pymc.smc.kernels.SMC_KERNEL.__init__

SMC_KERNEL.__init__(draws=2000, start=None, model=None, random_seed=None, threshold=0.5)

Initialize the SMC_kernel class.

Parameters:

drawsint, default 2000

The number of samples to draw from the posterior (i.e. last stage). Also the number of independent chains. Defaults to 2000.

startdict, or array of dict, default None

Starting point in parameter space. It should be a list of dict with length chains. When None (default) the starting point is sampled from the prior distribution, except for parameters with a custom initval, in which case that value is used.

modelModel (optional if in with context).

random_seedint, array_like of int, RandomState or Generator, optional

Value used to initialize the random number generator.

thresholdfloat, default 0.5

Determines the change of beta from stage to stage, i.e.indirectly the number of stages, the higher the value of threshold the higher the number of stages. Defaults to 0.5. It should be between 0 and 1.

Attributes:

self.var_infodict

Dictionary that contains information about model variables shape and size.