dts: Create Train and Test Datasets

In Ehsan-F/R-Mixtape: Mixtape: Enhancing the model making process

Description

By using a Train Index, this function easily creates train and test datasets.

Usage

dts(x = "TrainIndex", df = ds, df_train = ds1, df_test = ds2)

Arguments

x	The Train Index variable indicating which observations belong to the train sample.
df	The entire data.
df_train	The desired name for the train dataset.
df_test	The desired name for the test dataset.

Examples

set.seed(123)
TrainIndex <- createDataPartition(ds$t_Paid, p = 0.7,  # create a balanced partitions with respect to the target
                                  list = FALSE,       # change p for training data size
                                  times = 1)

ds$TrainIndex <- 0
ds$TrainIndex[TrainIndex] <- 1
dts(x = "TrainIndex", df = ds, df_train = ds1, df_test = ds2)

# 'ds' is the entire data
# 'ds1' is the chosen name for the train sample
# 'ds2' is the chosen name for the test sample

Ehsan-F/R-Mixtape documentation built on June 24, 2020, 12:22 a.m.