Category:Pre-processing in R: Missing Values

1.) Some useful how-to's in R: > notMissing <- data[data$myAttribute != "?"] # filter by non-missing entries > data$myAttribute[data$myAttribute == "?"] <- newValue

2.) R packages that handle missing values:
There are different packages in R that replace missing values: na.roughfix, Amelia, Mice and mitool. The first package works the same way as Weka methods (replace with median and mode). but the other three packages do multiple imputation (which considers the uncertainty in the data) which means that they consider different valid values for missing values. An example of na.roughfix: > dataNoMissingValue <- na.roughfix(data)