sampling: Random sampling of data into train and test

In scorecardModelUtils: Credit Scorecard Modelling Utils

Description

The function does random sampling of the data and split it into train and test datasets. Training base percentage and seed value(optional) is taken as arguments. If seed value is not specified, random seed will be generated on different iterations.

Usage

sampling(base, train_perc = 0.7, seed = NA, replace = F)

Arguments

base	input dataframe
train_perc	(optional) percentage of total base to be kept as training sample, to be provided as decimal/fraction (default percentage is 0.7)
seed	(optional) seed value (if not given random seed is generated)
replace	(optional) whether replacement will e with or without replacement (default is FALSE ie. without replacement)

Value

An object of class "sampling" is a list containing the following components:

train_sample	training sample as a dataframe
test_sample	test sample as a dataframe
seed	seed used

Author(s)

Arya Poddar <[e-mail]>

Examples

data <- iris
sampling_list <- sampling(base = data,train_perc = 0.7,seed = 1234)
sampling_list$train
sampling_list$test
sampling_list$seed

scorecardModelUtils documentation built on May 2, 2019, 9:59 a.m.