o
    h?                     @   sd   d dl Z d dlm  mZ d dlmZ d dlmZ d dl	m
Z
mZmZmZ dgZG dd deZdS )    N)constraints)Distribution)broadcast_alllazy_propertylogits_to_probsprobs_to_logitsNegativeBinomialc                       s   e Zd ZdZededdejdZej	Z
d fdd	Zd  fd	d
	Zdd Zedd Zedd Zedd Zedd Zedd Zedd Zedd Ze fddZdd Z  ZS )!r   ao  
    Creates a Negative Binomial distribution, i.e. distribution
    of the number of successful independent and identical Bernoulli trials
    before :attr:`total_count` failures are achieved. The probability
    of success of each Bernoulli trial is :attr:`probs`.

    Args:
        total_count (float or Tensor): non-negative number of negative Bernoulli
            trials to stop, although the distribution is still valid for real
            valued count
        probs (Tensor): Event probabilities of success in the half open interval [0, 1)
        logits (Tensor): Event log-odds for probabilities of success
    r                 ?)total_countprobslogitsNc                    s   |d u |d u krt d|d ur"t||\| _| _| j| j| _nt||\| _| _| j| j| _|d ur:| jn| j| _| j }t j	||d d S )Nz;Either `probs` or `logits` must be specified, but not both.validate_args)

ValueErrorr   r   r   type_asr   _paramsizesuper__init__)selfr   r   r   r   batch_shape	__class__ [/var/www/html/ai/venv/lib/python3.10/site-packages/torch/distributions/negative_binomial.pyr   $   s$   
zNegativeBinomial.__init__c                    s   |  t|}t|}| j||_d| jv r"| j||_|j|_d| jv r2| j	||_	|j	|_t
t|j|dd | j|_|S )Nr   r   Fr   )_get_checked_instancer   torchSizer   expand__dict__r   r   r   r   r   _validate_args)r   r   	_instancenewr   r   r   r   :   s   


zNegativeBinomial.expandc                 O   s   | j j|i |S N)r   r#   )r   argskwargsr   r   r   _newH   s   zNegativeBinomial._newc                 C   s   | j t| j S r$   )r   r   expr   r   r   r   r   meanK   s   zNegativeBinomial.meanc                 C   s    | j d | j   jddS )N   r	   )min)r   r   r(   floorclampr)   r   r   r   modeO   s    zNegativeBinomial.modec                 C   s   | j t| j  S r$   )r*   r   sigmoidr   r)   r   r   r   varianceS   s   zNegativeBinomial.variancec                 C      t | jddS NT)	is_binary)r   r   r)   r   r   r   r   W      zNegativeBinomial.logitsc                 C   r2   r3   )r   r   r)   r   r   r   r   [   r5   zNegativeBinomial.probsc                 C   s
   | j  S r$   )r   r   r)   r   r   r   param_shape_   s   
zNegativeBinomial.param_shapec                 C   s   t jj| jt | j ddS )NF)concentrationrater   )r   distributionsGammar   r(   r   r)   r   r   r   _gammac   s
   zNegativeBinomial._gammac                 C   sD   t   | jj|d}t |W  d    S 1 sw   Y  d S )N)sample_shape)r   no_gradr;   samplepoisson)r   r<   r8   r   r   r   r>   l   s   
$zNegativeBinomial.samplec                 C   s~   | j r| | | jt| j  |t| j  }t| j|  td|  t| j }|| j| dkd}|| S )Nr
   r	   )	r!   _validate_sampler   F
logsigmoidr   r   lgammamasked_fill)r   valuelog_unnormalized_problog_normalizationr   r   r   log_probq   s"   

zNegativeBinomial.log_prob)NNNr$   )__name__
__module____qualname____doc__r   greater_than_eqhalf_open_intervalrealarg_constraintsnonnegative_integersupportr   r   r'   propertyr*   r/   r1   r   r   r   r6   r;   r   r   r>   rH   __classcell__r   r   r   r   r      s4    







)r   torch.nn.functionalnn
functionalrA   torch.distributionsr    torch.distributions.distributionr   torch.distributions.utilsr   r   r   r   __all__r   r   r   r   r   <module>   s    