impute_missing: impute_missing

In Yuliaxis/LinkHD: LinkHD: a versatile framework to explore and integrate heterogeneous data

Description

This function imputes missing abundances by sampling the known values from each variable. The values are sampled according to the distribution, i.e. the most frequent values have a better chance of being chosen.

Usage

impute_missing(data)

Arguments

data	Data frame with abundances values. Note that this function should only be used with the raw data (counts)

Value

Imputed data frame

Author(s)

Laura M Zingatetti

Examples

{
# toy example. To simulate 13 missing columns with less than 50 \% of missing values in each.

data('Ruminotypes')
Data<-Ruminotypes$`16_S`
#13 indicates the number of columns with missing values.
Columns<-sample(1:ncol(Data),13)
for (i in Columns){
n<-sample(1:30,1)
Data[sample(1:nrow(Data),n),i]<-NA
}

A<-impute_missing(Data)
#check precision of imputed data
cor(A[,Columns[1]],Ruminotypes$`16_S`[,Columns[1]])
cor(A[,Columns[2]],Ruminotypes$`16_S`[,Columns[2]])
}

Yuliaxis/LinkHD documentation built on April 9, 2020, 12:02 a.m.