUtm+OBSflHOrAp+AfnvD19ycjIffPCBVtmaNWu0Oh2YmJjIJFp5gE5JYfv27XzzzTcymJUAnjcnrl+/noCAAMzMzDhx4gSDBw/m5s2baDQaHBwc8PPzk+Eu8hEzMzNOnTpFhQoV1LKXR7c9duwYH330UQ5HJt6WTk80ly1blrJly2Z3LCIPWL9+PQsWLKBDhw6UKFECRVEYNWoUGo2GsLAwfv/9d2JjY1m2bJm+QxU5qEOHDixatIjExMQMlycmJrJkyRI6d+6cw5GJt6XTlcKUKVOYMmUKffr0oXz58umm3WzcuHG2BCdyn40bN+Lr60uXLl2A573UYmJiGDdunDrl6pAhQ/D19WXYsGF6jFTkpAEDBhAREUGHDh3o378/jRs3pkSJEty/f58jR46wevVqypUrh7u7u75DFW+gU1I4ceIEp06dynD+ZF2Hzhbvh4sXL2Jtba2+j4yMRKPRaA2MV7VqVW7evKmH6IS+FC5cmHXr1rF06VICAwOZO3cuGo0GRVEoUaIErq6ueHl5YWhoqO9QxRvolBSWLVvGiBEj6NmzZ7obzSJ/KVy4MI8ePVLfR0ZGYmZmptVWHBsbS/HixfUQndCnwoULM3LkSEaMGMGVK1dISEigWLFiVKlSRZJBHqLTPQUjIyNat25N0aJFMTQ0TPdP5B/NmjUjKCgIgL/++oujR4/i4OCgLlcUhYCAAK2rCfH+O3v2rPpao9FQuXJl6tevj7m5+Sv/Rpw5cyanwhNvQacrhZEjRzJ79mzGjh2LmZkZBQpof+zlewzi/TVy5Eg+++wzrK2tSUpKonr16uoQC1u2bGHFihXcvHmT9evX6zlSkZOmT5+OqakpvXv3VsfEepX9+/ezfv16EhIS5MHYXEinpLBo0SJu377N7t27M1yeX+8p5MfBzypVqkR4eDj79u3DwMCAZs2aqV1Pk5KSaNq0KZ999tlbT8cp8rY1a9bw008/MXHiRB48eEDTpk3VIdWfPXvGnTt3OHv2LH/99RcmJiZ8+eWXuLi46DtskQGdksLcuXOzO44slRMDn0HOD36WWwY+K1SoEK1atUpX3q1bNz1EI3ILZ2dnnJ2d+fPPP/njjz/4888/SUhIwMDAgFKlSlGrVi169epF06ZN3zhmWkBAACEhIRgZGeHg4MDgwYPVZbt27WLq1Kla83qLrPPKpJCQkEDJkiUB3YfFvX37NqVKlcqayN5BTgx8Bjk7+FluGfgMIC4ujrCwMHr06EGxYsVISUlhwYIF7Nu3j5IlS/LFF19gZ2en7zCFntjY2GBjY5Ppz0dGRhIWFkZISAjGxsZ4enqyY8cO2rVrx61bt5g9e3YWRpsz8lKrwiuTwoABA2jevDm9evV644NrV65cYf369ezbt4+ffvopy4PMjOwe+AxydvCz3DLw2alTp3B3d6dEiRI4OjpSrFgxJkyYQHh4OH379uWDDz5g9OjRzJ07F3t7e32HK/KgkydPYmtrq84J37x5c3bu3Em7du3w8fHhq6++Yv78rPt9kCHVtb0yKWzYsIFVq1bRuXNnKlWqxCeffIK5ubnaRvjitIuXL1/G3d2dTZs2vdNO5DX5cfCzRYsW0bZtW2bOnIlGo+H69ev88ssvdOvWDW9vbwBKly7NsmXLJCmITLGysmLmzJl4eHhgbGxMREQEiqKwZs0aatWqRb169bJ0ezKkurZXJgUjIyMGDx5M//792bJlC3/88QdhYWEkJCSg0WgoXbo0tWrVws3NjY4dO1KkSJFMB5FX5cfBz/766y++//57tU04rfPBi9O1NmrUSGbrE5lmY2ODi4uLekVqY2PDwYMH2bFjB6tXryYuLi7LtylDqv/rjTeaCxcuzKeffsqnn36a6Y2I90dycrLWCcCff/5JkSJF0nVDlG7KIrMSExNp27Yt/fv3ByAwMJDq1auzZ88eXF1defLkCTdv3qRXr16sW7dOz9HqJi+1Kug8n4IQADVq1ODvv//GzMyMxMRE9u7di52dndazK9u3b1fHQRL504ULFzh+/DhPnz5NN6/Cm04wr169ire3NyEhISQlJREcHMy0adPw8fFRl/ft2zfPJATIW60KkhTEW/niiy/w8fHh77//5ujRo6SkpDBgwAAA9f7C8uXLmTp1qp4jFfoSEBDAggULKF68OEWLFtVaptFo3pgUatasSbt27ejcuTOpqan069ePRo0aZWfI2S7u6qV0VwrlzKroO6wMSVIQb6Vjx44YGxvz448/Ur58eUaPHk3t2rUB+O677/j5558ZMWKEOoqqyH8CAwMZM2YMX3zxRabr8PT0xNPTM8NlZmZmWfqMQgODe9QrfodSkd9kWZ0v6/ztek5cuArAiSP7CRzdnZ+9embb9iyKX+coH2bqs5IUxFtr0aKF1qioaUaOHMmECRPkfkI+9+TJE9q1a6fvMHKVQxevv/Z9bqJzUrh16xbBwcHExMQwduxYDhw4gLm5OTVr1szO+EQudebMGQoUKIC5ubnaEyntBvTp06eZMmWKjH+UTzk7OxMUFIS3t/cbn1zODY48K87OB8Y4dR2ebduwrLdHq/eRZb0m/Nks+7YXtno+pT4wonsmPqtTUjh27Bj9+vXDysqKv/76i6+++oqDBw8ybtw4li5diq2tbSY2LfKic+fO8dVXX3H58mXg+Y3nFStWUL58eRITE5k/fz4bN26UsY/ysTt37rBjxw7CwsKoWLEiBQsW1FqeNspufvLe9T6aNWsWX375JR4eHjRo0ACAyZMnU7p0aebPny9JIR+ZPn06JiYmBAUFUbBgQb799lumT5+Ol5cXHh4e3Lt3j2HDhqndCUX+Y25uzqBBg/QdRq7y3vU+OnnyZIYPIzk7O7Ny5cosD0rkXsePH+e///2venIwc+ZM2rRpw+nTp6lRowZTp06lfPnyeo5S6NNXX32l7xDEO9ApKZQqVYoLFy5QuXJlrfKoqCjKlCmTLYGJ3Onhw4daTUOmpqYA2NraMmXKFH2FJfRs7NixTJo0CRMTE8aOHfvadefMmZNDUYnM0CkpDBw4kK+//pqBAweiKAr79u0jNjaWNWvWMHr06OyOUeQyL988NDAwoG/fvnqKRuQGL86uJrMx5m06JYUePXpgamrKqlWrKFy4MPPnz6dq1arMmDFDaypGkX+lTbQj8qcXm5dl3Ku8Tecuqa1atcpwYhWR/6xcuRJjY2P1/ZMnT1i9ejXFihXTWm/YsGE5HZoQ4h3plBQeP35MSEgI0dHRpKSkH+512rRpWR6YyJ0aN27MiRMntMoaNGigNXE7pG9iEkLkDTolhREjRnDo0CGaNGlC4cKFszsmkYvJROtCvN90Sgr79+9n5cqVWFtbZ3c8Qggh9EinpFC1alVSU1OzOxYhRB63b98+Nm7cyNGjR7lz5w4lSpSgXr169OjRQx5yzSN0fqJ52LBhODo6UqFChXQDnsmImEKIadOmsX79epo3b07v3r0pXrw4N2/e5NixYwwYMIA+ffqocyKI3EunpLB582YuXrzI2rVr091T0Gg0khREOs+ePZPRUvOR0NBQtmzZwvr16zOcQ/nIkSMMGTKEunXr0rlzZz1EKHSl02/thg0bmDt3LgcOHGD37t1a/3bt2pXNIYrcasKECSQmJqYrP378OC4uLnqISOjLunXrGDVqVIYJAZ73UBs5cmS+HAwvr9EpKXz44YdYWlpmdywijzl+/DiOjo788ccfwPOuy35+fvTo0QNzc3M9Rydy0rlz57CxsXntOs2aNePcuXM5FJHILJ2aj3x8fPD19WXIkCGYmZlpzccLyDDJ+dSPP/7IypUrGTp0KK1ateLYsWMYGRmxcuVKmjVrpu/wsl1GUywCeWbaxaxkZGTEw4cPX7vOgwcP1Dk3RO6lU1IYMmQI8HwMJPj3wSRFUdBoNJw6dUqnjS1evJjw8HAA7O3ttQbOCgoKYtu2bdIPPg8pUKAA7u7unDp1ivDwcAoUKMD06dPfOiEkJibi5ubG8uXLMTMz48iRI/j5+fHw4UMsLS2ZNWtWlgyj0cDgHrWM4yi4edQ71wUwde1vnLgSDzyfYjHAy/n565fKvndvnSXby4iXcTInDcplW/26sra2ZtOmTUyaNOmV6wQHB9O4ceMcjEpkhk5J4bfffnvnDUVGRrJ37142b96MRqNhwIAB/Prrr7Rt25bz58+zYsUKqlR5/8+o3ichISEsWLCAkiVLEhQUxOnTp5k2bRqhoaFMmjRJpyako0eP4uPjQ0xMDPA8QQwdOpT//ve/1KxZk5EjR/LDDz/Qq1evd463TBlTity++871pPn72q3Xvk8rK2iYfU93FzQuTJlSptlWv648PT3p2bMnH3zwAQMHDsTExERddvfuXfz9/dm8eTObNm3SY5RCF69MCi/2HsmK8fFNTU0ZN26cesZXrVo1rl+/TkpKCpMmTWLYsGGEhoa+83ZEzpk8eTIeHh54eHhQsGBBGjVqRJs2bZgyZQrOzs4cO3bsjXVs2rQJX19f9apx37591K9fX53m1cfHJ8uekWkxLGuH9m62+5p6PwWgmW1zgHRl5lPXZel2X5Yb7t7UqlWLZcuW4e3tzapVq6hatSrFihXj1q1bXLt2jXLlyrFixQqqV6+u71DFG7wyKVhZWbF3715KlSpFrVq1XjuWjS7NRzVq1FBfx8TEsHXrVjZs2MD8+fNxdXXFzMzsLUMX+hYaGkq1atW0ysqWLcvSpUvZsWOHTnXMmDFD6/2lS5coUqQInp6eXL58GWtra8aNG5dlMWelwMBA+vfvT2RkJM2aNSMwMBCAFi1acO3aNWxtbdWy/KBZs2b89ttvREREcPz4ce7du0eDBg2oX78+dnZ2MpJuHvHKpPC///2P4sWLA7BmzZos2+C5c+fw8PDA29uba9euERsby/jx4zlw4ECWbUPkjGrVqhEfH8/Fixe1zuZTUlK4cOFCpupMTU1l7969bNy4kQoVKjBx4kQCAgIYOnRoVoWdZczNzdm9e3e68t69ewP5cwhpIyMjOnToQIcOHfQdisikVyaFJk2aqK83b97MxIkTtdoJAe7du8fXX3+tte7rREVF4eXlxYQJE3B0dGT8+PGcO3cOZ2dnHj16xK1btxg+fDjffPNN5vZG5Kj169czY8YMnj59ikajQVEU4HlHhHr16jF48OC3rrN06dLUq1dP7dHWsWNHvv/++yyNW2S9RYsW6byuDKmeu70yKRw+fFi9+RcaGkrNmjUpWrSo1jrR0dHs27dPpw3Fxsbi6enJwoUL1f7ML55JHThwgMWLF0tCyENWrlzJoEGD+PLLL2nVqhXBwcE8fPiQsWPH0rp15nrc2Nra4u/vT2xsLOXLl+f333/HysoqiyMXWe3w4cM6rSdDqud+r0wKJiYmLFu2DEVRUBSFwMBArWELNBoNRYoUeeN8rGlWrVpFcnIys2bNUsvc3Nzo2bPnO4Qv9OnmzZt06dIFIyMjrKysOHLkCA4ODkyYMIEJEyaoXZjfRvny5Zk6dSqDBg0iOTmZjz/+GG9v72yIXmQl6Ur+/nhlUqhZs6baFdXd3Z3Fixer9xgyw8fH57WDYTVt2pSmTZtmun6R80qVKkVCQgJmZmaYm5tz6tQpHBwcKFu2LDdv3nyruiIiItTXLVq0oEWLFlkcrRBCFzoNc7F27dp3Sgji/eTg4IC3tzdRUVE0b96ckJAQtm7dyrfffivPnAiRR+k8R7MQLxs1ahTFihXj7t27tG7dmm7dujF16lRKlCjBzJkz9R2eECITJCmITCtQoIBWD6MRI0YwYsQIPUYkhHhXkhREpj1+/JiQkBCio6NJSUlJt3zatGl6iEoI8S4kKYhMGzFiBIcOHaJJkybpJl8S4l0FBAQQEhKCkZERDg4ODB48mJ07d+Lv74+iKJiZmeHn5yf3O7OYJAWRafv372flypVYW1vrOxTxnomMjCQsLIyQkBCMjY3x9PQkNDSUefPmERISQtmyZVm0aBH+/v4yxWcWk/kSRaZVrVo1ywarE+JFJ0+exNbWFhMTEwwNDWnevDnbtm1j8uTJlC1bFgBLS0tiY2P1HOn7R64URKbNmjWLYcOG4ejoSIUKFdLNySxzd4vMsrKyYubMmXh4eGBsbExERASKotCmTRvg+f2sgIAA3N3d9RypbjKakCm3Tr4kSUFk2ubNm7l48SJr165Nd09Bo9FIUhCZZmNjg4uLC+7u7pQoUQIbGxuOHj0KPJ/BbciQIdSsWZOuXbtmyfZux10hbPX8LKkrI2GbNxJ3/RrwfPKlSZ7dcOraI9u2dzvuCqU+qPbmFTMgSUFk2oYNG5g7dy5OTk76DkW8ZxITE2nbti39+/cHng9TXqlSJW7evMkXX3zBf/7zHyZMmJAl28qJ+cRvxsWme1/qg+wbSrzUB9UyvV+SFESmffjhh1haWuo7DPEeunr1Kt7e3oSEhJCUlERwcDDTpk1j0KBBdOzYUZ0iOCtkZoyutxUZGak1+ZKtrW2uHVpdkoLINB8fH3x9fRkyZAhmZmYUKKB9OKUNfy3E26pZsybt2rWjc+fOpKam0q9fPxISEjh58iSpqals374dgNq1a6ebqCk3CgwMzDOTL0lSEJmWdraWdqaVNiyyoihoNBqdZuQT4lU8PT3x9PTUKjt9+rSeonk35ubmeWbyJUkKOnhVz4G81KMgO6SNoiuEeH+8l0mhgcE9ahnHUXDzqCypb+ra3zhxJR543nMgwMuZ791bv7I8O3gZJ3PSoFy21J1ZFStW1HcIQogs9l4mhTJlTCly+26W1ff3tVvp3hc01LyyPDsUNC5MmVKm2VK3EEKkeS+TQothU7K0vma7r2n1HGhm2xzzqeteWZ5dsr/jnBAiv5NhLnQQGBiInZ0dBQoUwM7OTu05kNZ32sDAQKtcCCHyqvfySiGrmZubs3v37gzL80qPAiGE0IVcKQghhFBJUhBCCKGSpCCEENksOjqaoKAg5syZg729PdHR0foO6ZUkKQghRDbr378/V65c4dmzZ/zxxx/qQH+5kSQFIYTIZpGRka99n5tIUhBCiGzWrFmz177PTSQpCCFENnvVs065kTynIIQQ2exVzzrlRnKlIIQQQiVJQQghhEqSghBCCJUkhXeQlx5IEUIIXUhSeAd56YEUIYTQhSSFd5CXHkgRQghdSFJ4B3npgRQhhNCFJIV3kJceSBFCCF3Iw2vvIC89kCKEELqQKwUhhBAqSQpCCCFUkhSEEEKoJCkIIYRQSVIQQgihkqQghBBCJUlBCCGESpKCEEIIlSQFIYQQqhxNCosXL8bR0RFHR0fmzJkDwMaNG+nUqRNOTk6MHz+elJSUnAxJ5AKJiYl06tSJq1evapUHBQXh7u6up6iEvgUEBNC+fXucnJxYtmwZAKdOncLV1ZX27dszceJEnj59quco3z85lhQiIyPZu3cvmzdvJjQ0lBMnThAQEMCqVavYsGEDP//8M8+ePWPdunU5FZLIBY4ePUrPnj2JiYnRKj9//jwrVqzQT1DvQObYyBqRkZGEhYUREhJCaGgoR48eZceOHYwZM4avv/6a7du3oygKmzZt0neo750cSwqmpqaMGzcOIyMjChYsSLVq1UhJSWHy5MmYmJig0WiwsLDg+vXrORWSyAU2bdqEr68vZcqUUctSUlKYNGkSw4YN02NkmSNzbGSNkydPYmtri4mJCYaGhjRv3py1a9fy+PFj6tevD4CLiwvbtm3Tb6DvoRwbEK9GjRrq65iYGLZu3cqGDRv46KOPAEhISCAoKAg/Pz+d6ktNTQUgLi4uy2PNa9K+g7TvJC+ZMWNGurL58+fj6uqKmZnZW9en7+Miozk2Xm4Wyyl5+biwsrJi5syZeHh4YGxsTEREBAUKFMDU1FRdx9TUlBs3buhUn76Pi9zkTcdFjo+Seu7cOTw8PPD29lYTwo0bNxgwYACurq40bdpUp3ri4+MB6N27d3aFmufEx8dTpUoVfYfxTvbt20dsbCzjx4/nwIEDb/15fR8XBQsW1GrnLliwIK1bt9ZLLGny4nFhY2ODi4sL7u7ulChRAhsbG/bv359uPY1Go1N9+j4ucqNXHRc5mhSioqLw8vJiwoQJODo6AnDhwgUGDhxInz59+Pzzz3Wuq3bt2gQFBWFqaoqhoWF2hZwnpKamEh8fT+3atfUdyjvbsmUL586dw9nZmUePHnHr1i2GDx/ON998o9Pn5bj4V14+LhITE2nbtq3a/BYYGEjFihU5fPiwuk58fLxWs+PryHHxrzceF0oOuX79utK0aVMlMjJSLXvw4IFib2+vhIaG5lQYIpdq2bKlcuXKFa2y/fv3K3369NFTREKfTp06pXTu3Fl58uSJcv/+faVjx47K4cOHFUdHR+Xw4cOKoijKxIkTlZUrV+o50vdPjl0prFq1iuTkZGbNmqWWOTg4cOvWLb777ju+++47AFq1apUnbzAKIbJOzZo1adeuHZ07dyY1NZV+/frRqFEj5s2bh4+PDw8fPqRWrVr07dtX36G+dzSKoij6DkIIIUTuIE80CyGEUElSEEIIoZKkIIQQQiVJQQghhEqSghBCCJUkBSGEECpJCkIIIVSSFIQQQqj+D6cQVB4PadyTAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x216 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# add ylims \n",
    "fig, axes = plt.subplots(1, 3, figsize=(6, 3), sharey=False)\n",
    "fig.suptitle('performance benchmarks', size=20, y=1.06)\n",
    "\n",
    "xlabels=['time (min)', 'max RSS (Gb)', 'IO in (Gb)']\n",
    "ylabels=['min', 'Gb', 'Gb']\n",
    "ylim= [(22,28), (14,20), (90,98)]\n",
    "\n",
    "for n,val in enumerate(val_vars):\n",
    "    sns.boxplot(ax=axes[n], data=replicates_df, y=val_vars[n],\n",
    "                medianprops={\"color\": \"coral\"},showcaps=False,\n",
    "                boxprops={\"facecolor\": (.4, .6, .8, .5)})#, notch=True)\n",
    "    sns.swarmplot(ax=axes[n], data=replicates_df, y=val_vars[n], color='black')\n",
    "    axes[n].yaxis.set_major_locator(ticker.MultipleLocator(2))\n",
    "    #axes[n].set_title(xlabels[n], size=14)\n",
    "\n",
    "    #axes[n].set_xlabel(xlabels[n], size=14)\n",
    "    axes[n].set_ylabel(xlabels[n], size=14)\n",
    "    axes[n].set_ylim(ylim[n])\n",
    "plt.subplots_adjust(wspace=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "63063e16",
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_threads_benchmark_and_annotate(filename, n_threads):\n",
    "    df = pd.read_csv(filename.format(threads=n_threads), sep='\\t')\n",
    "    df['threads'] = n_threads\n",
    "    \n",
    "    return df\n",
    "\n",
    "\n",
    "x = [ load_threads_benchmark_and_annotate(f'benchmarks/a_vs_a_1000_t{t}.txt', t) for t in (4,8,16) ]\n",
    "x.append(load_threads_benchmark_and_annotate(f'benchmarks/a_vs_a_1000.txt', 32))\n",
    "\n",
    "threads_df = pd.concat(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "67c6b9ad",
   "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>s</th>\n",
       "      <th>h:m:s</th>\n",
       "      <th>max_rss</th>\n",
       "      <th>max_vms</th>\n",
       "      <th>max_uss</th>\n",
       "      <th>max_pss</th>\n",
       "      <th>io_in</th>\n",
       "      <th>io_out</th>\n",
       "      <th>mean_load</th>\n",
       "      <th>cpu_time</th>\n",
       "      <th>threads</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>814.4291</td>\n",
       "      <td>0:13:34</td>\n",
       "      <td>2246.58</td>\n",
       "      <td>2380.33</td>\n",
       "      <td>2243.58</td>\n",
       "      <td>2244.08</td>\n",
       "      <td>89.14</td>\n",
       "      <td>0.11</td>\n",
       "      <td>341.55</td>\n",
       "      <td>2782.05</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>484.0369</td>\n",
       "      <td>0:08:04</td>\n",
       "      <td>2661.74</td>\n",
       "      <td>3352.14</td>\n",
       "      <td>2661.75</td>\n",
       "      <td>2662.25</td>\n",
       "      <td>1189.09</td>\n",
       "      <td>0.11</td>\n",
       "      <td>649.08</td>\n",
       "      <td>3142.19</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>329.2121</td>\n",
       "      <td>0:05:29</td>\n",
       "      <td>6144.40</td>\n",
       "      <td>7038.87</td>\n",
       "      <td>6169.71</td>\n",
       "      <td>6170.21</td>\n",
       "      <td>7404.52</td>\n",
       "      <td>0.10</td>\n",
       "      <td>1137.80</td>\n",
       "      <td>3746.52</td>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>154.1568</td>\n",
       "      <td>0:02:34</td>\n",
       "      <td>8026.29</td>\n",
       "      <td>8707.47</td>\n",
       "      <td>8042.18</td>\n",
       "      <td>8042.25</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.11</td>\n",
       "      <td>2112.52</td>\n",
       "      <td>3256.75</td>\n",
       "      <td>32</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          s    h:m:s  max_rss  max_vms  max_uss  max_pss    io_in  io_out  \\\n",
       "0  814.4291  0:13:34  2246.58  2380.33  2243.58  2244.08    89.14    0.11   \n",
       "0  484.0369  0:08:04  2661.74  3352.14  2661.75  2662.25  1189.09    0.11   \n",
       "0  329.2121  0:05:29  6144.40  7038.87  6169.71  6170.21  7404.52    0.10   \n",
       "0  154.1568  0:02:34  8026.29  8707.47  8042.18  8042.25     0.07    0.11   \n",
       "\n",
       "   mean_load  cpu_time  threads  \n",
       "0     341.55   2782.05        4  \n",
       "0     649.08   3142.19        8  \n",
       "0    1137.80   3746.52       16  \n",
       "0    2112.52   3256.75       32  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "threads_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "138afb24",
   "metadata": {},
   "outputs": [],
   "source": [
    "threads_df['mins'] = threads_df['s']/60\n",
    "threads_df['max_rss_gb'] = threads_df['max_rss']/1000\n",
    "threads_df['cpu_time_min'] = threads_df['cpu_time']/60"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "5510d1fe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x360 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# seaborn version (doesn't change much)\n",
    "fig, axs = plt.subplots(1, 3, figsize=(18, 5))\n",
    "\n",
    "ax = axs[0]\n",
    "sns.lineplot(data=threads_df,x=\"threads\", y=\"mins\",ax=ax, marker='o', label='observed times')\n",
    "#ax.plot(threads_df.threads, threads_df.s, '-p', label='observed times')\n",
    "ax.plot([4, 8, 16, 32], [(484 * 2)/60, 484/60, (484 / 2)/60, (484 / 4)/60], '--', label='perfect scaling')\n",
    "ax.axis(xmin=0, ymin=0)\n",
    "ax.set_xlabel('num threads', size=13, labelpad=8)\n",
    "ax.set_ylabel('walltime (s)', size=13)\n",
    "ax.set_title('walltime', size=16, pad=12)\n",
    "ax.legend(loc='upper right')\n",
    "\n",
    "ax = axs[1]\n",
    "sns.regplot(data=threads_df,x=\"threads\", y=\"cpu_time_min\",ax=ax, marker='o')\n",
    "#ax.plot(threads_df.threads, threads_df.cpu_time, '-p')\n",
    "ax.axis(xmin=0, ymin=0, ymax=70)\n",
    "ax.set_xlabel('num threads', size=13, labelpad=8)\n",
    "ax.set_ylabel('total CPU time (s)', size=13)\n",
    "ax.set_title('total compute', size=16, pad=12)\n",
    "\n",
    "ax = axs[2]\n",
    "#ax.plot(threads_df.threads, threads_df.max_rss, '-p')\n",
    "sns.regplot(data=threads_df,x=\"threads\", y=\"max_rss_gb\",ax=ax, marker='o')\n",
    "ax.axis(xmin=0, ymin=0)\n",
    "ax.set_xlabel('num threads', size=13, labelpad=8)\n",
    "ax.set_ylabel('max RSS (GB)', size=13)\n",
    "ax.set_title('memory usage', size=16, pad=12)\n",
    "\n",
    "plt.savefig('basic_benchmarks.svg')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f7c114bf",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "py38arm",
   "language": "python",
   "name": "py38arm"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}