Skip to content

Latest commit

 

History

History

preprocessing

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 
 
 
 
 

Polars friendly pre-processing models.

A collection of pre-processing (outlier detection and normalization) models that can be fit via polars expressions, either directly on a dataframe or in a groupby context. All only work with univariate data at present.

StandardScaler

Computes the mean and standard deviation of the data. Upon predict, subtracts the mean and divides by the standard deviation.

StddevCutoff

Removes all values that occur more than a specified threshold of standard deviations away from the mean.