Can anyone explain me StandardScaler?

Question

I am unable to understand the page of the StandardScaler in the documentation of sklearn.

Can anyone explain this to me in simple terms?

Answer 1

We apply StandardScalar() on a row basis.

So, for each row in a column (I am assuming that you are working with a Pandas DataFrame):

x_new = (x_original - mean_of_distribution) / std_of_distribution

Few points -

1. It is called Standard Scalar as we are dividing it by the standard deviation of the distribution (distr. of the feature). Similarly, you can guess for MinMaxScalar().

2. The original distribution remains the same after applying StandardScalar(). It is a common misconception that the distribution gets changed to a Normal Distribution. We are just squashing the range into [0, 1].