I have a fine randomForest classification model which I would like to use in an application that predicts the class of a new case. The new case has inevitably missing values. Predict won't work as such for NAs. How should I do this then?

# create first the new case with missing values

iris.rf <- randomForest(Species ~ ., data=iris[-na.row,])
# print(iris.rf)

myrf.pred <- predict(iris.rf, case.na[-5], type="response")
[1] <NA>

I tried missForest. I combined the original data and the new case, shaked it with missForest, and got imputed values for NAs in my new case. Too heavy computing though.

data.imp <- missForest(data.with.na)

But there must be a way to use rf-model to predict a new case with missing values, right?

EDIT: in my application the prediction is done on one new case. Simple imputation with mean|median values from the training set is a straightforward solution, but imposes bias on the new case prediction.

  • 4
    $\begingroup$ There are many ways missing values can be handled in decision trees, but the randomForest package in R only has the imputation method you described. If you want to stay in a similar environment, gbm has a somewhat smoother method of handling missing values in new data (it's not perfect, but it is useful). $\endgroup$ Jun 18, 2013 at 14:03
  • $\begingroup$ I think that party package deals better with missing values $\endgroup$
    – Simone
    Jun 19, 2013 at 23:54
  • $\begingroup$ Dear @Simone, how does party package work with NAs in the test set? I couldn't find a trace of imputing in party manuals or examples. $\endgroup$
    – hermo
    Aug 14, 2013 at 11:08
  • $\begingroup$ @hermo try to have a look at party's paper citeseerx.ist.psu.edu/viewdoc/summary?doi= it seems the algorithm works like CART - it looks for surrogate splits. $\endgroup$
    – Simone
    Aug 14, 2013 at 23:40
  • $\begingroup$ Try using "na.action = na.roughfix". $\endgroup$
    – user57134
    Oct 7, 2014 at 20:37

2 Answers 2


You have no choice but to impute the values or to change models. A good choice could be aregImpute in the Hmisc package. I think its less heavy than rfimpute which is what is detaining you, first package example (there are others):

# Check that aregImpute can almost exactly estimate missing values when
# there is a perfect nonlinear relationship between two variables
# Fit restricted cubic splines with 4 knots for x1 and x2, linear for x3
x1 <- rnorm(200)
x2 <- x1^2
x3 <- runif(200)
m <- 30
x2[1:m] <- NA
a <- aregImpute(~x1+x2+I(x3), n.impute=5, nk=4, match='closest')
matplot(x1[1:m]^2, a$imputed$x2)
abline(a=0, b=1, lty=2)


# Multiple imputation and estimation of variances and covariances of
# regression coefficient estimates accounting for imputation
# Example 1: large sample size, much missing data, no overlap in
# NAs across variables
x1 <- factor(sample(c('a','b','c'),1000,TRUE))
x2 <- (x1=='b') + 3*(x1=='c') + rnorm(1000,0,2)
x3 <- rnorm(1000)
y  <- x2 + 1*(x1=='c') + .2*x3 + rnorm(1000,0,2)
orig.x1 <- x1[1:250]
orig.x2 <- x2[251:350]
x1[1:250] <- NA
x2[251:350] <- NA
d <- data.frame(x1,x2,x3,y)
# Find value of nk that yields best validating imputation models
# tlinear=FALSE means to not force the target variable to be linear
f <- aregImpute(~y + x1 + x2 + x3, nk=c(0,3:5), tlinear=FALSE,
                data=d, B=10) # normally B=75
# Try forcing target variable (x1, then x2) to be linear while allowing
# predictors to be nonlinear (could also say tlinear=TRUE)
f <- aregImpute(~y + x1 + x2 + x3, nk=c(0,3:5), data=d, B=10)

# Use 100 imputations to better check against individual true values
f <- aregImpute(~y + x1 + x2 + x3, n.impute=100, data=d)
modecat <- function(u) {
 tab <- table(u)
table(orig.x1,apply(f$imputed$x1, 1, modecat))
plot(orig.x2, apply(f$imputed$x2, 1, mean))
fmi <- fit.mult.impute(y ~ x1 + x2 + x3, lm, f, 
fcc <- lm(y ~ x1 + x2 + x3)
summary(fcc)   # SEs are larger than from mult. imputation

You mention that you have many new observations that have missing values on the independant variables. Even though you have many cases like this, if for each new observation there is only missings in one or two of its variables and your amount of variables is not tiny maybe just filling the holes up with a median or average (are they continuous?) could work.

Another thing that could be interesting is to do a minor variable importance analysis. The random forest R implementation calculates two importance measures and respective plots:

varImpPlot(yourRandomForestModel) # yourRandomForestModel must have the argument importance=TRUE 

And you can play around with just including "important" variables in the model training, till the prediction accuracy isn't all that affected in comparison to the "full model". Maybe you keep variables with a low number of missings. It could help you reduce the size of your problem.


You can use na.roughfix when you are predicting more than 1 sample (i.e. predicting for >1 row of data).

For example:


# Make some of the data missing

# Look at NAs
#     Sepal.Length Sepal.Width Petal.Length Petal.Width   Species
# 145          6.7         3.3          5.7         2.5 virginica
# 146          6.7         3.0           NA         2.3      <NA>
# 147          6.3         2.5          5.0         1.9 virginica
# 148          6.5         3.0          5.2         2.0 virginica
# 149          6.2         3.4          5.4         2.3 virginica
# 150          5.9         3.0           NA         1.8      <NA>

# Run Random Forest on a subset of iris
iris.rf <- randomForest(Species ~ ., data=iris[1:120,]) #use argument na.action="na.roughfix" if there's NAs in your training set
# Call:
# randomForest(formula = Species ~ ., data = iris[1:120, ]) 
# Type of random forest: classification
# Number of trees: 500
# No. of variables tried at each split: 2
# OOB estimate of  error rate: 4.17%
# Confusion matrix:
#            setosa versicolor virginica class.error
# setosa         50          0         0        0.00
# versicolor      0         47         3        0.06
# virginica       0          2        18        0.10

# Run prediction on test set, using na.roughfix()
myrf.pred <- predict(iris.rf, na.roughfix(iris[121:150,]), type="response")
#       121        122        123        124        125        126        127        128        129        130        131        132        133        134        135        136 
# virginica versicolor  virginica versicolor  virginica  virginica versicolor versicolor  virginica versicolor  virginica  virginica  virginica versicolor versicolor  virginica 
#       137        138        139        140        141        142        143        144        145        146        147        148        149        150 
# virginica  virginica versicolor  virginica  virginica  virginica  virginica  virginica  virginica  virginica  virginica  virginica  virginica  virginica 
# Levels: setosa versicolor virginica

You can see it dealt fine with the NAs.

  • $\begingroup$ This is indeed a good in-built imputation solution for applications where imputation can be run on larger prediction set (>> 1 sample). From the randomForest documentation of na.roughfix: "A completed data matrix or data frame. For numeric variables, NAs are replaced with column medians. For factor variables, NAs are replaced with the most frequent levels (breaking ties at random). If object contains no NAs, it is returned unaltered." $\endgroup$
    – hermo
    Oct 9, 2022 at 7:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.