I\'m having difficulty transferring the logistic regression gradient into numpy operations and getting the intended result. Somehow when I print my theta it is composed of a