o
    yi]                     @   sp   d dl mZmZmZ d dlZd dlmZ d dlmZ d dlm	Z	m
Z
 d dlmZ d dlmZ G dd	 d	eZdS )
    )AnyOptionalUnionN)Tensor)Literal)_tschuprows_t_compute_tschuprows_t_update)_nominal_input_validation)Metricc                       s   e Zd ZU dZdZeed< dZeed< dZeed< e	ed< 				dd
e
deded deee
ef  def
 fddZde	de	ddfddZde	fddZ  ZS )TschuprowsTa  Compute `Tschuprow's T`_ statistic measuring the association between two categorical (nominal) data series.

    .. math::
        T = \sqrt{\frac{\chi^2 / n}{\sqrt{(r - 1) * (k - 1)}}}

    where

    .. math::
        \chi^2 = \sum_{i,j} \ frac{\left(n_{ij} - \frac{n_{i.} n_{.j}}{n}\right)^2}{\frac{n_{i.} n_{.j}}{n}}

    where :math:`n_{ij}` denotes the number of times the values :math:`(A_i, B_j)` are observed
    with :math:`A_i, B_j` represent frequencies of values in ``preds`` and ``target``, respectively.

    Tschuprow's T is a symmetric coefficient, i.e. :math:`T(preds, target) = T(target, preds)`.

    The output values lies in [0, 1] with 1 meaning the perfect association.

    Args:
        num_classes: Integer specifing the number of classes
        bias_correction: Indication of whether to use bias correction.
        nan_strategy: Indication of whether to replace or drop ``NaN`` values
        nan_replace_value: Value to replace ``NaN``s when ``nan_strategy = 'replace'``
        kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info.

    Returns:
        Tschuprow's T statistic

    Raises:
        ValueError:
            If `nan_strategy` is not one of `'replace'` and `'drop'`
        ValueError:
            If `nan_strategy` is equal to `'replace'` and `nan_replace_value` is not an `int` or `float`

    Example:
        >>> from torchmetrics import TschuprowsT
        >>> _ = torch.manual_seed(42)
        >>> preds = torch.randint(0, 4, (100,))
        >>> target = torch.round(preds + torch.randn(100)).clamp(0, 4)
        >>> tschuprows_t = TschuprowsT(num_classes=5)
        >>> tschuprows_t(preds, target)
        tensor(0.4930)
    Ffull_state_updateis_differentiableThigher_is_betterconfmatreplace        num_classesbias_correctionnan_strategy)r   dropnan_replace_valuekwargsc                    sP   t  jdi | || _|| _t|| || _|| _| jdt	||dd d S )Nr   sum)dist_reduce_fx )
super__init__r   r   r	   r   r   	add_statetorchzeros)selfr   r   r   r   r   	__class__r   S/home/ubuntu/.local/lib/python3.10/site-packages/torchmetrics/nominal/tschuprows.pyr   J   s   
zTschuprowsT.__init__predstargetreturnNc                 C   s(   t ||| j| j| j}|  j|7  _dS )a  Update state with predictions and targets.

        Args:
            preds: 1D or 2D tensor of categorical (nominal) data:

                - 1D shape: (batch_size,)
                - 2D shape: (batch_size, num_classes)

            target: 1D or 2D tensor of categorical (nominal) data:

                - 1D shape: (batch_size,)
                - 2D shape: (batch_size, num_classes)
        N)r   r   r   r   r   )r    r$   r%   r   r   r   r#   update\   s   zTschuprowsT.updatec                 C   s   t | j| jS )z!Computer Tschuprow's T statistic.)r   r   r   )r    r   r   r#   computem   s   zTschuprowsT.compute)Tr   r   )__name__
__module____qualname____doc__r   bool__annotations__r   r   r   intr   r   r   floatr   r   r'   r(   __classcell__r   r   r!   r#   r      s,   
 +r   )typingr   r   r   r   r   typing_extensionsr   *torchmetrics.functional.nominal.tschuprowsr   r   %torchmetrics.functional.nominal.utilsr	   torchmetrics.metricr
   r   r   r   r   r#   <module>   s   