bellini.distributions.TransformedDistribution

class bellini.distributions.TransformedDistribution(args, op, **kwargs)

Bases: bellini.distributions.Distribution

A transformed distribution from one base distribution. While originally intended as a generalized form of distribution transformations such as log or exp, this type is practically a wrapper for any numpy operation on a stochastic value. For example, this allows for the concatenation of Distributions and Quantities.

NOTE: This is only designed to work with np function that output one output. Behavior hasn’t been tested for np functions that return multiple outputs.

__init__(args, op, **kwargs)

Parameters

• args (list of Distribution or Quantity) – The Distributions/Quantities needed as arguments for op

• op (str) – Name of the numpy function to use. See bellini.api.functional for more information

• **kwargs – Any keyword args needed for op

Methods

__init__(args, op, **kwargs)	Parameters args (list of Distribution or Quantity) -- The Distributions/Quantities needed as arguments for op
to_units(new_units[, force])	Return self converted to units new_units
unitless()	Return a unitless version of self

Attributes

dimensionality	The Distribution's dimensionality (e.g.
g	A networkx graph describing the computation graph used to create this Distribution
internal_units	The units associated with internal computations in Bellini for this Distribution
magnitude	A quantity (array-like or scalar) with the same shape as expected samples from this Distribution
name	A string representing the name of self
shape	The shape of samples from this Distribution
units	The units of this Distribution (e.g.

property dimensionality

The Distribution’s dimensionality (e.g. ‘length’ for meter)

property g

A networkx graph describing the computation graph used to create this Distribution

property internal_units

The units associated with internal computations in Bellini for this Distribution

property magnitude

A quantity (array-like or scalar) with the same shape as expected samples from this Distribution

property name

A string representing the name of self

property shape

The shape of samples from this Distribution

to_units(new_units, force=False)

Return self converted to units new_units

unitless()

Return a unitless version of self

property units

The units of this Distribution (e.g. molar)