pymc.dims.ZeroSumNormal#

class pymc.dims.ZeroSumNormal(*args, core_dims=None, dims=None, default_transform=UNSET, observed=None, **kwargs)[source]#

Zero-sum multivariate normal distribution.

Parameters:
sigmaxtensor_like, optional

The standard deviation of the underlying unconstrained normal distribution. Defaults to 1.0. It cannot have core dimensions.

core_dimsSequence of str, optional

The axes along which the zero-sum constraint is applied.

**kwargs

Additional keyword arguments used to define the distribution.

Returns:
XTensorVariable

An xtensor variable representing the zero-sum multivariate normal distribution.

Methods

ZeroSumNormal.dist([sigma, core_dims])