Hi all, I'm trying to use kNNimpute in the imputation package to fill in missing precipitation data for a data frame I have. Example is: okee: Date rainfall 1997-05-01 0 1997-05-02 0 1997-05-03 NA 1997-05-04 0 1997-05-05 0 ..................... .. 2007-04-01 NA 2007-04-02 NA 2007-04-03 NA 2007-04-04 NA 2007-04-05 0 ................... .. where there are large swatches (30 days) of data missing in the ten year time series. I tried newokee<-kNNImpute(okee, k=30, verbose = T) hoping it would impute data for the rows with NA values according to weighted means of closest 30 non-NA neighbours and I got the following message back: [1] "imputing on 270 missing values with matrix size 7304" [1] "Computing distance matrix..." [1] "Distance matrix complete" [1] "Imputing row 70" Error in which(missing.matrix[rowIndex, ]) : subscript out of bounds In addition: Warning message: In dist(x, upper = T) : NAs introduced by coercion What syntax do I need to impute the precipitation data? Failing that, do you have another recommendation of a method to use? Statistics is not my strong point. thank you for any help you are able to give, Aimee. [[alternative HTML version deleted]]