i)&ddlmZddlmZmZmZmZddlZddlmZddl m Z m Z ddl m Z ddlmZddlmZmZesd gZ dd ejd ejd ejd ejdejdejdejdejdedeejejejejejejejejffdZGdde Zy))Sequence)AnyListOptionalUnionN)Tensor)_pearson_corrcoef_compute_pearson_corrcoef_update)Metric)_MATPLOTLIB_AVAILABLE)_AX_TYPE_PLOT_OUT_TYPEzPearsonCorrCoef.plotmeans_xmeans_y maxs_abs_x maxs_abs_yvars_xvars_ycorrs_xynbsepsreturnc t|dk(r"|d|d|d|d|d|d|d|dfS|d} |d} |d} |d} |d} |d}|d}|d}tdt|D]}||}||}||}||}||}||}||}||}tjtj||||z|}|| z||zz|z }|| z||zz|z }||z|z }|| z }|| z }| |z||dzzz} ||z||dzzz}!||z||z|zz}"tj | |}#tj | |}$|} |} |#} |$} | } |!}|"}|}#$ !"fS)aAggregate the statistics from multiple devices. Formula taken from here: `Parallel algorithm for calculating variance `_ We use `eps` to avoid division by zero when `n1` and `n2` are both zero. Generally, the value of `eps` should not matter, as if `n1` and `n2` are both zero, all the states will also be zero. r)lenrangetorchwhere logical_ormaximum)%rrrrrrrrrmx1my1max1may1vx1vy1cxy1n1imx2my2max2may2vx2vy2cxy2n2nbmean_xmean_yn12_bdelta_xdelta_yvar_xvar_ycorr_xy max_abs_dev_x max_abs_dev_ys% t/Volumes/fast/ai/experiments/voice-extract-mac/.venv/lib/python3.12/site-packages/torchmetrics/regression/pearson.py_final_aggregationr?s, 7|qqz71:z!}jmVAYPVWXPY[cde[fhklmhnnn !*C !*C a=D a=D )C )C A;D QB 1c'l #ajaj!}!}QiQi{ V [[))"b127C @s(R#X%+s(R#X%+R" ))c EGQJ..c EGQJ..+' 99 dD1  dD1  G$H 6=-wPR RRcFeZdZUdZdZeed<dZeeed<dZ eed<dZ e ed<d Z e ed <e eed <e eed <eed <eed<eed<eed<eed<eed<eed<eed< ddededdffd Zd ed eddfdZdefdZ ddeeeeefdeedefdZxZS) PearsonCorrCoefaCompute `Pearson Correlation Coefficient`_. .. math:: P_{corr}(x,y) = \frac{cov(x,y)}{\sigma_x \sigma_y} Where :math:`y` is a tensor of target values, and :math:`x` is a tensor of predictions. As input to ``forward`` and ``update`` the metric accepts the following input: - ``preds`` (:class:`~torch.Tensor`): either single output float tensor with shape ``(N,)`` or multioutput float tensor of shape ``(N,d)`` - ``target`` (:class:`~torch.Tensor`): either single output tensor with shape ``(N,)`` or multioutput tensor of shape ``(N,d)`` As output of ``forward`` and ``compute`` the metric returns the following output: - ``pearson`` (:class:`~torch.Tensor`): A tensor with the Pearson Correlation Coefficient Args: num_outputs: Number of outputs in multioutput setting kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info. Example (single output regression): >>> from torchmetrics.regression import PearsonCorrCoef >>> target = torch.tensor([3, -0.5, 2, 7]) >>> preds = torch.tensor([2.5, 0.0, 2, 8]) >>> pearson = PearsonCorrCoef() >>> pearson(preds, target) tensor(0.9849) Example (multi output regression): >>> from torchmetrics.regression import PearsonCorrCoef >>> target = torch.tensor([[3, -0.5], [2, 7]]) >>> preds = torch.tensor([[2.5, 0.0], [2, 8]]) >>> pearson = PearsonCorrCoef(num_outputs=2) >>> pearson(preds, target) tensor([1., 1.]) Tis_differentiableNhigher_is_betterfull_state_updategplot_lower_boundg?plot_upper_boundpredstargetr4r5r<r=r9r:r;n_total num_outputskwargsrc t|d i|t|ts|dkr t d||_|j dtj|j d|j dtj|j d|j dtj|j d|j dtj|j d|j dtj|j d|j d tj|j d|j d tj|j d|j d tj|j dy) NrzQExpected argument `num_outputs` to be an int larger than 0, but got {num_outputs}r4)defaultdist_reduce_fxr5r<r=r9r:r;rJ) super__init__ isinstanceint ValueErrorrK add_staterzeros)selfrKrL __class__s r>rRzPearsonCorrCoef.__init__sI "6"+s+ apq q& xT5E5E)FW[\ xT5E5E)FW[\  DupdatezPearsonCorrCoef.updates %;;;;,,,,****LLll((  K K     J J L Lr@c 4|jdk(r|jjdkDs(|jdkDr|jjdkDrmt |j|j |j |j|j|j|j|j\}}}}}}}}nH|j }|j}|j}|j}|j}|j}t||||||S)z3Compute pearson correlation coefficient over state.r)rrrrrrrr) rKr4numelndimr?r5r<r=r9r:r;rJr )rX_r<r=r9r:r;rJs r>computezPearsonCorrCoef.computes    !dkk&7&7&9A&=4CSCSVWCW\`\g\g\l\lop\pQc  ----zzzzLL R NAq-ugw!..M ..MJJEJJEllGllG( ueU\^effr@valaxc&|j||S)a Plot a single or multiple values from the metric. Args: val: Either a single result from calling `metric.forward` or `metric.compute` or a list of these results. If no value is provided, will automatically call `metric.compute` and plot that result. ax: An matplotlib axis object. If provided will add plot to that axis Returns: Figure and Axes object Raises: ModuleNotFoundError: If `matplotlib` is not installed .. plot:: :scale: 75 >>> from torch import randn >>> # Example plotting a single value >>> from torchmetrics.regression import PearsonCorrCoef >>> metric = PearsonCorrCoef() >>> metric.update(randn(10,), randn(10,)) >>> fig_, ax_ = metric.plot() .. plot:: :scale: 75 >>> from torch import randn >>> # Example plotting multiple values >>> from torchmetrics.regression import PearsonCorrCoef >>> metric = PearsonCorrCoef() >>> values = [] >>> for _ in range(10): ... values.append(metric(randn(10,), randn(10,))) >>> fig, ax = metric.plot(values) )_plot)rXrbrcs r>plotzPearsonCorrCoef.plotsPzz#r""r@)r)NN)__name__ __module__ __qualname____doc__rCbool__annotations__rDrrErFfloatrGrrrTrrRr\rarrr rrf __classcell__)rYs@r>rBrBds&P#t"'+htn+"t""e"!e! < L N N M M O O^^^  ^& F F t 2gg0_c(#E&(6*:":;<(#IQRZI[(# (#r@rB)g|=)collections.abcrtypingrrrrrr*torchmetrics.functional.regression.pearsonr r torchmetrics.metricr torchmetrics.utilities.importsr torchmetrics.utilities.plotr r__doctest_skip__rmtupler?rBrPr@r>rws%-- j&@@./DS \\DS \\DS DS DS LL DS LL DSllDS DS DS  LL%,, ellELL%,,X]XdXdfkfrfrrDSNd#fd#r@