pyfaust.proj.skperm Class Reference

Functor for the SKPERM projector. More...

Public Member Functions
def	__init__ (self, shape, k, normalized=True, pos=False)
	Projector constructor. More...
Public Member Functions inherited from pyfaust.proj.proj_gen
def	__init__ (self, shape)
def	__call__ (self, M)

Public Attributes
	constraint
Public Attributes inherited from pyfaust.proj.proj_gen
	shape

Detailed Description

Functor for the SKPERM projector.

Examples

>>> from pyfaust.proj import skperm
>>> import numpy as np
>>> from numpy import array
>>> k = 2
>>> M = array([[-0.04440802, -0.17569296, -0.02557815, -0.15559154], \

[-0.0083095, -3.38725936, -0.78484126, -0.4883618 ], \ [-1.48942563, -1.71787215, -0.84000212, -3.71752454], \ [-0.88957883, -0.19107863, -5.92900636, -6.51064175]])

>>> p = skperm(M.shape, k, normalized=False)
>>> p(M)
array([[-0.04440802,  0.        , -0.02557815,  0.        ],
[-0.0083095 , -3.38725936,  0.        ,  0.        ],
[ 0.        , -1.71787215,  0.        , -3.71752454],
[ 0.        ,  0.        , -5.92900636, -6.51064175]])

Reference

[1] Quoc-Tung Le, Rémi Gribonval. Structured Support Exploration For Multilayer Sparse Matrix Fac- torization. ICASSP 2021 - IEEE International Conference on Acoustics, Speech and Signal Processing, Jun 2021, Toronto, Ontario, Canada. pp.1-5 hal-03132013.

Constructor & Destructor Documentation

__init__()

def pyfaust.proj.skperm.__init__	(		self,
			shape,
			k,
			normalized = True,
			pos = False
	)

Projector constructor.

Parameters

shape	(tuple(int,int)) the size of the input matrix.
k	(int) the integer sparsity (number of nonzeros) targeted per-row and per-column.
normalized	(bool) True to normalize the projection image according to its Frobenius norm.
pos	(bool) True to skip negative values (replaced by zero) of the matrix to project.

Member Data Documentation

constraint

pyfaust.proj.skperm.constraint

---

The documentation for this class was generated from the following file:

• proj.py