fig, axes = plt.subplots(nrows=5, ncols=1, sharex=True)
w = workload.Workload.from_csv(arguments["<swf_file1>"])
w0 = workload.Workload.from_csv(arguments["<swf_file2>"])
visu.plot_series_comparison(
{
arguments["<swf_file1>"]: w.utilisation.load,
arguments["<swf_file2>"]: w0.utilisation.load
}, axes[0], "Utilisation comparison")
visu.plot_series_comparison(
{
arguments["<swf_file1>"]: w.queue.load,
arguments["<swf_file2>"]: w0.queue.load
}, axes[1], "Queue comparison")
visu.plot_job_details(w.df, 100, axes[2], arguments["<swf_file1>"])
visu.plot_job_details(w0.df, 100, axes[3], arguments["<swf_file2>"])
visu.plot_series("waiting_time", {
arguments["<swf_file1>"]: w,
arguments["<swf_file2>"]: w0
}, axes[4])
plt.tight_layout(True)
plt.legend()
plt.show()
from evalys import workload
from docopt import docopt
#retrieving arguments
arguments = docopt(__doc__, version='1.0.0rc2')
fig, axes = plt.subplots(nrows=5, ncols=1, sharex=True)
w = workload.Workload.from_csv(arguments["<swf_file1>"])
w0 = workload.Workload.from_csv(arguments["<swf_file2>"])
visu.plot_series_comparison({arguments["<swf_file1>"]: w.utilisation,
arguments["<swf_file2>"]: w0.utilisation},
axes[0],
"Utilisation comparison")
visu.plot_series_comparison({arguments["<swf_file1>"]: w.queue,
arguments["<swf_file2>"]: w0.queue},
axes[1],
"Queue comparison")
visu.plot_job_details(w.df, 100, axes[2], arguments["<swf_file1>"])
visu.plot_job_details(w0.df, 100, axes[3], arguments["<swf_file2>"])
visu.plot_series("waiting_time", {arguments["<swf_file1>"]: w, arguments["<swf_file2>"]: w0}, axes[4])
plt.tight_layout(True)
plt.legend()
plt.show()
def main():
parser = argparse.ArgumentParser(description='visualisation tool for scheduling trace and Batsim output file.')
parser.add_argument('inputCSV', nargs='+', help='The input CSV file(s)')
parser.add_argument('--gantt', '-g',
action='store_true',
default=False, help='Generate Gantt charts')
parser.add_argument('--series', '-s',
nargs='?',
default=None,
const='all',
help='Generate timeseries on cumulative metrics. Available metrics are: {}'.format(available_series))
parser.add_argument('--output', '-o',
nargs='?',
help='The output Gantt chart file depending on the extension (PDF format is RECOMMENDED). For example: figure.pdf')
parser.add_argument('--gantt_diff', '-d',
action='store_true',
default=False,
help='Generate a gantt diff comparison between inputs (no more than 3 recommended')
args = parser.parse_args()
if NO_GRAPHICS and not args.output:
print("No available display: please provide an output using the -o,--output option")
exit(1)
# generate subplot
nb_subplot = 0
if args.gantt:
nb_subplot += len(args.inputCSV)
if args.gantt_diff:
nb_subplot += 1
if args.series:
nb_subplot += 1
if not args.gantt and not args.gantt_diff and not args.series:
print("You must select at least one option (use -h to see available options)")
exit(1)
fig, ax_list = plt.subplots(nb_subplot, sharex=True,)
#sharey=True)
# manage unique ax probleme
try:
iter(ax_list)
except:
ax_list = [ax_list]
else:
ax_list = list(ax_list)
# backup all ax
all_ax = list(ax_list)
# reserve last plot for series
if args.series:
ax_series = ax_list[-1:][0]
ax_list = ax_list[:-1]
# reserve last remaining plot for gantt diff
if args.gantt_diff:
ax_shape = ax_list[-1:][0]
ax_list = ax_list[:-1]
# generate josets from CSV inputs
jobsets = {}
index = 0
for inputCSV in sorted(args.inputCSV):
js = JobSet.from_csv(inputCSV)
file_name = os.path.basename(inputCSV)
file_name = unique_file_name(jobsets, file_name)
jobsets[file_name] = js
if args.gantt:
js.gantt(ax_list[index], file_name)
index += 1
if args.gantt_diff:
plot_gantt_general_shape(jobsets, ax_shape)
if args.series:
plot_series(args.series, jobsets, ax_series)
# set axes and resources
x_axes_min_value = min({m.df.submission_time.min()
for m in jobsets.values()})
x_axes_max_value = max({m.df.finish_time.max() for m in jobsets.values()})
y_axes_min_value = min([js.res_bounds[0] for js in jobsets.values()])
y_axes_max_value = max([js.res_bounds[1] for js in jobsets.values()])
x_size = x_axes_max_value - x_axes_min_value
y_size = y_axes_max_value - y_axes_min_value
print("x = ({},{})".format(x_axes_min_value, x_axes_max_value))
print("y = ({},{})".format(y_axes_min_value, y_axes_max_value))
print("x size = {}".format(x_size))
print("y size = {}".format(y_size))
for ax in all_ax:
ax.set_xlim((x_axes_min_value, x_axes_max_value))
#ax.set_ylim((y_axes_min_value, y_axes_max_value))
# Layout and cosmetic changes
# plt.subplots_adjust(left=0.0, right=1.0, bottom=0.0, top=1.0)
fig.set_tight_layout(True)
y_inches = max(y_size * len(all_ax) * 0.15, 8)
if args.gantt_diff:
y_inches += y_size * 0.15
fig.set_size_inches(y_inches * 1.7,
y_inches,
forward=True)
if args.output is not None:
plt.savefig(args.output)
else:
plt.show()
def main():
parser = argparse.ArgumentParser(
description=
'visualisation tool for scheduling trace and Batsim output file.')
parser.add_argument('inputCSV', nargs='+', help='The input CSV file(s)')
parser.add_argument('--gantt',
'-g',
action='store_true',
default=False,
help='Generate Gantt charts')
parser.add_argument(
'--series',
'-s',
nargs='?',
default=None,
const='all',
help=
'Generate timeseries on cumulative metrics. Available metrics are: {}'
.format(available_series))
parser.add_argument(
'--output',
'-o',
nargs='?',
help=
'The output Gantt chart file depending on the extension (PDF format is RECOMMENDED). For example: figure.pdf'
)
parser.add_argument(
'--gantt_diff',
'-d',
action='store_true',
default=False,
help=
'Generate a gantt diff comparison between inputs (no more than 3 recommended'
)
args = parser.parse_args()
if NO_GRAPHICS and not args.output:
print(
"No available display: please provide an output using the -o,--output option"
)
exit(1)
# generate subplot
nb_subplot = 0
if args.gantt:
nb_subplot += len(args.inputCSV)
if args.gantt_diff:
nb_subplot += 1
if args.series:
nb_subplot += 1
if not args.gantt and not args.gantt_diff and not args.series:
print(
"You must select at least one option (use -h to see available options)"
)
exit(1)
fig, ax_list = plt.subplots(
nb_subplot,
sharex=True,
)
#sharey=True)
# manage unique ax probleme
try:
iter(ax_list)
except:
ax_list = [ax_list]
else:
ax_list = list(ax_list)
# backup all ax
all_ax = list(ax_list)
# reserve last plot for series
if args.series:
ax_series = ax_list[-1:][0]
ax_list = ax_list[:-1]
# reserve last remaining plot for gantt diff
if args.gantt_diff:
ax_shape = ax_list[-1:][0]
ax_list = ax_list[:-1]
# generate josets from CSV inputs
jobsets = {}
index = 0
for inputCSV in sorted(args.inputCSV):
js = JobSet.from_csv(inputCSV)
file_name = os.path.basename(inputCSV)
file_name = unique_file_name(jobsets, file_name)
jobsets[file_name] = js
if args.gantt:
js.gantt(ax_list[index], file_name)
index += 1
if args.gantt_diff:
plot_gantt_general_shape(jobsets, ax_shape)
if args.series:
plot_series(args.series, jobsets, ax_series)
# set axes and resources
x_axes_min_value = min(
{m.df.submission_time.min()
for m in jobsets.values()})
x_axes_max_value = max({m.df.finish_time.max() for m in jobsets.values()})
y_axes_min_value = min([js.res_bounds[0] for js in jobsets.values()])
y_axes_max_value = max([js.res_bounds[1] for js in jobsets.values()])
x_size = x_axes_max_value - x_axes_min_value
y_size = y_axes_max_value - y_axes_min_value
print("x = ({},{})".format(x_axes_min_value, x_axes_max_value))
print("y = ({},{})".format(y_axes_min_value, y_axes_max_value))
print("x size = {}".format(x_size))
print("y size = {}".format(y_size))
for ax in all_ax:
ax.set_xlim((x_axes_min_value, x_axes_max_value))
#ax.set_ylim((y_axes_min_value, y_axes_max_value))
# Layout and cosmetic changes
# plt.subplots_adjust(left=0.0, right=1.0, bottom=0.0, top=1.0)
fig.set_tight_layout(True)
y_inches = max(y_size * len(all_ax) * 0.15, 8)
if args.gantt_diff:
y_inches += y_size * 0.15
fig.set_size_inches(y_inches * 1.7, y_inches, forward=True)
if args.output is not None:
plt.savefig(args.output)
else:
plt.show()
