LinearCov

class glimix_core.cov.LinearCov(X)[source]

Linear covariance function, K = s⋅XXᵀ.

The mathematical representation is s⋅XXᵀ, for a n×r matrix X provided by the user and a scaling parameter s.

Example

>>> from glimix_core.cov import LinearCov
>>> from numpy import dot
>>> from numpy.random import RandomState
>>>
>>> X = RandomState(0).randn(2, 3)
>>> cov = LinearCov(X)
>>> cov.scale = 1.3
>>> cov.name = "K"
>>> print(cov)
LinearCov(): K
  scale: 1.3
__init__(X)[source]

Constructor.

Parameters

X (array_like) – Matrix X from K = s⋅XXᵀ.

Methods

__init__(X)

Constructor.

fix()

Prevent s update during optimization.

gradient()

Derivative of the covariance matrix over log(s).

unfix()

Enable s update during optimization.

value()

Covariance matrix.

Attributes

X

Matrix X from K = s⋅XXᵀ.

name

Name of this function.

scale

Scale parameter.