Transform_Data: Automated data transformation and variable selection

Description Usage Arguments Value

Description

The function reads the processed data from Prepare_Data, creates transformed versions of selected variables, selects the transformed version with the ideal correlation to the target and writes out the final train and test splits of the data which can be directly fed to the Bayesian network creation module or any other models.

Usage

1
Transform_Data(ls_dt_sub, vec_time, dt_scp, wd_path, prm = F)

Arguments

ls_dt_sub

The "Data" object returned by Prepare_Data

vec_time

The "Period" object returned by Prepare_Data

dt_scp

The "Scope" object returned by Prepare_Data

wd_path

The "Working Directory" object returned by Prepare_Data

prm

Logical. If set to TRUE, the functions filters the data for an additional category. For the US market, use prm = TRUE to filter for both Retailer and Promo Family. For the India market, can be used to filter for both Brand and Cluster in case consolidated data for multiple brands is provided in a single file.

Value

All outputs are generated within specific folders for the different Markets/Retailers and Promos


supratim1121992/BBNAutoPrep documentation built on Dec. 1, 2019, 4:59 p.m.