Distributions

Degree distribution sampling for the configuration model.

distributions

Stub/degree distributions for network generation.

Distribution

Bases: Protocol

Protocol for discrete distributions used in stub allocation.

sample

sample(n: int, rng: Generator) -> NDArray[int_]

Sample n values from the distribution.

Parameters:

Name	Type	Description	Default
n	int	Number of values to sample.	required
rng	Generator	Random number generator.	required

Returns:

Type	Description
NDArray[int_]	Array of n non-negative integers.

Poisson dataclass

Poisson(mu: float)

Poisson distribution with mean mu.

sample

sample(n: int, rng: Generator) -> NDArray[int_]

Sample n values from Poisson(mu).

Fixed dataclass

Fixed(value: int)

Fixed value (all nodes get the same stub count).

sample

sample(n: int, rng: Generator) -> NDArray[int_]

Return array of n identical values.

Empirical dataclass

Empirical(values: tuple[int, ...])

Empirical distribution from observed values.

sample

sample(n: int, rng: Generator) -> NDArray[int_]

Sample n values with replacement from observed values.

prepare_stubs

prepare_stubs(
    distribution: Distribution,
    n: int,
    rng: Generator,
    method: CorrectionMethod = "resample_one",
) -> NDArray[int_]

Sample stubs from distribution and ensure even sum.

Parameters:

Name	Type	Description	Default
distribution	Distribution	Distribution to sample from.	required
n	int	Number of nodes.	required
rng	Generator	Random number generator.	required
method	CorrectionMethod	How to correct odd sums: - "adjust": Add or subtract 1 from a random node - "resample_one": Resample one node until sum is even - "resample_all": Resample all nodes until sum is even	'resample_one'

Returns:

Type	Description
NDArray[int_]	Array of stub counts with even sum.