def test_NamedStandardScaler(self):
self.scaler = NamedStandardScaler()
self.assertEqual(self.scaler.name(), 'standard_scaler')
self.assertEqual(self.scaler.with_mean, True)
self.assertEqual(self.scaler.with_std, True)
self.assertEqual(NamedStandardScaler.fit, StandardScaler.fit)
self.assertEqual(NamedStandardScaler.transform,
StandardScaler.transform)
def test_GiveMeSomeCreditExperiment(self):
self.experiment = GiveMeSomeCreditExperiment(
fixed_random_seed=0xabcd,
train_data_sampler=CompleteData(),
missing_value_handler=CompleteCaseAnalysis(),
numeric_attribute_scaler=NamedStandardScaler(),
learners=[NonTunedLogisticRegression(),
NonTunedDecisionTree()],
pre_processors=[NoPreProcessing()],
post_processors=[NoPostProcessing()])
self.experiment.run()
def calculate_metrics(seed, learner, pre_processor, post_processor):
'''
Experiment function to run the experiments
'''
exp = GermanCreditDatasetSexExperiment(
fixed_random_seed=seed,
train_data_sampler=CompleteData(),
missing_value_handler=CompleteCaseAnalysis(),
numeric_attribute_scaler=NamedStandardScaler(),
learners=[learner],
pre_processors=[pre_processor],
post_processors=[post_processor])
exp.run()
def calculate_metrics(seed, learner, missing_value_imputer, pre_processor,
post_processor):
'''
Experiment function to run the experiments
'''
exp = AdultDatasetWhiteExperiment(
fixed_random_seed=seed,
train_data_sampler=CompleteData(),
missing_value_handler=missing_value_imputer,
numeric_attribute_scaler=NamedStandardScaler(),
learners=[learner],
pre_processors=[pre_processor],
post_processors=[post_processor],
optimal_validation_strategy=[])
exp.run()
def calculate_metrics(seed, learners, pre_processors, post_processors,
filter_val_strategy):
'''
Experiment function to run the experiments with multiple combinations of learners and processors in the input
'''
exp = GermanCreditDatasetSexExperiment(
fixed_random_seed=seed,
train_data_sampler=CompleteData(),
missing_value_handler=CompleteCaseAnalysis(),
numeric_attribute_scaler=NamedStandardScaler(),
learners=learners,
pre_processors=pre_processors,
post_processors=post_processors,
optimal_validation_strategy=filter_val_strategy)
exp.run()
return exp.generate_file_path()