Multivariate Normal from Bijector¤
distreqx.distributions.mvn_from_bijector.MultivariateNormalFromBijector (AbstractMultivariateNormalFromBijector)
¤
Multivariate normal distribution on R^k
.
The multivariate normal over x
is characterized by an invertible affine
transformation x = f(z) = A @ z + b
, where z
is a random variable that
follows a standard multivariate normal on R^k
, i.e., p(z) = N(0, I_k)
,
A
is a k x k
transformation matrix, and b
is a k
-dimensional vector.
The resulting PDF on x
is a multivariate normal, p(x) = N(b, C)
, where
C = A @ A.T
is the covariance matrix.
The transformation x = f(z)
must be specified by a linear scale bijector
implementing the operation A @ z
and a shift (or location) term b
.
__init__(self, loc: Array, scale: AbstractLinearBijector)
¤
Initializes the distribution.
Arguments:
loc
: The termb
, i.e., the mean of the multivariate normal distribution.scale
: The bijector specifying the linear transformationA @ z
, as described in the class docstring.