featcopilot 0.1.0__tar.gz

featcopilot-0.1.0/PKG-INFO

@@ -0,0 +1,218 @@
+Metadata-Version: 2.4
+Name: featcopilot
+Version: 0.1.0
+Summary: Next-generation LLM-powered auto feature engineering framework with GitHub Copilot SDK
+Author: FeatCopilot Contributors
+License: MIT
+Project-URL: Homepage, https://github.com/thinkall/featcopilot
+Project-URL: Documentation, https://github.com/thinkall/featcopilot#readme
+Project-URL: Repository, https://github.com/thinkall/featcopilot
+Keywords: machine-learning,feature-engineering,automl,llm,copilot,data-science
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+Requires-Dist: numpy>=1.21.0
+Requires-Dist: pandas>=1.3.0
+Requires-Dist: scipy>=1.7.0
+Requires-Dist: scikit-learn>=1.0.0
+Requires-Dist: pydantic>=2.0.0
+Requires-Dist: joblib>=1.1.0
+Provides-Extra: llm
+Requires-Dist: github-copilot-sdk>=0.1.0; extra == "llm"
+Provides-Extra: timeseries
+Requires-Dist: statsmodels>=0.13.0; extra == "timeseries"
+Provides-Extra: full
+Requires-Dist: github-copilot-sdk>=0.1.0; extra == "full"
+Requires-Dist: statsmodels>=0.13.0; extra == "full"
+Provides-Extra: dev
+Requires-Dist: pytest>=7.0.0; extra == "dev"
+Requires-Dist: pytest-asyncio>=0.21.0; extra == "dev"
+Requires-Dist: pytest-cov>=4.0.0; extra == "dev"
+Requires-Dist: black>=23.0.0; extra == "dev"
+Requires-Dist: ruff>=0.1.0; extra == "dev"
+Requires-Dist: mypy>=1.0.0; extra == "dev"
+Requires-Dist: pre-commit>=3.6.0; extra == "dev"
+
+# FeatCopilot 🚀
+
+**Next-Generation LLM-Powered Auto Feature Engineering with GitHub Copilot SDK**
+
+FeatCopilot is a unified feature engineering framework that combines the best approaches from existing libraries (Featuretools, TSFresh, AutoFeat, OpenFE) with novel LLM-powered capabilities via GitHub Copilot SDK.
+
+## 📊 Benchmark Highlights
+
+### Tabular Engine (Fast Mode - <1s)
+
+| Task Type | Average Improvement | Best Case |
+|-----------|--------------------:|----------:|
+| **Text Classification** | **+12.44%** | +49.02% (News Headlines) |
+| Time Series | +1.51% | +12.12% (Retail Demand) |
+| Classification | +0.54% | +4.35% |
+| Regression | +0.65% | +5.57% |
+
+### LLM Engine (With Copilot - 30-60s)
+
+| Task Type | Average Improvement | Best Case |
+|-----------|--------------------:|----------:|
+| **Regression** | **+7.79%** | +19.66% (Retail Demand) |
+| Classification | +2.38% | +2.87% |
+
+- ✅ **12/12 wins** on text classification (tabular mode)
+- 🧠 **+19.66% max improvement** with LLM-powered features
+- ⚡ **<1 second** (tabular) or **30-60s** (with LLM) processing time
+- 📈 Largest gains with simple models (LogisticRegression, Ridge)
+
+[View Full Benchmark Results](https://thinkall.github.io/featcopilot/user-guide/benchmarks/)
+
+## Key Features
+
+- 🔧 **Multi-Engine Architecture**: Tabular, time series, relational, and text feature engines
+- 🤖 **LLM-Powered Intelligence**: Semantic feature discovery, domain-aware generation, and code synthesis
+- 📊 **Intelligent Selection**: Statistical testing, importance ranking, and redundancy elimination
+- 🔌 **Scikit-learn Compatible**: Drop-in replacement for sklearn transformers
+- 📝 **Interpretable**: Every feature comes with human-readable explanations
+
+## Installation
+
+```bash
+# Basic installation
+pip install featcopilot
+
+# With LLM capabilities (requires GitHub Copilot)
+pip install featcopilot[llm]
+
+# Full installation
+pip install featcopilot[full]
+```
+
+## Quick Start
+
+### Fast Mode (Tabular Only)
+
+```python
+from featcopilot import AutoFeatureEngineer
+
+# Sub-second feature engineering
+engineer = AutoFeatureEngineer(
+    engines=['tabular'],
+    max_features=50
+)
+
+X_transformed = engineer.fit_transform(X, y)  # <1 second
+print(f"Features: {X.shape[1]} -> {X_transformed.shape[1]}")
+```
+
+### LLM Mode (With Copilot)
+
+```python
+from featcopilot import AutoFeatureEngineer
+
+# LLM-powered semantic features (+19.66% max improvement)
+engineer = AutoFeatureEngineer(
+    engines=['tabular', 'llm'],
+    max_features=50
+)
+
+X_transformed = engineer.fit_transform(
+    X, y,
+    column_descriptions={
+        'age': 'Customer age in years',
+        'income': 'Annual household income in USD',
+        'tenure': 'Months as customer',
+    },
+    task_description="Predict customer churn"
+)  # 30-60 seconds
+
+# Get LLM-generated explanations
+for feature, explanation in engineer.explain_features().items():
+    print(f"{feature}: {explanation}")
+```
+
+## Engines
+
+### Tabular Engine
+Generates polynomial features, interaction terms, and mathematical transformations.
+
+```python
+from featcopilot.engines import TabularEngine
+
+engine = TabularEngine(
+    polynomial_degree=2,
+    interaction_only=False,
+    include_transforms=['log', 'sqrt', 'square']
+)
+```
+
+### Time Series Engine
+Extracts statistical, frequency, and temporal features from time series data.
+
+```python
+from featcopilot.engines import TimeSeriesEngine
+
+engine = TimeSeriesEngine(
+    features=['mean', 'std', 'skew', 'autocorr', 'fft_coefficients']
+)
+```
+
+### LLM Engine
+Uses GitHub Copilot SDK for intelligent feature generation.
+
+```python
+from featcopilot.llm import SemanticEngine
+
+engine = SemanticEngine(
+    model='gpt-5',
+    max_suggestions=20,
+    validate_features=True
+)
+```
+
+## Feature Selection
+
+```python
+from featcopilot.selection import FeatureSelector
+
+selector = FeatureSelector(
+    methods=['mutual_info', 'importance', 'correlation'],
+    max_features=30,
+    correlation_threshold=0.95
+)
+
+X_selected = selector.fit_transform(X, y)
+```
+
+## Comparison with Existing Libraries
+
+| Feature | FeatCopilot | Featuretools | TSFresh | AutoFeat | OpenFE | CAAFE |
+|---------|-------------|--------------|---------|----------|--------|-------|
+| Tabular Features | ✅ | ❌ | ❌ | ✅ | ✅ | ✅ |
+| Time Series | ✅ | ❌ | ✅ | ❌ | ❌ | ❌ |
+| Relational | ✅ | ✅ | ❌ | ❌ | ❌ | ❌ |
+| LLM-Powered | ✅ | ❌ | ❌ | ❌ | ❌ | ✅ |
+| Semantic Understanding | ✅ | ❌ | ❌ | ❌ | ❌ | ⚠️ |
+| Code Generation | ✅ | ❌ | ❌ | ❌ | ❌ | ⚠️ |
+| Sklearn Compatible | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
+| Interpretable | ✅ | ⚠️ | ⚠️ | ⚠️ | ❌ | ✅ |
+
+## Documentation
+
+📖 **Full Documentation**: [https://thinkall.github.io/featcopilot/](https://thinkall.github.io/featcopilot/)
+
+## Requirements
+
+- Python 3.9+
+- NumPy, Pandas, Scikit-learn
+- GitHub Copilot CLI (for LLM features)
+
+## License
+
+MIT License

featcopilot-0.1.0/README.md

@@ -0,0 +1,174 @@
+# FeatCopilot 🚀
+
+**Next-Generation LLM-Powered Auto Feature Engineering with GitHub Copilot SDK**
+
+FeatCopilot is a unified feature engineering framework that combines the best approaches from existing libraries (Featuretools, TSFresh, AutoFeat, OpenFE) with novel LLM-powered capabilities via GitHub Copilot SDK.
+
+## 📊 Benchmark Highlights
+
+### Tabular Engine (Fast Mode - <1s)
+
+| Task Type | Average Improvement | Best Case |
+|-----------|--------------------:|----------:|
+| **Text Classification** | **+12.44%** | +49.02% (News Headlines) |
+| Time Series | +1.51% | +12.12% (Retail Demand) |
+| Classification | +0.54% | +4.35% |
+| Regression | +0.65% | +5.57% |
+
+### LLM Engine (With Copilot - 30-60s)
+
+| Task Type | Average Improvement | Best Case |
+|-----------|--------------------:|----------:|
+| **Regression** | **+7.79%** | +19.66% (Retail Demand) |
+| Classification | +2.38% | +2.87% |
+
+- ✅ **12/12 wins** on text classification (tabular mode)
+- 🧠 **+19.66% max improvement** with LLM-powered features
+- ⚡ **<1 second** (tabular) or **30-60s** (with LLM) processing time
+- 📈 Largest gains with simple models (LogisticRegression, Ridge)
+
+[View Full Benchmark Results](https://thinkall.github.io/featcopilot/user-guide/benchmarks/)
+
+## Key Features
+
+- 🔧 **Multi-Engine Architecture**: Tabular, time series, relational, and text feature engines
+- 🤖 **LLM-Powered Intelligence**: Semantic feature discovery, domain-aware generation, and code synthesis
+- 📊 **Intelligent Selection**: Statistical testing, importance ranking, and redundancy elimination
+- 🔌 **Scikit-learn Compatible**: Drop-in replacement for sklearn transformers
+- 📝 **Interpretable**: Every feature comes with human-readable explanations
+
+## Installation
+
+```bash
+# Basic installation
+pip install featcopilot
+
+# With LLM capabilities (requires GitHub Copilot)
+pip install featcopilot[llm]
+
+# Full installation
+pip install featcopilot[full]
+```
+
+## Quick Start
+
+### Fast Mode (Tabular Only)
+
+```python
+from featcopilot import AutoFeatureEngineer
+
+# Sub-second feature engineering
+engineer = AutoFeatureEngineer(
+    engines=['tabular'],
+    max_features=50
+)
+
+X_transformed = engineer.fit_transform(X, y)  # <1 second
+print(f"Features: {X.shape[1]} -> {X_transformed.shape[1]}")
+```
+
+### LLM Mode (With Copilot)
+
+```python
+from featcopilot import AutoFeatureEngineer
+
+# LLM-powered semantic features (+19.66% max improvement)
+engineer = AutoFeatureEngineer(
+    engines=['tabular', 'llm'],
+    max_features=50
+)
+
+X_transformed = engineer.fit_transform(
+    X, y,
+    column_descriptions={
+        'age': 'Customer age in years',
+        'income': 'Annual household income in USD',
+        'tenure': 'Months as customer',
+    },
+    task_description="Predict customer churn"
+)  # 30-60 seconds
+
+# Get LLM-generated explanations
+for feature, explanation in engineer.explain_features().items():
+    print(f"{feature}: {explanation}")
+```
+
+## Engines
+
+### Tabular Engine
+Generates polynomial features, interaction terms, and mathematical transformations.
+
+```python
+from featcopilot.engines import TabularEngine
+
+engine = TabularEngine(
+    polynomial_degree=2,
+    interaction_only=False,
+    include_transforms=['log', 'sqrt', 'square']
+)
+```
+
+### Time Series Engine
+Extracts statistical, frequency, and temporal features from time series data.
+
+```python
+from featcopilot.engines import TimeSeriesEngine
+
+engine = TimeSeriesEngine(
+    features=['mean', 'std', 'skew', 'autocorr', 'fft_coefficients']
+)
+```
+
+### LLM Engine
+Uses GitHub Copilot SDK for intelligent feature generation.
+
+```python
+from featcopilot.llm import SemanticEngine
+
+engine = SemanticEngine(
+    model='gpt-5',
+    max_suggestions=20,
+    validate_features=True
+)
+```
+
+## Feature Selection
+
+```python
+from featcopilot.selection import FeatureSelector
+
+selector = FeatureSelector(
+    methods=['mutual_info', 'importance', 'correlation'],
+    max_features=30,
+    correlation_threshold=0.95
+)
+
+X_selected = selector.fit_transform(X, y)
+```
+
+## Comparison with Existing Libraries
+
+| Feature | FeatCopilot | Featuretools | TSFresh | AutoFeat | OpenFE | CAAFE |
+|---------|-------------|--------------|---------|----------|--------|-------|
+| Tabular Features | ✅ | ❌ | ❌ | ✅ | ✅ | ✅ |
+| Time Series | ✅ | ❌ | ✅ | ❌ | ❌ | ❌ |
+| Relational | ✅ | ✅ | ❌ | ❌ | ❌ | ❌ |
+| LLM-Powered | ✅ | ❌ | ❌ | ❌ | ❌ | ✅ |
+| Semantic Understanding | ✅ | ❌ | ❌ | ❌ | ❌ | ⚠️ |
+| Code Generation | ✅ | ❌ | ❌ | ❌ | ❌ | ⚠️ |
+| Sklearn Compatible | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
+| Interpretable | ✅ | ⚠️ | ⚠️ | ⚠️ | ❌ | ✅ |
+
+## Documentation
+
+📖 **Full Documentation**: [https://thinkall.github.io/featcopilot/](https://thinkall.github.io/featcopilot/)
+
+## Requirements
+
+- Python 3.9+
+- NumPy, Pandas, Scikit-learn
+- GitHub Copilot CLI (for LLM features)
+
+## License
+
+MIT License

featcopilot-0.1.0/featcopilot/__init__.py

@@ -0,0 +1,29 @@
+"""
+FeatCopilot - Next-Generation LLM-Powered Auto Feature Engineering
+
+A unified feature engineering framework combining traditional approaches
+with novel LLM-powered capabilities via GitHub Copilot SDK.
+"""
+
+__version__ = "0.1.0"
+__author__ = "FeatCopilot Contributors"
+
+from featcopilot.core.base import BaseEngine, BaseSelector
+from featcopilot.core.feature import Feature, FeatureSet
+from featcopilot.transformers.sklearn_compat import (
+    AutoFeatureEngineer,
+    FeatureEngineerTransformer,
+)
+
+__all__ = [
+    # Core
+    "BaseEngine",
+    "BaseSelector",
+    "Feature",
+    "FeatureSet",
+    # Main API
+    "AutoFeatureEngineer",
+    "FeatureEngineerTransformer",
+    # Version
+    "__version__",
+]

featcopilot-0.1.0/featcopilot/core/__init__.py

@@ -0,0 +1,13 @@
+"""Core module containing base classes and interfaces."""
+
+from featcopilot.core.base import BaseEngine, BaseSelector
+from featcopilot.core.feature import Feature, FeatureSet
+from featcopilot.core.registry import FeatureRegistry
+
+__all__ = [
+    "BaseEngine",
+    "BaseSelector",
+    "Feature",
+    "FeatureSet",
+    "FeatureRegistry",
+]

featcopilot-0.1.0/featcopilot/core/base.py

@@ -0,0 +1,195 @@
+"""Base classes for feature engineering engines and selectors."""
+
+from abc import ABC, abstractmethod
+from typing import Any, Optional, Union
+
+import numpy as np
+import pandas as pd
+from pydantic import BaseModel, Field
+
+
+class EngineConfig(BaseModel):
+    """Configuration for feature engineering engines."""
+
+    name: str = Field(description="Engine name")
+    enabled: bool = Field(default=True, description="Whether engine is enabled")
+    max_features: Optional[int] = Field(default=None, description="Max features to generate")
+    verbose: bool = Field(default=False, description="Verbose output")
+
+
+class BaseEngine(ABC):
+    """
+    Abstract base class for feature engineering engines.
+
+    All engines (tabular, timeseries, relational, llm) inherit from this class.
+    """
+
+    def __init__(self, config: Optional[EngineConfig] = None, **kwargs):
+        self.config = config or EngineConfig(name=self.__class__.__name__, **kwargs)
+        self._is_fitted = False
+        self._feature_names: list[str] = []
+        self._feature_metadata: dict[str, Any] = {}
+
+    @property
+    def is_fitted(self) -> bool:
+        """Check if engine has been fitted."""
+        return self._is_fitted
+
+    @abstractmethod
+    def fit(
+        self,
+        X: Union[pd.DataFrame, np.ndarray],
+        y: Optional[Union[pd.Series, np.ndarray]] = None,
+        **kwargs,
+    ) -> "BaseEngine":
+        """
+        Fit the engine to the data.
+
+        Parameters
+        ----------
+        X : DataFrame or ndarray
+            Input features
+        y : Series or ndarray, optional
+            Target variable
+        **kwargs : dict
+            Additional parameters
+
+        Returns
+        -------
+        self : BaseEngine
+            Fitted engine
+        """
+        pass
+
+    @abstractmethod
+    def transform(self, X: Union[pd.DataFrame, np.ndarray], **kwargs) -> pd.DataFrame:
+        """
+        Transform data to generate new features.
+
+        Parameters
+        ----------
+        X : DataFrame or ndarray
+            Input features
+        **kwargs : dict
+            Additional parameters
+
+        Returns
+        -------
+        X_transformed : DataFrame
+            Transformed features
+        """
+        pass
+
+    def fit_transform(
+        self,
+        X: Union[pd.DataFrame, np.ndarray],
+        y: Optional[Union[pd.Series, np.ndarray]] = None,
+        **kwargs,
+    ) -> pd.DataFrame:
+        """Fit and transform in one step."""
+        return self.fit(X, y, **kwargs).transform(X, **kwargs)
+
+    def get_feature_names(self) -> list[str]:
+        """Get names of generated features."""
+        return self._feature_names.copy()
+
+    def get_feature_metadata(self) -> dict[str, Any]:
+        """Get metadata for generated features."""
+        return self._feature_metadata.copy()
+
+    def _validate_input(self, X: Union[pd.DataFrame, np.ndarray]) -> pd.DataFrame:
+        """Convert input to DataFrame and validate."""
+        if isinstance(X, np.ndarray):
+            X = pd.DataFrame(X, columns=[f"feature_{i}" for i in range(X.shape[1])])
+        elif not isinstance(X, pd.DataFrame):
+            raise TypeError(f"Expected DataFrame or ndarray, got {type(X)}")
+        return X
+
+
+class SelectorConfig(BaseModel):
+    """Configuration for feature selectors."""
+
+    max_features: Optional[int] = Field(default=None, description="Max features to select")
+    min_importance: float = Field(default=0.0, description="Minimum importance threshold")
+    correlation_threshold: float = Field(default=0.95, description="Threshold for correlation-based elimination")
+
+
+class BaseSelector(ABC):
+    """
+    Abstract base class for feature selection.
+
+    Handles selection of most important/relevant features from generated set.
+    """
+
+    def __init__(self, config: Optional[SelectorConfig] = None, **kwargs):
+        self.config = config or SelectorConfig(**kwargs)
+        self._is_fitted = False
+        self._selected_features: list[str] = []
+        self._feature_scores: dict[str, float] = {}
+
+    @property
+    def is_fitted(self) -> bool:
+        """Check if selector has been fitted."""
+        return self._is_fitted
+
+    @abstractmethod
+    def fit(self, X: Union[pd.DataFrame, np.ndarray], y: Union[pd.Series, np.ndarray], **kwargs) -> "BaseSelector":
+        """
+        Fit the selector to determine feature importance.
+
+        Parameters
+        ----------
+        X : DataFrame or ndarray
+            Input features
+        y : Series or ndarray
+            Target variable
+        **kwargs : dict
+            Additional parameters
+
+        Returns
+        -------
+        self : BaseSelector
+            Fitted selector
+        """
+        pass
+
+    @abstractmethod
+    def transform(self, X: Union[pd.DataFrame, np.ndarray], **kwargs) -> pd.DataFrame:
+        """
+        Transform data to keep only selected features.
+
+        Parameters
+        ----------
+        X : DataFrame or ndarray
+            Input features
+        **kwargs : dict
+            Additional parameters
+
+        Returns
+        -------
+        X_selected : DataFrame
+            Data with only selected features
+        """
+        pass
+
+    def fit_transform(
+        self, X: Union[pd.DataFrame, np.ndarray], y: Union[pd.Series, np.ndarray], **kwargs
+    ) -> pd.DataFrame:
+        """Fit and transform in one step."""
+        return self.fit(X, y, **kwargs).transform(X, **kwargs)
+
+    def get_selected_features(self) -> list[str]:
+        """Get names of selected features."""
+        return self._selected_features.copy()
+
+    def get_feature_scores(self) -> dict[str, float]:
+        """Get importance scores for all features."""
+        return self._feature_scores.copy()
+
+    def _validate_input(self, X: Union[pd.DataFrame, np.ndarray]) -> pd.DataFrame:
+        """Convert input to DataFrame and validate."""
+        if isinstance(X, np.ndarray):
+            X = pd.DataFrame(X, columns=[f"feature_{i}" for i in range(X.shape[1])])
+        elif not isinstance(X, pd.DataFrame):
+            raise TypeError(f"Expected DataFrame or ndarray, got {type(X)}")
+        return X