block-size-analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import common\n",
    "import plotting\n",
    "import importlib\n",
    "from matplotlib.lines import Line2D\n",
    "from matplotlib.patches import Patch\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import os\n",
    "import pandas as pd\n",
    "import re\n",
    "import socket\n",
    "import matplotlib.ticker as ticker\n",
    "\n",
    "import matplotlib.style\n",
    "matplotlib.style.use('classic')\n",
    "\n",
    "# Show all columns and rows in a dataframe\n",
    "pd.set_option('display.max_columns', None)\n",
    "pd.set_option('display.max_rows', None)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Analytics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>algorithm</th>\n",
       "      <th>bfs</th>\n",
       "      <th>cdlp</th>\n",
       "      <th>lcc</th>\n",
       "      <th>pagerank</th>\n",
       "      <th>sssp</th>\n",
       "      <th>wcc</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>library</th>\n",
       "      <th>graph</th>\n",
       "      <th>block_size</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">sorted_vector_al.6</th>\n",
       "      <th>graph500-24</th>\n",
       "      <th>1024</th>\n",
       "      <td>0.169685</td>\n",
       "      <td>16.213700</td>\n",
       "      <td>41.796077</td>\n",
       "      <td>2.125752</td>\n",
       "      <td>3.786768</td>\n",
       "      <td>0.636032</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>uniform-24</th>\n",
       "      <th>1024</th>\n",
       "      <td>0.163313</td>\n",
       "      <td>39.476652</td>\n",
       "      <td>3.412015</td>\n",
       "      <td>3.969665</td>\n",
       "      <td>8.901694</td>\n",
       "      <td>1.090635</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"14\" valign=\"top\">sortledton.3</th>\n",
       "      <th>com-friendster</th>\n",
       "      <th>512</th>\n",
       "      <td>5.000507</td>\n",
       "      <td>260.627118</td>\n",
       "      <td>63.085961</td>\n",
       "      <td>34.759500</td>\n",
       "      <td>49.860027</td>\n",
       "      <td>17.116932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dota-league</th>\n",
       "      <th>512</th>\n",
       "      <td>0.001022</td>\n",
       "      <td>1.170734</td>\n",
       "      <td>49.754592</td>\n",
       "      <td>0.127522</td>\n",
       "      <td>0.277204</td>\n",
       "      <td>0.027643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>graph500-22</th>\n",
       "      <th>512</th>\n",
       "      <td>0.044506</td>\n",
       "      <td>7.326248</td>\n",
       "      <td>8.387854</td>\n",
       "      <td>0.447248</td>\n",
       "      <td>0.824945</td>\n",
       "      <td>0.137254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"8\" valign=\"top\">graph500-24</th>\n",
       "      <th>16</th>\n",
       "      <td>0.165297</td>\n",
       "      <td>31.975175</td>\n",
       "      <td>185.775184</td>\n",
       "      <td>4.925753</td>\n",
       "      <td>8.091575</td>\n",
       "      <td>1.526682</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>0.158436</td>\n",
       "      <td>23.623083</td>\n",
       "      <td>111.866535</td>\n",
       "      <td>3.559308</td>\n",
       "      <td>6.010531</td>\n",
       "      <td>1.102882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>0.157659</td>\n",
       "      <td>21.913115</td>\n",
       "      <td>77.417981</td>\n",
       "      <td>2.820035</td>\n",
       "      <td>5.131479</td>\n",
       "      <td>0.813570</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>0.157581</td>\n",
       "      <td>22.783623</td>\n",
       "      <td>64.812515</td>\n",
       "      <td>2.476586</td>\n",
       "      <td>4.268784</td>\n",
       "      <td>0.688123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>256</th>\n",
       "      <td>0.156484</td>\n",
       "      <td>23.093207</td>\n",
       "      <td>55.991248</td>\n",
       "      <td>2.299985</td>\n",
       "      <td>3.951504</td>\n",
       "      <td>0.658264</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <td>0.159607</td>\n",
       "      <td>23.521098</td>\n",
       "      <td>47.649419</td>\n",
       "      <td>2.236379</td>\n",
       "      <td>3.835369</td>\n",
       "      <td>0.624513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1024</th>\n",
       "      <td>0.157453</td>\n",
       "      <td>32.824404</td>\n",
       "      <td>44.241582</td>\n",
       "      <td>2.201741</td>\n",
       "      <td>3.824666</td>\n",
       "      <td>0.627702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2048</th>\n",
       "      <td>0.153184</td>\n",
       "      <td>31.337447</td>\n",
       "      <td>43.026160</td>\n",
       "      <td>2.190112</td>\n",
       "      <td>3.715067</td>\n",
       "      <td>0.626452</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>graph500-26</th>\n",
       "      <th>512</th>\n",
       "      <td>0.598947</td>\n",
       "      <td>149.688249</td>\n",
       "      <td>276.867575</td>\n",
       "      <td>11.900861</td>\n",
       "      <td>24.281894</td>\n",
       "      <td>3.098358</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>uniform-24</th>\n",
       "      <th>512</th>\n",
       "      <td>0.171751</td>\n",
       "      <td>43.936870</td>\n",
       "      <td>3.748761</td>\n",
       "      <td>4.021060</td>\n",
       "      <td>8.303376</td>\n",
       "      <td>1.061033</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>uniform-26</th>\n",
       "      <th>512</th>\n",
       "      <td>0.633141</td>\n",
       "      <td>173.083578</td>\n",
       "      <td>15.913968</td>\n",
       "      <td>20.835272</td>\n",
       "      <td>43.973580</td>\n",
       "      <td>5.534944</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "algorithm                                          bfs        cdlp  \\\n",
       "library            graph          block_size                         \n",
       "sorted_vector_al.6 graph500-24    1024        0.169685   16.213700   \n",
       "                   uniform-24     1024        0.163313   39.476652   \n",
       "sortledton.3       com-friendster 512         5.000507  260.627118   \n",
       "                   dota-league    512         0.001022    1.170734   \n",
       "                   graph500-22    512         0.044506    7.326248   \n",
       "                   graph500-24    16          0.165297   31.975175   \n",
       "                                  32          0.158436   23.623083   \n",
       "                                  64          0.157659   21.913115   \n",
       "                                  128         0.157581   22.783623   \n",
       "                                  256         0.156484   23.093207   \n",
       "                                  512         0.159607   23.521098   \n",
       "                                  1024        0.157453   32.824404   \n",
       "                                  2048        0.153184   31.337447   \n",
       "                   graph500-26    512         0.598947  149.688249   \n",
       "                   uniform-24     512         0.171751   43.936870   \n",
       "                   uniform-26     512         0.633141  173.083578   \n",
       "\n",
       "algorithm                                            lcc   pagerank  \\\n",
       "library            graph          block_size                          \n",
       "sorted_vector_al.6 graph500-24    1024         41.796077   2.125752   \n",
       "                   uniform-24     1024          3.412015   3.969665   \n",
       "sortledton.3       com-friendster 512          63.085961  34.759500   \n",
       "                   dota-league    512          49.754592   0.127522   \n",
       "                   graph500-22    512           8.387854   0.447248   \n",
       "                   graph500-24    16          185.775184   4.925753   \n",
       "                                  32          111.866535   3.559308   \n",
       "                                  64           77.417981   2.820035   \n",
       "                                  128          64.812515   2.476586   \n",
       "                                  256          55.991248   2.299985   \n",
       "                                  512          47.649419   2.236379   \n",
       "                                  1024         44.241582   2.201741   \n",
       "                                  2048         43.026160   2.190112   \n",
       "                   graph500-26    512         276.867575  11.900861   \n",
       "                   uniform-24     512           3.748761   4.021060   \n",
       "                   uniform-26     512          15.913968  20.835272   \n",
       "\n",
       "algorithm                                          sssp        wcc  \n",
       "library            graph          block_size                        \n",
       "sorted_vector_al.6 graph500-24    1024         3.786768   0.636032  \n",
       "                   uniform-24     1024         8.901694   1.090635  \n",
       "sortledton.3       com-friendster 512         49.860027  17.116932  \n",
       "                   dota-league    512          0.277204   0.027643  \n",
       "                   graph500-22    512          0.824945   0.137254  \n",
       "                   graph500-24    16           8.091575   1.526682  \n",
       "                                  32           6.010531   1.102882  \n",
       "                                  64           5.131479   0.813570  \n",
       "                                  128          4.268784   0.688123  \n",
       "                                  256          3.951504   0.658264  \n",
       "                                  512          3.835369   0.624513  \n",
       "                                  1024         3.824666   0.627702  \n",
       "                                  2048         3.715067   0.626452  \n",
       "                   graph500-26    512         24.281894   3.098358  \n",
       "                   uniform-24     512          8.303376   1.061033  \n",
       "                   uniform-26     512         43.973580   5.534944  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# graphalytics inserts\n",
    "sortledton_block_sizeonly = \"num_threads_read == 0 and (hostname == 'scyper21' or hostname == 'scyper22') and library == 'sortledton.3' or library == 'sorted_vector_al.6' and graph in ('graph500-22', 'graph500-24', 'graph500-26', 'uniform-24')\"\n",
    "\n",
    "\n",
    "data = common.import_gfe(\"view_graphalytics_inserts\").query(sortledton_block_sizeonly).copy() # data from the experiments\n",
    "data[\"build_frequency\"].fillna(pd.Timedelta(0), inplace=True) # replace NaT with 0, otherwise the records are ignored in the group by\n",
    "\n",
    "data = data.groupby([\"library\", \"graph\", \"block_size\", \"algorithm\"]) \\\n",
    "    .agg(completion_time=(\"median_secs\", \"median\"), count=(\"median_secs\", \"count\"))\n",
    "data = data.unstack(\"algorithm\")[(\"completion_time\")]\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Normalize against a sorted vector. Blocksize for the vector is only set as default, it means nothing for the vector.\n",
    "normalized = data.loc[\"sortledton.3\", \"graph500-24\", slice(None)] / data.loc[\"sorted_vector_al.6\", \"graph500-24\", 1024]\n",
    "normalized = normalized.sort_values(axis=1, by='algorithm')\n",
    "normalized = normalized.query(\"block_size != 2048\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 912x563.647 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "## FIGURE 4b\n",
    "visualize = normalized.copy()\n",
    "visualize = visualize.droplevel(\"library\")\n",
    "visualize = visualize.droplevel(\"graph\")\n",
    "visualize = visualize.drop(\"cdlp\", axis=1)\n",
    "\n",
    "colormap=common.set_algorithm_colormap()\n",
    "\n",
    "plotting.latexify(columns=3)\n",
    "\n",
    "p = visualize.plot.bar(color=colormap, linewidth=0.2)\n",
    "plt.xlabel(\"Block Size [edges]\")\n",
    "plt.ylabel(\"Slowdown\")\n",
    "plt.xticks(rotation=0)\n",
    "\n",
    "p.set_ylim(0, 5)\n",
    "\n",
    "legend_labels = [\"BFS\", \"LCC\", \"PR\", \"SSSP\", \"WCC\"]\n",
    "plt.legend(loc='upper right', labels=legend_labels).get_frame().set_linewidth(0.0)\n",
    "\n",
    "plotting.add_line_on_1(plt)\n",
    "plotting.format_axes(p, categorical_x=True)\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig(plotting.base_path + \"block-sizes-analytics-new.png\", dpi=400)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Inserts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": true
   },
   "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></th>\n",
       "      <th>edges_per_second</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>graph</th>\n",
       "      <th>block_size</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>graph500-22</th>\n",
       "      <th>512</th>\n",
       "      <td>4.104741e+06</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"8\" valign=\"top\">graph500-24</th>\n",
       "      <th>16</th>\n",
       "      <td>1.663785e+06</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>3.561707e+06</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>4.231400e+06</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>4.352898e+06</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>256</th>\n",
       "      <td>4.268190e+06</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <td>3.836104e+06</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1024</th>\n",
       "      <td>3.175241e+06</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2048</th>\n",
       "      <td>2.245332e+06</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>graph500-26</th>\n",
       "      <th>512</th>\n",
       "      <td>3.713227e+06</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>uniform-24</th>\n",
       "      <th>512</th>\n",
       "      <td>4.737711e+06</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        edges_per_second  count\n",
       "graph       block_size                         \n",
       "graph500-22 512             4.104741e+06      6\n",
       "graph500-24 16              1.663785e+06      6\n",
       "            32              3.561707e+06      6\n",
       "            64              4.231400e+06      6\n",
       "            128             4.352898e+06      5\n",
       "            256             4.268190e+06      5\n",
       "            512             3.836104e+06     11\n",
       "            1024            3.175241e+06      5\n",
       "            2048            2.245332e+06      5\n",
       "graph500-26 512             3.713227e+06      6\n",
       "uniform-24  512             4.737711e+06      7"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sortledton_block_sizeonly = \"(hostname == 'scyper21' or hostname == 'scyper22') and library == 'sortledton.3' and graph in ('graph500-22', 'graph500-24', 'graph500-26', 'uniform-24') and block_size != -1\"\n",
    "\n",
    "edges_per_graph = pd.DataFrame({\n",
    "    \"graph\": [\"friendster\", \"dota-league\", \"graph500-22\", \n",
    "              \"graph500-24\", \"graph500-26\", \"uniform-22\", \n",
    "              \"uniform-24\", \"uniform-26\"\n",
    "              ],\n",
    "    \"edges\": [1806067135, 50870313, 64155735, 260379520, 1051922853, 64155735, 260379520, 1051922853]\n",
    "})\n",
    "\n",
    "data = common.import_gfe(\"View_Inserts\").query(sortledton_block_sizeonly).copy() # data from the experiments\n",
    "data[\"build_frequency\"].fillna(pd.Timedelta(0), inplace=True) # replace NaT with 0, otherwise the records are ignored in the group by\n",
    "data = data.merge(edges_per_graph, on=\"graph\")\n",
    "data[\"edges_per_second\"] = data[\"edges\"] / data[\"completion_time_secs\"]\n",
    "\n",
    "data_grouped = data.groupby([\"graph\", \"block_size\"]) \\\n",
    "    .agg(edges_per_second=(\"edges_per_second\", \"median\"), count=(\"edges_per_second\", \"count\"))\n",
    "#data = data.unstack(\"algorithm\")[(\"completion_time\")]\n",
    "data_grouped"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1356x472.441 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "## FIGURE 7\n",
    "\n",
    "visualize = data_grouped.loc[\"graph500-24\"].copy()\n",
    "visualize = visualize.drop(\"count\", axis=1)\n",
    "\n",
    "plotting.latexify(columns=2, fig_height=plotting.cm2inch(3))\n",
    "\n",
    "p = visualize.plot.bar(color='royalblue',  linewidth=0.5)\n",
    "\n",
    "plt.xlabel(\"Block Size\")\n",
    "plt.ylabel(\"meps\")\n",
    "\n",
    "plt.xticks(rotation=0)\n",
    "\n",
    "p.get_legend().remove()\n",
    "\n",
    "scale_y = 1e6\n",
    "p.yaxis.set_major_locator(ticker.MultipleLocator(scale_y))\n",
    "p.yaxis.set_major_formatter(ticker.FuncFormatter(lambda x, pos: round(x / scale_y)))\n",
    "\n",
    "plotting.format_axes(p, categorical_x=True)\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig(plotting.base_path + \"block-sizes-insertions-new.png\", dpi=400)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}