def test_should_equal_n_features_length_of_feature_names(self):
self.assertEqual(len(gsd.generate_synthetic_dataset(n_features=2,feature_names=['A','B'], feature_weights=[0.5,0.5]).columns),3)
def test_should_throw_error_if_n_samples_none(self):
with self.assertRaises(ValueError):
gsd.generate_synthetic_dataset(n_samples=None)
def test_should_return_dataframe(self):
self.assertTrue(isinstance(gsd.generate_synthetic_dataset(), pd.DataFrame))
def test_should_equal_n_samples(self):
self.assertEqual(len(gsd.generate_synthetic_dataset(n_samples=10)),10)
def test_should_throw_error_if_n_features_does_not_equal_length_of_feature_weights(self):
with self.assertRaises(ValueError):
gsd.generate_synthetic_dataset(n_features=3, feature_names=['A','B','C'], feature_weights=[0.0,0.0])
def test_should_throw_error_if_feature_names_is_None(self):
with self.assertRaises(ValueError):
gsd.generate_synthetic_dataset(feature_names=None)
def test_should_throw_error_if_class_weights_dont_add_up_to_1(self):
with self.assertRaises(ValueError):
gsd.generate_synthetic_dataset(class_weights=[0,0])
def test_should_throw_error_if_n_features_is_less_than_1(self):
with self.assertRaises(ValueError):
gsd.generate_synthetic_dataset(n_features=0)
import generate_synthetic_dataset as gsd
import json
import sys
with open(f'{sys.argv[1]}') as f:
json_from_file = json.load(f)
data_model = gsd.SyntheticDataModel(**json_from_file)
dataset = gsd.generate_synthetic_dataset(
n_samples=data_model.n_samples,
n_classes=data_model.n_classes,
class_weights=data_model.class_weights,
class_names=data_model.class_names,
y_column_name=data_model.y_column_name,
n_features=data_model.n_features,
feature_names=data_model.feature_names,
feature_types=data_model.feature_types,
feature_number_range=data_model.feature_number_range,
feature_positive_class_ratio=data_model.feature_positive_class_ratio,
feature_negative_class_ratio=data_model.feature_negative_class_ratio)
dataset.to_csv(f"{data_model.dataset_name}.csv", index=False)