Regression Summary

To get a summary report of the regression model based on the performance metrics:

from dmba import regressionSummary

To evaluate regression models:

# training 
regressionSummary(train_y, reg.predict(train_X)) 
# validation 
regressionSummary(valid_y, reg.predict(valid_X))

Sample output:

Regression statistics  (training)
Mean Error (ME) : 0.0000 
Root Mean Squared Error (RMSE) : 1121.0606 
Mean Absolute Error (MAE) : 811.6770 
Mean Percentage Error (MPE) : -0.8630 
Mean Absolute Percentage Error (MAPE) : 8.0054 

Regression statistics (testing)
Mean Error (ME) : 97.1891 
Root Mean Squared Error (RMSE) : 1382.0352
Mean Absolute Error (MAE) : 880.1396 
Mean Percentage Error (MPE) : 0.0138 
Mean Absolute Percentage Error (MAPE) : 8.8744

It appears that the training dataset works better on the training dataset than the validation dataset
i.e. MAE and MAPE scores are better for training dataset (lower error scores), but still overall works well