I have a large dataset and when I run it it takes very longtime. Generally speaking I find that the only way to avoid it we have to vectorize it using numpy. Or I maybe be w