In addition to Model based Imputation Methods (see
VIM package also
presents an iterative imputation method.
This vignette showcases the function
IRMI is short for Iterative
Imputation. This method can be used to generate
imputations for several variables in a dataset.
irmi() mimics the functionality of IVEWARE (Raghunathan
et al., 2001), but there are several improvements with respect to
the stability of the initialized values, or the robustness of the
imputed values. In contrast to other imputation methods, the IRMI
algorithm does not require at least one fully observed variable. In each
step of the iteration, one variable is used as a response variable and
the remaining variables serve as the regressors. Thus the “whole”
multivariate information will be used for imputation in the response
variable. For more details, please see IRMI
The following example demonstrates the functionality of
irmi() using a subset of
sleep. The columns
have been selected deliberately to include some interactions between the
library(VIM) library(magrittr) <- sleep[, c("Dream", "NonD", "BodyWgt", "Span")] dataset $BodyWgt <- log(dataset$BodyWgt) dataset$Span <- log(dataset$Span) datasetaggr(dataset)
The plot indicates several missing values in
sapply(dataset, function(x) sum(is.na(x))) #> Dream NonD BodyWgt Span #> 12 14 0 4
The call of the function is straightforward and the algorithm usually converges in a few iterations.
<- irmi(dataset) imp_irmi aggr(imp_irmi, delimiter = "_imp")
We can see that
irmi() imputed all missing values for
all variables in our dataset.
As we can see in the next plot, for imputing missing values in
Bodygt plays an important role. The
original data structure of
preserved by the
irmi() imputation method.
c("NonD", "BodyWgt", "NonD_imp")] %>% imp_irmi[, marginplot(delimiter = "_imp")
The same is true for the data structure of
c("Span", "BodyWgt", "Span_imp")] %>% imp_irmi[, marginplot(delimiter = "_imp")
In order to validate the performance of
irmi() and to
highlight the ability to impute different datatypes the
iris dataset is used. Firstly, some values are randomly set
data(iris) <- iris df colnames(df) <- c("S.Length", "S.Width", "P.Length", "P.Width", "Species") # randomly produce some missing values in the data set.seed(1) <- 50 nbr_missing <- data.frame(row = sample(nrow(iris), size = nbr_missing, replace = TRUE), y col = sample(ncol(iris), size = nbr_missing, replace = TRUE)) <- y[!duplicated(y), ] y as.matrix(y)] <- NA df[ aggr(df)
sapply(df, function(x) sum(is.na(x))) #> S.Length S.Width P.Length P.Width Species #> 10 9 8 10 12
We can see that there are missings in all variables and some observations reveal missing values on several points.
<- irmi(df) imp_irmi aggr(imp_irmi, delimiter = "imp")
The plot indicates that all missing values have been imputed by the IRMI algorithm. The following table displays the rounded first five results of the imputation for all variables.