multimodel-analysis 0.2.0__tar.gz → 0.3.0__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 (25) hide show
  1. multimodel_analysis-0.3.0/CHANGELOG.md +48 -0
  2. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/MANIFEST.in +1 -0
  3. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/PKG-INFO +1 -1
  4. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis/__init__.py +1 -1
  5. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis/core/base.py +467 -377
  6. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis/core/registry.py +94 -90
  7. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis/supervised/classification.py +21 -8
  8. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis/supervised/regression.py +22 -9
  9. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis.egg-info/PKG-INFO +1 -1
  10. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis.egg-info/SOURCES.txt +1 -0
  11. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/LICENSE +0 -0
  12. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis/core/__init__.py +0 -0
  13. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis/core/persistence.py +0 -0
  14. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis/core/preprocessing.py +0 -0
  15. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis/core/utils.py +0 -0
  16. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis/multimodel.py +0 -0
  17. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis/supervised/__init__.py +0 -0
  18. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis.egg-info/dependency_links.txt +0 -0
  19. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis.egg-info/requires.txt +0 -0
  20. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/multimodel_analysis.egg-info/top_level.txt +0 -0
  21. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/pyproject.toml +0 -0
  22. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/readme.md +0 -0
  23. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/requirements.txt +0 -0
  24. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/setup.cfg +0 -0
  25. {multimodel_analysis-0.2.0 → multimodel_analysis-0.3.0}/setup.py +0 -0
@@ -0,0 +1,48 @@
1
+ # Changelog
2
+
3
+ ## 0.3.0 - Phase 4: hyper-parameter tuning
4
+
5
+ - `tune_models(include=, exclude=, search='random'|'grid'|'halving', n_iter=20, cv=5, scoring=, param_grids=)`: tunes each model on the training split with leak-free CV (preprocessing is refitted inside every fold) and returns the same report tuples as `run_all_models`, so `show_tabular_report`, `predict`, plots and `save_model` work on tuned models.
6
+ - Every built-in classifier and regressor ships a default `search_space` (`ModelSpec.search_space`); pass `param_grids={'knn': {...}}` to override one.
7
+ - New attributes `best_params_` (per model) and `tuning_results_` (best CV score, parameters, candidates tried, search time). Failures go to `failed_models_`; `n_iter` is capped at the grid size.
8
+ - New notebook `examples/phase4_tuning.ipynb` with executed output.
9
+ - Repo cleanup: removed stale example notebooks/CSV and an old PyPI screenshot; `.gitignore` and `MANIFEST.in` tidied.
10
+
11
+ ## 0.2.0 - restructure, registry, preprocessing, prediction, cross-validation
12
+
13
+ ### Phase 3: preprocessing, prediction, persistence, cross-validation
14
+
15
+ - `preprocess=True` (classifier and regressor): impute missing values, one-hot encode text columns and scale, fitted on the training split only. Default stays strict.
16
+ - `predict(X_new, model=None)`, `export_model()`, `save_model(path)` and `load_model(path)` (joblib); a saved `TrainedModel` carries its preprocessing and returns original class labels.
17
+ - `cross_validate_models(cv=5, scoring=..., include=, exclude=, custom_models=)`: leak-free CV on the training split, results in `cv_results_`.
18
+ - `class_weight='balanced'` for classification (Logistic Regression, SVM, Decision Tree, Random Forest).
19
+ - `Fit Time (s)` column in both reports.
20
+
21
+ ### Phase 2: package restructure
22
+
23
+ - Code split into `multimodel_analysis/core` (base, registry, utils) and `multimodel_analysis/supervised` (classification, regression). `multimodel_analysis.multimodel` remains as a compatibility shim; all old imports and method names still work.
24
+ - Model registry (`ModelSpec`, `ModelRegistry`): every built-in model is one registry entry (key, display name, factory, defaults).
25
+ - New: `run_model(name, **params)`, `available_models()`, `get_model(name)`, `run_all_models(include=, exclude=)`, `estimators_` (fitted models) and `fit_times_`.
26
+
27
+ ## 0.1.4 - Phase 0/1: bug fixes and packaging
28
+
29
+ ### Behaviour changes (read these)
30
+ - `scaled_data` now defaults to `True` (features are standardised, fit on the training split only).
31
+ - `Support_vector_model` now defaults to `kernel='rbf'` (was `'linear'`).
32
+ - The `ROC AUC` column is numeric (`NaN` when unavailable) instead of the string `"N/A"`.
33
+ - `run_all_models()` no longer silences all warnings: they are stored in `warnings_` and the important ones are printed.
34
+ - Models that fail are listed in `failed_models_` and shown under the report table.
35
+ - `MultiModelRegressor.run_all_models()` now includes a KNN regressor (9 models); `SVR` standardises the target internally.
36
+
37
+ ### Fixed
38
+ - Confusion matrix / classification report crashed or had the wrong shape when the test split missed a class.
39
+ - Multiclass ROC AUC was `NaN` whenever one class was absent from the test split; it is now computed on the classes present.
40
+ - Continuous targets, NaN/inf or text columns in `X`, X/y length mismatch and invalid `test_size` now raise clear errors.
41
+ - Seaborn palette `FutureWarning` in `plot_comparison`.
42
+ - `LogisticRegression(n_jobs=...)` deprecation warning on scikit-learn >= 1.8.
43
+ - `.xlsx` export without `openpyxl` gives an install hint; legacy `.xls` is rejected clearly.
44
+ - Removed dead code (unused imports, broken `__main__` check).
45
+
46
+ ### Added
47
+ - `verbose`, `warnings_`, `failed_models_`, `sort_by=` in `show_tabular_report`.
48
+ - `pyproject.toml`, GitHub Actions CI (tests + ruff), pytest suite (25 tests), `examples/phase1_results.ipynb`.
@@ -1,6 +1,7 @@
1
1
  include readme.md
2
2
  include LICENSE
3
3
  include requirements.txt
4
+ include CHANGELOG.md
4
5
  prune dist
5
6
  prune examples
6
7
  prune tests
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: multimodel_analysis
3
- Version: 0.2.0
3
+ Version: 0.3.0
4
4
  Summary: A Python package for automatic multi-model analysis (classification, regression and clustering)
5
5
  Author-email: Uditya Narayan Tiwari <tiwarimerit@gmail.com>
6
6
  License: Apache-2.0
@@ -5,7 +5,7 @@ from .core.utils import save_report
5
5
  from .multimodel import MultiModelRegressior
6
6
  from .supervised import MultiModelClassifier, MultiModelRegressor
7
7
 
8
- __version__ = '0.2.0'
8
+ __version__ = '0.3.0'
9
9
  __author__ = 'Uditya Narayan Tiwari'
10
10
 
11
11
  __all__ = [