def create(self, validated_data):
with transaction.atomic():
# prepare ValidationRun model
new_val_run = ValidationRun(start_time=timezone.now())
new_val_run.interval_from = validated_data.get('validation_period').get('interval_from', None)
new_val_run.interval_to = validated_data.get('validation_period').get('interval_to', None)
new_val_run.anomalies = validated_data.get('anomalies').get('method')
new_val_run.anomalies_from = validated_data.get('anomalies').get('anomalies_from', None)
new_val_run.anomalies_to = validated_data.get('anomalies').get('anomalies_to', None)
new_val_run.min_lat = validated_data.get('spatial_subsetting').get('min_lat', None)
new_val_run.min_lon = validated_data.get('spatial_subsetting').get('min_lon', None)
new_val_run.max_lat = validated_data.get('spatial_subsetting').get('max_lat', None)
new_val_run.max_lon = validated_data.get('spatial_subsetting').get('max_lon', None)
new_val_run.scaling_method = validated_data.get('scaling').get('method', None)
for metric in validated_data.get('metrics'):
if metric.get('id') == 'tcol':
new_val_run.tcol = metric.get('value')
new_val_run.save()
# prepare DatasetConfiguration models
reference_config = None
dataset_config_models = []
configs_to_save = [validated_data.get('reference_config')]
print('Reference config:')
print(configs_to_save)
configs_to_save.extend(validated_data.get('dataset_configs'))
for config in configs_to_save:
config_model = DatasetConfiguration.objects.create(validation=new_val_run,
dataset_id=config.get('dataset_id'),
version_id=config.get('version_id'),
variable_id=config.get('variable_id'))
config_model.save()
filter_models = []
for filter_id in config.get('basic_filters'):
filter_models.append(DataFilter.objects.get(id=filter_id))
for filter_model in filter_models:
config_model.filters.add(filter_model)
config_model.save()
dataset_config_models.append(config_model)
new_val_run.reference_configuration = dataset_config_models[0]
new_val_run.save()
return new_val_run
def test_validation_run_clean(self):
run = ValidationRun()
## default object should be valid
run.clean()
## object with just a from date should be invalid
run.interval_from = datetime(2000, 1, 1)
with pytest.raises(ValidationError):
run.clean()
## object with just a to start date should be invalid
run.interval_from = None
run.interval_to = datetime(2000, 1, 1)
with pytest.raises(ValidationError):
run.clean()
## object with from date after to date should be invalid
run.interval_from = datetime(2001, 1, 1)
with pytest.raises(ValidationError):
run.clean()
## object with from date before to date should be valid
run.interval_to = datetime(2005, 1, 1)
run.clean()
## object with no spatial subsetting should be valid
run.min_lat = None
run.max_lat = None
run.min_lon = None
run.max_lon = None
run.clean()
## spatial subsetting with only two coords should be invalid
run.min_lat = -45.0
run.max_lat = +45.0
with pytest.raises(ValidationError):
run.clean()
## spatial subsetting with four coords should be valid
run.min_lon = -120.0
run.max_lon = +120.0
run.clean()
## climatology with moving average should be valid without time period
run.anomalies = ValidationRun.MOVING_AVG_35_D
run.clean()
## climatology without anomalies time period should be invalid
run.anomalies = ValidationRun.CLIMATOLOGY
with pytest.raises(ValidationError):
run.clean()
## climatology with broken time period should be invalid
run.anomalies_from = datetime(2005, 1, 1)
run.anomalies_to = datetime(2000, 1, 1)
with pytest.raises(ValidationError):
run.clean()
## climatology with correct time period should be invalid
run.anomalies_from = datetime(2000, 1, 1)
run.anomalies_to = datetime(2005, 1, 1)
run.clean()
## climatology with moving average should be invalid with time period
run.anomalies = ValidationRun.MOVING_AVG_35_D
with pytest.raises(ValidationError):
run.clean()