def test_experiment_settings(self, project, ctx, monkeypatch):
config = ctx.config
config.set('config', 'target', 'species')
config.set('config', 'experiment/cross_validation_folds', 55)
config.set('config', 'experiment/max_total_time', 606)
config.set('config', 'experiment/max_eval_time', 55)
config.set('config', 'experiment/max_n_trials', 101)
config.set('config', 'experiment/use_ensemble', False)
PAYLOAD = {
'get_experiment': EXPERIMENT,
'get_project_file': PROJECT_FILE
}
interceptor(PAYLOAD, monkeypatch)
config, model_type = AugerExperimentApi(
ctx, 'project-api', 'iris-1.csv-experiment', '1234').\
get_experiment_settings()
assert config['evaluation_options']['crossValidationFolds'] == 55
assert config['evaluation_options']['max_total_time_mins'] == 606
assert config['evaluation_options']['max_eval_time_mins'] == 55
assert config['evaluation_options']['max_n_trials'] == 101
assert config['evaluation_options']['use_ensemble'] == False
# dataset
assert config['evaluation_options']['targetFeature'] == 'species'
assert config['evaluation_options']['featureColumns'] == \
['sepal_length', 'sepal_width', 'petal_length', 'petal_width']
assert config['evaluation_options']['categoricalFeatures'] == \
['species']
assert config['evaluation_options']['timeSeriesFeatures'] == []
assert config['evaluation_options']['binaryClassification'] == False
assert config['evaluation_options']['labelEncodingFeatures'] == []
assert config['evaluation_options']['classification'] == True
assert config['evaluation_options']['scoring'] == 'f1_macro'
def train(self):
# verify avalability of auger credentials
self.credentials.verify()
self.start_project()
data_set_name = self.ctx.config['auger'].get('dataset')
if data_set_name is None:
raise AugerException(
'Plese specify DataSet name in auger.yaml/dataset')
experiment_api = AugerExperimentApi(self.ctx, self.project_api)
experiment_api.create(data_set_name)
self.ctx.log('Created Experiment %s ' % experiment_api.object_name)
experiment_session_id = experiment_api.run()
self.ctx.log('Started Experiment %s training.' %
experiment_api.object_name)
AugerConfig(self.ctx).set_experiment(experiment_api.object_name,
experiment_session_id)
def test_exclude_setting(self, project, ctx, monkeypatch):
config = ctx.get_config('config')
config.exclude = ['sepal_length']
PAYLOAD = {
'get_experiment': EXPERIMENT,
'get_project_file': PROJECT_FILE
}
interceptor(PAYLOAD, monkeypatch)
config, model_type = AugerExperimentApi(
ctx, 'project-api', 'iris-1.csv-experiment', '1234').\
get_experiment_settings()
assert config['evaluation_options']['targetFeature'] == 'species'
assert config['evaluation_options']['featureColumns'] == \
['sepal_width', 'petal_length', 'petal_width']
assert config['evaluation_options']['categoricalFeatures'] == \
['species']
def test_model_type_setting(self, project, ctx, monkeypatch):
ctx.config.set('config', 'target', 'species')
ctx.config.set('config', 'model_type', 'regression')
ctx.config.set('auger', 'experiment/metric', None)
PAYLOAD = {
'get_experiment': EXPERIMENT,
'get_project_file': PROJECT_FILE
}
interceptor(PAYLOAD, monkeypatch)
config, model_type = AugerExperimentApi(
ctx, 'project-api', 'iris-1.csv-experiment', '1234').\
get_experiment_settings()
assert config['evaluation_options']['timeSeriesFeatures'] == []
assert config['evaluation_options']['binaryClassification'] == False
assert config['evaluation_options']['labelEncodingFeatures'] == []
assert config['evaluation_options']['classification'] == False
assert config['evaluation_options']['scoring'] == 'r2'