I have a dataframe with missing values in some columns. I want to fill these missing values with the mean of non-missing entries in the same column. It\'s easy to show that
