I wonder if it is allowed to log a variable which is already logged (transformed) in R studio. When I do this, the skewness of my model goes to 0.05 which is quite good. Let