def test_load_dataset_with_unknowns_nominals(self):
al = ArffLoader("tests/dataset_with_unknowns.arff")
expected_instances = numpy.asarray([[1.0, 0.0, 0.0, 1.2],
[0.0, 0.0, 0.0, 0.3],
[0.0, 1.0, 0.0, 3.0],
[0.0, 0.0, 1.0, 4.0]])
instances, _ = al.load_dataset()
numpy.testing.assert_array_equal(expected_instances, instances)
def test_load_dataset_with_label_not_last_attribute(self):
al = ArffLoader("tests/dataset.arff", label_attribute="first")
expected_instances = numpy.asarray([[1.2, 0.0], [0.3, 0.0], [3.0, 1.0],
[4.0, 1.0]])
expected_labels = numpy.asarray(["sheep", "cannon", "egg", "cannon"])
instances, labels = al.load_dataset()
numpy.testing.assert_array_equal(expected_instances, instances)
numpy.testing.assert_array_equal(expected_labels, labels)
def test_load_dataset_standard(self):
al = ArffLoader("tests/dataset.arff")
expected_instances = numpy.asarray([[1.0, 0.0, 0.0, 1.2],
[0.0, 0.0, 1.0, 0.3],
[0.0, 1.0, 0.0, 3.0],
[0.0, 0.0, 1.0, 4.0]])
expected_labels = numpy.asarray(["POS", "POS", "NEG", "NEG"])
instances, labels = al.load_dataset()
numpy.testing.assert_array_equal(expected_instances, instances)
numpy.testing.assert_array_equal(expected_labels, labels)
def test_load_dataset_with_unknown_feature(self):
al = ArffLoader("tests/dataset_with_unknown_feature.arff")
expected_instances = numpy.asarray([
[1.2],
[0.3],
[3.0],
[4.0]
])
instances, _ = al.load_dataset()
numpy.testing.assert_array_equal(expected_instances, instances)
def test_load_dataset_with_useless_real_values(self):
al = ArffLoader("tests/dataset_with_useless_real_values.arff")
expected_instances = numpy.asarray([
[1.2],
[0.3],
[3.0],
[4.0]
])
instances, _ = al.load_dataset()
numpy.testing.assert_array_almost_equal(expected_instances, instances)
def test_load_dataset_with_unknowns_nominals(self):
al = ArffLoader("tests/dataset_with_unknowns.arff")
expected_instances = numpy.asarray([
[1.0, 0.0, 0.0, 1.2],
[0.0, 0.0, 0.0, 0.3],
[0.0, 1.0, 0.0, 3.0],
[0.0, 0.0, 1.0, 4.0]
])
instances, _ = al.load_dataset()
numpy.testing.assert_array_equal(expected_instances, instances)
def test_load_dataset_with_label_not_last_attribute(self):
al = ArffLoader("tests/dataset.arff", label_attribute="first")
expected_instances = numpy.asarray([
[1.2, 0.0],
[0.3, 0.0],
[3.0, 1.0],
[4.0, 1.0]
])
expected_labels = numpy.asarray([
"sheep",
"cannon",
"egg",
"cannon"
])
instances, labels = al.load_dataset()
numpy.testing.assert_array_equal(expected_instances, instances)
numpy.testing.assert_array_equal(expected_labels, labels)
def test_load_dataset_standard(self):
al = ArffLoader("tests/dataset.arff")
expected_instances = numpy.asarray([
[1.0, 0.0, 0.0, 1.2],
[0.0, 0.0, 1.0, 0.3],
[0.0, 1.0, 0.0, 3.0],
[0.0, 0.0, 1.0, 4.0]
])
expected_labels = numpy.asarray([
"POS",
"POS",
"NEG",
"NEG"
])
instances, labels = al.load_dataset()
numpy.testing.assert_array_equal(expected_instances, instances)
numpy.testing.assert_array_equal(expected_labels, labels)
def main():
for dataset_name in evaluation.dataset_names():
dataset_path = "evaluation/datasets/{}.arff".format(dataset_name)
X, y = ArffLoader(dataset_path).load_dataset()
X = preprocessing.MinMaxScaler().fit_transform(X)
n, n_feats = X.shape
n_classes = len(numpy.unique(y))
precisions = pow(X.var(axis=0), -1)
mean_precision = numpy.mean(precisions)
precision_var = numpy.var(precisions)
numpy.set_printoptions(precision=3, suppress=True)
print dataset_name
print "\t size:{: 5d} ; features: {: 6d} ; classes: {}".format(n, n_feats, n_classes)
print "\tmean precision: {: 8.3f}; precision variance: {:10.3f}".format(mean_precision, precision_var)
print "\tprecisions:\n{}".format(precisions)
print "\n\n"
def main():
for dataset_name in evaluation.dataset_names():
dataset_path = "evaluation/datasets/{}.arff".format(dataset_name)
X, y = ArffLoader(dataset_path).load_dataset()
X = preprocessing.MinMaxScaler().fit_transform(X)
n, n_feats = X.shape
n_classes = len(numpy.unique(y))
precisions = pow(X.var(axis=0), -1)
mean_precision = numpy.mean(precisions)
precision_var = numpy.var(precisions)
numpy.set_printoptions(precision=3, suppress=True)
print dataset_name
print "\t size:{: 5d} ; features: {: 6d} ; classes: {}".format(
n, n_feats, n_classes)
print "\tmean precision: {: 8.3f}; precision variance: {:10.3f}".format(
mean_precision, precision_var)
print "\tprecisions:\n{}".format(precisions)
print "\n\n"
def test_load_dataset_with_useless_real_values(self):
al = ArffLoader("tests/dataset_with_useless_real_values.arff")
expected_instances = numpy.asarray([[1.2], [0.3], [3.0], [4.0]])
instances, _ = al.load_dataset()
numpy.testing.assert_array_almost_equal(expected_instances, instances)
def test_load_dataset_with_unknown_feature(self):
al = ArffLoader("tests/dataset_with_unknown_feature.arff")
expected_instances = numpy.asarray([[1.2], [0.3], [3.0], [4.0]])
instances, _ = al.load_dataset()
numpy.testing.assert_array_equal(expected_instances, instances)