I have a few Independent variables that's normal and a Dependent variables that's skewed , I pick log(feature+SHIFT) to correct skewness. The procedure I follow to get prediction is just take exp(predictions)-SHIFT.

Now how do I get my predictions for the following cases.

1) I have Normal - dependent variable and the independent variables are skewed , I transform only a few of them using same log(feature+SHIFT) where the shifts are unique for every feature.

2) Transformations on both dependent and independent variables with different SHIFT values.

Suitable code examples will be highly appreciated. Thanks


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.