iq 8ddlmZddlZddlmZddlmZddlmZdedee ddfd Z d edefd Z d ed e defd Z d edefdZdededededef dZdedededeeeffdZdededededeeeeff dZ ddedededdee deeeff dZdeddfdZy))OptionalN)Tensor)Literalrank_zero_warn nan_strategynan_replace_valuereturnc||dvrtd||dk(r%t|ttfstd|yy)NreplacedropzPArgument `nan_strategy` is expected to be one of `['replace', 'drop']`, but got r zlArgument `nan_replace` is expected to be of a type `int` or `float` when `nan_strategy = 'replace`, but got ) ValueError isinstancefloatint)rr s z/Volumes/fast/ai/experiments/voice-extract-mac/.venv/lib/python3.12/site-packages/torchmetrics/functional/nominal/utils.py_nominal_input_validationrsb..^_k^l m  y 4Es|)T () +  *U confmatc|jd|jd}}tjd|||jz S)zDCompute the expected frequenceis from the provided confusion matrix.rz r, c -> rc)sumtorcheinsum)rmargin_sum_rowsmargin_sum_colss r_compute_expected_freqsr#s9'.{{1~w{{1~_O << o G'++- WWrbias_correctionct|}|jt|jz |jzdz }|dk(r!t j d|jS|dk(rW|rU||z }|j}||t jdt j|z|jzz }t j||z dz|z S)zChi-square test of independenc of variables in a confusion matrix table. Adapted from: https://github.com/scipy/scipy/blob/v1.9.2/scipy/stats/contingency.py. rrdeviceg?) rnumelrshapendimrtensorr#signminimum ones_likeabs)rrexpected_freqsdfdiff directions r_compute_chi_squaredr1)s -W5N    #n&:&:"; ;n>Q>Q QTU UB Qw||C77 Qw?'IIK 9u}}S5??93M-My}}___ 99g.14~E FFrcf||jddk7}|dd|jddk7fS)aDrop all rows and columns containing only zeros. Example: >>> from torch import randint >>> from torchmetrics.functional.nominal.utils import _drop_empty_rows_and_cols >>> matrix = randint(10, size=(4, 3)) >>> matrix[1, :] = matrix[:, 1] = 0 >>> matrix tensor([[2, 0, 6], [0, 0, 0], [0, 0, 0], [3, 0, 4]]) >>> _drop_empty_rows_and_cols(matrix) tensor([[2, 6], [3, 4]]) rrN)r)rs r_drop_empty_rows_and_colsr3=s8$gkk!n)*G 1gkk!n)) **r phi_squarednum_rowsnum_cols confmat_sumctjtjd|j||dz |dz z|dz z z S)z#Compute bias-corrected Phi Squared.r!r"r)rmaxr(r#)r4r5r6r7s r_compute_phi_squared_correctedr:SsG 99 S!3!341 A6;?KK rcN||dz dz|dz z z }||dz dz|dz z z }||fS)z2Compute bias-corrected number of rows and columns.rr$)r5r6r7rows_correctedcols_correcteds r _compute_rows_and_cols_correctedr?`sEA! 3{Q GGNA! 3{Q GGN > ))rcHt||||}t|||\}}|||fS)zBCompute bias-corrected Phi Squared and number of rows and columns.)r:r?)r4r5r6r7phi_squared_correctedr=r>s r_compute_bias_corrected_valuesrBgs9;;RZ\gh%EhPXZe%f"NN .. @@rpredstargetr c|dk(r"|j||j|fStj|j|j}||||fS)a0Handle ``NaN`` values in input data. If ``nan_strategy = 'replace'``, all ``NaN`` values are replaced with ``nan_replace_value``. If ``nan_strategy = 'drop'``, all rows containing ``NaN`` in any of two vectors are dropped. Args: preds: 1D tensor of categorical (nominal) data target: 1D tensor of categorical (nominal) data nan_strategy: Indication of whether to replace or drop ``NaN`` values nan_replace_value: Value to replace ``NaN`s when ``nan_strategy = 'replace``` Returns: Updated ``preds`` and ``target`` tensors which contain no ``Nan`` Raises: ValueError: If ``nan_strategy`` is not from ``['replace', 'drop']``. ValueError: If ``nan_strategy = replace`` and ``nan_replace_value`` is not of a type ``int`` or ``float``. r ) nan_to_numr logical_orisnan)rCrDrr rows_contain_nans r_handle_nan_in_datarJpsk2y  12F4E4EFW4XXX'' v||~F "" #V-=,=%> >>r metric_namec"td|dy)NzUnable to compute zG using bias correction. Please consider to set `bias_correction=False`.r)rKs r&_unable_to_use_bias_correction_warningrMs [M)pqr)r r!)typingrrrtyping_extensionsrtorchmetrics.utilities.printsrstrrrrboolr1r3rr:tupler?rBrJrMr<rrrTs~ %8  C  HUO  X\  XVXX G&G4GFG(+v+&+,       *s*c*PV*[`agioao[p*AA#&A25ADJA 666 !"A09), ? ? ?+,?  ?  66> ?>r