def test_plot_experiment(self):
"""
Tests the plot_experiment method.
"""
datasets = [
self.datafile("bolts.arff"),
self.datafile("bodyfat.arff"),
self.datafile("autoPrice.arff")
]
cls = [
classifiers.Classifier("weka.classifiers.trees.REPTree"),
classifiers.Classifier(
"weka.classifiers.functions.LinearRegression"),
classifiers.Classifier("weka.classifiers.functions.SMOreg"),
]
outfile = self.tempfile("results-rs.arff")
exp = experiments.SimpleRandomSplitExperiment(classification=False,
runs=10,
percentage=66.6,
preserve_order=False,
datasets=datasets,
classifiers=cls,
result=outfile)
exp.setup()
exp.run()
# evaluate
loader = converters.loader_for_file(outfile)
data = loader.load_file(outfile)
matrix = experiments.ResultMatrix(
"weka.experiment.ResultMatrixPlainText")
tester = experiments.Tester("weka.experiment.PairedCorrectedTTester")
tester.resultmatrix = matrix
comparison_col = data.attribute_by_name(
"Correlation_coefficient").index
tester.instances = data
tester.header(comparison_col)
tester.multi_resultset_full(0, comparison_col)
# plot
plot.plot_experiment(matrix,
title="Random split (w/ StdDev)",
measure="Correlation coefficient",
show_stdev=True,
wait=False)
plot.plot_experiment(matrix,
title="Random split",
measure="Correlation coefficient",
wait=False)
def test_plot_experiment(self):
"""
Tests the plot_experiment method.
"""
datasets = [self.datafile("bolts.arff"), self.datafile("bodyfat.arff"), self.datafile("autoPrice.arff")]
cls = [
classifiers.Classifier("weka.classifiers.trees.REPTree"),
classifiers.Classifier("weka.classifiers.functions.LinearRegression"),
classifiers.Classifier("weka.classifiers.functions.SMOreg"),
]
outfile = self.tempfile("results-rs.arff")
exp = experiments.SimpleRandomSplitExperiment(
classification=False,
runs=10,
percentage=66.6,
preserve_order=False,
datasets=datasets,
classifiers=cls,
result=outfile)
exp.setup()
exp.run()
# evaluate
loader = converters.loader_for_file(outfile)
data = loader.load_file(outfile)
matrix = experiments.ResultMatrix("weka.experiment.ResultMatrixPlainText")
tester = experiments.Tester("weka.experiment.PairedCorrectedTTester")
tester.resultmatrix = matrix
comparison_col = data.attribute_by_name("Correlation_coefficient").index
tester.instances = data
tester.header(comparison_col)
tester.multi_resultset_full(0, comparison_col)
# plot
plot.plot_experiment(matrix, title="Random split (w/ StdDev)", measure="Correlation coefficient", show_stdev=True, wait=False)
plot.plot_experiment(matrix, title="Random split", measure="Correlation coefficient", wait=False)
def main():
"""
Just runs some example code.
"""
print(helper.get_data_dir())
# cross-validation + classification
helper.print_title("Experiment: Cross-validation + classification")
datasets = [
helper.get_data_dir() + os.sep + "iris.arff",
helper.get_data_dir() + os.sep + "anneal.arff"
]
classifiers = [
Classifier("weka.classifiers.rules.ZeroR"),
Classifier("weka.classifiers.trees.J48")
]
outfile = tempfile.gettempdir() + os.sep + "results-cv.arff"
exp = SimpleCrossValidationExperiment(classification=True,
runs=10,
folds=10,
datasets=datasets,
classifiers=classifiers,
result=outfile)
exp.setup()
exp.run()
# evaluate
loader = converters.loader_for_file(outfile)
data = loader.load_file(outfile)
matrix = ResultMatrix("weka.experiment.ResultMatrixPlainText")
tester = Tester("weka.experiment.PairedCorrectedTTester")
tester.resultmatrix = matrix
comparison_col = data.attribute_by_name("Percent_correct").index
tester.instances = data
print(tester.header(comparison_col))
print(tester.multi_resultset_full(0, comparison_col))
# random split + regression
helper.print_title("Experiment: Random split + regression")
datasets = [
helper.get_data_dir() + os.sep + "bolts.arff",
helper.get_data_dir() + os.sep + "bodyfat.arff"
]
classifiers = [
Classifier("weka.classifiers.rules.ZeroR"),
Classifier("weka.classifiers.functions.LinearRegression")
]
outfile = tempfile.gettempdir() + os.sep + "results-rs.arff"
exp = SimpleRandomSplitExperiment(classification=False,
runs=10,
percentage=66.6,
preserve_order=False,
datasets=datasets,
classifiers=classifiers,
result=outfile)
exp.setup()
exp.run()
# evaluate
loader = converters.loader_for_file(outfile)
data = loader.load_file(outfile)
matrix = ResultMatrix("weka.experiment.ResultMatrixPlainText")
tester = Tester("weka.experiment.PairedCorrectedTTester")
tester.resultmatrix = matrix
comparison_col = data.attribute_by_name("Correlation_coefficient").index
tester.instances = data
print(tester.header(comparison_col))
print(tester.multi_resultset_full(0, comparison_col))
# plot
plot_exp.plot_experiment(matrix,
title="Random split",
measure="Correlation coefficient",
wait=True)
def main():
"""
Run sample code.
"""
print(helper.getDataDir())
# cross-validation + classification
helper.printTitle("Experiment: Cross-validation + classification")
datasets = [helper.getDataDir() + os.sep + "iris.arff", helper.getDataDir() + os.sep + "anneal.arff"]
classifiers = [Classifier("weka.classifiers.rules.ZeroR"), Classifier("weka.classifiers.trees.J48")]
outfile = tempfile.gettempdir() + os.sep + "results-cv.arff"
exp = SimpleCrossValidationExperiment(
classification=True,
runs=10,
folds=10,
datasets=datasets,
classifiers=classifiers,
result=outfile)
exp.setup()
exp.run()
# evaluate
loader = converters.loader_for_file(outfile)
data = loader.load_file(outfile)
matrix = ResultMatrix("weka.experiment.ResultMatrixPlainText")
tester = Tester("weka.experiment.PairedCorrectedTTester")
tester.resultmatrix = matrix
comparison_col = data.attribute_by_name("Percent_correct").index
tester.instances = data
print(tester.header(comparison_col))
print(tester.multi_resultset_full(0, comparison_col))
# random split + regression
helper.printTitle("Experiment: Random split + regression")
datasets = [helper.getDataDir() + os.sep + "bolts.arff", helper.getDataDir() + os.sep + "bodyfat.arff"]
classifiers = [
Classifier("weka.classifiers.rules.ZeroR"),
Classifier("weka.classifiers.functions.LinearRegression")
]
outfile = tempfile.gettempdir() + os.sep + "results-rs.arff"
exp = SimpleRandomSplitExperiment(
classification=False,
runs=10,
percentage=66.6,
preserve_order=False,
datasets=datasets,
classifiers=classifiers,
result=outfile)
exp.setup()
exp.run()
# evaluate
loader = converters.loader_for_file(outfile)
data = loader.load_file(outfile)
matrix = ResultMatrix("weka.experiment.ResultMatrixPlainText")
tester = Tester("weka.experiment.PairedCorrectedTTester")
tester.resultmatrix = matrix
comparison_col = data.attribute_by_name("Correlation_coefficient").index
tester.instances = data
print(tester.header(comparison_col))
print(tester.multi_resultset_full(0, comparison_col))
# plot
plot_exp.plot_experiment(matrix, title="Random split", measure="Correlation coefficient", wait=True)