def test_dummy_generator(self):
features = pd.DataFrame({"a": np.arange(10)})
targets = pd.DataFrame({"a": np.arange(10)})
one = DataGenerator(DummySplitter(1), features,
targets).complete_samples()
two = DataGenerator(DummySplitter(2), features,
targets).complete_samples()
sampled1 = [s for s in one.sample()]
sampled2 = [s for s in two.sample()]
self.assertEqual(1, len(sampled1))
self.assertEqual(2, len(sampled2))
def test_wrapping_cross_validation(self):
features = pd.DataFrame({"a": np.arange(10)})
labels = pd.DataFrame({
"a": np.arange(10),
"b": np.arange(10) + 10,
"c": np.arange(10) + 20,
"d": np.arange(10) + 30,
})
s = DataGenerator(DummySplitter(2), features, labels, None, None,
None).train_test_sampler()
s1 = ExtractMultiMultiModelSampler(0, 2, s)
s2 = ExtractMultiMultiModelSampler(1, 2, s)
train, test = next(s.sample())
y = train[1]
self.assertEqual(4, y.shape[1])
self.assertEqual(780, y.sum())
train, test = next(s1.sample())
y = train[1]
self.assertEqual(2, y.shape[1])
self.assertEqual(190, y.sum())
train, test = next(s2.sample())
y = train[1]
self.assertEqual(2, y.shape[1])
self.assertEqual(590, y.sum())
for s in s1.sample():
print("----")
print(s[0][0])
self.assertEqual(20, sum([len(s[0][0]) for s in s1.sample()]))
self.assertEqual(20, sum([len(s[0][0]) for s in s2.sample()]))
def predict(df: pd.DataFrame,
model: Model,
tail: int = None,
samples: int = 1,
**kwargs) -> pd.DataFrame:
min_required_samples = model.features_and_labels.min_required_samples
if tail is not None:
if min_required_samples is not None:
# just use the tail for feature engineering
df = df[-(abs(tail) + (min_required_samples - 1)):]
else:
_log.warning(
"could not determine the minimum required data from the model")
kwargs = merge_kwargs(model.features_and_labels.kwargs, model.kwargs,
kwargs)
columns, features, targets = extract(model.features_and_labels, df,
extract_features, **kwargs)
if samples > 1:
print(f"draw {samples} samples")
sampler = DataGenerator(DummySplitter(samples), features, None, targets,
None).complete_samples()
predictions = model.predict(sampler, **kwargs)
y_hat = to_pandas(predictions, index=features.index, columns=columns)
return _assemble_result_frame(targets, y_hat, None, None, None, features)
def backtest(df: pd.DataFrame, model: Model, summary_provider: Callable[[pd.DataFrame], Summary] = None, **kwargs) -> Summary:
kwargs = merge_kwargs(model.features_and_labels.kwargs, model.kwargs, kwargs)
(features, _), labels, targets, weights, gross_loss =\
extract(model.features_and_labels, df, extract_feature_labels_weights, **kwargs)
sampler = DataGenerator(DummySplitter(1), features, labels, targets, None).complete_samples()
predictions = model.predict(sampler, **kwargs)
y_hat = to_pandas(predictions, index=features.index, columns=labels.columns)
df_backtest = _assemble_result_frame(targets, y_hat, labels, gross_loss, weights, features)
return (summary_provider or model.summary_provider)(df_backtest, model, **kwargs)