Data cleaning in numpys library

If EDA was to be done on the loan data set would it be okay to replace those missing numerical data with the mean, minimum and maximum values or row's with such missing data should be dropped