def test_plot_logreg(self):
from matplotlib import pyplot as plt
temp = get_temp_folder(__file__, "temp_plot_logreg")
img = os.path.join(temp, "plot_logreg.png")
data = os.path.join(temp, "..", "data",
"onnxruntime_LogisticRegression.perf.csv")
df = pandas.read_csv(data)
ax = plot_bench_results(df, row_cols='N', col_cols='method',
x_value='dim', hue_cols='fit_intercept',
title="unittest")
fig = ax[0, 0].get_figure()
fig.savefig(img)
if __name__ == "__main__":
plt.show()
plt.close('all')
self.assertExists(img)
def test_plot_polypf(self):
from matplotlib import pyplot as plt
temp = get_temp_folder(__file__, "temp_plot_polypf")
img = os.path.join(temp, "plot_polypf.png")
data = os.path.join(temp, "..", "data",
"plot_bench_polynomial_features_partial_fit.perf.csv")
df = pandas.read_csv(data)
ax = plot_bench_results(df, row_cols='N', col_cols=None,
x_value='dim', hue_cols=None,
cmp_col_values='test',
title="unittest")
fig = ax[0].get_figure()
fig.savefig(img)
if __name__ == "__main__":
plt.show()
plt.close('all')
self.assertExists(img)
print(dfi)
else:
df = pandas.read_csv("%s.perf.csv" % filename)
print(df.head())
#############################
# Plot the results
# ++++++++++++++++
def label_fct(la):
la = la.replace("-lib=", "")
la = la.replace("rt=", "-")
return la
plot_bench_results(
df,
row_cols=('rt', ),
col_cols=('dataset', ),
label_fct=label_fct,
x_value='N',
hue_cols=('lib', ),
cmp_col_values='lib',
title="Numerical datasets\nBenchmark scikit-learn, xgboost, lightgbm")
plt.savefig("%s.curve.png" % filename)
import sys
if "--quiet" not in sys.argv:
plt.show()
'onnxruntime', 'onnx', 'mlprodict']
dfi = pandas.DataFrame(machine_information(pkgs))
dfi.to_csv("%s.time.csv" % filename, index=False)
print(dfi)
#############################
# Plot the results
# ++++++++++++++++
def label_fct(la):
la = la.replace("onxpython_compiled", "opy")
la = la.replace("onxpython", "opy")
la = la.replace("onxonnxruntime1", "ort")
la = la.replace("fit_intercept", "fi")
la = la.replace("True", "1")
la = la.replace("False", "0")
la = la.replace("max_depth", "mxd")
return la
plot_bench_results(df, row_cols=['N', 'max_depth', 'onnx_options'], col_cols='method',
x_value='dim', hue_cols=['n_estimators'],
title="%s\nBenchmark scikit-learn / onnxruntime" % model_name,
label_fct=label_fct)
plt.savefig("%s.png" % filename)
import sys
if "--quiet" not in sys.argv:
plt.show()
if has_dask:
cluster.close()
#########################
# Extract information about the machine used
# ++++++++++++++++++++++++++++++++++++++++++
from pymlbenchmark.context import machine_information
pkgs = ['numpy', 'pandas', 'sklearn']
dfi = pandas.DataFrame(machine_information(pkgs))
dfi.to_csv("%s.time.csv" % filename, index=False)
print(dfi)
#############################
# Plot the results
# ++++++++++++++++
from pymlbenchmark.plotting import plot_bench_results
print(df.columns)
plot_bench_results(df,
row_cols=['N'],
col_cols=['n_jobs'],
x_value='dim',
hue_cols=['test'],
cmp_col_values='test',
title="GridSearchCV\nBenchmark caching strategies")
import sys
if "--quiet" not in sys.argv:
plt.show()
def label_fct(la):
la = la.replace("onxpython_compiled", "opy")
la = la.replace("onxpython", "opy")
la = la.replace("onxonnxruntime1", "ort")
la = la.replace("fit_intercept", "fi")
la = la.replace("True", "1")
la = la.replace("False", "0")
la = la.replace("max_depth", "mxd")
return la
from pymlbenchmark.plotting import plot_bench_results
plot_bench_results(df,
row_cols='N',
col_cols='dim',
x_value='nbnode',
title="%s\nBenchmark scikit-learn / onnxruntime" %
"Cascade Add",
label_fct=label_fct)
plt.savefig("%s.node.png" % filename)
##################################
# Plot one ONNX graph
# +++++++++++++++++++
for nbnode in (2, 4):
onx = generate_onnx_graph(5, nbnode)[0]
from onnx.tools.net_drawer import GetPydotGraph, GetOpNodeProducer
pydot_graph = GetPydotGraph(onx.graph,
df.to_csv("bench_re2.csv", index=False)
print(df.head())
#########################
# Extract information about the machine used
# ++++++++++++++++++++++++++++++++++++++++++
from pymlbenchmark.context import machine_information
pkgs = ['numpy', 'pandas', 'sklearn']
dfi = pandas.DataFrame(machine_information(pkgs))
print(dfi)
#############################
# Plot the results
# ++++++++++++++++
from pymlbenchmark.plotting import plot_bench_results
print(df.columns)
plot_bench_results(df,
row_cols=['N'],
col_cols=None,
x_value='dim',
hue_cols=None,
cmp_col_values='test',
title="re2\nBenchmark re / re2")
import matplotlib.pyplot as plt
import sys
if "--quiet" not in sys.argv:
plt.show()
la = la.replace("method=proba", "prob")
la = la.replace(
"onnx_options={<class 'sklearn.gaussian_process._gpr.GaussianProcessRegressor'>: {'optim': 'cdist'}}",
'cdist')
la = la.replace("onnx_options=nan", '-')
return la
name = "bench_plot_onnxruntime_gpr.perf.csv"
df = pandas.read_csv(name)
plot_bench_results(df,
row_cols=('alpha', 'N'),
col_cols='onnx_options',
hue_cols='dtype',
cmp_col_values=('lib', 'skl'),
x_value='dim',
y_value='mean',
title=None,
label_fct=label_fct,
ax=None,
box_side=4)
plt.suptitle(
"Acceleration onnxruntime / scikit-learn for GaussianProcessRegressor")
plt.savefig("%s.png" % filename)
import sys
if "--quiet" not in sys.argv:
plt.show()
plot_bench_xtime(
df[df.norm],
col_cols='dataset',
hue_cols='model',
title=
"Numerical datasets - norm=False\nBenchmark scikit-learn / onnxruntime")
plt.savefig("%s.normT.time.png" % filename)
# plt.show()
plot_bench_xtime(
df[~df.norm],
col_cols='dataset',
hue_cols='model',
title=
"Numerical datasets - norm=False\nBenchmark scikit-learn / onnxruntime")
plt.savefig("%s.normF.time.png" % filename)
# plt.show()
plot_bench_results(
df,
row_cols='model',
col_cols=('dataset', 'norm'),
x_value='N',
title="Numerical datasets\nBenchmark scikit-learn / onnxruntime")
plt.savefig("%s.curve.png" % filename)
import sys
if "--quiet" not in sys.argv:
plt.show()
'onnxruntime', 'onnx', 'mlprodict']
dfi = pandas.DataFrame(machine_information(pkgs))
dfi.to_csv("%s.time.csv" % filename, index=False)
print(dfi)
#############################
# Plot the results
# ++++++++++++++++
def label_fct(la):
la = la.replace("onxpython_compiled", "opy")
la = la.replace("onxpython", "opy")
la = la.replace("onxonnxruntime1", "ort")
la = la.replace("fit_intercept", "fi")
la = la.replace("True", "1")
la = la.replace("False", "0")
la = la.replace("max_depth", "mxd")
return la
plot_bench_results(df, row_cols=['N', 'onnx_options'], col_cols='method',
x_value='dim', hue_cols='fit_intercept',
title="%s\nBenchmark scikit-learn / onnxruntime" % model_name,
label_fct=label_fct)
plt.savefig("%s.png" % filename)
import sys
if "--quiet" not in sys.argv:
plt.show()
##############################
# Run the benchmark
# +++++++++++++++++
df = run_bench(verbose=True)
df.to_csv("plot_bench_polynomial_features_partial_fit.perf.csv", index=False)
print(df.head())
#########################
# Extract information about the machine used
# ++++++++++++++++++++++++++++++++++++++++++
pkgs = ['numpy', 'pandas', 'sklearn']
dfi = pandas.DataFrame(machine_information(pkgs))
dfi.to_csv("plot_bench_polynomial_features_partial_fit.time.csv", index=False)
print(dfi)
#############################
# Plot the results
# ++++++++++++++++
print(df.columns)
plot_bench_results(df, row_cols='N', col_cols=None,
x_value='dim', hue_cols=None,
cmp_col_values='test',
title="PolynomialFeatures + partial_fit\n"
"Benchmark scikit-learn PR13290")
plt.show()
#############################
# Plot the results
# ++++++++++++++++
def label_fct(la):
la = la.replace("onxpython_compiled", "opy")
la = la.replace("onxpython", "opy")
la = la.replace("onxonnxruntime1", "ort")
la = la.replace("fit_intercept", "fi")
la = la.replace("True", "1")
la = la.replace("False", "0")
la = la.replace("max_depth", "mxd")
return la
plot_bench_results(df,
row_cols=['n_neighbors', 'metric', 'onnx_options'],
col_cols=['leaf_size', 'method'],
x_value='N',
hue_cols=['dim'],
title="%s\nBenchmark scikit-learn / onnxruntime" %
model_name,
label_fct=label_fct)
plt.savefig("%s.png" % filename)
import sys
if "--quiet" not in sys.argv:
plt.show()
dfi.to_csv("%s.time.csv" % filename, index=False)
print(dfi)
#############################
# Plot the results by number of nodes
# +++++++++++++++++++++++++++++++++++
def label_fct(la):
la = la.replace("onxpython_compiled", "opy")
la = la.replace("onxpython", "opy")
la = la.replace("onxonnxruntime1", "ort")
la = la.replace("edims=(", "edims=(N, ")
return la
from pymlbenchmark.plotting import plot_bench_results
plot_bench_results(df,
row_cols='edims',
col_cols='axes',
x_value='N',
cmp_col_values=('lib', 'npy'),
title="Benchmark ReduceSum",
label_fct=label_fct)
plt.savefig("%s.node.png" % filename)
import sys
if "--quiet" not in sys.argv:
plt.show()
# Plot the results
# ++++++++++++++++
def label_fct(la):
la = la.replace("onxpython_compiled", "opy")
la = la.replace("onxpython", "opy")
la = la.replace("onxonnxruntime1", "ort")
la = la.replace("fit_intercept", "fi")
la = la.replace("True", "1")
la = la.replace("False", "0")
la = la.replace("max_depth", "mxd")
return la
plot_bench_results(df,
row_cols=['N', 'hidden_layer_sizes'],
col_cols='method',
hue_cols='activation',
cmp_col_values=('lib', 'skl'),
x_value='dim',
y_value='mean',
title="%s\nBenchmark scikit-learn / onnxruntime" %
model_name,
label_fct=label_fct)
plt.savefig("%s.png" % filename)
import sys
if "--quiet" not in sys.argv:
plt.show()
'onnxruntime', 'onnx', 'mlprodict']
dfi = pandas.DataFrame(machine_information(pkgs))
dfi.to_csv("%s.time.csv" % filename, index=False)
print(dfi)
#############################
# Plot the results
# ++++++++++++++++
def label_fct(la):
la = la.replace("onxpython_compiled", "opy")
la = la.replace("onxpython", "opy")
la = la.replace("onxonnxruntime1", "ort")
la = la.replace("fit_intercept", "fi")
la = la.replace("True", "1")
la = la.replace("False", "0")
la = la.replace("max_depth", "mxd")
return la
plot_bench_results(df, row_cols='N', col_cols='max_depth',
x_value='dim',
title="%s\nBenchmark scikit-learn / onnxruntime" % model_name,
label_fct=label_fct)
plt.savefig("%s.png" % filename)
import sys
if "--quiet" not in sys.argv:
plt.show()
# ++++++++++++++++
def label_fct(la):
la = la.replace("onxpython_compiled", "opy")
la = la.replace("onxpython", "opy")
la = la.replace("onxonnxruntime1", "ort")
la = la.replace("fit_intercept", "fi")
la = la.replace("True", "1")
la = la.replace("False", "0")
return la
def color_fct(la, col):
if "onxpython_compiled" in la:
return "red"
if "onxpython" in la:
return "red"
return col
plot_bench_results(
df,
row_cols=['N', 'fit_intercept'],
col_cols='method',
x_value='dim',
title="LogisticRegression\nBenchmark scikit-learn / onnxruntime",
label_fct=label_fct,
color_fct=color_fct)
plt.show()
