What bugs me about pandas is that it is so copy heavy. I just wanted to know if there was some pythonic way to get performance without just writing normal SQL