def __init__(self,
explainer: SmartExplainer,
project_info_file: str,
x_train: Optional[pd.DataFrame] = None,
y_train: Optional[pd.DataFrame] = None,
y_test: Optional[pd.DataFrame] = None,
config: Optional[dict] = None):
self.explainer = explainer
self.metadata = load_yml(path=project_info_file)
self.x_train_init = x_train
if x_train is not None:
self.x_train_pre = inverse_transform(x_train,
self.explainer.preprocessing)
if self.explainer.postprocessing:
self.x_train_pre = apply_postprocessing(
self.x_train_pre, self.explainer.postprocessing)
else:
self.x_train_pre = None
self.x_pred = self.explainer.x_pred
self.config = config if config is not None else dict()
self.col_names = list(self.explainer.columns_dict.values())
self.df_train_test = self._create_train_test_df(test=self.x_pred,
train=self.x_train_pre)
self.y_pred = self.explainer.model.predict(self.explainer.x_init)
self.y_test, target_name_test = self._get_values_and_name(
y_test, 'target')
self.y_train, target_name_train = self._get_values_and_name(
y_train, 'target')
self.target_name = target_name_train or target_name_test
if 'title_story' in self.config.keys():
self.title_story = config['title_story']
elif self.explainer.title_story != '':
self.title_story = self.explainer.title_story
else:
self.title_story = 'Shapash report'
self.title_description = self.config[
'title_description'] if 'title_description' in self.config.keys(
) else ''
print_css_style()
print_javascript_misc()
if 'metrics' in self.config.keys():
if not isinstance(self.config['metrics'], list) or not isinstance(
self.config['metrics'][0], dict):
raise ValueError(
"The metrics parameter expects a list of dict.")
for metric in self.config['metrics']:
for key in metric:
if key not in ['path', 'name', 'use_proba_values']:
raise ValueError(
f"Unknown key : {key}. Key should be in ['path', 'name', 'use_proba_values']"
)
if key == 'use_proba_values' and not isinstance(
metric['use_proba_values'], bool):
raise ValueError(
'"use_proba_values" metric key expects a boolean value.'
)
def apply_postprocessing(self, postprocessing=None):
"""
Modifies x_pred Dataframe according to postprocessing modifications, if exists.
Parameters
----------
postprocessing: Dict
Dictionnary of postprocessing modifications to apply in x_pred.
Returns
-------
pandas.Dataframe
Returns x_pred if postprocessing is empty, modified dataframe otherwise.
"""
if postprocessing:
return apply_postprocessing(self.x_pred, postprocessing)
else:
return self.x_pred