Data containing missing values. Should be a matrix of integers.
initialvalues
The initial values for the start-up process of the imputation. Should be "integer" and length(initialvalues)==1 | length(initialvalues)==dim(DATA)[2]. The default of 0 is not normally a good value.
navalues
NA code for each variable that should be imputed. Should be "integer" and length(initialvalues)==1 | length(initialvalues)==dim(DATA)[2]. Default is R's NA value.
modifyinplace
Should DATA be modified in place? (See the Section: Warning.) If not, a copy is made.
Details
This function imputes the missing values in any variable by replicating the most recently observed value in that variable.
Value
An imputed data matrix the same size as the input DATA.
Warning
If modifyinplace == FALSEDATA or rather the variable supplied is edited directly! This is significantly faster if the data set is large.
Note
This is by far the fastest imputation method. Only one pass of the data is needed. However, no covariate information is used, thus only leads to good results if the data are missing MCAR.
Hanson, R.H. (1978) The Current Population Survey: Design and Methodology. Technical Paper No. 40 . U.S. Bureau of the Census.
Joenssen, D.W. (2015) Hot-Deck-Verfahren zur Imputation fehlender Daten – Auswirkungen des Donor-Limits. Ilmenau: Ilmedia. [in German, Dissertation]
Joenssen, D.W. and Bankhofer, U. (2012) Donor Limited Hot Deck Imputation: Effects on Parameter Estimation. Journal of Theoretical and Applied Computer Science. 6, 58–70.
Joenssen, D.W. and Muellerleile, T. (2014) Fehlende Daten bei Data-Mining. HMD Praxis der Wirtschaftsinformatik. 51, 458–468, 2014. doi: 10.1365/s40702-014-0038-8 [in German]
See Also
impute.CPS_SEQ_HD, impute.mean, impute.NN_HD
Examples
#Set the random seed to an arbitrary number
set.seed(421)
n<-1000
m<-5
pmiss<-.1
#Generate matrix of random integers
Y<-matrix(sample(0:9,replace=TRUE,size=n*m),nrow=n)
#generate 6 missing values, MCAR, in all but the first row
Y[-1,][sample(1:length(Y[-1,]),size=floor(pmiss*length(Y[-1,])))]<-NA
#perform the sequential imputation Y
impute.SEQ_HD(DATA=Y,initialvalues=0, navalues=NA, modifyinplace = FALSE)
####an example highlighting the modifyinplace option
#using cbind to show the results of the function and the intial data next to another
cbind(impute.SEQ_HD(DATA=Y,initialvalues=0, navalues=NA, modifyinplace = FALSE),Y)
#notice that columns 6-10 (representing Y) still have missing data
#same procedure, except modifyinplace is set to TRUE
cbind(impute.SEQ_HD(DATA=Y,initialvalues=0, navalues=NA, modifyinplace = TRUE),Y)
#notice that columns 6-10 (representing Y) are identical to columns 1-5,
#Y has (and any Variables pointing to the same object have) been directly modified.
Results
R version 3.3.1 (2016-06-21) -- "Bug in Your Hair"
Copyright (C) 2016 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)
R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.
R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.
Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.
> library(HotDeckImputation)
Error in library(HotDeckImputation) :
there is no package called 'HotDeckImputation'
Execution halted