TabularPreprocessing
struct defined in module
FastTabular
TabularPreprocessing <: Encoding
Encodes a
TableRow by applying the following preprocessing steps:
DataAugmentation.NormalizeRow (for normalizing a row of data for continuous columns)
DataAugmentation.FillMissing (for filling missing values)
DataAugmentation.Categorify (for label encoding categorical columns, which can be later used for indexing into embedding matrices) or a sequence of these transformations.
There are
2
methods for FastTabular.TabularPreprocessing:
The following pages link back here:
FastTabular.jl , encodings/tabularpreprocessing.jl , tasks/classification.jl , tasks/regression.jl