zjddlmZddlZddlmZddlZddlmZmZer ddl m Z ddlm Z Gddej Z dS) ) annotationsN) TYPE_CHECKING) dirichletexponential_family)Sequence)TensorceZdZUdZded<ded<dfd Zedd Zedd Zdd Z dd Z gfddZ ddZ eddZ ddZxZS)Betaa Beta distribution parameterized by alpha and beta. In probability theory and statistics, the beta distribution is a family of continuous probability distributions defined on the interval [0, 1] parameterized by two positive shape parameters, denoted by alpha and beta, that appear as exponents of the random variable and control the shape of the distribution. The generalization to multiple variables is called a Dirichlet distribution. The probability density function (pdf) is .. math:: f(x; \alpha, \beta) = \frac{1}{B(\alpha, \beta)}x^{\alpha-1}(1-x)^{\beta-1} where the normalization, B, is the beta function, .. math:: B(\alpha, \beta) = \int_{0}^{1} t^{\alpha - 1} (1-t)^{\beta - 1}\mathrm{d}t Args: alpha (float|Tensor): Alpha parameter. It supports broadcast semantics. The value of alpha must be positive. When the parameter is a tensor, it represents multiple independent distribution with a batch_shape(refer to ``Distribution`` ). beta (float|Tensor): Beta parameter. It supports broadcast semantics. The value of beta must be positive(>0). When the parameter is tensor, it represent multiple independent distribution with a batch_shape(refer to ``Distribution`` ). Examples: .. code-block:: python >>> import paddle >>> # scale input >>> beta = paddle.distribution.Beta(alpha=0.5, beta=0.5) >>> print(beta.mean) Tensor(shape=[], dtype=float32, place=Place(cpu), stop_gradient=True, 0.50000000) >>> print(beta.variance) Tensor(shape=[], dtype=float32, place=Place(cpu), stop_gradient=True, 0.12500000) >>> print(beta.entropy()) Tensor(shape=[], dtype=float32, place=Place(cpu), stop_gradient=True, -0.24156499) >>> # tensor input with broadcast >>> beta = paddle.distribution.Beta(alpha=paddle.to_tensor([0.2, 0.4]), beta=0.6) >>> print(beta.mean) Tensor(shape=[2], dtype=float32, place=Place(cpu), stop_gradient=True, [0.25000000, 0.40000001]) >>> print(beta.variance) Tensor(shape=[2], dtype=float32, place=Place(cpu), stop_gradient=True, [0.10416666, 0.12000000]) >>> print(beta.entropy()) Tensor(shape=[2], dtype=float32, place=Place(cpu), stop_gradient=True, [-1.91923141, -0.38095081]) ralphabetafloat | TensorreturnNonect|tjrtjg|}t|tjrtjg|}tj||g\|_|_tj tj |j|jgd|_ t |j jdS)N)shape fill_value) isinstancenumbersRealpaddlefullbroadcast_tensorsr r r Dirichletstack _dirichletsuper__init__ _batch_shape)selfr r __class__s h/lsinfo/ai/hellotax_ai/data_center/backend/venv/lib/python3.11/site-packages/paddle/distribution/beta.pyrz Beta.__init__ds eW\ * * <KbU;;;E dGL ) ) :;RD999D & 8% G G DI#- L$*di0" 5 5   566666c0|j|j|jzz S)zMean of beta distribution.r r r s r"meanz Beta.meansszTZ$)344r#cv|j|jz}|j|jz|d|dzzz S)zVariance of beat distribution)r r pow)r sums r"variancez Beta.variancexs9j49$zDI%sQw)?@@r#valuecPtj||S)zProbability density function evaluated at value Args: value (Tensor): Value to be evaluated. Returns: Tensor: Probability. )rexplog_probr r.s r"probz Beta.prob~s z$--..///r#cf|jtj|d|z gdS)zLog probability density function evaluated at value Args: value (Tensor): Value to be evaluated Returns: Tensor: Log probability. g?r)rr1rrr2s r"r1z Beta.log_probs/'' eS5[5I2(N(NOOOr#r Sequence[int]ct|tr|nt|}tj|j|ddS)zSample from beta distribution with sample shape. Args: shape (Sequence[int], optional): Sample shape. Returns: Tensor, Sampled data with shape `sample_shape` + `batch_shape` + `event_shape`. ).rr)axis)rtuplersqueezersample)r rs r"r:z Beta.samplesL$E511CuU||~do44U;;FC"MMMMr#c4|jS)zYEntropy of dirichlet distribution Returns: Tensor: Entropy. )rentropyr&s r"r<z Beta.entropys &&(((r#tuple[Tensor, Tensor]c|j|jfSNr%r&s r"_natural_parameterszBeta._natural_parameterss DI&&r#xyctj|tj|ztj||zz Sr?)rlgamma)r rArBs r"_log_normalizerzBeta._log_normalizers4}Q&-"2"22V]1q55I5IIIr#)r r r r rr)rr)r.rrr)rr5rr)rr=)rArrBrrr)__name__ __module__ __qualname____doc____annotations__rpropertyr'r-r3r1r:r<r@rE __classcell__)r!s@r"r r sCBBHMMMLLL 7 7 7 7 7 7555X5AAAXA 0 0 0 0 P P P P-/ N N N N N))))'''X'JJJJJJJJr#r ) __future__rrtypingrrpaddle.distributionrrcollections.abcrrExponentialFamilyr r#r"rSs#""""" ========((((((QJQJQJQJQJ  /QJQJQJQJQJr#