From the following image, it is obvious that while the values start to converge after a while, close to zero they are unstable.

What method of imputation would you choose for a distribution like this, given the fact that the missing values are always close to zero?

It seems logical that the most frequent value should be used (or some other way that captures the convergence), but close to zero maybe the mean value is a better indicator?




Your Answer

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