Speed up Pandas code with Numpy | Towards Data Science

But I can’t vectorise this, can I? …. yes, you probably can!