plottingfunctions-jvms.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from itertools import islice\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from scipy import stats\n",
    "from  qgrid import show_grid as show \n",
    "import plotly.express as px\n",
    "import math\n",
    "import seaborn as sns\n",
    "from matplotlib import pyplot as plt \n",
    "import matplotlib as mpl\n",
    "from matplotlib.lines import Line2D \n",
    "from matplotlib.patches import Patch\n",
    "from itertools import cycle\n",
    "# pd.options.plotting.backend = \"plotly\"\n",
    "pd.options.plotting.backend ='matplotlib'\n",
    "\n",
    "\n",
    "# mpl.rcParams['boxplot.boxprops.color'] ='red'\n",
    "\n",
    "# mpl.rcParams['boxplot.boxprops.linewidth']=2\n",
    "# mpl.rcParams['patch.edgecolor']='black'\n",
    "\n",
    "# mpl.rcParams['boxplot.patchartist']= \"True\"\n",
    "# mpl.rcParams['boxplot.flierprops.marker'] = 'o'\n",
    "# mpl.rcParams['boxplot.flierprops.color'] = 'k'\n",
    "# mpl.rcParams['boxplot.flierprops.markerfacecolor'] = \"white\"\n",
    "# mpl.rcParams['boxplot.medianprops.color'] = 'red'\n",
    "#### axes \n",
    "mpl.rcParams['axes.titlesize'] = 30 \n",
    "mpl.rcParams['axes.titlepad'] = 22 \n",
    "mpl.rcParams['axes.labelsize']=30\n",
    "mpl.rcParams['axes.labelpad']=20\n",
    "mpl.rcParams['xtick.labelsize']= 25\n",
    "mpl.rcParams['ytick.labelsize']= 25\n",
    "mpl.rcParams['axes.grid']= True\n",
    "\n",
    "# labelpad\n",
    "# mpl.rc('axes',titlesize= 30)\n",
    "\n",
    "# mpl.rc('xtick', labelsize=22) \n",
    "# mpl.rc('ytick', labelsize=22) \n",
    "mpl.rcParams['text.usetex'] = True\n",
    "\n",
    "palette=[\"#D6CADD\",\n",
    "   \"#3B0D6E\",\n",
    "    \"#3240AB\",\n",
    "    \"#0095B6\",\n",
    "    \"#72A0C1\",\n",
    "    \"#7EECE5\",\n",
    "    \"#05B881\",\n",
    "    \"#49796B\",\n",
    "    \"#E2DC4C\",\n",
    "    \"#E0542A\",\n",
    "    \"#A83731\",\n",
    "    \"#2E3544\",\n",
    "    \"#AC2269\"\n",
    "]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def get_recap(x):\n",
    "    recap=pd.DataFrame()\n",
    "    mins=x.min()\n",
    "    maxs=x.max()\n",
    "    means=x.mean()\n",
    "    stds=x.std()\n",
    "    \n",
    "    mini=x.loc[x[\"duration\"]==mins[\"duration\"]]\n",
    "    mini[\"Metric\"]=\"Execution Time\"\n",
    "    recap=recap.append(mini)\n",
    "    mini=x.loc[x[\"CPU\"]==mins[\"CPU\"]]\n",
    "    mini[\"Metric\"]=\"Energy PKG\"\n",
    "    recap=recap.append(mini)\n",
    "    mini=x.loc[x[\"DRAM\"]==mins[\"DRAM\"]]\n",
    "    mini[\"Metric\"]=\"Energy DRAM\"\n",
    "    recap=recap.append(mini)\n",
    "    recap[\"Type\"]=\"Best\"\n",
    "    \n",
    "    maxi=x.loc[x[\"duration\"]==maxs[\"duration\"]]\n",
    "    maxi[\"Metric\"]=\"Execution Time\"\n",
    "    maxi[\"Type\"]=\"Worst\"\n",
    "    recap=recap.append(maxi)\n",
    "    maxi=x.loc[x[\"CPU\"]==maxs[\"CPU\"]]\n",
    "    maxi[\"Metric\"]=\"Energy PKG\"\n",
    "    maxi[\"Type\"]=\"Worst\"\n",
    "    recap=recap.append(maxi)\n",
    "    maxi=x.loc[x[\"DRAM\"]==maxs[\"DRAM\"]]\n",
    "    maxi[\"Metric\"]=\"Energy DRAM\"\n",
    "    maxi[\"Type\"]=\"Worst\"\n",
    "    recap=recap.append(maxi)\n",
    "    recap.reset_index(inplace=True)  \n",
    "    means[\"Type\"]=\"Mean\"\n",
    "    stds[\"Type\"]=\"STD\"\n",
    "    means[\"Metric\"]=\"All\"\n",
    "    stds[\"Metric\"]=\"All\"\n",
    "    recap=recap.append(means,ignore_index=True)\n",
    "    recap=recap.append(stds,ignore_index=True)\n",
    "    recap=recap.set_index([\"Type\",\"Metric\"]).loc[:,[\"duration\",\"CPU\",\"DRAM\",\"jvm\",\"options\"]]\n",
    "    recap[\"duration\"]=recap[\"duration\"]\n",
    "    recap[\"CPU\"]=recap[\"CPU\"]\n",
    "    recap[\"DRAM\"]=recap[\"DRAM\"]\n",
    "    recap.columns=[\"Execution Time (s)\",\"Energy CPU (J)\",\"Energy DRAM(J)\", \"JVM\",\"Execution Flags\"]\n",
    "    recap=recap.fillna('')\n",
    "    return recap\n",
    "\n",
    "def highlight_max(s):\n",
    "    is_max = np.logical_and(s == s[:-1].max()  ,s.apply(lambda x:type(x) is float))\n",
    "    return ['color: red' if v else '' for v in is_max]\n",
    "\n",
    "def highlight_min(s):\n",
    "\n",
    "    is_min =np.logical_and(s == s[:-1].min()  ,s.apply(lambda x:type(x) is float)) \n",
    "    return ['color: green' if v else '' for v in is_min]\n",
    "\n",
    "\n",
    "\n",
    "def read_data(path):\n",
    "    data=pd.read_csv(path,sep=\";\",dtype={\"jvm\":str,\"options\":str,\"iteration\":\"Int64\",\"exitcode\":\"Int64\"\n",
    "                                     ,\"duration\":\"Float64\",\n",
    "                                     \"DRAM\":\"Float64\",\"CPU\":\"Float64\"})\n",
    "    data=data.loc[data[\"exitcode\"]==0]\n",
    "    data[\"duration\"]=data[\"duration\"]/1000000\n",
    "    data[\"DRAM\"]=data[\"DRAM\"]/1000000\n",
    "    data[\"CPU\"]=data[\"CPU\"]/1000000\n",
    "    data[\"distribution\"]=data[\"jvm\"].apply(get_distribution)\n",
    "    data[\"version\"]=data[\"jvm\"].apply(get_version)\n",
    "    data=data.dropna()\n",
    "    return data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def plot_energy_cpu(df,benchmark,machine=\"\",save=False):\n",
    "    title=f\"Energy consumption of the benchmark {benchmark} within different JVMs\"\n",
    "    fig, axes = plt.subplots(1, 1, figsize=(32, 14))\n",
    "\n",
    "    fig=df.loc[df[\"benchmark\"]==benchmark].boxplot(by=\"jvm\",column=\"energy_pkg\",ax=axes,rot=-90,return_type=\"both\",showfliers=False)\n",
    "\n",
    "    fig[0].ax.set_ylabel(ylabel=\"Energy Consumption (J)\",labelpad=22,fontsize=\"30\")\n",
    "    fig[0].ax.get_figure().suptitle('')\n",
    "    fig[0].ax.set_xlabel(xlabel='')\n",
    "    ########## legends \n",
    "    legends=[Line2D([0], [0], marker='o', color='k', label='J9',\n",
    "                          markerfacecolor='#3498DB', markersize=30),\n",
    "        Line2D([0], [0], marker='o', color='k', label='GraalVM',\n",
    "                          markerfacecolor='#16A085', markersize=30),\n",
    "        Line2D([0], [0], marker='o', color='k', label='OpenJDK',\n",
    "                          markerfacecolor='#F0B27A', markersize=30),\n",
    "            Line2D([0], [0], marker='o', color='k', label='Others',\n",
    "                          markerfacecolor='w', markersize=30)\n",
    "            ]\n",
    "    \n",
    "    fig[0].ax.legend(handles=legends, loc='lower left',fontsize='xx-large',labelspacing=1)\n",
    "    \n",
    "    ############## inside the boxplots \n",
    "    for i in range(len(fig[0][1]['boxes'])): \n",
    "        color = get_filling_color(fig[0][0].get_xticklabels()[i].get_text())\n",
    "        fig[0][1]['boxes'][i].set(color=\"b\",edgecolor='black',facecolor=color,linewidth=1)\n",
    "        fig[0][1]['medians'][i].set(color=\"red\",linewidth=0.5)\n",
    "        fig[0][1]['whiskers'][i].set(color=\"black\",linewidth=1)\n",
    "    \n",
    "    plt.subplots_adjust(left=0, bottom=0, right=1, top=1, wspace=0, hspace=0)\n",
    "    if save :\n",
    "        fig[0].ax.set_title(\"\",fontsize=\"30\",pad=50  )\n",
    "        plt.savefig(f\"energy_pkg_all_{benchmark}{machine}.pdf\",format=\"pdf\")\n",
    "    else:\n",
    "        fig[0].ax.set_title(title,fontsize=\"30\",pad=50  )\n",
    "        plt.show()\n",
    "    return fig\n",
    "\n",
    "def plot_execution_time(df,benchmark,machine=\"\",save=False):\n",
    "    title=f\"Excution time of the benchmark {benchmark} within different JVMs\"\n",
    "    fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "#     plt.subplots_adjust(wspace=0.25,top=4,left = 0,right = 1)\n",
    "    fig=df.loc[df[\"benchmark\"]==benchmark].boxplot(by=\"jvm\",column=\"execution_time\",ax=axes,rot=-90,return_type=\"both\",showfliers=False)\n",
    "\n",
    "    \n",
    "    fig[0].ax.set_ylabel(ylabel=\"Execution time (s)\",labelpad=22,fontsize=\"30\")\n",
    "    fig[0].ax.get_figure().suptitle('')\n",
    "    fig[0].ax.set_xlabel(xlabel='')\n",
    "    \n",
    "    legends=[Line2D([0], [0], marker='o', color='k', label='J9',\n",
    "                          markerfacecolor='#3498DB', markersize=30),\n",
    "        Line2D([0], [0], marker='o', color='k', label='GraalVM',\n",
    "                          markerfacecolor='#16A085', markersize=30),\n",
    "        Line2D([0], [0], marker='o', color='k', label='OpenJDK',\n",
    "                          markerfacecolor='#F0B27A', markersize=30),\n",
    "            Line2D([0], [0], marker='o', color='k', label='Others',\n",
    "                          markerfacecolor='w', markersize=30)\n",
    "            \n",
    "            ]\n",
    "    \n",
    "    fig[0].ax.legend(handles=legends, loc='lower left',fontsize='xx-large',labelspacing=1)\n",
    "    \n",
    "    for i in range(len(fig[0][1]['boxes'])): \n",
    "        color = get_filling_color(fig[0][0].get_xticklabels()[i].get_text())\n",
    "        fig[0][1]['boxes'][i].set(color=\"b\",edgecolor='black',facecolor=color,linewidth=1)\n",
    "        fig[0][1]['medians'][i].set(color=\"red\",linewidth=0.5)\n",
    "        fig[0][1]['whiskers'][i].set(color=\"black\",linewidth=1)\n",
    "    if save :\n",
    "        fig[0].ax.set_title(\"\",fontsize=\"30\",pad=50  )\n",
    "        plt.savefig(f\"execution_time_all_{benchmark}{machine}.pdf\",format=\"pdf\")\n",
    "    else:\n",
    "        fig[0].ax.set_title(title,fontsize=\"30\",pad=50  )\n",
    "        plt.show()\n",
    "    return fig\n",
    "\n",
    "\n",
    "def plot_energy_dram(df,benchmark,machine=\"\",save=False):\n",
    "    title=f\"DRAM Energy consumption of the benchmark {benchmark} within different JVMs\"\n",
    "    fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "    plt.tight_layout()\n",
    "    \n",
    "    fig=df.loc[df[\"benchmark\"]==benchmark].boxplot(by=\"jvm\",column=\"energy_dram\",ax=axes,rot=-90,return_type=\"both\",showfliers=False)\n",
    "\n",
    "\n",
    "    \n",
    "    fig[0].ax.set_ylabel(ylabel=\"Energy Consumption (J)\",labelpad=22,fontsize=\"30\")\n",
    "    fig[0].ax.get_figure().suptitle('')\n",
    "    fig[0].ax.set_xlabel(xlabel='')\n",
    "    ########## legends \n",
    "    legends=[Line2D([0], [0], marker='o', color='k', label='J9',\n",
    "                          markerfacecolor='#3498DB', markersize=30),\n",
    "        Line2D([0], [0], marker='o', color='k', label='GraalVM',\n",
    "                          markerfacecolor='#16A085', markersize=30),\n",
    "        Line2D([0], [0], marker='o', color='k', label='OpenJDK',\n",
    "                          markerfacecolor='#F0B27A', markersize=30),\n",
    "            Line2D([0], [0], marker='o', color='k', label='Others',\n",
    "                          markerfacecolor='w', markersize=30)\n",
    "            ]\n",
    "    \n",
    "    fig[0].ax.legend(handles=legends, loc='upper left',fontsize='xx-large',labelspacing=1)\n",
    "    \n",
    "    ############## inside the boxplots \n",
    "    for i in range(len(fig[0][1]['boxes'])): \n",
    "        color = get_filling_color(fig[0][0].get_xticklabels()[i].get_text())\n",
    "        fig[0][1]['boxes'][i].set(color=\"b\",edgecolor='black',facecolor=color,linewidth=1)\n",
    "        fig[0][1]['medians'][i].set(color=\"red\",linewidth=0.5)\n",
    "        fig[0][1]['whiskers'][i].set(color=\"black\",linewidth=1)\n",
    "    plt.subplots_adjust(left=0, bottom=0, right=1, top=1, wspace=0, hspace=0)\n",
    "    if save :\n",
    "        fig[0].ax.set_title(\"\",fontsize=\"30\",pad=50  )\n",
    "        plt.savefig(f\"energy_dram_all_{benchmark}{machine}.pdf\",format=\"pdf\")\n",
    "    else:\n",
    "        fig[0].ax.set_title(title,fontsize=\"30\",pad=50  )\n",
    "        plt.show()\n",
    "    return fig\n",
    "\n",
    "\n",
    "def plot_power_cpu(df,benchmark,machine=\"\",save=False):\n",
    "    title=f\"Average power  the benchmark {benchmark} within different JVMs\"\n",
    "    fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "    fig=df.loc[df[\"benchmark\"]==benchmark].boxplot(by=\"jvm\",column=\"av_power_cpu\",ax=axes,rot=-90,return_type=\"both\",showfliers=False)\n",
    "\n",
    "\n",
    "    \n",
    "    fig[0].ax.set_ylabel(ylabel=\"Average Power Consumption (W)\",labelpad=22,fontsize=\"30\")\n",
    "    fig[0].ax.get_figure().suptitle('')\n",
    "    fig[0].ax.set_xlabel(xlabel='')\n",
    "    ########## legends \n",
    "    legends=[Line2D([0], [0], marker='o', color='k', label='J9',\n",
    "                          markerfacecolor='#3498DB', markersize=30),\n",
    "        Line2D([0], [0], marker='o', color='k', label='GraalVM',\n",
    "                          markerfacecolor='#16A085', markersize=30),\n",
    "        Line2D([0], [0], marker='o', color='k', label='OpenJDK',\n",
    "                          markerfacecolor='#F0B27A', markersize=30),\n",
    "            Line2D([0], [0], marker='o', color='k', label='Others',\n",
    "                          markerfacecolor='w', markersize=30)\n",
    "            ]\n",
    "    \n",
    "    fig[0].ax.legend(handles=legends, loc='upper left',fontsize='xx-large',labelspacing=1)\n",
    "    \n",
    "    ############## inside the boxplots \n",
    "    for i in range(len(fig[0][1]['boxes'])): \n",
    "        color = get_filling_color(fig[0][0].get_xticklabels()[i].get_text())\n",
    "        fig[0][1]['boxes'][i].set(color=\"b\",edgecolor='black',facecolor=color,linewidth=1)\n",
    "        fig[0][1]['medians'][i].set(color=\"red\",linewidth=0.5)\n",
    "        fig[0][1]['whiskers'][i].set(color=\"black\",linewidth=1)\n",
    "    if save :\n",
    "        fig[0].ax.set_title(\"\",fontsize=\"30\",pad=50  )\n",
    "        plt.savefig(f\"power_pkg_all_{benchmark}{machine}.pdf\",format=\"pdf\")\n",
    "    else:\n",
    "        fig[0].ax.set_title(title,fontsize=\"30\",pad=50  )\n",
    "        plt.show()\n",
    "    return fig\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#BOX Plots \n",
    "\n",
    "def boxplot_cpu(df,benchmark,axes):\n",
    "    \n",
    "    title=f\"Energy consumption of CPU for the benchmark {benchmark} within different JVMs\"\n",
    "    fig =plot_box(df,\"CPU\",axes)\n",
    "    fig[0].ax.set_ylabel(ylabel=\"Energy Consumption (J)\")\n",
    "    fig[0].ax.get_figure().suptitle('')\n",
    "    fig[0].ax.set_xlabel(xlabel='')\n",
    "    fig[0].ax.set_title(title  )\n",
    "    \n",
    "def boxplot_dram(df,benchmark,axes):\n",
    "    \n",
    "    title=f\"Energy consumption of DRAM for the benchmark  {benchmark} within different JVMs\"\n",
    "    fig =plot_box(df,\"DRAM\",axes)\n",
    "    fig[0].ax.set_ylabel(ylabel=\"Energy Consumption (J)\")\n",
    "    fig[0].ax.get_figure().suptitle('')\n",
    "    fig[0].ax.set_xlabel(xlabel='')\n",
    "    fig[0].ax.set_title(title  )\n",
    "    \n",
    "def boxplot_duration(df,benchmark,axes):\n",
    "    title=f\"Excution time of the benchmark {benchmark} within different JVMs\"\n",
    "    fig =plot_box(df,\"duration\",axes)\n",
    "    fig[0].ax.set_ylabel(ylabel=\"Execution Time (s)\")\n",
    "    fig[0].ax.get_figure().suptitle('')\n",
    "    fig[0].ax.set_xlabel(xlabel='')\n",
    "    fig[0].ax.set_title(title  )\n",
    "    \n",
    "def plot_box(df,metric,axes):\n",
    "    colors=dict(zip(df[\"distribution\"].unique(),cycle(palette)))\n",
    "    legends=[Patch(facecolor=color, label=distribution) for distribution,color in colors.items()]\n",
    "    \n",
    "    fig=df.boxplot(by=\"name\",column=metric,ax=axes,rot=-90,return_type=\"both\",showfliers=False,backend='matplotlib')\n",
    "   ############# inside the boxplots \n",
    "    \n",
    "    for i in range(len(fig[0][1]['boxes'])): \n",
    "        jvm=fig[0][0].get_xticklabels()[i].get_text().split(\" \")[1]\n",
    "        distribution=get_distribution(jvm)\n",
    "        color = colors[distribution]\n",
    "        fig[0][1]['boxes'][i].set(color=color,linewidth=2)\n",
    "        fig[0][1]['medians'][i].set(color=\"red\",linewidth=1)\n",
    "    for i in range(len(fig[0][1]['whiskers'])): \n",
    "        fig[0][1]['whiskers'][i].set(color=color,linewidth=1)\n",
    "        fig[0][1]['caps'][i].set(color=color,linewidth=1) \n",
    "    axes.legend(handles=legends,fontsize=25,loc='center left', bbox_to_anchor=(1.0, 0.5))\n",
    "    \n",
    "    axes.set_xticklabels([ x.get_text().split(' ')[1] for x in axes.get_xticklabels()])\n",
    "    return fig\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.distribution.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#BarPlot\n",
    "\n",
    "\n",
    "def plot_stacked_bar(full_data,metric,base,axes):\n",
    "    full_data=data.sort_values([\"name\",\"benchmark\"]).reset_index()\n",
    "    benchmarks=full_data[\"benchmark\"].unique()\n",
    "    l2={bench:full_data.loc[full_data[\"benchmark\"]==bench]for bench in benchmarks}\n",
    "    jvms=full_data[\"name\"].unique()\n",
    "    means={bench:df.groupby([\"name\"]).mean()[metric]  for bench,df in l2.items() }\n",
    "    means=[df/base.loc[bench,metric]*100 / len(benchmarks) for  bench,df in means.items()]\n",
    "    for i in range(len(benchmarks)) :\n",
    "        axes.bar(jvms,means[i],bottom=sum(means[:i]),color=palette[i],label=benchmarks[i])\n",
    "   \n",
    "    axes.legend(fontsize=25,loc='center left', bbox_to_anchor=(1.0, 0.5))\n",
    "    axes.yaxis.set_major_formatter(mpl.ticker.PercentFormatter())\n",
    "    axes.tick_params(axis='x', labelrotation = -90)\n",
    "#     axes.set_xticks([ x.get_text().split(' ')[1] for x in axes.get_xticklabels()])\n",
    "    \n",
    "    return axes\n",
    "\n",
    "\n",
    "def plot_stacked_cpu(full_data,basename,base,axes):\n",
    "    title=f\"Relative Energy consumption of CPU for all benchmarks based on {basename}\"\n",
    "    axes=plot_stacked_bar(full_data,\"CPU\",base,axes)\n",
    "    axes.set_ylabel(ylabel=\"Energy Consumption ratio (\\%)\")\n",
    "    axes.set_xlabel(xlabel=\"JVM Distribution\")\n",
    "    axes.set_title(title)\n",
    "    return axes\n",
    "\n",
    "def plot_stacked_dram(full_data,basename,base,axes):\n",
    "    title=f\"Relative Energy consumption of DRAM for all benchmarks based on {basename}\"\n",
    "    axes=plot_stacked_bar(full_data,\"DRAM\",base,axes)\n",
    "    axes.set_ylabel(ylabel=\"Energy Consumption ratio (\\%)\")\n",
    "    axes.set_xlabel(xlabel=\"JVM Distribution\")\n",
    "    axes.set_title(title)\n",
    "    return axes\n",
    "def plot_stacked_duration(full_data,basename,base,axes):\n",
    "    title=f\"Relative Execution time for all benchmarks based on {basename}\"\n",
    "    axes=plot_stacked_bar(full_data,\"DRAM\",base,axes)\n",
    "    axes.set_ylabel(ylabel=\"Execution Time ratio (\\%)\")\n",
    "    axes.set_xlabel(xlabel=\"JVM Distribution\")\n",
    "    axes.set_title(title)\n",
    "    return axes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "jvm_labels={\n",
    "'open' : 'HotSpot' ,\n",
    "'zulu' : 'Zulu' ,\n",
    "'albba' : 'Dragonwell (LTS)' ,\n",
    "'amzn' : 'Corretto (MJR)' ,\n",
    "'librca' : 'Librca' ,\n",
    "'sapmchn' : 'SapMachine' ,\n",
    "'trava' : 'HotSpotTrava (LTS)' ,\n",
    "'hs-adpt' : 'HotSpotAdopt' ,\n",
    "'j9-adpt' : 'J9' ,\n",
    "'grl-19' : 'GraalVM-19 (LTS)' ,\n",
    "'mandrel' : 'Mandrel' ,\n",
    "'grl-20' : 'GraalVM-20 (LTS)',\n",
    "    }\n",
    "\n",
    "def get_distribution(jvm):\n",
    "    s=jvm.split(\"-\")[-1]\n",
    "    if s == \"grl\":\n",
    "        version=jvm.split(\".\")[0]\n",
    "        s=f\"grl-{version}\"\n",
    "    elif s == \"adpt\" : \n",
    "        s=jvm.split(\".\")[-1]\n",
    "\n",
    "    return s\n",
    "\n",
    "def get_version(jvm):\n",
    "    s=jvm.split(\"-\")[-1]\n",
    "    jvm=jvm.split(\"-\")[0]\n",
    "    if s == \"grl\" :\n",
    "        s=jvm.split(\".\")[-1].replace(\"r\",\"\")\n",
    "    elif s ==\"albba\" : \n",
    "        s=jvm.split(\".\")[0].split(\"u\")[0]\n",
    "    elif s ==\"mandrel\" :\n",
    "        s=11\n",
    "    else :\n",
    "        s = jvm.split(\".\")[0]\n",
    "    try : \n",
    "        s= int(s)\n",
    "    except : \n",
    "        s= 0\n",
    "    return s\n",
    "\n",
    "def plot_lineplots_global(data,metric,base,axes):\n",
    "    full_data=data.sort_values([\"jvm\",\"benchmark\"])\n",
    "    benchmarks=full_data[\"benchmark\"].unique()\n",
    "    l2={bench:full_data.loc[full_data[\"benchmark\"]==bench]for bench in benchmarks}\n",
    "    jvms=full_data[\"jvm\"].unique()\n",
    "    means={bench:df.groupby([\"jvm\"]).mean()[metric]  for bench,df in l2.items() }\n",
    "    means=[df/base.loc[bench,metric]*100 / len(benchmarks) for  bench,df in means.items()]\n",
    "    full_data=pd.concat(means).reset_index()\n",
    "    full_data[\"version\"]=full_data[\"jvm\"].apply(get_version)\n",
    "    full_data[\"distribution\"]=full_data[\"jvm\"].apply(get_distribution)\n",
    "    full_data=full_data.groupby([\"jvm\",\"version\",\"distribution\"])[metric].sum().reset_index()\n",
    "    full_data=full_data.sort_values(\"version\")   \n",
    "    for i in full_data[\"distribution\"].unique() : \n",
    "        x= full_data.loc[full_data[\"distribution\"]==i]      \n",
    "        axes.plot(x[\"version\"],x[metric])\n",
    "        axes.scatter(x[\"version\"],x[metric],s=100,label=jvm_labels[i]) \n",
    "    axes.legend(fontsize=25,loc='center left', bbox_to_anchor=(1.0, 0.5))\n",
    "    axes.yaxis.set_major_formatter(mpl.ticker.PercentFormatter())\n",
    "    axes.set_xlabel(xlabel=\"Java version \")\n",
    "    return axes\n",
    "\n",
    "def plot_lineplots_global_cpu(full_data,basename, base,axes,benchname=\"\"):\n",
    "    title=f\"The evolution of average Energy consumption of CPU through different java versions based on {basename} {benchname}\"\n",
    "    axes=plot_lineplots_global(full_data,\"CPU\",base,axes)\n",
    "    axes.set_ylabel(ylabel=\"Energy Consumption ratio (\\%)\")\n",
    "    axes.set_xlabel(xlabel=\"JVM Distribution\")\n",
    "    axes.set_title(title)\n",
    "    return axes\n",
    "\n",
    "def plot_lineplots_global_dram(full_data,basename,base,axes,benchname=\"\"):\n",
    "    title=f\"The evolution of average Energy consumption of DRAM through different java versions based on {basename} {benchname}\"\n",
    "    axes=plot_lineplots_global(full_data,\"DRAM\",base,axes)\n",
    "    axes.set_ylabel(ylabel=\"Energy Consumption ratio (\\%)\")\n",
    "    axes.set_xlabel(xlabel=\"JVM Distribution\")\n",
    "    axes.set_title(title)\n",
    "    return axes\n",
    "\n",
    "def plot_lineplots_global_duration(full_data,basename,base,axes,benchname=\"\"):\n",
    "    title=f\"The evolution of average Execution time through different java versions based on {basename} {benchname} \"\n",
    "    axes=plot_lineplots_global(full_data,\"duration\",base,axes)\n",
    "    axes.set_ylabel(ylabel=\"Execution Time ratio (\\%)\")\n",
    "    axes.set_xlabel(xlabel=\"JVM Distribution\")\n",
    "    axes.set_title(title)\n",
    "    return axes\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_evolution(data2,metric,axes):\n",
    "    base=data2.loc[data2[\"version\"]==min(data2[\"version\"])]\n",
    "    base=base.groupby(\"benchmark\").mean()\n",
    "    data2=data2.groupby([\"version\",\"benchmark\"])[metric].mean().reset_index()\n",
    "    data2=data2.sort_values([\"version\",\"benchmark\"])\n",
    "    for i in data2[\"benchmark\"].unique() : \n",
    "        x= data2.loc[data2[\"benchmark\"]==i]\n",
    "        x[\"relative_metric\"]=x[metric]/base.loc[i,metric]*100\n",
    "        axes.plot(x[\"version\"],x[\"relative_metric\"])\n",
    "        axes.scatter(x[\"version\"],x[\"relative_metric\"],s=100,label=i)\n",
    "    \n",
    "    axes.legend(fontsize=25,loc='center left', bbox_to_anchor=(1.0, 0.5),)\n",
    "    axes.set_xlabel(xlabel=\"Java version \")\n",
    "    axes.yaxis.set_major_formatter(mpl.ticker.PercentFormatter())\n",
    "    return axes \n",
    "    \n",
    "\n",
    "def plot_evolution_cpu(full_data,axes):\n",
    "    title=f\"The evolution of average Energy consumption of CPU through different  java versions \"\n",
    "    axes=plot_evolution(full_data,\"CPU\",axes)\n",
    "    axes.set_ylabel(ylabel=\"Energy Consumption ratio (\\%)\")\n",
    "    axes.set_title(title)\n",
    "    return axes\n",
    "\n",
    "def plot_evolution_dram(full_data,axes):\n",
    "    title=f\"The evolution of average Energy consumption of DRAM through different java versions \"\n",
    "    axes=plot_evolution(full_data,\"DRAM\",axes)\n",
    "    axes.set_ylabel(ylabel=\"Energy Consumption ratio (\\%)\")\n",
    "    axes.set_title(title)\n",
    "    return axes\n",
    "\n",
    "def plot_evolution_duration(full_data,axes):\n",
    "    title=f\"The evolution of average Execution time through different java versions \"\n",
    "    axes=plot_evolution(full_data,\"duration\",axes)\n",
    "    axes.set_ylabel(ylabel=\"Execution Time ratio (\\%)\")\n",
    "    axes.set_title(title)\n",
    "    return axes\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def bar_plot_energy(df,save=False):\n",
    "    title=f\"Energy consumption within different JVMs\"\n",
    "    fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "    y=df.pivot(columns=\"jvm\",index='benchmark', values='CPU')\n",
    "    y=y.loc[:,[\"20.2.0.r11-grl\",\"15.0.1.j9-adpt\",\"15.0.1-open\",\"8.0.265-open\"]]\n",
    "\n",
    "   \n",
    "    y[\"20.2.0.r11-grl\"]=y.apply(lambda row : row[\"20.2.0.r11-grl\"]/row[\"8.0.265-open\"]*100,axis=1)\n",
    "    y[\"15.0.1.j9-adpt\"]=y.apply(lambda row : row[\"15.0.1.j9-adpt\"]/row[\"8.0.265-open\"]*100,axis=1)\n",
    "    y[\"15.0.1-open\"]=y.apply(lambda row : row[\"15.0.1-open\"]/row[\"8.0.265-open\"]*100,axis=1)\n",
    "    y[\"8.0.265-open\"]=y.apply(lambda row : row[\"8.0.265-open\"]/row[\"8.0.265-open\"]*100,axis=1)\n",
    "    \n",
    "    fig= y.plot.bar(align='center' ,rot=0,ax=axes,color={\"15.0.1.j9-adpt\":\"#3498DB\",\"20.2.0.r11-grl\":\"#16A085\",\"15.0.1-open\":\"#F0B27A\",\"8.0.265-open\":\"#ea5f5b\"},grid=True)\n",
    "    \n",
    "    fig.set_ylabel(ylabel=\"Energy Consumption ratio (\\%)\")\n",
    "    fig.get_figure().suptitle('')\n",
    "    fig.set_xlabel(xlabel='')\n",
    "    fig.yaxis.set_major_formatter(mpl.ticker.PercentFormatter())\n",
    "    ########## legends \n",
    "    \n",
    "    legends=[\n",
    "        Line2D([0], [0], marker='o', color='k', label='GraalVM',\n",
    "                          markerfacecolor='#16A085', markersize=30),\n",
    "        Line2D([0], [0], marker='o', color='k', label='J9',\n",
    "                          markerfacecolor='#3498DB', markersize=30),\n",
    "        Line2D([0], [0], marker='o', color='k', label='HotSpot-8',\n",
    "                          markerfacecolor='#ea5f5b', markersize=30),\n",
    "        Line2D([0], [0], marker='o', color='k', label='HotSpot-15',\n",
    "                          markerfacecolor='#F0B27A', markersize=30),\n",
    "              \n",
    "#             Line2D([0], [0], marker='o', color='k', label='Others',\n",
    "#                           markerfacecolor='w', markersize=30)\n",
    "            ]\n",
    "    \n",
    "    fig.legend(handles=legends, loc='upper left',fontsize='25',labelspacing=1)\n",
    "    \n",
    "    plt.xlim([-0.5, 11.65])\n",
    "    \n",
    "    plt.tight_layout()\n",
    "    if save :\n",
    "        fig.set_title(\"\",fontsize=\"30\",pad=50  )\n",
    "        plt.savefig(f\"bar_plot_all_basedon8.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n",
    "    else:\n",
    "        fig.set_title(title,fontsize=\"30\",pad=50  )\n",
    "        plt.show()\n",
    "    return fig"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "data=read_data(\"alldata_dacapo_renaissance.csv\")\n",
    "targetdir=\"results_all_data/dacapo_renaissance\"\n",
    "# data=read_data(\"data_gc.csv\")\n",
    "# data=read_data(\"data_jit.csv\")\n",
    "# data=read_data(\"jvm-recommender/zip4j_1_chetemi_14.csv\")\n",
    "\n",
    "data[\"name\"]=data.apply(lambda row : f\"{(row['distribution'])} {row['jvm']}\",axis=1)\n",
    "data.sort_values([\"distribution\",\"jvm\"], inplace=True)\n",
    "basename=\"8.0.265-open\"\n",
    "base=data.loc[data[\"jvm\"]==basename]\n",
    "base=base.groupby(\"benchmark\").mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "fig=plot_stacked_cpu(data,basename,base,axes)\n",
    "plt.savefig(f\"{targetdir}/bar_all_cpu_basedon_8.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n",
    "fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "fig=plot_stacked_dram(data,basename,base,axes)\n",
    "plt.savefig(f\"{targetdir}/bar_all_dram_basedon_8.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n",
    "fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "fig=plot_stacked_duration(data,basename,base,axes)\n",
    "plt.savefig(f\"{targetdir}/bar_all_duration_basedon_8.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "# fig=plot_lineplots_global_cpu(data,basename,base,axes)\n",
    "# plt.savefig(f\"{targetdir}/global_evolution_cpu_basedon_8.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n",
    "# fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "# fig=plot_lineplots_global_dram(data,basename,base,axes)\n",
    "# plt.savefig(f\"{targetdir}/global_evolution_dram_basedon_8.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n",
    "# fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "# fig=plot_lineplots_global_duration(data,basename,base,axes)\n",
    "# plt.savefig(f\"{targetdir}/global_evolution_duration_basedon_8.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# benchmarks=data.benchmark.unique()\n",
    "# for bench in benchmarks: \n",
    "#     df=data.loc[data[\"benchmark\"]==bench]\n",
    "#     fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "#     _=plot_lineplots_global_cpu(df,basename,base,axes,f\"for {bench}\")\n",
    "#     plt.savefig(f\"{targetdir}/evolution_cpu_basedon_8_{bench}.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n",
    "#     plt.close(fig)\n",
    "#     fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "#     _=plot_lineplots_global_dram(df,basename,base,axes,f\"for {bench}\")\n",
    "#     plt.savefig(f\"{targetdir}/evolution_dram_basedon_8_{bench}.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n",
    "#     plt.close(fig)\n",
    "#     fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "#     _=plot_lineplots_global_duration(df,basename,base,axes,f\"for {bench}\")\n",
    "#     plt.savefig(f\"{targetdir}/evolution_duration_basedon_8_{bench}.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n",
    "#     plt.close(fig)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# distributions=data[\"distribution\"].unique()\n",
    "# for dist in distributions: \n",
    "#     df=data.loc[data[\"distribution\"]==dist]\n",
    "#     fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "#     _=plot_evolution_cpu(df,axes)\n",
    "#     plt.savefig(f\"{targetdir}/evolution_cpu_{dist}.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n",
    "#     plt.close(fig)\n",
    "#     fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "#     _=plot_evolution_dram(df,axes)\n",
    "#     plt.savefig(f\"{targetdir}/evolution_dram_{dist}.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n",
    "#     plt.close(fig)\n",
    "#     fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "#     _=plot_evolution_duration(df,axes)\n",
    "#     plt.savefig(f\"{targetdir}/evolution_duration_{dist}.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n",
    "#     plt.close(fig)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "benchmarks=data.benchmark.unique()\n",
    "\n",
    "for bench in benchmarks: \n",
    "    df=data.loc[data[\"benchmark\"]==bench]\n",
    "    fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "    _=boxplot_cpu(df,bench,axes)\n",
    "    plt.savefig(f\"{targetdir}/boxplot_cpu_{bench}.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n",
    "    plt.close(fig)\n",
    "    fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "    _=boxplot_dram(df,bench,axes)\n",
    "    plt.savefig(f\"{targetdir}/boxplot_dram_{bench}.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n",
    "    plt.close(fig)\n",
    "    fig, axes = plt.subplots(1, 1, figsize=(32, 12))\n",
    "    _=boxplot_duration(df,bench,axes)\n",
    "    plt.savefig(f\"{targetdir}/boxplot_duration_{bench}.pdf\",format=\"pdf\",bbox_inches='tight',pad_inches=0)\n",
    "    plt.close(fig)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.6"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}