kfc-procedure 0.1.0__tar.gz

kfc_procedure-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 Ougi
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

kfc_procedure-0.1.0/PKG-INFO

@@ -0,0 +1,169 @@
+Metadata-Version: 2.4
+Name: kfc-procedure
+Version: 0.1.0
+Summary: Clusterwise predictive modeling library
+Author-email: OUGI POV <ougipov113@gmail.com>
+License-Expression: MIT
+Project-URL: Homepage, https://github.com/Ougi3ay/kfc-procedure
+Classifier: Programming Language :: Python :: 3
+Classifier: Operating System :: OS Independent
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.11
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: pandas
+Requires-Dist: numpy
+Requires-Dist: scikit-learn
+Requires-Dist: matplotlib
+Requires-Dist: xgboost
+Provides-Extra: core
+Requires-Dist: numpy; extra == "core"
+Requires-Dist: scikit-learn; extra == "core"
+Provides-Extra: cobra
+Requires-Dist: numba; extra == "cobra"
+Requires-Dist: faiss-cpu; extra == "cobra"
+Requires-Dist: xgboost; extra == "cobra"
+Requires-Dist: pandas; extra == "cobra"
+Requires-Dist: numpy; extra == "cobra"
+Requires-Dist: scikit-learn; extra == "cobra"
+Requires-Dist: matplotlib; extra == "cobra"
+Requires-Dist: plotly; extra == "cobra"
+Provides-Extra: dev
+Requires-Dist: pytest; extra == "dev"
+Requires-Dist: build; extra == "dev"
+Requires-Dist: twine; extra == "dev"
+Requires-Dist: jupyter; extra == "dev"
+Provides-Extra: all
+Requires-Dist: kfc-procedure[cobra]; extra == "all"
+Requires-Dist: kfc-procedure[dev]; extra == "all"
+Dynamic: license-file
+
+# KFC-Model: A Python Implementation of the KFC Procedure
+
+KFC-Model is a modular Python library for clusterwise predictive modeling using the KFC procedure (K-step, F-step, C-step). It combines multiple clustering divergences, local models, and aggregation strategies for regression and classification tasks.
+
+## Features
+
+- KFC meta-estimator for clusterwise learning
+- Modular `KStep`, `FStep`, and `CStep` components
+- Support for Bregman K-Means divergences
+- Local model factories for regression and classification
+- Aggregation strategies including mean, stacking, and GradientCOBRA
+- Easy extension with custom components
+
+## Installation
+
+Requirements:
+
+- Python 3.11+
+- `numpy`
+- `pandas`
+- `scikit-learn`
+- `xgboost`
+- `matplotlib`
+
+Create and activate a virtual environment:
+
+```bash
+python3 -m venv .venv
+source .venv/bin/activate
+```
+
+Install dependencies:
+
+```bash
+python3 -m pip install -r requirements.txt
+```
+
+Install the package locally:
+
+```bash
+python3 -m pip install -e .
+```
+
+## Quick Start
+
+```python
+import numpy as np
+from kfc_procedure.kfc import KFCRegressor, KFCClassifier
+
+# Example data
+X = np.random.randn(200, 5)
+y_reg = X[:, 0] * 2 + np.random.randn(200) * 0.1
+y_clf = (y_reg > 0).astype(int)
+
+# Regression example
+model = KFCRegressor(
+    divergences=["euclidean", "kl"],
+    local_model="linear",
+    aggregation="mean",
+    random_state=42,
+)
+model.fit(X, y_reg)
+y_pred = model.predict(X)
+
+# Classification example
+clf = KFCClassifier(
+    divergences=["euclidean"],
+    local_model="logistic",
+    aggregation="majority_vote",
+    random_state=42,
+)
+clf.fit(X, y_clf)
+y_pred_clf = clf.predict(X)
+proba = clf.predict_proba(X)
+```
+
+## Core Components
+
+- `KStep`: fits clustering models using one or more Bregman divergences
+- `FStep`: trains local models for each cluster and divergence
+- `CStep`: aggregates local predictions into final outputs
+- `KFCRegressor` / `KFCClassifier`: full meta-estimators exposing `fit`, `predict`, and `predict_proba`
+
+## Configuration
+
+### `divergences`
+
+The divergences parameter accepts:
+
+- a list of divergence names, e.g. `['euclidean', 'kl']`
+- a list of config dictionaries, e.g. `[{ 'name': 'euclidean', 'n_clusters': 4 }]`
+
+Available divergences: `'euclidean'`, `'kl'`, `'gkl'`, `'is'`, `'logistic'`
+
+### `local_model`
+
+The local_model parameter accepts:
+
+- a model name string, e.g. `'linear'`, `'ridge'`
+
+Supported regression models include: `linear`, `ridge`, `lasso`, `decision_tree`, `random_forest`.
+Supported classification models include: `logistic`, `decision_tree`, `random_forest`.
+
+### `aggregation`
+
+The aggregation parameter accepts:
+
+- an aggregation strategy name string, e.g. `'mean'`, `'stacking'`
+
+Supported aggregators:
+
+- Regression: `mean`, `weighted_mean`, `stacking`
+- Classification: `majority_vote`, `stacking`, `combine_classifier`
+
+## Project Structure
+
+- `src/kfc_procedure/`: main package code
+- `src/kfc_procedure/core/`: factories, clustering, ML wrappers, and aggregation strategies
+- `src/kfc_procedure/steps/`: KFC step implementations
+- `src/kfc_procedure/utils/`: resolution and validation helpers
+
+## Contributing
+
+Contributions, bug reports, and improvements are welcome. Use `pytest` for testing and follow the existing package layout for new components.
+
+## License
+
+MIT License
+

kfc_procedure-0.1.0/README.md

@@ -0,0 +1,129 @@
+# KFC-Model: A Python Implementation of the KFC Procedure
+
+KFC-Model is a modular Python library for clusterwise predictive modeling using the KFC procedure (K-step, F-step, C-step). It combines multiple clustering divergences, local models, and aggregation strategies for regression and classification tasks.
+
+## Features
+
+- KFC meta-estimator for clusterwise learning
+- Modular `KStep`, `FStep`, and `CStep` components
+- Support for Bregman K-Means divergences
+- Local model factories for regression and classification
+- Aggregation strategies including mean, stacking, and GradientCOBRA
+- Easy extension with custom components
+
+## Installation
+
+Requirements:
+
+- Python 3.11+
+- `numpy`
+- `pandas`
+- `scikit-learn`
+- `xgboost`
+- `matplotlib`
+
+Create and activate a virtual environment:
+
+```bash
+python3 -m venv .venv
+source .venv/bin/activate
+```
+
+Install dependencies:
+
+```bash
+python3 -m pip install -r requirements.txt
+```
+
+Install the package locally:
+
+```bash
+python3 -m pip install -e .
+```
+
+## Quick Start
+
+```python
+import numpy as np
+from kfc_procedure.kfc import KFCRegressor, KFCClassifier
+
+# Example data
+X = np.random.randn(200, 5)
+y_reg = X[:, 0] * 2 + np.random.randn(200) * 0.1
+y_clf = (y_reg > 0).astype(int)
+
+# Regression example
+model = KFCRegressor(
+    divergences=["euclidean", "kl"],
+    local_model="linear",
+    aggregation="mean",
+    random_state=42,
+)
+model.fit(X, y_reg)
+y_pred = model.predict(X)
+
+# Classification example
+clf = KFCClassifier(
+    divergences=["euclidean"],
+    local_model="logistic",
+    aggregation="majority_vote",
+    random_state=42,
+)
+clf.fit(X, y_clf)
+y_pred_clf = clf.predict(X)
+proba = clf.predict_proba(X)
+```
+
+## Core Components
+
+- `KStep`: fits clustering models using one or more Bregman divergences
+- `FStep`: trains local models for each cluster and divergence
+- `CStep`: aggregates local predictions into final outputs
+- `KFCRegressor` / `KFCClassifier`: full meta-estimators exposing `fit`, `predict`, and `predict_proba`
+
+## Configuration
+
+### `divergences`
+
+The divergences parameter accepts:
+
+- a list of divergence names, e.g. `['euclidean', 'kl']`
+- a list of config dictionaries, e.g. `[{ 'name': 'euclidean', 'n_clusters': 4 }]`
+
+Available divergences: `'euclidean'`, `'kl'`, `'gkl'`, `'is'`, `'logistic'`
+
+### `local_model`
+
+The local_model parameter accepts:
+
+- a model name string, e.g. `'linear'`, `'ridge'`
+
+Supported regression models include: `linear`, `ridge`, `lasso`, `decision_tree`, `random_forest`.
+Supported classification models include: `logistic`, `decision_tree`, `random_forest`.
+
+### `aggregation`
+
+The aggregation parameter accepts:
+
+- an aggregation strategy name string, e.g. `'mean'`, `'stacking'`
+
+Supported aggregators:
+
+- Regression: `mean`, `weighted_mean`, `stacking`
+- Classification: `majority_vote`, `stacking`, `combine_classifier`
+
+## Project Structure
+
+- `src/kfc_procedure/`: main package code
+- `src/kfc_procedure/core/`: factories, clustering, ML wrappers, and aggregation strategies
+- `src/kfc_procedure/steps/`: KFC step implementations
+- `src/kfc_procedure/utils/`: resolution and validation helpers
+
+## Contributing
+
+Contributions, bug reports, and improvements are welcome. Use `pytest` for testing and follow the existing package layout for new components.
+
+## License
+
+MIT License
+

kfc_procedure-0.1.0/pyproject.toml

@@ -0,0 +1,85 @@
+[build-system]
+requires = ["setuptools>=68", "wheel"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "kfc-procedure"
+version = "0.1.0"
+description = "Clusterwise predictive modeling library"
+readme = "README.md"
+requires-python = ">=3.11"
+license = "MIT"
+license-files = ["LICEN[CS]E*"]
+
+authors = [
+  { name = "OUGI POV", email = "ougipov113@gmail.com" }
+]
+
+dependencies = [
+  "pandas",
+  "numpy",
+  "scikit-learn",
+  "matplotlib",
+  "xgboost"
+]
+
+classifiers = [
+  "Programming Language :: Python :: 3",
+  "Operating System :: OS Independent",
+  "Topic :: Scientific/Engineering :: Artificial Intelligence"
+]
+
+[project.urls]
+Homepage = "https://github.com/Ougi3ay/kfc-procedure"
+
+# --------------------------
+# Optional dependencies (extras)
+# --------------------------
+[project.optional-dependencies]
+
+core = [
+  "numpy",
+  "scikit-learn"
+]
+
+cobra = [
+  "numba",
+  "faiss-cpu",
+  "xgboost",
+  "pandas",
+  "numpy",
+  "scikit-learn",
+  "matplotlib",
+  "plotly"
+]
+
+dev = [
+  "pytest",
+  "build",
+  "twine",
+  "jupyter"
+]
+
+all = [
+  "kfc-procedure[cobra]",
+  "kfc-procedure[dev]"
+]
+
+# --------------------------
+# setuptools config (src layout)
+# --------------------------
+[tool.setuptools]
+package-dir = {"" = "src"}
+
+[tool.setuptools.packages.find]
+where = ["src"]
+include = ["kfc_procedure*"]
+
+# --------------------------
+# pytest config
+# --------------------------
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+python_files = ["test_*.py"]
+python_functions = ["test_*"]
+addopts = "-vv --disable-warnings --strict-markers"

kfc_procedure-0.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

kfc_procedure-0.1.0/src/kfc_procedure/__init__.py

@@ -0,0 +1,10 @@
+"""
+kfc_procedure
+
+Meta-estimators and modular pipeline components for the KFC algorithm.
+
+This package implements the K-step, F-step, and C-step stages used to build
+local-model ensembles and aggregate their predictions into final outputs.
+"""
+
+from .kfc import KFCProcedure, KFCRegressor, KFCClassifier

kfc_procedure-0.1.0/src/kfc_procedure/cobra/__init__.py

@@ -0,0 +1,13 @@
+
+from .core import *
+from .mixcobra import MixCOBRARegressor
+from .gradientcobra import GradientCOBRA
+from .combined_classifier import CombinedClassifier
+from .superlearner import SuperLearner
+
+__all__ = [
+    "MixCOBRARegressor",
+    "GradientCOBRA",
+    "CombinedClassifier",
+    "SuperLearner",
+]