§
    ’©ºjã#  ã            	       óB  — d dl Z d dlmZ d dlmZ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 d dlmZmZ d dlmZ d	Zd
eeee         f         defd„Zd
edefd„Zd
edefd„Zd
edefd„Zd
edefd„Zd
edefd„Zd
edeeef         fd„Z	 d-dedee         defd„Z d.d
edededefd„Z!d.dedededefd„Z"d/d
ededefd„Z#d
edefd„Z$d edefd!„Z%d-d
ed"ee         defd#„Z&d0d
edee         d$eej'                 defd%„Z(d
edefd&„Z)d'ed(edefd)„Z*d
ed*ed+edefd,„Z+dS )1é    N)ÚSequence)ÚAnyÚListÚOptionalÚUnion)Úapply_to_collection)ÚTensor)ÚTorchMetricsUserWarning)Ú_TORCH_LESS_THAN_2_6Ú_XLA_AVAILABLE)Úrank_zero_warng�íµ ÷Æ°>ÚxÚreturnc                 ó    — t          | t          j        ¦  «        r| S d„ | D ¦   «         } | st          d¦  «        ‚t          j        | d¬¦  «        S )z'Concatenation along the zero dimension.c                 ó‚   — g | ]<}|                      ¦   «         d k    r |j        dk    r|                     d¦  «        n|‘Œ=S )é   r   )ÚnumelÚndimÚ	unsqueeze)Ú.0Úys     úk/root/private/project/pallet_detection_v2/.venv/lib/python3.11/site-packages/torchmetrics/utilities/data.pyú
<listcomp>z dim_zero_cat.<locals>.<listcomp>!   s@   € ÐLÐLÐLÀQ˜1Ÿ7š7™9œ9¨š>˜>¨a¬f¸ªk¨kˆ�Š�Q‰Œˆ¸qÐLÐLÐLó    zNo samples to concatenater   ©Údim)Ú
isinstanceÚtorchr	   Ú
ValueErrorÚcat©r   s    r   Údim_zero_catr"      sZ   € å�!•U”\Ñ"Ô"ð ØˆØLÐLÈ!ÐLÑLÔL€AØð 6ÝÐ4Ñ5Ô5Ð5ÝŒ9�Q˜AÐÑÔÐr   c                 ó.   — t          j        | d¬¦  «        S )z#Summation along the zero dimension.r   r   )r   Úsumr!   s    r   Údim_zero_sumr%   '   s   € åŒ9�Q˜AÐÑÔÐr   c                 ó.   — t          j        | d¬¦  «        S )z!Average along the zero dimension.r   r   )r   Úmeanr!   s    r   Údim_zero_meanr(   ,   s   € åŒ:�a˜QÐÑÔÐr   c                 ó8   — t          j        | d¬¦  «        j        S )zMax along the zero dimension.r   r   )r   ÚmaxÚvaluesr!   s    r   Údim_zero_maxr,   1   ó   € åŒ9�Q˜AÐÑÔÔ%Ð%r   c                 ó8   — t          j        | d¬¦  «        j        S )zMin along the zero dimension.r   r   )r   Úminr+   r!   s    r   Údim_zero_minr0   6   r-   r   c                 ó   — d„ | D ¦   «         S )z&Flatten list of list into single list.c                 ó   — g | ]	}|D ]}|‘ŒŒ
S © r3   )r   ÚsublistÚitems      r   r   z_flatten.<locals>.<listcomp>=   s%   € Ð6Ð6Ð6�W¨gÐ6Ð6 dˆDÐ6Ð6Ð6Ð6r   r3   r!   s    r   Ú_flattenr6   ;   s   € à6Ð6 Ð6Ñ6Ô6Ð6r   c                 óÒ   — i }d}|                       ¦   «         D ]K\  }}t          |t          ¦  «        r&|                      ¦   «         D ]\  }}||v rd}|||<   ŒŒ@||v rd}|||<   ŒL||fS )zYFlatten dict of dicts into single dict and checking for duplicates in keys along the way.FT)Úitemsr   Údict)r   Únew_dictÚ
duplicatesÚkeyÚvalueÚkÚvs          r   Ú_flatten_dictr@   @   s˜   € à€HØ€JØ—g’g‘i”ið 	"ð 	"‰
ˆˆUÝ�e�TÑ"Ô"ð 	"ØŸš™œð  ð  ‘��1Ø˜�=�=Ø!%�JØ�˜‘�ð ð
 �hˆˆØ!�
Ø!ˆH�S‰MˆMØ�ZÐÐr   Úlabel_tensorÚnum_classesc                 óª  — |€Ht          |                      ¦   «                              ¦   «                              ¦   «         dz   ¦  «        }t	          j        | j        d         |g| j        dd…         ¢R | j        | j        dœŽ}|  	                    ¦   «          
                    d¦  «                             |¦  «        }|                     d|d¦  «        S )a¥  Convert  a dense label tensor to one-hot format.

    Args:
        label_tensor: dense label tensor, with shape [N, d1, d2, ...]
        num_classes: number of classes C

    Returns:
        A sparse label tensor with shape [N, C, d1, d2, ...]

    Example:
        >>> x = torch.tensor([1, 2, 3])
        >>> to_onehot(x)
        tensor([[0, 1, 0, 0],
                [0, 0, 1, 0],
                [0, 0, 0, 1]])

    Nr   r   )ÚdtypeÚdeviceç      ð?)Úintr*   Údetachr5   r   ÚzerosÚshaperD   rE   Úlongr   Ú	expand_asÚscatter_)rA   rB   Útensor_onehotÚindexs       r   Ú	to_onehotrP   Q   sÙ   € ð* ÐÝ˜,×*Ò*Ñ,Ô,×3Ò3Ñ5Ô5×:Ò:Ñ<Ô<¸qÑ@ÑAÔAˆå”KØÔ˜1ÔØðð 
Ô	˜A˜B˜BÔ	ðð ð Ô ØÔ"ðð ð €Mð ×ÒÑÔ×)Ò)¨!Ñ,Ô,×6Ò6°}ÑEÔE€EØ×!Ò! ! U¨CÑ0Ô0Ð0r   r   r>   r   c                 óô   — | j         t          j        k    rH| j        sAt          j        | |d¬¦  «                             |¦  «        }|                     |d|¦  «        S |                      ||¬¦  «        j        S )z3torch.top_k does not support half precision on CPU.T)r   Ústabler   ©r>   r   )	rD   r   ÚhalfÚis_cudaÚargsortÚflipÚnarrowÚtopkÚindices)r   r>   r   Úidxs       r   Ú"_top_k_with_half_precision_supportr\   t   sk   € à„w•%”*ÒÐ Q¤YÐÝŒm˜A 3¨tÐ4Ñ4Ô4×9Ò9¸#Ñ>Ô>ˆØ�zŠz˜#˜q !Ñ$Ô$Ð$Ø�6Š6�A˜3ˆ6ÑÔÔ'Ð'r   Úprob_tensorrY   c                 ó  — t          j        | t           j        ¬¦  «        }|dk    r-|                     ||                      |d¬¦  «        d¦  «         n'|                     |t          | ||¬¦  «        d¦  «         |                     ¦   «         S )aw  Convert a probability tensor to binary by selecting top-k the highest entries.

    Args:
        prob_tensor: dense tensor of shape ``[..., C, ...]``, where ``C`` is in the
            position defined by the ``dim`` argument
        topk: number of the highest entries to turn into 1s
        dim: dimension on which to compare entries

    Returns:
        A binary tensor of the same shape as the input tensor of type ``torch.int32``

    Example:
        >>> x = torch.tensor([[1.1, 2.0, 3.0], [2.0, 1.0, 0.5]])
        >>> select_topk(x, topk=2)
        tensor([[0, 1, 1],
                [1, 1, 0]], dtype=torch.int32)

    ©rD   r   T)r   ÚkeepdimrF   rS   )r   Ú
zeros_likerG   rM   Úargmaxr\   )r]   rY   r   Útopk_tensors       r   Úselect_topkrd   |   s�   € õ& Ô" ;µe´iÐ@Ñ@Ô@€KØˆq‚y€yØ×Ò˜S +×"4Ò"4¸ÀdÐ"4Ñ"KÔ"KÈSÑQÔQÐQÐQà×Ò˜SÕ"DÀ[ÐTXÐ^aÐ"bÑ"bÔ"bÐdgÑhÔhÐhØ�?Š?ÑÔÐr   Ú
argmax_dimc                 ó.   — t          j        | |¬¦  «        S )aw  Convert  a tensor of probabilities to a dense label tensor.

    Args:
        x: probabilities to get the categorical label [N, d1, d2, ...]
        argmax_dim: dimension to apply

    Return:
        A tensor with categorical labels [N, d2, ...]

    Example:
        >>> x = torch.tensor([[0.2, 0.5], [0.9, 0.1]])
        >>> to_categorical(x)
        tensor([1, 0])

    r   )r   rb   )r   re   s     r   Úto_categoricalrg   —   s   € õ  Œ<˜˜zÐ*Ñ*Ô*Ð*r   c                 ó^   — |                       ¦   «         dk    r|                      ¦   «         n| S )Nr   )r   Úsqueezer!   s    r   Ú_squeeze_scalar_element_tensorrj   ª   s$   € ØŸ'š'™)œ) qš.˜.ˆ1�9Š9‰;Œ;ˆ;¨aÐ/r   Údatac                 ó8   — t          | t          t          ¦  «        S ©N)r   r	   rj   )rk   s    r   Ú_squeeze_if_scalarrn   ®   s   € Ý˜t¥VÕ-KÑLÔLÐLr   Ú	minlengthc                 ó¨  — |€!t          t          j        | ¦  «        ¦  «        }t          j        ¦   «         st          s| j        ryt          j        || j        ¬¦  «                             t          | ¦  «        d¦  «        }t          j	        |  
                    dd¦  «        |¦  «                             d¬¦  «        S t          j        | |¬¦  «        S )a   Implement custom bincount.

    PyTorch currently does not support ``torch.bincount`` when running in deterministic mode on GPU or when running
    MPS devices or when running on XLA device. This implementation therefore falls back to using a combination of
    `torch.arange` and `torch.eq` in these scenarios. A small performance hit can expected and higher memory consumption
    as `[batch_size, mincount]` tensor needs to be initialized compared to native ``torch.bincount``.

    Args:
        x: tensor to count
        minlength: minimum length to count

    Returns:
        Number of occurrences for each unique element in x

    Example:
        >>> x = torch.tensor([0,0,0,1,1,2,2,2,2])
        >>> _bincount(x, minlength=3)
        tensor([3, 2, 4])

    N)rE   r   éÿÿÿÿr   r   ©ro   )Úlenr   ÚuniqueÚ$are_deterministic_algorithms_enabledr   Úis_mpsÚarangerE   ÚrepeatÚeqÚreshaper$   Úbincount)r   ro   Úmeshs      r   Ú	_bincountr}   ²   s²   € ð* ÐÝ�œ Q™œÑ(Ô(ˆ	åÔ1Ñ3Ô3ð ;µ~ð ;ÈÌð ;ÝŒ|˜I¨a¬hÐ7Ñ7Ô7×>Ò>½sÀ1¹v¼vÀqÑIÔIˆÝŒx˜Ÿ	š	 " aÑ(Ô(¨$Ñ/Ô/×3Ò3¸Ð3Ñ:Ô:Ð:åŒ>˜! yÐ1Ñ1Ô1Ð1r   rD   c                 ój  — t          j        ¦   «         o| j        o|                      ¦   «         }t          rh|rft
          j        dk    rVt          dt          ¦  «         |  	                    ¦   «          
                    ||¬¦  «                             | j        ¦  «        S t          j
        | ||¬¦  «        S )z\Implement custom cumulative summation for Torch versions which does not support it natively.Úwin32zåYou are trying to use a metric in deterministic mode on GPU that uses `torch.cumsum`, which is currently not supported. The tensor will be copied to the CPU memory to compute it and then copied back to GPU. Expect some slowdowns.)r   rD   )r   ru   rU   Úis_floating_pointr   ÚsysÚplatformr   r
   ÚcpuÚcumsumÚtorE   )r   r   rD   Úis_cuda_fp_deterministics       r   Ú_cumsumr‡   Ñ   s­   € å$ÔIÑKÔKÐsÐPQÔPYÐsÐ^_×^qÒ^qÑ^sÔ^sÐÝð AÐ 8ð A½S¼\ÈWÒ=TÐ=TÝð&õ $ñ		
ô 	
ð 	
ð �uŠu‰wŒw�~Š~ #¨Uˆ~Ñ3Ô3×6Ò6°q´xÑ@Ô@Ð@ÝŒ<˜˜s¨%Ð0Ñ0Ô0Ð0r   c                 ó´   — | |                       ¦   «         z
  } t          j        | ¦  «        }t          | t          j        |¦  «        dz   ¬¦  «        }||         S )zÎSimilar to `_bincount`, but works also with tensor that do not contain continuous values.

    Args:
        x: tensor to count

    Returns:
        Number of occurrences for each unique element in x

    r   rr   )r/   r   rt   r}   r*   )r   Úunique_xÚoutputs      r   Ú_flexible_bincountr‹   ß   sN   € ð 	
ˆA�EŠE‰GŒG‰€AÝŒ|˜A‰Œ€Hå�q¥E¤I¨hÑ$7Ô$7¸!Ñ$;Ð<Ñ<Ô<€Fà�(ÔÐr   Útensor1Útensor2c                 ó‚   — | j         |j         k    r|                     | j         ¬¦  «        }t          j        | |¦  «        S )z:Wrap torch.allclose to be robust towards dtype difference.r_   )rD   r…   r   Úallclose)rŒ   r�   s     r   r�   r�   ò   s8   € à„}˜œÒ%Ð%Ø—*’* 7¤=�*Ñ1Ô1ˆÝŒ>˜' 7Ñ+Ô+Ð+r   ÚxpÚfpc                 óV  — t          j        |¦  «        }||         }||         }|dd…         |dd…         z
  |dd…         |dd…         z
  z  }t          j        || ¦  «        dz
  }t          j        |dt	          |¦  «        dz
  ¦  «        }||         ||         | ||         z
  z  z   S )zàInterpolation function comparable to numpy.interp.

    Args:
        x: x-coordinates where to evaluate the interpolated values
        xp: x-coordinates of the data points
        fp: y-coordinates of the data points

    r   Nrq   r   )r   rV   ÚsearchsortedÚclamprs   )r   r�   r‘   Úsorted_indicesÚslopesrZ   s         r   Úinterpr—   ù   sµ   € õ ”] 2Ñ&Ô&€NØ	ˆNÔ	€BØ	ˆNÔ	€Bð ���Œf�r˜#˜2˜#”wÑ 2 a b b¤6¨B¨s°¨s¬GÑ#3Ñ4€Fõ Ô   QÑ'Ô'¨!Ñ+€GÝŒk˜' 1¥c¨&¡k¤k°A¡oÑ6Ô6€Gð ˆgŒ;˜ œ¨A°°7´©OÑ<Ñ<Ð<r   rm   )r   r   )r   )r   N),r�   Úcollections.abcr   Útypingr   r   r   r   r   Úlightning_utilitiesr   r	   Ú!torchmetrics.utilities.exceptionsr
   Útorchmetrics.utilities.importsr   r   Útorchmetrics.utilities.printsr   Ú
METRIC_EPSr"   r%   r(   r,   r0   Úlistr6   r9   ÚtupleÚboolr@   rG   rP   r\   rd   rg   rj   rn   r}   rD   r‡   r‹   r�   r—   r3   r   r   ú<module>r¢      sÐ  ðð €
€
€
Ø $Ð $Ð $Ð $Ð $Ð $Ø -Ð -Ð -Ð -Ð -Ð -Ð -Ð -Ð -Ð -Ð -Ð -à €€€Ø 3Ð 3Ð 3Ð 3Ð 3Ð 3Ø Ð Ð Ð Ð Ð à EÐ EÐ EÐ EÐ EÐ EØ OÐ OÐ OÐ OÐ OÐ OÐ OÐ OØ 8Ð 8Ð 8Ð 8Ð 8Ð 8à€
ð�E˜& $ v¤,Ð.Ô/ð °Fð ð ð ð ð�Fð ˜vð ð ð ð ð
 �Vð   ð  ð  ð  ð  ð
&�Fð &˜vð &ð &ð &ð &ð
&�Fð &˜vð &ð &ð &ð &ð
7�ð 7˜Tð 7ð 7ð 7ð 7ð
 �Tð  ˜e D¨$ JÔ/ð  ð  ð  ð  ð& "&ð 1ð  1Øð 1à˜#”ð 1ð ð 1ð  1ð  1ð  1ðF(ð (¨&ð (°Sð (À3ð (Èvð (ð (ð (ð (ðð ˜Vð ¨3ð ¸ð ÀVð ð ð ð ð6+ð +�fð +¨#ð +°fð +ð +ð +ð +ð&0 fð 0°ð 0ð 0ð 0ð 0ðM˜Sð M Sð Mð Mð Mð Mð2ð 2�ð 2 H¨S¤Mð 2¸Vð 2ð 2ð 2ð 2ð>1ð 1ˆvð 1˜H SœMð 1°h¸u¼{Ô6Kð 1ÐW]ð 1ð 1ð 1ð 1ð˜&ð  Vð ð ð ð ð&,�fð , vð ,°$ð ,ð ,ð ,ð ,ð=ˆfð =˜&ð = fð =°ð =ð =ð =ð =ð =ð =r   