Sometimes when working with (struggling with!) missing values, you find that it is not missing at all! Sometimes someone has been helpful (!) and entered some placeholder value like "THISISMISSING" when that happens, what could you do (choose the best answer, it won't be the ONLY possible answer, just the best one here) np. NaN corrected reads "THISISMISSING" Replace "THISISMISSING" with "MISSING" Replace "THISISMISSING" with a missing value (np.NaN) for all values np.NaN, impute 0 's