h3SrSSKJr SSKJr SSKJr SSKJr SSK J r SSK J r SSK rSS jrSS jr"S S 5r\"S S55r"SS5rg)aModule for managing and transforming label data during augmentation. This module provides utilities for encoding, decoding, and tracking metadata for labels during the augmentation process. It includes classes for managing label transformations, preserving data types, and ensuring consistent handling of categorical, numerical, and mixed label types. The module supports automatic encoding of string labels to numerical values and restoration of original data types after processing. ) annotations) defaultdict)Sequence) dataclass)Real)AnyNcN[U[5(aSU4$S[U54$)a:Sort items by type then value for consistent label ordering. This function is used to sort labels in a consistent order, prioritizing numerical values before string values. All numerical values are given priority 0, while string values are given priority 1, ensuring numerical values are sorted first. Args: item (Any): Item to be sorted, can be either a numeric value or any other type. Returns: tuple[int, Real | str]: A tuple with sort priority (0 for numbers, 1 for others) and the value itself (or string representation for non-numeric values). r) isinstancerstr)items [/var/www/fran/franai/venv/lib/python3.13/site-packages/albumentations/core/label_manager.py custom_sortrs'#4..At9BQD NBcr/n/nUH+n[U[5(aUOURU5 M- X4$N)r rappend)labelsnumeric_labels string_labelslabels r_categorize_labelsr's;!#N!M%eT22 MMeT  ((rcT\rSrSrSrS SjrS SjrS SjrS SjrS Sjr S Sjr S r g ) LabelEncoder0aEncodes labels into integer indices. This class handles the conversion between original label values and their numerical representations. It supports both numerical and categorical labels. Args: classes_ (dict[str | Real, int]): Mapping from original labels to encoded indices. inverse_classes_ (dict[int, str | Real]): Mapping from encoded indices to original labels. num_classes (int): Number of unique classes. is_numerical (bool): Whether the original labels are numerical. c<0Ul0UlSUlSUlg)NrT)classes_inverse_classes_ num_classes is_numericalselfs r__init__LabelEncoder.__init__>s!/1 79 !"&rc[U[R5(aUR5R 5nU(d SUlU$[ SU55UlUR (aU$[[U5[S9nUHZnX0R;dMURURU'X0RUR'U=RS- sl M\ U$)zFit the encoder to the input labels. Args: y (Sequence[Any] | np.ndarray): Input labels to fit the encoder. Returns: LabelEncoder: The fitted encoder instance. Fc3B# UHn[U[5v M g7fr)r r.0rs r #LabelEncoder.fit..VsGQE 5$ 7 7Qskeyr ) r npndarrayflattentolistr allsortedsetrrrr)r"y unique_labelsrs rfitLabelEncoder.fitDs a $ $ ""$A %D KGQGG   Ks1v;7 "EMM)'+'7'7 e$:?%%d&6&67  A% #  rc:[U[R5(aUR5R 5nUR (a[R "U5$[R "UVs/sHo RUPM sn5$s snf)zTransform labels to encoded integer indices. Args: y (Sequence[Any] | np.ndarray): Input labels to transform. Returns: np.ndarray: Encoded integer indices. )r r-r.r/r0r arrayrr"r4rs r transformLabelEncoder.transformcsh a $ $ ""$A   88A; xx1=1%u-1=>>=9BcFURU5 URU5$)zFit the encoder and transform the input labels in one step. Args: y (Sequence[Any] | np.ndarray): Input labels to fit and transform. Returns: np.ndarray: Encoded integer indices. )r6r;)r"r4s r fit_transformLabelEncoder.fit_transformus  ~~a  rc:[U[R5(aUR5R 5nUR (a[R "U5$[R "UVs/sHo RUPM sn5$s snf)zTransform encoded indices back to original labels. Args: y (Sequence[Any] | np.ndarray): Encoded integer indices. Returns: np.ndarray: Original labels. )r r-r.r/r0r r9rr:s rinverse_transformLabelEncoder.inverse_transformsj a $ $ ""$A   88A; xx1E1%..u51EFFEr=crUR(aU$[U[R5(aUR 5R 5nO>[U[ 5(a![U[5(d [U5nOUcU$U/n[URR55n[U5U- nU(dU$[U5upV[U5[U[S9-nUH@nURn XRU'XRU 'U=RS- slMB U$)aUpdate the encoder with new labels encountered after initial fitting. This method identifies labels in the input sequence that are not already known to the encoder and adds them to the internal mapping. It does not change the encoding of previously seen labels. Args: y (Sequence[Any] | np.ndarray): A sequence or array of potentially new labels. Returns: LabelEncoder: The updated encoder instance. r+r )r r r-r.r/r0rr listr3rkeysrr2rr) r"r4 input_labelscurrent_labels_setnew_unique_labelsrrsorted_new_labelsrnew_ids rupdateLabelEncoder.updates   K a $ $99;--/L 8 $ $Z3-?-?7L YK3L!!3!3!56 -0BB K);;L(M%">2VMs5SS&E%%F#)MM% ,1 ! !& )    !  '  r)rrr rNreturnNone)r4Sequence[Any] | np.ndarrayrOr)r4rQrO np.ndarray) __name__ __module__ __qualname____firstlineno____doc__r#r6r;r?rBrL__static_attributes__rrrr0s& ' >?$ !G$0rrcJ\rSrSr%SrS\S'S\S'SrS\S 'SrS \S 'S rg) LabelMetadataz$Stores metadata about a label field.type input_typeboolr Nznp.dtype | NonedtypeLabelEncoder | NoneencoderrY) rSrTrUrVrW__annotations__r`rbrXrYrrr[r[s&.!E?!#'G 'rr[cr\rSrSrSrSSjrSSjrSSjrSSjrSSjr SSjr SS jr SS jr SS jr S rg ) LabelManageraManages label encoding and decoding across multiple data fields. This class handles the encoding, decoding, and type management for label fields. It maintains metadata about each field to ensure proper conversion between original and encoded representations. Args: metadata (dict[str, dict[str, LabelMetadata]]): Dictionary mapping data types and label fields to their metadata. c,[[5Ulgr)rdictmetadatar!s rr#LabelManager.__init__s=H=N rcJXR;abX RU;aPURUUnUR(d,UR(aURRU5 O"UR U5nX@RUU'UR X45$)aPProcess a label field, store metadata, and encode. If the field has been processed before (metadata exists), this will update the existing LabelEncoder with any new labels found in `field_data` before encoding. Otherwise, it analyzes the input, creates metadata, and fits the encoder. Args: data_name (str): The name of the main data type (e.g., 'bboxes', 'keypoints'). label_field (str): The specific label field being processed (e.g., 'class_labels'). field_data (Any): The actual label data for this field. Returns: np.ndarray: The encoded label data as a numpy array. )rir rbrL_analyze_input _encode_data)r" data_name label_field field_dataris r process_fieldLabelManager.process_fields  %+y9Q*Q}}Y/ [U5n[U[R5(a UROSn[U[R5(a@UR S:a0[R "UR[R5nOd[U[5(a/[U[5(dU(a[SU55nO [U[[45(aSnOSn[UUUS9nU(d[5UlU$)z'Analyze input data and create metadata.Nrc3N# UHn[U[[45v M g7fr)r intfloatr's rr).LabelManager._analyze_input.. sWJ5z%#u>>Js#%TF)r^r r`)r]r r-r.r`size issubdtypenumberrr r1r|r}r[rrb)r"rpr^r`r ris rrlLabelManager._analyze_inputs*% $.z2::$F$F   D j"** - -*//A2E==)9)9299EL  H - -jS6Q6QV`WJWWL  S%L 1 1L!L !%  +~H rcUR(a[U[R5(a/UR SS5R [R 5$[R"U[R S9R SS5$URc [S5eURRU5R SS5$)z!Encode field data for processing.r )r`z.Encoder not initialized for non-numerical data) r r r-r.reshapeastypefloat32r9rb ValueErrorr?)r"rpris rrmLabelManager._encode_datas  *bjj11!))"a077 CC88Jbjj9AA"aH H    #MN N--j9AA"aHHrcBUR(a8URbURUR5$UR5$URc [ S5eURR UR[55nURS5$)zDecode processed data.z(Encoder not found for non-numerical datar) r r`rr/rbrrBr|r)r"rvridecodeds rrtLabelManager._decode_data,s  ~~)#**8>>::'') )    #GH H""44\5H5H5MNr""rc[UR[5(a5[UR[[ 45(aUR 5$[UR[5(aS[UR[R5(a*URbURUR5$U$UR 5$)z"Restore data to its original type.) r r^r] issubclassrErr0r-r.r`r)r"rwris rruLabelManager._restore_type9s h))4 0 0Z@S@SVZ\dUe5f5f&&( ( h))4 0 0Z@S@SUWU_U_5`5`~~)#**8>>:: ""$$rc/$)zHandle empty data case.rYr!s rhandle_empty_dataLabelManager.handle_empty_dataHs rcXR;aEX RU;a3URUURn[U[5(aU$g)zFRetrieves the fitted LabelEncoder for a specific data and label field.N)rirbr r)r"rnrorbs r get_encoderLabelManager.get_encoderLsH  %+y9Q*QmmI.{;CCG'<00r)riNrN)rnr ror rprrOrR)rnr ror rvrRrOr)rprrOr[)rprrir[rOrR)rvrRrir[rOrR)rwrRrir[rOr)rOz list[Any])rnr ror rOra)rSrTrUrVrWr#rqrxrlrmrtrurrrXrYrrreres6 O7:: 6 I # %rre)r rrOztuple[int, Real | str])rzset[Any]rOztuple[list[Real], list[str]])rW __future__r collectionsrcollections.abcr dataclassesrnumbersrtypingrnumpyr-rrrr[rerYrrrs^##$!C$)TTn (( (BBr