def expe_printer(res_file, comparison_metric):
loader = weka.core.converters.loader_for_file(res_file)
data = loader.load_file(res_file)
matrix = ResultMatrix(classname="weka.experiment.ResultMatrixPlainText")
tester = Tester(classname="weka.experiment.PairedCorrectedTTester")
tester.resultmatrix = matrix
comparison_col = data.attribute_by_name(comparison_metric).index
tester.instances = data
print(tester.header(comparison_col))
print(tester.multi_resultset_full(0, comparison_col))
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)
classifiers = [Classifier(classname="weka.classifiers.trees.J48")]
outfile = tempfile.gettempdir() + os.sep + "results-cv.arff"
exp = SimpleCrossValidationExperiment(datasets=datasets,
classifiers=classifiers,
folds=10,
runs=10,
result=outfile)
exp.setup()
exp.run()
# evaluate experiment
loader = converters.loader_for_file(outfile)
data = loader.load_file(outfile)
matrix = ResultMatrix(classname="weka.experiment.ResultMatrixPlainText",
options=[
"-print-row-names", "-print-col-names",
"-enum-col-names", "-show-stddev"
])
tester = Tester(classname="weka.experiment.PairedCorrectedTTester",
options=["-V", "-S", "0.05"])
comparison_col = data.attribute_by_name("Percent_correct").index
tester.instances = data
tester.resultmatrix = matrix
print(tester.header(comparison_col))
print(tester.multi_resultset_full(0, comparison_col))
# setup random-split experiment and run it
outfile = tempfile.gettempdir() + os.sep + "results-rs.csv"
exp = SimpleRandomSplitExperiment(datasets=datasets,
classifiers=classifiers,
percentage=90,
Classifier(classname="weka.classifiers.rules.OneR"),
Classifier(classname="weka.classifiers.trees.J48"),
Classifier(classname="weka.classifiers.bayes.NaiveBayes"),
Classifier(classname="weka.classifiers.lazy.IBk"),
Classifier(classname="weka.classifiers.functions.SMO"),
Classifier(classname="weka.classifiers.meta.AdaBoostM1"),
# handles only 2-class problems: Classifier(classname="weka.classifiers.functions.VotedPerceptron")
]
outfile = tempfile.gettempdir(
) + os.sep + "results-cv.arff" # store results for later analysis
exp = SimpleCrossValidationExperiment(classification=True,
runs=10,
folds=10,
datasets=datasets,
classifiers=classifiers,
result=outfile)
exp.setup()
exp.run()
# evaluate previous run
loader = converters.loader_for_file(outfile)
data = loader.load_file(outfile)
matrix = ResultMatrix(classname="weka.experiment.ResultMatrixPlainText")
tester = Tester(classname="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))
jvm.stop()