validmind.lending_club

compute_scores

defcompute_scores(probabilities:np.ndarray) → np.ndarray:

feature_engineering

deffeature_engineering(df:pd.DataFrame,verbose:bool=True) → pd.DataFrame:

get_demo_test_config

defget_demo_test_config(x_test:Optional[np.ndarray]=None,y_test:Optional[np.ndarray]=None) → Dict[str, Any]:

Get demo test configuration.

Arguments

  • x_test: Test features DataFrame
  • y_test: Test target Series

Returns

  • Test configuration dictionary

init_vm_objects

definit_vm_objects(scorecard):

load_data

defload_data(source:str='online',verbose:bool=True) → pd.DataFrame:

Load data from either an online source or offline files, automatically dropping specified columns for offline data.

Arguments

  • source: 'online' for online data, 'offline' for offline files. Defaults to 'online'.

Returns

  • DataFrame containing the loaded data.

load_scorecard

defload_scorecard():

load_test_config

defload_test_config(scorecard):

preprocess

defpreprocess(df:pd.DataFrame,verbose:bool=True) → pd.DataFrame:

split

defsplit(df:pd.DataFrame,validation_split:Optional[float]=None,test_size:float=0.2,add_constant:bool=False,verbose:bool=True) → Tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]:

Split dataset into train, validation (optional), and test sets.

Arguments

  • df: Input DataFrame
  • validation_split: If None, returns train/test split. If float, returns train/val/test split
  • test_size: Proportion of data for test set (default: 0.2)
  • add_constant: Whether to add constant column for statsmodels (default: False)

Returns

  • If validation_size is None: train_df, test_df If validation_size is float: train_df, validation_df, test_df

woe_encoding

defwoe_encoding(df:pd.DataFrame,verbose:bool=True) → pd.DataFrame: