def _fit_transformer_with_progress_logging(transformer: TransformerMixin,
X,
logger: logging.Logger,
message_prefix: str,
unit: str,
message_suffx: str = ': '):
if isinstance(transformer, Pipeline):
steps = transformer.steps
if len(steps) == 1 and isinstance(steps[0][1], FeatureUnion):
feature_union = steps[0][1]
for name, union_transformer in feature_union.transformer_list:
X = logging_tqdm(
iterable=X,
logger=logger,
desc=f'{message_prefix}.{name}{message_suffx}',
unit=unit)
union_transformer.fit(X)
return
X = logging_tqdm(iterable=X,
logger=logger,
desc=message_prefix + message_suffx,
unit=unit)
transformer.fit(X)
def fit_transformer(self, transformer: TransformerMixin) -> None:
"""
Fit a transformer on this dataset.
"""
transformer.fit(self.XY)
def fit_transform_scaler(
scaler: TransformerMixin,
x: ArrayLike) -> tuple[TransformerMixin, ArrayLike]:
fitted = scaler.fit(x)
return fitted, apply_transform(fitted, x)
def fit_on(self, predictor: TransformerMixin):
"""
:return:
"""
predictor.fit(self.sample())
return predictor
