search

jimthompson5802 / model-stacking-workbench

Framework for model stacking that can be applied to Kaggle competitions
MIT License
3 stars 1 forks source link

Refactor code to join train and test data frames using config ID_VARS instead of dataframe index values. #8

Closed jimthompson5802 closed 6 years ago

jimthompson5802 commented 6 years ago

For the moment will keep the current method of rebuilding the test and training data frames based on the pandas data frame index value, i.e., joining back the id, target and predictor data frames

Will use ID_VARS if need to join with data outside of the train and test data sets.

May revisit this in future.