Lava: Split the benchmark plots for clarification

[lttng-ci.git] / scripts / lttng-baremetal-tests / generate-plots.py

# Copyright (C) 2017 - Francis Deslauriers <francis.deslauriers@efficios.com>
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program.  If not, see <http://www.gnu.org/licenses/>.


import os, sys
import numpy as np
import pandas as pd

#Set Matplotlib to use the PNG non interactive backend
import matplotlib as mpl
mpl.use('Agg')

import matplotlib.pyplot as plt
from matplotlib.ticker import MaxNLocator
from cycler import cycler
from collections import OrderedDict

def rename_cols(df):
    new_cols = {'baseline_1thr_peritermean': 'basel_1thr',
            'baseline_2thr_peritermean': 'basel_2thr',
            'baseline_4thr_peritermean': 'basel_4thr',
            'baseline_8thr_peritermean': 'basel_8thr',
            'baseline_16thr_peritermean': 'basel_16thr',
            'lttng_1thr_peritermean': 'lttng_1thr',
            'lttng_2thr_peritermean': 'lttng_2thr',
            'lttng_4thr_peritermean': 'lttng_4thr',
            'lttng_8thr_peritermean': 'lttng_8thr',
            'lttng_16thr_peritermean': 'lttng_16thr',
            'baseline_1thr_periterstdev': 'basel_1thr_stdev',
            'baseline_2thr_periterstdev': 'basel_2thr_stdev',
            'baseline_4thr_periterstdev': 'basel_4thr_stdev',
            'baseline_8thr_periterstdev': 'basel_8thr_stdev',
            'baseline_16thr_periterstdev': 'basel_16thr_stdev',
            'lttng_1thr_periterstdev': 'lttng_1thr_stdev',
            'lttng_2thr_periterstdev': 'lttng_2thr_stdev',
            'lttng_4thr_periterstdev': 'lttng_4thr_stdev',
            'lttng_8thr_periterstdev': 'lttng_8thr_stdev',
            'lttng_16thr_periterstdev': 'lttng_16thr_stdev'
            }
    df.rename(columns=new_cols, inplace=True)
    return df

def convert_us_to_ns(df):
    cols = [col for col in df.columns if 'periter' in col]
    df[cols] = df[cols].apply(lambda x: x*1000)
    return df

def create_plot(df, graph_type):
    # We map all test configurations and their
    # respective color
    conf_to_color = OrderedDict([
        ('basel_1thr','lightcoral'),
        ('lttng_1thr','red'),
        ('basel_2thr','gray'),
        ('lttng_2thr','black'),
        ('basel_4thr','chartreuse'),
        ('lttng_4thr','forestgreen'),
        ('basel_8thr','deepskyblue'),
        ('lttng_8thr','mediumblue'),
        ('basel_16thr','orange'),
        ('lttng_16thr','saddlebrown')])

    # We create a list for each of the subplots
    baseline = [x for x in conf_to_color.keys() if 'basel' in x]
    lttng = [x for x in conf_to_color.keys() if 'lttng' in x]
    one_thr = [x for x in conf_to_color.keys() if '_1thr' in x]
    two_thr = [x for x in conf_to_color.keys() if '_2thr' in x]
    four_thr = [x for x in conf_to_color.keys() if '_4thr' in x]
    eight_thr = [x for x in conf_to_color.keys() if '_8thr' in x]
    sixteen_thr = [x for x in conf_to_color.keys() if '_16thr' in x]

    plots = [baseline, lttng, one_thr, two_thr, four_thr, eight_thr, sixteen_thr]

    title='Meantime per syscalls for {} testcase'.format(graph_type)

    # Create a axe object for each sub-plots
    f, arrax = plt.subplots(len(plots), sharex=True, figsize=(16, 25))
    f.suptitle(title, fontsize=20)

    for (ax, data_cols)  in zip(arrax, plots):
        curr_df = df[data_cols]

        stdev_cols = ['{}_stdev'.format(x) for x in data_cols]
        # Extract the color for each configuration
        colors = [conf_to_color[x] for x in data_cols]

        # set the color cycler for this plot
        ax.set_prop_cycle(cycler('color', colors))

        # Plot each line and its errorbars
        for (data, stdev) in  zip(data_cols, stdev_cols):
            ax.errorbar(x=df.index.values, y=df[data], yerr=df[stdev], marker='o')

        ax.set_ylim(0)
        ax.grid()
        ax.set_xlabel('Jenkins Build ID')
        ax.set_ylabel('Meantime per syscall [us]')

        ax.xaxis.set_major_locator(MaxNLocator(integer=True))

        ax.legend(prop={'family': 'monospace'},
                    labels=curr_df.columns.values, bbox_to_anchor=(1.2,1))

    plt.subplots_adjust(top=0.95)
    plt.savefig('{}.png'.format(graph_type), bbox_inches='tight')

# Writes a file that contains commit id of all configurations shown in the
# plots
def create_metadata_file(res_dir):
    list_ = []
    for dirname, dirnames, res_files in os.walk('./'+res_dir):
        if len(dirnames) > 0:
            continue
        metadata = pd.read_csv(os.path.join(dirname, 'metadata.csv'))
        list_.append(metadata)

    df = pd.concat(list_)
    df.index=df.build_id
    df.sort_index(inplace=True)
    df.to_csv('metadata.csv', index=False)

#Iterates over a result directory and creates the plots for the different
#testcases
def create_plots(res_dir):
    df = pd.DataFrame()
    metadata_df = pd.DataFrame()
    list_ = []
    for dirname, dirnames, res_files in os.walk('./'+res_dir):
        if len(dirnames) > 0:
            continue
        metadata = pd.read_csv(os.path.join(dirname, 'metadata.csv'))

        for res in res_files:
            if res in 'metadata.csv':
                continue
            tmp = pd.read_csv(os.path.join(dirname, res))
            #Use the build id as the index for the dataframe for filtering
            tmp.index = metadata.build_id
            #Add the testcase name to the row for later filtering
            tmp['testcase'] = res.split('.')[0]
            list_.append(tmp)

        df = pd.concat(list_)
        df = convert_us_to_ns(df)
        df = rename_cols(df)
        df.sort_index(inplace=True)

    #Go over the entire dataframe by testcase and create a plot for each type
    for testcase in df.testcase.unique():
        df_testcase  = df.loc[df['testcase'] == testcase]
        create_plot(df=df_testcase, graph_type=testcase)

def main():
    res_path = sys.argv[1]
    create_plots(os.path.join(res_path))
    create_metadata_file(os.path.join(res_path))

if __name__ == '__main__':
    main()