I would like an algorithm that does the following - say n is the number of columns in a pandas dataframe
compare column 1 to all of n (n columns) compare column 2 to all