pydiodon.scale

pydiodon.scale(A)

Scales a matrix columnwise

Parameters

Aa 2D numpy array, n x p

the matrix to be scaled

Returns

Asa 2D numpy array

matrix A with scaled columns

aa 1D numpy array, p values

the norm of each column of A

Notes

It is recommended to center A columnwise before scaling.

Each coordinates \(a_{ij}\) of A is replaced by \(a_{ij} / ||a_{.j}||\) where \(||a_{.j}||\) is the norm of column j: \(||a_{.j}||^2 = \sum_i a_{ij}^2\)

Example

>>> import pydiodon as dio
>>> import numpy as np
>>> n = 4 ; p = 3
>>> A = np.arange(n*p)
>>> A.shape =(n,p)
>>> Ac, m = dio.center_col(A)
>>> As, s = dio.scale(Ac)

af, 21.02.19, 22.10.13