labelfree 1.0.0__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (42) hide show
  1. labelfree-1.0.0/.github/workflows/docs.yml +39 -0
  2. labelfree-1.0.0/.github/workflows/release.yml +45 -0
  3. labelfree-1.0.0/.github/workflows/tests.yml +37 -0
  4. labelfree-1.0.0/.gitignore +179 -0
  5. labelfree-1.0.0/.pre-commit-config.yaml +7 -0
  6. labelfree-1.0.0/CHANGELOG.md +9 -0
  7. labelfree-1.0.0/CITATION.cff +12 -0
  8. labelfree-1.0.0/CONTRIBUTING.md +13 -0
  9. labelfree-1.0.0/LICENSE +21 -0
  10. labelfree-1.0.0/PKG-INFO +82 -0
  11. labelfree-1.0.0/README.md +46 -0
  12. labelfree-1.0.0/SECURITY.md +8 -0
  13. labelfree-1.0.0/docs/api.md +133 -0
  14. labelfree-1.0.0/docs/examples.md +59 -0
  15. labelfree-1.0.0/docs/index.md +47 -0
  16. labelfree-1.0.0/docs/metrics.md +66 -0
  17. labelfree-1.0.0/examples/metric_signal_analysis.ipynb +1114 -0
  18. labelfree-1.0.0/examples/metric_tuning.ipynb +452 -0
  19. labelfree-1.0.0/labelfree/__init__.py +5 -0
  20. labelfree-1.0.0/labelfree/metrics/__init__.py +48 -0
  21. labelfree-1.0.0/labelfree/metrics/asoi.py +125 -0
  22. labelfree-1.0.0/labelfree/metrics/autouad.py +94 -0
  23. labelfree-1.0.0/labelfree/metrics/consensus.py +95 -0
  24. labelfree-1.0.0/labelfree/metrics/ireos.py +356 -0
  25. labelfree-1.0.0/labelfree/metrics/laplacian.py +55 -0
  26. labelfree-1.0.0/labelfree/metrics/mass_volume.py +147 -0
  27. labelfree-1.0.0/labelfree/metrics/score_cluster.py +57 -0
  28. labelfree-1.0.0/labelfree/metrics/sireos.py +66 -0
  29. labelfree-1.0.0/labelfree/metrics/stability.py +126 -0
  30. labelfree-1.0.0/labelfree/utils/__init__.py +1 -0
  31. labelfree-1.0.0/labelfree/utils/validation.py +77 -0
  32. labelfree-1.0.0/pyproject.toml +52 -0
  33. labelfree-1.0.0/tests/test_asoi.py +47 -0
  34. labelfree-1.0.0/tests/test_autouad.py +67 -0
  35. labelfree-1.0.0/tests/test_consensus.py +80 -0
  36. labelfree-1.0.0/tests/test_ireos.py +180 -0
  37. labelfree-1.0.0/tests/test_laplacian.py +49 -0
  38. labelfree-1.0.0/tests/test_mass_volume.py +114 -0
  39. labelfree-1.0.0/tests/test_score_cluster.py +67 -0
  40. labelfree-1.0.0/tests/test_sireos.py +31 -0
  41. labelfree-1.0.0/tests/test_stability.py +69 -0
  42. labelfree-1.0.0/zensical.toml +28 -0
@@ -0,0 +1,39 @@
1
+ name: Documentation
2
+
3
+ on:
4
+ push:
5
+ branches: [main]
6
+ paths:
7
+ - "docs/**"
8
+ - "zensical.toml"
9
+ - ".github/workflows/docs.yml"
10
+ workflow_dispatch:
11
+
12
+ permissions:
13
+ contents: read
14
+ pages: write
15
+ id-token: write
16
+
17
+ concurrency:
18
+ group: pages
19
+ cancel-in-progress: true
20
+
21
+ jobs:
22
+ deploy:
23
+ environment:
24
+ name: github-pages
25
+ url: ${{ steps.deployment.outputs.page_url }}
26
+ runs-on: ubuntu-latest
27
+ steps:
28
+ - uses: actions/configure-pages@v5
29
+ - uses: actions/checkout@v5
30
+ - uses: actions/setup-python@v5
31
+ with:
32
+ python-version: "3.12"
33
+ - run: python -m pip install "zensical==0.0.47"
34
+ - run: zensical build --clean
35
+ - uses: actions/upload-pages-artifact@v4
36
+ with:
37
+ path: site
38
+ - uses: actions/deploy-pages@v4
39
+ id: deployment
@@ -0,0 +1,45 @@
1
+ name: Release
2
+
3
+ on:
4
+ release:
5
+ types: [published]
6
+
7
+ permissions:
8
+ contents: read
9
+
10
+ jobs:
11
+ build:
12
+ runs-on: ubuntu-latest
13
+ steps:
14
+ - uses: actions/checkout@v5
15
+ - uses: actions/setup-python@v6
16
+ with:
17
+ python-version: "3.12"
18
+ - run: python -m pip install ".[dev]" build twine
19
+ - name: Verify release tag
20
+ env:
21
+ RELEASE_TAG: ${{ github.event.release.tag_name }}
22
+ run: >-
23
+ python -c "import importlib.metadata as metadata, labelfree, os;
24
+ expected = os.environ['RELEASE_TAG'];
25
+ assert expected == f'v{labelfree.__version__}' == 'v' + metadata.version('labelfree')"
26
+ - run: pytest
27
+ - run: python -m build
28
+ - run: twine check dist/*
29
+ - uses: actions/upload-artifact@v4
30
+ with:
31
+ name: distributions
32
+ path: dist/
33
+
34
+ publish:
35
+ needs: build
36
+ runs-on: ubuntu-latest
37
+ environment: pypi
38
+ permissions:
39
+ id-token: write
40
+ steps:
41
+ - uses: actions/download-artifact@v4
42
+ with:
43
+ name: distributions
44
+ path: dist/
45
+ - uses: pypa/gh-action-pypi-publish@release/v1
@@ -0,0 +1,37 @@
1
+ name: Tests
2
+
3
+ on:
4
+ push:
5
+ pull_request:
6
+
7
+ permissions:
8
+ contents: read
9
+
10
+ jobs:
11
+ lint:
12
+ runs-on: ubuntu-latest
13
+ steps:
14
+ - uses: actions/checkout@v5
15
+ - uses: actions/setup-python@v6
16
+ with:
17
+ python-version: "3.12"
18
+ cache: pip
19
+ cache-dependency-path: pyproject.toml
20
+ - run: python -m pip install ".[dev]"
21
+ - run: pre-commit run --all-files --show-diff-on-failure
22
+
23
+ test:
24
+ runs-on: ubuntu-latest
25
+ strategy:
26
+ fail-fast: false
27
+ matrix:
28
+ python-version: ["3.12", "3.13", "3.14"]
29
+ steps:
30
+ - uses: actions/checkout@v5
31
+ - uses: actions/setup-python@v6
32
+ with:
33
+ python-version: ${{ matrix.python-version }}
34
+ cache: pip
35
+ cache-dependency-path: pyproject.toml
36
+ - run: python -m pip install ".[dev]"
37
+ - run: pytest
@@ -0,0 +1,179 @@
1
+ # Project
2
+ .idea
3
+ .claude
4
+ CLAUDE.md
5
+
6
+ # Byte-compiled / optimized / DLL files
7
+ __pycache__/
8
+ *.py[cod]
9
+ *$py.class
10
+
11
+ # C extensions
12
+ *.so
13
+
14
+ # Distribution / packaging
15
+ .Python
16
+ build/
17
+ develop-eggs/
18
+ dist/
19
+ downloads/
20
+ eggs/
21
+ .eggs/
22
+ lib/
23
+ lib64/
24
+ parts/
25
+ sdist/
26
+ var/
27
+ wheels/
28
+ share/python-wheels/
29
+ *.egg-info/
30
+ .installed.cfg
31
+ *.egg
32
+ MANIFEST
33
+
34
+ # PyInstaller
35
+ # Usually these files are written by a python script from a template
36
+ # before PyInstaller builds the exe, so as to inject date/other infos into it.
37
+ *.manifest
38
+ *.spec
39
+
40
+ # Installer logs
41
+ pip-log.txt
42
+ pip-delete-this-directory.txt
43
+
44
+ # Unit test / coverage reports
45
+ htmlcov/
46
+ .tox/
47
+ .nox/
48
+ .coverage
49
+ .coverage.*
50
+ .cache
51
+ nosetests.xml
52
+ coverage.xml
53
+ *.cover
54
+ *.py,cover
55
+ .hypothesis/
56
+ .pytest_cache/
57
+ cover/
58
+
59
+ # Translations
60
+ *.mo
61
+ *.pot
62
+
63
+ # Django stuff:
64
+ *.log
65
+ local_settings.py
66
+ db.sqlite3
67
+ db.sqlite3-journal
68
+
69
+ # Flask stuff:
70
+ instance/
71
+ .webassets-cache
72
+
73
+ # Scrapy stuff:
74
+ .scrapy
75
+
76
+ # Sphinx documentation
77
+ docs/_build/
78
+
79
+ # PyBuilder
80
+ .pybuilder/
81
+ target/
82
+
83
+ # Jupyter Notebook
84
+ .ipynb_checkpoints
85
+
86
+ # IPython
87
+ profile_default/
88
+ ipython_config.py
89
+
90
+ # pyenv
91
+ # For a library or package, you might want to ignore these files since the code is
92
+ # intended to run in multiple environments; otherwise, check them in:
93
+ # .python-version
94
+
95
+ # pipenv
96
+ # According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
97
+ # However, in case of collaboration, if having platform-specific dependencies or dependencies
98
+ # having no cross-platform support, pipenv may install dependencies that don't work, or not
99
+ # install all needed dependencies.
100
+ #Pipfile.lock
101
+
102
+ # UV
103
+ # Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control.
104
+ # This is especially recommended for binary packages to ensure reproducibility, and is more
105
+ # commonly ignored for libraries.
106
+ uv.lock
107
+
108
+ # poetry
109
+ # Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
110
+ # This is especially recommended for binary packages to ensure reproducibility, and is more
111
+ # commonly ignored for libraries.
112
+ # https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
113
+ #poetry.lock
114
+
115
+ # pdm
116
+ # Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
117
+ #pdm.lock
118
+ # pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
119
+ # in version control.
120
+ # https://pdm.fming.dev/latest/usage/project/#working-with-version-control
121
+ .pdm.toml
122
+ .pdm-python
123
+ .pdm-build/
124
+
125
+ # PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
126
+ __pypackages__/
127
+
128
+ # Celery stuff
129
+ celerybeat-schedule
130
+ celerybeat.pid
131
+
132
+ # SageMath parsed files
133
+ *.sage.py
134
+
135
+ # Environments
136
+ .env
137
+ .venv
138
+ env/
139
+ venv/
140
+ ENV/
141
+ env.bak/
142
+ venv.bak/
143
+
144
+ # Spyder project settings
145
+ .spyderproject
146
+ .spyproject
147
+
148
+ # Rope project settings
149
+ .ropeproject
150
+
151
+ # mkdocs documentation
152
+ /site
153
+
154
+ # mypy
155
+ .mypy_cache/
156
+ .dmypy.json
157
+ dmypy.json
158
+
159
+ # Pyre type checker
160
+ .pyre/
161
+
162
+ # pytype static type analyzer
163
+ .pytype/
164
+
165
+ # Cython debug symbols
166
+ cython_debug/
167
+
168
+ # PyCharm
169
+ # JetBrains specific template is maintained in a separate JetBrains.gitignore that can
170
+ # be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
171
+ # and can be added to the global gitignore or merged into this file. For a more nuclear
172
+ # option (not recommended) you can uncomment the following to ignore the entire idea folder.
173
+ #.idea/
174
+
175
+ # Ruff stuff:
176
+ .ruff_cache/
177
+
178
+ # PyPI configuration file
179
+ .pypirc
@@ -0,0 +1,7 @@
1
+ repos:
2
+ - repo: https://github.com/astral-sh/ruff-pre-commit
3
+ rev: v0.15.21
4
+ hooks:
5
+ - id: ruff-check
6
+ args: [--fix]
7
+ - id: ruff-format
@@ -0,0 +1,9 @@
1
+ # Changelog
2
+
3
+ ## 1.0.0 - 2026-07-10
4
+
5
+ - First stable functional API for label-free anomaly-detector metrics.
6
+ - Source-checked implementations with explicit score polarity and metric direction.
7
+ - Zensical documentation, tests, examples, CI, and trusted PyPI publishing.
8
+ - Faithful relative IREOS with kernel-logistic separability and adaptive
9
+ integration, replacing the former distance heuristic.
@@ -0,0 +1,12 @@
1
+ cff-version: 1.2.0
2
+ message: "If you use labelfree, please cite this software."
3
+ title: "labelfree"
4
+ type: software
5
+ authors:
6
+ - family-names: "Hennhoefer"
7
+ given-names: "Oliver"
8
+ version: 1.0.0
9
+ date-released: 2026-07-10
10
+ repository-code: "https://github.com/OliverHennhoefer/label-free"
11
+ url: "https://oliverhennhoefer.github.io/label-free/"
12
+ license: MIT
@@ -0,0 +1,13 @@
1
+ # Contributing
2
+
3
+ Open an issue before large changes. For a focused fix:
4
+
5
+ ```bash
6
+ python -m pip install -e ".[dev]"
7
+ pre-commit install
8
+ pre-commit run --all-files
9
+ pytest
10
+ ```
11
+
12
+ Keep pull requests small, add one regression test for changed behavior, and cite
13
+ the paper or author implementation when changing a published metric.
@@ -0,0 +1,21 @@
1
+ MIT License
2
+
3
+ Copyright (c) 2025 Oliver Hennhöfer
4
+
5
+ Permission is hereby granted, free of charge, to any person obtaining a copy
6
+ of this software and associated documentation files (the "Software"), to deal
7
+ in the Software without restriction, including without limitation the rights
8
+ to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
9
+ copies of the Software, and to permit persons to whom the Software is
10
+ furnished to do so, subject to the following conditions:
11
+
12
+ The above copyright notice and this permission notice shall be included in all
13
+ copies or substantial portions of the Software.
14
+
15
+ THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16
+ IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17
+ FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18
+ AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19
+ LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20
+ OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21
+ SOFTWARE.
@@ -0,0 +1,82 @@
1
+ Metadata-Version: 2.4
2
+ Name: labelfree
3
+ Version: 1.0.0
4
+ Summary: Label-free metrics for unsupervised anomaly detector evaluation
5
+ Project-URL: Documentation, https://oliverhennhoefer.github.io/label-free/
6
+ Project-URL: Issues, https://github.com/OliverHennhoefer/label-free/issues
7
+ Project-URL: Repository, https://github.com/OliverHennhoefer/label-free
8
+ Author-email: Oliver Hennhoefer <oliver.hennhoefer@mail.de>
9
+ License-Expression: MIT
10
+ License-File: LICENSE
11
+ Keywords: anomaly-detection,model-selection,unsupervised-learning
12
+ Classifier: Development Status :: 5 - Production/Stable
13
+ Classifier: Intended Audience :: Science/Research
14
+ Classifier: Programming Language :: Python :: 3
15
+ Classifier: Programming Language :: Python :: 3.12
16
+ Classifier: Programming Language :: Python :: 3.13
17
+ Classifier: Programming Language :: Python :: 3.14
18
+ Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
19
+ Requires-Python: >=3.12
20
+ Requires-Dist: numpy>=1.24
21
+ Requires-Dist: scikit-learn>=1.3
22
+ Requires-Dist: scipy>=1.10
23
+ Provides-Extra: dev
24
+ Requires-Dist: pre-commit>=4.2; extra == 'dev'
25
+ Requires-Dist: pytest>=8; extra == 'dev'
26
+ Requires-Dist: ruff>=0.12; extra == 'dev'
27
+ Provides-Extra: docs
28
+ Requires-Dist: zensical==0.0.47; extra == 'docs'
29
+ Provides-Extra: examples
30
+ Requires-Dist: numba>=0.66; extra == 'examples'
31
+ Requires-Dist: oddball>=1.4; extra == 'examples'
32
+ Requires-Dist: optuna>=4; extra == 'examples'
33
+ Requires-Dist: pandas>=2.2; extra == 'examples'
34
+ Requires-Dist: pyod>=1.1; extra == 'examples'
35
+ Description-Content-Type: text/markdown
36
+
37
+ # labelfree
38
+
39
+ [![Tests](https://github.com/OliverHennhoefer/label-free/actions/workflows/tests.yml/badge.svg)](https://github.com/OliverHennhoefer/label-free/actions/workflows/tests.yml)
40
+ [![Documentation](https://github.com/OliverHennhoefer/label-free/actions/workflows/docs.yml/badge.svg)](https://oliverhennhoefer.github.io/label-free/)
41
+ [![PyPI](https://img.shields.io/pypi/v/labelfree.svg)](https://pypi.org/project/labelfree/)
42
+
43
+ Label-free metrics for comparing unsupervised anomaly detectors and
44
+ hyperparameters when labeled anomalies are unavailable.
45
+
46
+ ## Install
47
+
48
+ ```bash
49
+ pip install labelfree
50
+ ```
51
+
52
+ ## Quick start
53
+
54
+ ```python
55
+ from labelfree.metrics import score_cluster_metrics
56
+
57
+ scores = [0.1, 0.2, 0.3, 3.8, 4.1]
58
+ metrics = score_cluster_metrics(scores, n_outliers=2)
59
+ print(metrics["silhouette"]) # higher is better
60
+ ```
61
+
62
+ The package includes score-cluster, ASI/ASOI, AutoUAD, Laplacian,
63
+ consensus, ranking-stability, IREOS, SIREOS, and Excess-Mass/Mass-Volume metrics.
64
+ Every raw-score API requires an explicit score polarity when larger values
65
+ mean more normal.
66
+
67
+ Label-free metrics are model-selection signals, not substitutes for labeled
68
+ evaluation. Check each metric's assumptions and direction before comparing
69
+ results.
70
+
71
+ See the [documentation](https://oliverhennhoefer.github.io/label-free/) for
72
+ metric selection, API details, and examples.
73
+
74
+ ## Development
75
+
76
+ ```bash
77
+ python -m pip install -e ".[dev]"
78
+ pytest
79
+ ```
80
+
81
+ Contributions are welcome; see [CONTRIBUTING.md](CONTRIBUTING.md). Released
82
+ under the [MIT License](LICENSE).
@@ -0,0 +1,46 @@
1
+ # labelfree
2
+
3
+ [![Tests](https://github.com/OliverHennhoefer/label-free/actions/workflows/tests.yml/badge.svg)](https://github.com/OliverHennhoefer/label-free/actions/workflows/tests.yml)
4
+ [![Documentation](https://github.com/OliverHennhoefer/label-free/actions/workflows/docs.yml/badge.svg)](https://oliverhennhoefer.github.io/label-free/)
5
+ [![PyPI](https://img.shields.io/pypi/v/labelfree.svg)](https://pypi.org/project/labelfree/)
6
+
7
+ Label-free metrics for comparing unsupervised anomaly detectors and
8
+ hyperparameters when labeled anomalies are unavailable.
9
+
10
+ ## Install
11
+
12
+ ```bash
13
+ pip install labelfree
14
+ ```
15
+
16
+ ## Quick start
17
+
18
+ ```python
19
+ from labelfree.metrics import score_cluster_metrics
20
+
21
+ scores = [0.1, 0.2, 0.3, 3.8, 4.1]
22
+ metrics = score_cluster_metrics(scores, n_outliers=2)
23
+ print(metrics["silhouette"]) # higher is better
24
+ ```
25
+
26
+ The package includes score-cluster, ASI/ASOI, AutoUAD, Laplacian,
27
+ consensus, ranking-stability, IREOS, SIREOS, and Excess-Mass/Mass-Volume metrics.
28
+ Every raw-score API requires an explicit score polarity when larger values
29
+ mean more normal.
30
+
31
+ Label-free metrics are model-selection signals, not substitutes for labeled
32
+ evaluation. Check each metric's assumptions and direction before comparing
33
+ results.
34
+
35
+ See the [documentation](https://oliverhennhoefer.github.io/label-free/) for
36
+ metric selection, API details, and examples.
37
+
38
+ ## Development
39
+
40
+ ```bash
41
+ python -m pip install -e ".[dev]"
42
+ pytest
43
+ ```
44
+
45
+ Contributions are welcome; see [CONTRIBUTING.md](CONTRIBUTING.md). Released
46
+ under the [MIT License](LICENSE).
@@ -0,0 +1,8 @@
1
+ # Security policy
2
+
3
+ Security fixes are supported for the latest 1.x release.
4
+
5
+ Please report vulnerabilities privately through GitHub's security advisory
6
+ form. If private reporting is unavailable, email
7
+ `oliver.hennhoefer@mail.de`. Do not open a public issue before a fix is
8
+ available.
@@ -0,0 +1,133 @@
1
+ # API reference
2
+
3
+ Import public functions from `labelfree.metrics`.
4
+
5
+ ## Partition metrics
6
+
7
+ ```python
8
+ score_cluster_metrics(
9
+ scores, *, n_outliers=None, contamination=None,
10
+ score_polarity="higher_is_anomalous"
11
+ ) -> dict[str, float]
12
+
13
+ asi_score(
14
+ X, scores, *, n_outliers=None, contamination=None,
15
+ score_polarity="higher_is_anomalous"
16
+ ) -> float
17
+
18
+ asoi_score(
19
+ X, scores, *, n_outliers=None, contamination=None,
20
+ score_polarity="higher_is_anomalous",
21
+ alpha=0.5314, beta=0.4686
22
+ ) -> float
23
+ ```
24
+
25
+ Pass exactly one of `n_outliers` and `contamination`. Contamination uses
26
+ `ceil(n_samples * contamination)`.
27
+
28
+ ## Score-distribution metrics
29
+
30
+ ```python
31
+ relative_top_median_score(
32
+ scores, *, top_fraction=0.05,
33
+ score_polarity="higher_is_anomalous", eps=1e-6
34
+ ) -> float
35
+
36
+ expected_anomaly_gap_score(
37
+ scores, *, top_fraction=0.2,
38
+ score_polarity="higher_is_anomalous", eps=1e-6
39
+ ) -> float
40
+
41
+ normalized_pseudo_discrepancy_score(
42
+ validation_scores, generated_scores, *,
43
+ score_polarity="higher_is_anomalous", eps=1e-6
44
+ ) -> float
45
+ ```
46
+
47
+ All three return higher-is-better scalars.
48
+
49
+ ## Feature-space metrics
50
+
51
+ ```python
52
+ laplacian_score(
53
+ X, scores, *, n_neighbors=5,
54
+ score_polarity="higher_is_anomalous"
55
+ ) -> float
56
+
57
+ sireos_score(
58
+ X, scores, *, score_polarity="higher_is_anomalous",
59
+ kernel_width=None, kernel_quantile=0.01
60
+ ) -> float
61
+
62
+ ireos_score(
63
+ X, outlier_probabilities, *, gamma_max=None,
64
+ max_clump_size=1, penalty_cost=100.0,
65
+ integration_tol=0.005
66
+ ) -> float
67
+
68
+ ireos_scores(
69
+ X, outlier_probability_matrix, *, gamma_max=None,
70
+ max_clump_size=1, penalty_cost=100.0,
71
+ integration_tol=0.005
72
+ ) -> np.ndarray
73
+ ```
74
+
75
+ IREOS is higher-is-better; SIREOS and Laplacian Score are lower-is-better.
76
+ IREOS requires detector-specific outlier probabilities in `[0, 1]`, not raw
77
+ scores. The plural function uses one shared automatically selected `gamma_max`
78
+ for all rows. Automatic selection requires at least one probability above
79
+ `0.5`; otherwise, pass `gamma_max` explicitly. If `kernel_width` is omitted,
80
+ SIREOS uses the requested quantile
81
+ of nonzero pairwise distances.
82
+
83
+ ## Candidate consensus
84
+
85
+ ```python
86
+ model_centrality_scores(score_matrix, *, score_polarity="higher_is_anomalous")
87
+ average_rank_consensus_scores(score_matrix, *, score_polarity="higher_is_anomalous")
88
+ hits_model_scores(
89
+ score_matrix, *, score_polarity="higher_is_anomalous",
90
+ max_iter=100, tol=1e-10
91
+ )
92
+ ```
93
+
94
+ `score_matrix` has shape `(n_models, n_samples)`. Each function returns one
95
+ higher-is-better value per model.
96
+
97
+ ## Ranking stability
98
+
99
+ ```python
100
+ ranking_stability_score(
101
+ score_matrix, *, contamination, psi=0.8,
102
+ score_polarity="higher_is_anomalous"
103
+ ) -> float
104
+
105
+ top_k_stability_score(
106
+ score_matrix, *, top_k=None, top_fraction=None,
107
+ score_polarity="higher_is_anomalous"
108
+ ) -> float
109
+ ```
110
+
111
+ `score_matrix` has shape `(n_runs, n_samples)`. Pass exactly one of
112
+ `top_k` and `top_fraction` to top-k stability.
113
+
114
+ ## Excess-Mass and Mass-Volume
115
+
116
+ ```python
117
+ excess_mass_curve(scores, reference_scores, *, support_volume, ...)
118
+ excess_mass_auc(scores, reference_scores, *, support_volume, ...)
119
+ mass_volume_curve(scores, reference_scores, *, support_volume, ...)
120
+ mass_volume_auc(scores, reference_scores, *, support_volume, ...)
121
+ bounding_box_volume(X, *, offset=1e-12) -> float
122
+ ```
123
+
124
+ Curve functions return `(axis, values)`. `reference_scores` are detector
125
+ scores on uniform samples from the support whose Lebesgue volume is
126
+ `support_volume`. `bounding_box_volume` is a convenience for an
127
+ axis-aligned support; it does not generate the reference samples.
128
+
129
+ ## Validation
130
+
131
+ Inputs are converted to finite floating-point arrays. Invalid dimensions,
132
+ non-finite values, inconsistent sample counts, invalid fractions, and unknown
133
+ score polarities raise `ValueError`.