"""Poisson multiplier distribution policy."""
from __future__ import annotations
import math
from typing import TYPE_CHECKING
from eclypse.policies.distribution._helpers import (
build_sampled_distribution_policy,
normalize_distributions,
validate_distributions,
)
from eclypse.utils.constants import MIN_FLOAT
_NORMAL_APPROXIMATION_THRESHOLD = 30
if TYPE_CHECKING:
from random import Random
from eclypse.policies._filters import (
EdgeFilter,
NodeFilter,
)
from eclypse.utils.types import UpdatePolicy
[docs]
def poisson(
*,
node_assets: str | list[str] | None = None,
edge_assets: str | list[str] | None = None,
node_distribution: float = 1.0,
edge_distribution: float | None = None,
node_asset_distributions: dict[str, float] | None = None,
edge_asset_distributions: dict[str, float] | None = None,
minimum: float = MIN_FLOAT,
node_ids: list[str] | None = None,
node_filter: NodeFilter | None = None,
edge_ids: list[tuple[str, str]] | None = None,
edge_filter: EdgeFilter | None = None,
) -> UpdatePolicy:
"""Sample Poisson multiplicative factors without NumPy.
Args:
node_assets (str | list[str] | None): Optional node asset key selector.
edge_assets (str | list[str] | None): Optional edge asset key selector.
node_distribution (float): Default Poisson lambda for selected node assets.
edge_distribution (float | None): Default lambda for selected edge assets. When omitted,
``node_distribution`` is reused.
node_asset_distributions (dict[str, float] | None): Optional per-node-asset lambdas.
edge_asset_distributions (dict[str, float] | None): Optional per-edge-asset lambdas.
minimum (float): Lower bound after applying the sampled multiplier.
node_ids (list[str] | None): Optional explicit node identifiers to mutate.
node_filter (NodeFilter | None): Optional predicate receiving ``(node_id, data)``.
edge_ids (list[tuple[str, str]] | None): Optional explicit edge identifiers to mutate.
edge_filter (EdgeFilter | None): Optional predicate receiving ``(source, target, data)``.
Returns:
Policy that multiplies selected numeric assets by Poisson samples.
"""
effective_edge_distribution = (
node_distribution if edge_distribution is None else edge_distribution
)
validate_distributions(
{
**normalize_distributions("node_distribution", node_distribution),
**normalize_distributions("edge_distribution", effective_edge_distribution),
**normalize_distributions(
"node_asset_distributions", node_asset_distributions
),
**normalize_distributions(
"edge_asset_distributions", edge_asset_distributions
),
},
checks=[(lambda distribution: distribution >= 0, "must be non-negative.")],
)
return build_sampled_distribution_policy(
kind="poisson",
node_assets=node_assets,
edge_assets=edge_assets,
node_distribution=node_distribution,
edge_distribution=effective_edge_distribution,
node_asset_distributions=node_asset_distributions,
edge_asset_distributions=edge_asset_distributions,
minimum=minimum,
node_ids=node_ids,
node_filter=node_filter,
edge_ids=edge_ids,
edge_filter=edge_filter,
sampler=_sample_poisson,
)
def _sample_poisson(rnd: Random, lam: float) -> int:
if lam == 0:
return 0
if lam > _NORMAL_APPROXIMATION_THRESHOLD:
return max(0, round(rnd.gauss(lam, math.sqrt(lam))))
threshold = math.exp(-lam)
product = 1.0
value = -1
while product > threshold:
value += 1
product *= rnd.random()
return value
