def test_rank(metric):
real_data, synthetic_data, metadata = load_timeseries_demo()
real_score = metric.compute(real_data, real_data, metadata)
synthetic_score = metric.compute(real_data, synthetic_data, metadata)
assert metric.min_value <= synthetic_score <= real_score <= metric.max_value
def test_rank(metric):
real_data, synthetic_data, metadata = load_timeseries_demo()
real_score = metric.compute(real_data, real_data, metadata)
synthetic_score = metric.compute(real_data, synthetic_data, metadata)
normalized_real_score = metric.normalize(real_score)
normalized_synthetic_score = metric.normalize(synthetic_score)
assert metric.min_value <= synthetic_score <= real_score <= metric.max_value
assert 0.0 <= normalized_synthetic_score <= normalized_real_score <= 1.0
def test_compute_all():
real_data, synthetic_data, metadata = load_timeseries_demo()
output = compute_metrics(
TimeSeriesMetric.get_subclasses(),
real_data,
synthetic_data,
metadata=metadata
)
assert not pd.isnull(output.score.mean())
scores = output[output.score.notnull()]
assert scores.score.between(scores.min_value, scores.max_value).all()