mvdist_normal

mvdist_normal(
    mean,
    covariance,
    seed=None,
    max_evaluations=None,
    abs_error=None,
    rel_error=None,
    decomposition='Cholesky',
)

Construct a multivariate normal distribution.

Mirrors the C# MultivariateNormal class of the Numerics library.

Parameters

Name	Type	Description	Default
mean	array-like of float	Means, length d.	required
covariance	array - like	d x d covariance matrix. It must be positive definite for "Cholesky" and may be positive semidefinite for "SingularValue".	required
seed	int	Seed for the Genz quasi-Monte-Carlo integrator behind :meth:MultivariateDistribution.cdf at dimension three and above. None uses the reproducible upstream default, 12345. Set a distinct seed when aggregating many instances so their quadrature errors do not share the same lattice shifts.	None
decomposition	('Cholesky', 'SingularValue')	Covariance factorization. The singular value option supports degenerate Gaussian densities on their affine support.	"Cholesky"
max_evaluations	optional	Integrator tuning; None (the default) for each leaves the ported upstream default untouched.	None
abs_error	optional	Integrator tuning; None (the default) for each leaves the ported upstream default untouched.	None
rel_error	optional	Integrator tuning; None (the default) for each leaves the ported upstream default untouched.	None

Returns

Name	Type	Description
	MultivariateDistribution	Family "MultivariateNormal".

Examples

>>> mv = mvdist_normal([0, 0], [[1, 0], [0, 1]])
>>> mv.pdf([0, 0]) > 0
True