nmar: NMAR: Not missing at random

In JerryTucay/mfdata: Creates missing and faulty data.

Description

nmar() alows the user to forceibly input missing values (NA) that replicates being not missing at random. This kind of missing depends strictly on the variable's value. If Xi > p% of max value for Xi then there is q% chance the data point is missing.

Usage

nmar(data, p, q, column)

Arguments

data	The data frame that you want to put missing values into.
p	Percent of the the max value in the specified column
q	Percent chance the Xi data point is missing given Xi> p% of max
column = NULL	When NULL the function will run through your whole data set Set a value to target specific columns

Details

Definintly play around with p and q if you want to get a specfic amount of data missing.

Examples

## Inputs missing data in all the columns in df

df<- data.frame(x=rnorm(100, 10, 2), y=rpois(100,4), z=rbinom(100, 1, .4))
df_missing<- nmar(df, .25)
sum(is.na(df_missing))/300

## Inputs missing data for columns one and three of df2

df2<- data.frame(x=rnorm(100, 10, 2), y=rpois(100,4), z=rbinom(100, 1, .4))
df_missing2<- nmar(df2, p=.5, q=.2 , c(1,3))
sum(is.na(df_missing2))/200


{ ~kwd1 }
{ ~kwd2 }

JerryTucay/mfdata documentation built on May 7, 2019, 6:56 p.m.