Automatic Model Evaluation

autoEval.autoevaluatemodel.auto_evaluate_model(model, train_predictions, test_predictions, feature_columns, target, prediction_column, modelling_task_type='regression')[source]

The function helps evaluate the model and provides below reports :

  1. Model Performance Metrics

  2. Model Performance Report

  3. Model Fairness Report

  4. Model Testing Report

Inputs :

modelobject

The model instance from sklearn/spark or pipeline object.

train_predictionsdataframe

The dataframe with feature columns, target, predicted column with train data.

test_predictionsdataframe

The dataframe with feature columns, target, predicted column with test data.

feature_columnslist

The list of feature columns used for training the model.

targetstr

The name of the actual target column available in the dataframe.

prediction_columnstr

The name of the predicted column containing the model predictions.

modelling_task_typestr, optional

The name of modelling task type. The default value is “regression”. It should be one of the below -

  1. Regression

  2. Classification

  3. Forecasting

ReturnsNone

The function performs the auto evaluation of the model and dislays the result.

Function Usage

Example usage of auto_evaluate_model function

from autoEval import autoevaluatemodel

autoevaluatemodel.auto_evaluate_model(
    model,
    train_output,
    test_output,
    feature_columns,
    target_column,
    predicted_column)