Yeo-Johnson

Yeo–Johnson Transformation

Applies the Yeo–Johnson power transform to a numeric column to make its distribution more Gaussian and stabilize variance.

Yeo–Johnson is an extension of Box–Cox: it behaves similarly on positive data but, unlike Box–Cox, it supports zero and negative values, so you don’t need to shift the data.

Python configuration:

pip install scipy

Input ports

1. Type: Table
   Input
   Table with the column to transform.

Output ports

1. Type: Flow Variable
   Lambda
   Lambda of the transformation.
2. Type: Table
   Output
   Table with new transformed column.

External resources

• scipy.stats. yeojohnson