Adding a na.rm option

[soerenkuenzel]

Often there are missing values in the data frames. It is annoying to manually delete those rows all the time, and it might be interesting to add an option na.rm. It could have the options:
remove: those rows are removed,
error: everything is stopped,
impute: these values are somehow imputed. (we could ignore this options for now)

[JasjeetSekhon]

Good idea. One simple way to impute: replace the missing values with the compete case mean, and add an indicator for missing observations for that variable. Problem: when there is lots of missingness in many different variables, there will be many indicator variables produced.

On Sat, Dec 8, 2018 at 10:21 PM soerenkuenzel ***@***.***> wrote: Often there are missing values in the data frames. It is annoying to manually delete those rows all the time, and it might be interesting to add an option na.rm. It could have the options: remove: those rows are removed, error: everything is stopped, impute: these values are somehow imputed. (we could ignore this options for now) — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#6>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AM9zoQ-v6dWuvmqu3zSlSQDeSXlotTzXks5u3IGwgaJpZM4ZJ7nM> .

-- Sent from Gmail Mobile