torch.distributions.gamma 的源代码

from numbers import Number

import torch
from torch.distributions import constraints
from torch.distributions.exp_family import ExponentialFamily
from torch.distributions.utils import broadcast_all

__all__ = ["Gamma"]


def _standard_gamma(concentration):
    return torch._standard_gamma(concentration)


class Gamma(ExponentialFamily):
    r"""
    创建一个由形状参数 :attr:`concentration` 和 :attr:`rate` 参数化的 Gamma 分布。

    示例::

        >>> # xdoctest: +IGNORE_WANT("非确定性")
        >>> m = Gamma(torch.tensor([1.0]), torch.tensor([1.0]))
        >>> m.sample()  # 浓度=1 和 速率=1 的 Gamma 分布
        tensor([ 0.1046])

    参数:
        concentration (float 或 Tensor): 分布的形状参数
            (通常称为 alpha)
        rate (float 或 Tensor): 速率 = 1 / 分布的尺度
            (通常称为 beta)
    """
    arg_constraints = {
        "concentration": constraints.positive,
        "rate": constraints.positive,
    }
    support = constraints.nonnegative
    has_rsample = True
    _mean_carrier_measure = 0

    @property
    def mean(self):
        return self.concentration / self.rate

    @property
    def mode(self):
        return ((self.concentration - 1) / self.rate).clamp(min=0)

    @property
    def variance(self):
        return self.concentration / self.rate.pow(2)

    def __init__(self, concentration, rate, validate_args=None):
        self.concentration, self.rate = broadcast_all(concentration, rate)
        if isinstance(concentration, Number) and isinstance(rate, Number):
            batch_shape = torch.Size()
        else:
            batch_shape = self.concentration.size()
        super().__init__(batch_shape, validate_args=validate_args)

    def expand(self, batch_shape, _instance=None):
        new = self._get_checked_instance(Gamma, _instance)
        batch_shape = torch.Size(batch_shape)
        new.concentration = self.concentration.expand(batch_shape)
        new.rate = self.rate.expand(batch_shape)
        super(Gamma, new).__init__(batch_shape, validate_args=False)
        new._validate_args = self._validate_args
        return new

    def rsample(self, sample_shape=torch.Size()):
        shape = self._extended_shape(sample_shape)
        value = _standard_gamma(self.concentration.expand(shape)) / self.rate.expand(
            shape
        )
        value.detach().clamp_(
            min=torch.finfo(value.dtype).tiny
        )  # 不要在 autograd 图中记录
        return value

    def log_prob(self, value):
        value = torch.as_tensor(value, dtype=self.rate.dtype, device=self.rate.device)
        if self._validate_args:
            self._validate_sample(value)
        return (
            torch.xlogy(self.concentration, self.rate)
            + <