dfBinningFun: Generate WOE Configuration List Automatically

In xxzcool/scoremodel: Credit Scoring Modeling Toolbox

Description

dfBinningFun is a wrapper and parallelization of smbinning2, so it can generate the woe configuration list by using optimal binning automatically.

Usage

dfBinningFun(mydf, binMethod = 1, p = 0.05, maxcat = 10,
  aliquots = 5)

Arguments

mydf	A dataframe only with Xs and Y variables, the last column must be Y. All character x variables must be converted to factors in advance.
binMethod	An integer from 1 to 4, indicates 4 different binning methods(see details smbinning2), default 1.
p	A numeric, means percentage of records per bin, from 0 to 0.5, default 0.05.
maxcat	An integer, specifies the maximum number of categories, default 10.
aliquots	An integer, specifies the number of bins for equal-frequency or equal-interval binning method, default 5.

Details

woe configuration list is a list object lengths as number of x variables in mydf, whose component is a dataframe consisting of two columns, named as 'vals' and 'woe'.

Value

A list corresponding to X variables in mydf

See Also

Other dataset binning and woe-encoding functions: convertCutPoints, executeBinFun_df, genConfigList, preBinningFun, smbinning2, woeEncodeFun_df

xxzcool/scoremodel documentation built on May 4, 2019, 10:56 a.m.