vtreat Learner Python

This component uses python package vtreat to create a transform object and automatically prepare the input table for ML problems.

Github: https://github.com/WinVector/pyvtreat

Data preparation for Machine Learning with KNIME and the Python “vtreat” package

https://medium.com/p/efcaf58fa783

----------------------------------------
Original component by https://hub.knime.com/ajc

https://hub.knime.com/-/spaces/-/~CS0Qx06ugC-1mkV_/current-state/

Input ports

1. Type: Table
   The data to be treated
   The data to be treated
2. Type: Flow Variable
   conda.environment
   A Flow Variable containing a conda.environment with the vtreat package KNIME and Python — Setting up and managing Conda environments https://medium.com/p/2ac217792539

Output ports

1. Type: Table
   1. Transformation Rules Table
   This table contains the description matrix generated by vtreat, which outlines the transformations applied to the input data. The description matrix is crucial for understanding what vtreat has done during its preprocessing steps. It includes information about each new variable created by vtreat
2. Type: Table
   2. Score Frame Table
   The score frame is another table generated by vtreat that provides detailed diagnostics on the variables created during the treatment. For each derived variable, it includes metrics and information
3. Type: Table
   3. Prepared Data Table
   This is the transformed dataset, where all the preprocessing steps defined by vtreat and specified in the script have been applied to the input data. It's ready for use in machine learning models or further analysis.
4. Type: Python
   vtreat Python model
   The pickled Python vtreat model to be stored and re-used

External resources

• Data preparation for Machine Learning with KNIME and the Python “vtreat” package