24912462866

Committed 24 Apr 2026 09:21PM UTC coverage: 76.362% (-3.1%) from 79.464%

Build # 24912462866

Build Type

push

github

Committed by

web-flow

Commit Message

Merge pull request #73 from rafaelpadilla/eduardo/refactor_data_operations

Refactor of data operations, trainers and models.

Coverage Stats

244 of 339 branches covered (71.98%)

Branch coverage included in aggregate %.

1317 of 1706 new or added lines in 50 files covered. (77.2%)

28 existing lines in 5 files now uncovered.

2124 of 2762 relevant lines covered (76.9%)

0.77 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

97.3

/toolkit/ThreeWToolkit/preprocessing/remap.py

from pydantic import Field, PrivateAttr
from ..core.base_dataset import BaseDataset
from ..core.base_preprocessing import BasePreprocessing, BasePreprocessingConfig
from ..core.dataset_outputs import DatasetOutputs


class RemapClassConfig(BasePreprocessingConfig):
    """Configuration for remapping class labels to integers starting from 0."""

    class_map: dict | None = Field(
        default=None,
        description="Optional mapping from original class labels to new labels. Auto-generated from data if None.",
    )
    _target: type = PrivateAttr(default_factory=lambda: RemapClass)


class RemapClass(BasePreprocessing):
    """
    Preprocessing step to remap class/label values according to a provided mapping.
    If class_map is not provided, it will be generated in fit() by collecting all
    unique classes across events. Stores the original mapping for convenience.
    """

    def __init__(self, config: RemapClassConfig):
        """Initialize the RemapClass preprocessing step with the given configuration.

        Args:
            config: RemapClassConfig object containing the configuration for this preprocessing step.
        """
        self.config: RemapClassConfig = config

    def fit(self, data: BaseDataset) -> None:
        """
        Collect all unique classes from the label Series if class_map is not provided.
        If class_map is provided, this method will check that all classes in the data are present in the class_map.

        Args:
            data: BaseDataset object from which to collect unique class labels if class_map is not provided.
        """
        if self.config.class_map is not None:
            self.class_map = self.config.class_map

            for event in data:
                _ = self.transform(event)  # check that all labels can be mapped
            return

        unique_classes: set[int] = set()
        # collect unique classes across all events
        for event in data:
            if event.label is not None:
                unique_classes.update(event.label.dropna().unique())
        self.class_map = {c: i for i, c in enumerate(sorted(unique_classes))}

    def transform(self, data: DatasetOutputs) -> DatasetOutputs:
        """
        Remap class labels according to the class_map.

        Args:
            data: DatasetOutputs object with label field

        Returns:
            DatasetOutputs: Transformed data with remapped labels
        """
        if self.class_map is None:
            raise ValueError("RemapClass: class_map is not set. Call fit first.")

        mapped_label = data.label
        if data.label is not None:
            mapped_label = data.label.map(self.class_map)
            if mapped_label.isna().any():
                # If there are any labels that were not in the class_map, they will be NaN after mapping.
                raise ValueError("Some labels were not in the class_map.")

        return DatasetOutputs(
            signal=data.signal,
            label=mapped_label,
            metadata=data.metadata.copy(),
        )

1	from pydantic import Field, PrivateAttr	1✔
2	from ..core.base_dataset import BaseDataset	1✔
3	from ..core.base_preprocessing import BasePreprocessing, BasePreprocessingConfig	1✔
4	from ..core.dataset_outputs import DatasetOutputs	1✔
5
6
7	class RemapClassConfig(BasePreprocessingConfig):	1✔
8	"""Configuration for remapping class labels to integers starting from 0."""
9
10	class_map: dict \| None = Field(	1✔
11	default=None,
12	description="Optional mapping from original class labels to new labels. Auto-generated from data if None.",
13	)
14	_target: type = PrivateAttr(default_factory=lambda: RemapClass)	1✔
15
16
17	class RemapClass(BasePreprocessing):	1✔
18	"""
19	Preprocessing step to remap class/label values according to a provided mapping.
20	If class_map is not provided, it will be generated in fit() by collecting all
21	unique classes across events. Stores the original mapping for convenience.
22	"""
23
24	def __init__(self, config: RemapClassConfig):	1✔
25	"""Initialize the RemapClass preprocessing step with the given configuration.
26
27	Args:
28	config: RemapClassConfig object containing the configuration for this preprocessing step.
29	"""
30	self.config: RemapClassConfig = config	1✔
31
32	def fit(self, data: BaseDataset) -> None:	1✔
33	"""
34	Collect all unique classes from the label Series if class_map is not provided.
35	If class_map is provided, this method will check that all classes in the data are present in the class_map.
36
37	Args:
38	data: BaseDataset object from which to collect unique class labels if class_map is not provided.
39	"""
40	if self.config.class_map is not None:	1✔
41	self.class_map = self.config.class_map	1✔
42
43	for event in data:	1✔
44	_ = self.transform(event) # check that all labels can be mapped	1✔
45	return	1✔
46
47	unique_classes: set[int] = set()	1✔
48	# collect unique classes across all events
49	for event in data:	1✔
50	if event.label is not None:	1✔
51	unique_classes.update(event.label.dropna().unique())	1✔
52	self.class_map = {c: i for i, c in enumerate(sorted(unique_classes))}	1✔
53
54	def transform(self, data: DatasetOutputs) -> DatasetOutputs:	1✔
55	"""
56	Remap class labels according to the class_map.
57
58	Args:
59	data: DatasetOutputs object with label field
60
61	Returns:
62	DatasetOutputs: Transformed data with remapped labels
63	"""
64	if self.class_map is None:	1✔
NEW 65	raise ValueError("RemapClass: class_map is not set. Call fit first.")	×
66
67	mapped_label = data.label	1✔
68	if data.label is not None:	1✔
69	mapped_label = data.label.map(self.class_map)	1✔
70	if mapped_label.isna().any():	1✔
71	# If there are any labels that were not in the class_map, they will be NaN after mapping.
72	raise ValueError("Some labels were not in the class_map.")	1✔
73
74	return DatasetOutputs(	1✔
75	signal=data.signal,
76	label=mapped_label,
77	metadata=data.metadata.copy(),
78	)

rafaelpadilla / 3W / 24912462866

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous