I\'ve taken from similar questions on here to process 50 datasets (which are all have identical 33 columns but different number of rows) at once, but I\'m having trouble cre