StatMeasures (version 1.0)

randomise: Order the rows of a data randomly

Description

Takes in data and seed, and returns the data with randomly ordered observations

Usage

randomise(data, seed = NULL)

Arguments

data

a matrix, data.frame or data.table

seed

an integer value

Value

data of same class as input with randomly ordered observations

Details

Some of the modeling algorithms pick top p percent of the observations for training the model, which could lead to skewed predictions. This function solves that problem by randomly ordering the observations so that the response variable has more or less the same distribution even if the algorithms don't pick training observations randomly.

See Also

factorise, rmdupkey, rmdupobs

Examples

# A 'data.frame'
df <- data.frame(x = c(1, 2, 3, 4, 5), y = c('a', 'b', 'c', 'd', 'e'))

# Change the order of the observations randomly
dfRan <- randomise(data = df)
dfRan <- randomise(data = df, seed = 150)