featcopilot 0.1.0__py3-none-any.whl → 0.2.0__py3-none-any.whl

featcopilot/utils/models.py

@@ -0,0 +1,287 @@
+"""Model utilities for Copilot client."""
+
+import asyncio
+from typing import Optional
+
+from featcopilot.utils.logger import get_logger
+
+logger = get_logger(__name__)
+
+# Cache for models fetched from Copilot
+_cached_models: Optional[list[dict]] = None
+
+# Default model
+DEFAULT_MODEL = "gpt-5.2"
+
+
+async def _fetch_models_from_copilot() -> list[dict]:
+    """Fetch available models from Copilot SDK."""
+    try:
+        from copilot import CopilotClient
+
+        client = CopilotClient()
+        await client.start()
+
+        # Get available models from Copilot
+        models = await client.list_models()
+        await client.stop()
+
+        return models
+
+    except ImportError:
+        logger.warning("copilot-sdk not installed. Cannot fetch models from Copilot.")
+        return []
+    except Exception as e:
+        logger.warning(f"Could not fetch models from Copilot: {e}")
+        return []
+
+
+def _get_event_loop():
+    """Get or create an event loop."""
+    try:
+        loop = asyncio.get_event_loop()
+        if loop.is_closed():
+            loop = asyncio.new_event_loop()
+            asyncio.set_event_loop(loop)
+    except RuntimeError:
+        loop = asyncio.new_event_loop()
+        asyncio.set_event_loop(loop)
+    return loop
+
+
+def fetch_models(force_refresh: bool = False) -> list[dict]:
+    """
+    Fetch available models from the Copilot client.
+
+    This function queries the Copilot SDK to get the current list of
+    supported models. Results are cached for subsequent calls.
+
+    Parameters
+    ----------
+    force_refresh : bool, default=False
+        If True, bypass cache and fetch fresh model list
+
+    Returns
+    -------
+    list[dict]
+        List of model information dictionaries. Each dict contains
+        model metadata from the Copilot API.
+
+    Examples
+    --------
+    >>> from featcopilot.utils import fetch_models
+    >>> models = fetch_models()
+    >>> for m in models:
+    ...     print(m.get('id') or m.get('name'))
+    """
+    global _cached_models
+
+    if _cached_models is not None and not force_refresh:
+        return _cached_models
+
+    loop = _get_event_loop()
+    models = loop.run_until_complete(_fetch_models_from_copilot())
+
+    if models:
+        _cached_models = models
+
+    return models
+
+
+def list_models(
+    provider: Optional[str] = None,
+    verbose: bool = False,
+    force_refresh: bool = False,
+) -> list[dict]:
+    """
+    List all supported models from the Copilot client.
+
+    Retrieves the current list of available models directly from
+    the Copilot SDK.
+
+    Parameters
+    ----------
+    provider : str, optional
+        Filter by provider (e.g., 'OpenAI', 'Anthropic', 'Google')
+    verbose : bool, default=False
+        If True, print model information to logger
+    force_refresh : bool, default=False
+        If True, bypass cache and fetch fresh model list
+
+    Returns
+    -------
+    list[dict]
+        List of model information dictionaries from Copilot API
+
+    Examples
+    --------
+    >>> from featcopilot.utils import list_models
+    >>> models = list_models()
+    >>> for m in models:
+    ...     print(m)
+
+    >>> # With verbose output
+    >>> list_models(verbose=True)
+
+    >>> # Filter by provider (if supported by returned data)
+    >>> openai_models = list_models(provider='OpenAI')
+    """
+    models = fetch_models(force_refresh=force_refresh)
+
+    # Apply provider filter if specified
+    if provider and models:
+        filtered = []
+        for m in models:
+            model_provider = m.get("provider", "") or m.get("vendor", "") or ""
+            if provider.lower() in model_provider.lower():
+                filtered.append(m)
+        models = filtered
+
+    if verbose:
+        _print_models(models)
+
+    return models
+
+
+def get_model_info(model_name: str, force_refresh: bool = False) -> Optional[dict]:
+    """
+    Get information about a specific model from Copilot.
+
+    Parameters
+    ----------
+    model_name : str
+        The model identifier
+    force_refresh : bool, default=False
+        If True, bypass cache and fetch fresh model list
+
+    Returns
+    -------
+    dict or None
+        Model information if found, None otherwise
+
+    Examples
+    --------
+    >>> from featcopilot.utils import get_model_info
+    >>> info = get_model_info('gpt-5.2')
+    >>> if info:
+    ...     print(info)
+    """
+    models = fetch_models(force_refresh=force_refresh)
+
+    for model in models:
+        # Check various possible name fields
+        if model.get("id") == model_name:
+            return model
+        if model.get("name") == model_name:
+            return model
+        if model.get("model") == model_name:
+            return model
+
+    return None
+
+
+def get_default_model() -> str:
+    """
+    Get the default model name.
+
+    Returns
+    -------
+    str
+        The default model identifier
+
+    Examples
+    --------
+    >>> from featcopilot.utils import get_default_model
+    >>> model = get_default_model()
+    >>> print(model)  # 'gpt-5.2'
+    """
+    return DEFAULT_MODEL
+
+
+def get_model_names(force_refresh: bool = False) -> list[str]:
+    """
+    Get list of available model names/identifiers.
+
+    Parameters
+    ----------
+    force_refresh : bool, default=False
+        If True, bypass cache and fetch fresh model list
+
+    Returns
+    -------
+    list[str]
+        List of model identifiers
+
+    Examples
+    --------
+    >>> from featcopilot.utils import get_model_names
+    >>> names = get_model_names()
+    >>> print(names)
+    """
+    models = fetch_models(force_refresh=force_refresh)
+    names = []
+
+    for m in models:
+        name = m.get("id") or m.get("name") or m.get("model")
+        if name:
+            names.append(name)
+
+    return names
+
+
+def is_valid_model(model_name: str, force_refresh: bool = False) -> bool:
+    """
+    Check if a model name is valid/supported.
+
+    Parameters
+    ----------
+    model_name : str
+        The model identifier to check
+    force_refresh : bool, default=False
+        If True, bypass cache and fetch fresh model list
+
+    Returns
+    -------
+    bool
+        True if model is supported, False otherwise
+
+    Examples
+    --------
+    >>> from featcopilot.utils import is_valid_model
+    >>> is_valid_model('gpt-5.2')
+    """
+    models = fetch_models(force_refresh=force_refresh)
+
+    if not models:
+        # If we couldn't fetch models, allow any model name
+        # (let the Copilot API validate it)
+        logger.warning("Could not validate model - Copilot unavailable")
+        return True
+
+    return model_name in get_model_names(force_refresh=False)
+
+
+def _print_models(models: list[dict]) -> None:
+    """Print models information."""
+    if not models:
+        logger.info("No models found. Copilot SDK may not be available.")
+        return
+
+    logger.info("Available models from Copilot:")
+    logger.info("-" * 60)
+
+    for model in models:
+        model_id = model.get("id") or model.get("name") or model.get("model") or "unknown"
+        description = model.get("description", "")
+        provider = model.get("provider") or model.get("vendor") or ""
+
+        line = f"  {model_id}"
+        if provider:
+            line += f" ({provider})"
+        if description:
+            line += f" - {description[:50]}..."
+
+        logger.info(line)
+
+    logger.info("-" * 60)
+    logger.info(f"Total: {len(models)} models")

featcopilot/utils/parallel.py

@@ -4,6 +4,10 @@ from typing import Any, Callable, Optional
 
 import pandas as pd
 
+from featcopilot.utils.logger import get_logger
+
+logger = get_logger(__name__)
+
 
 def parallel_apply(
     func: Callable,
@@ -51,7 +55,7 @@ def parallel_apply(
     except ImportError:
         # Fallback to sequential processing
         if verbose:
-            print("joblib not available, using sequential processing")
+            logger.warning("joblib not available, using sequential processing")
 
         return [func(row) for _, row in data.iterrows()]
 

{featcopilot-0.1.0.dist-info → featcopilot-0.2.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: featcopilot
-Version: 0.1.0
+Version: 0.2.0
 Summary: Next-generation LLM-powered auto feature engineering framework with GitHub Copilot SDK
 Author: FeatCopilot Contributors
 License: MIT
@@ -28,11 +28,26 @@ 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"
+Requires-Dist: nest_asyncio>=1.5.0; extra == "llm"
+Provides-Extra: litellm
+Requires-Dist: litellm>=1.0.0; extra == "litellm"
+Requires-Dist: nest_asyncio>=1.5.0; extra == "litellm"
 Provides-Extra: timeseries
 Requires-Dist: statsmodels>=0.13.0; extra == "timeseries"
+Provides-Extra: feast
+Requires-Dist: feast>=0.30.0; extra == "feast"
 Provides-Extra: full
 Requires-Dist: github-copilot-sdk>=0.1.0; extra == "full"
+Requires-Dist: litellm>=1.0.0; extra == "full"
 Requires-Dist: statsmodels>=0.13.0; extra == "full"
+Requires-Dist: feast>=0.30.0; extra == "full"
+Requires-Dist: nest_asyncio>=1.5.0; extra == "full"
+Provides-Extra: benchmark
+Requires-Dist: github-copilot-sdk>=0.1.0; extra == "benchmark"
+Requires-Dist: statsmodels>=0.13.0; extra == "benchmark"
+Requires-Dist: flaml[automl,blendsearch]>=2.0.0; extra == "benchmark"
+Requires-Dist: autogluon.tabular>=1.0.0; extra == "benchmark"
+Requires-Dist: h2o>=3.40.0; extra == "benchmark"
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0.0; extra == "dev"
 Requires-Dist: pytest-asyncio>=0.21.0; extra == "dev"
@@ -44,9 +59,9 @@ Requires-Dist: pre-commit>=3.6.0; extra == "dev"
 
 # FeatCopilot 🚀
 
-**Next-Generation LLM-Powered Auto Feature Engineering with GitHub Copilot SDK**
+**Next-Generation LLM-Powered Auto Feature Engineering Framework**
 
-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.
+FeatCopilot automatically generates, selects, and explains predictive features using semantic understanding. It analyzes column meanings, applies domain-aware transformations, and provides human-readable explanations—turning raw data into ML-ready features in seconds.
 
 ## 📊 Benchmark Highlights
 
@@ -59,7 +74,7 @@ FeatCopilot is a unified feature engineering framework that combines the best ap
 | Classification | +0.54% | +4.35% |
 | Regression | +0.65% | +5.57% |
 
-### LLM Engine (With Copilot - 30-60s)
+### LLM Engine (With LiteLLM - 30-60s)
 
 | Task Type | Average Improvement | Best Case |
 |-----------|--------------------:|----------:|
@@ -87,7 +102,7 @@ FeatCopilot is a unified feature engineering framework that combines the best ap
 # Basic installation
 pip install featcopilot
 
-# With LLM capabilities (requires GitHub Copilot)
+# With LLM capabilities
 pip install featcopilot[llm]
 
 # Full installation
@@ -111,7 +126,7 @@ X_transformed = engineer.fit_transform(X, y)  # <1 second
 print(f"Features: {X.shape[1]} -> {X_transformed.shape[1]}")
 ```
 
-### LLM Mode (With Copilot)
+### LLM Mode (With LiteLLM)
 
 ```python
 from featcopilot import AutoFeatureEngineer
@@ -164,16 +179,24 @@ engine = TimeSeriesEngine(
 ```
 
 ### LLM Engine
-Uses GitHub Copilot SDK for intelligent feature generation.
+Uses GitHub Copilot SDK (default) or LiteLLM (100+ providers) for intelligent feature generation.
 
 ```python
 from featcopilot.llm import SemanticEngine
 
+# Default: GitHub Copilot SDK
 engine = SemanticEngine(
-    model='gpt-5',
+    model='gpt-5.2',
     max_suggestions=20,
     validate_features=True
 )
+
+# Alternative: LiteLLM backend
+engine = SemanticEngine(
+    model='gpt-4o',
+    backend='litellm',
+    max_suggestions=20
+)
 ```
 
 ## Feature Selection
@@ -211,7 +234,7 @@ X_selected = selector.fit_transform(X, y)
 
 - Python 3.9+
 - NumPy, Pandas, Scikit-learn
-- GitHub Copilot CLI (for LLM features)
+- GitHub Copilot SDK (default) or LiteLLM (for 100+ LLM providers)
 
 ## License
 

featcopilot-0.2.0.dist-info/RECORD

@@ -0,0 +1,35 @@
+featcopilot/__init__.py,sha256=pHamq5ULkm-VFVvy0JQFu-x3CRNqY4mZXppVqV_W4f0,763
+featcopilot/core/__init__.py,sha256=TX_AsPRsVDY1ZKO-ApK7qBFhbgngQbYNamhzYU9P3P8,338
+featcopilot/core/base.py,sha256=lN1zfV6GHCNy2XSNV9OH2dXtvwrZZsu376kT_ExrZto,6090
+featcopilot/core/feature.py,sha256=BzzRJB6RH5RlaOszIvVqcRfQ2F0vfkP_sctfpYAlb-Y,7441
+featcopilot/core/registry.py,sha256=EK4lleLShhMkGRXI1evv6SsVO81rpLLkzJbjImHryJE,4030
+featcopilot/engines/__init__.py,sha256=bXY5eiEQTZ9IxPY1fTESw4v1-nXFkegSs8p4VYGr7TQ,360
+featcopilot/engines/relational.py,sha256=MUqOZgOwKmp7sMsE7lCMbLEjqekBPpcdLXeXkCEKzPM,8402
+featcopilot/engines/tabular.py,sha256=YPjqc9vxCI9-n9pzP3Bm4w2MZ5KHKE5XOk1BdI_nqSc,10878
+featcopilot/engines/text.py,sha256=sixWPg7iuf1NGCWo51gIVZ6DMEf0NUrKA5HECcLXF1k,7496
+featcopilot/engines/timeseries.py,sha256=FdwpOUJJToIkT1RR4VTpFIWYSvEfl_5pXJXf2qAuKm8,13644
+featcopilot/llm/__init__.py,sha256=NswbNiGISinrIzx0EKwaL8fZL4t0DDDD8CIKZ3v_vps,629
+featcopilot/llm/code_generator.py,sha256=p6qSSZ4ul8kxOBjmnAEAU_1dwtVMJWqIfDiM5UXkZqY,9814
+featcopilot/llm/copilot_client.py,sha256=PX7JKBvF0dpnTNMRn_eZcH3swkvZqGQ6pKRICoxwF6c,19034
+featcopilot/llm/explainer.py,sha256=5ukhm27zqyH3oHDOgCHrw3OeJMDEt9olIOYD_hTHqbA,6232
+featcopilot/llm/litellm_client.py,sha256=HSrzJ9-cvxaDdqv142cgoeAqJ2WcIcwVhgz1Jb-PPv8,20443
+featcopilot/llm/semantic_engine.py,sha256=6eg_9Ru_llo0W4wcdIPbMZOlx_p1MTdmCWLQ85g2v6o,14466
+featcopilot/selection/__init__.py,sha256=pjoos64ym3CR7Hk75qq2dY0NN4OT0Sn3hiBSG_h44SE,406
+featcopilot/selection/importance.py,sha256=GEUXfSwknLweAnbdaKX1OZuOn7tfhn-SPyIGiuBSvBg,5575
+featcopilot/selection/redundancy.py,sha256=2S-fP20nfTPPikgIQFc0dvfJU3SprThake5waDdINF4,5398
+featcopilot/selection/statistical.py,sha256=dey-vX-WKT9rrDtHPLYyksfJLHQqdREjSnrXJIUPT0I,6269
+featcopilot/selection/unified.py,sha256=7H8USqe59FLA6cCXJC2etP8Cp6YMLJW3rwaNjIS_iUI,6358
+featcopilot/stores/__init__.py,sha256=w_Ap3jA8pbV7eij3jK6TxAapG1O1tdKFAmUvAPcSdZA,439
+featcopilot/stores/base.py,sha256=VzHUA1IYMUiP4-cE4SXt_xmgkk2raIz09-9aZ6szOxQ,4395
+featcopilot/stores/feast_store.py,sha256=ej6b05JRMwXqQ8pTorxKH8LQ3JTlwzGAGNckZPzq7d8,19032
+featcopilot/transformers/__init__.py,sha256=pHHSivxNVyC9AmKzdA8UJS0B-b9V1B4eI1QgTP0Z8uw,234
+featcopilot/transformers/sklearn_compat.py,sha256=rf9tapjbiRT6Bx-hU_t-7UMncrKGwTaKn45qHcYWfUI,13637
+featcopilot/utils/__init__.py,sha256=aWRMS_4sZHdw8eTHBYT_O4qMYuM9EHsPGPlo5a-PzaE,489
+featcopilot/utils/cache.py,sha256=CYIVXQU9dpVdrLkemhjVAoCzXRMRzuQ-U36-q_8zDoc,6489
+featcopilot/utils/logger.py,sha256=qJe4OT13TPanF-vebfs8BILmwkuV8zi2KSHlAUDaSnU,1334
+featcopilot/utils/models.py,sha256=X9amoyLjy8umaCvbp0G6BkmzmasvAYZc2xVKvD6yeAA,7317
+featcopilot/utils/parallel.py,sha256=k0ePtnI48aXNomD0lBBptla0RlhCX1-5xyEDD9zbkDQ,2963
+featcopilot-0.2.0.dist-info/METADATA,sha256=9p7mmGYqFf4gerx16TnIxOLUueLxU5DBATWivlapS9k,7763
+featcopilot-0.2.0.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
+featcopilot-0.2.0.dist-info/top_level.txt,sha256=Lf6lB8VBDX71TBzSXW3qq44RO5hIUKQ6DMi_dL8G_Ak,12
+featcopilot-0.2.0.dist-info/RECORD,,

featcopilot-0.1.0.dist-info/RECORD

@@ -1,29 +0,0 @@
-featcopilot/__init__.py,sha256=Ljn4cK5Emw_Rh8rg-T1EqDoyuWz_GXCAkiaZ_amz-cA,708
-featcopilot/core/__init__.py,sha256=TX_AsPRsVDY1ZKO-ApK7qBFhbgngQbYNamhzYU9P3P8,338
-featcopilot/core/base.py,sha256=lN1zfV6GHCNy2XSNV9OH2dXtvwrZZsu376kT_ExrZto,6090
-featcopilot/core/feature.py,sha256=YGE-xDj7cRZDA50Oa5W-USXiEZ1NrX3WagZueRQiZbk,7361
-featcopilot/core/registry.py,sha256=EK4lleLShhMkGRXI1evv6SsVO81rpLLkzJbjImHryJE,4030
-featcopilot/engines/__init__.py,sha256=bXY5eiEQTZ9IxPY1fTESw4v1-nXFkegSs8p4VYGr7TQ,360
-featcopilot/engines/relational.py,sha256=0zr8OhCRg-jvch1HRaxR6RgQFti3McDPvdco-onsUTc,8311
-featcopilot/engines/tabular.py,sha256=M5cynbfNzkFilxV2rPG5kIcEwni-wc_HtjGNuGiSXGU,10781
-featcopilot/engines/text.py,sha256=I9OGN0Rwf2SW1V9shaJDmmud1eymnx5xLdSmLfCzKhU,7396
-featcopilot/engines/timeseries.py,sha256=76YE5QAD-yzXdajeP3wP3cfqtRSOZglz-E-OVSnvvs8,13553
-featcopilot/llm/__init__.py,sha256=75MgMNw8cAggNOUAJV8ib1uV_UCTjA04tdYbqWNwumU,468
-featcopilot/llm/code_generator.py,sha256=dAn5n3fp-sWj3DhQ9I-ZME5bYSXRVSe4Blx5P-W484Q,9715
-featcopilot/llm/copilot_client.py,sha256=BLAENfNjkyOvgeVWQt-94mkTtxcdz5esk0ro6iaUCPM,17328
-featcopilot/llm/explainer.py,sha256=fztidSmLX5dmbPRMcLmgXNv7AvDEJjXM5OzD6NFtQjQ,6142
-featcopilot/llm/semantic_engine.py,sha256=U4CQ-OsWEr_c2Q8HF6yN7jXhKpTaSR6lft-lEacBzXo,12554
-featcopilot/selection/__init__.py,sha256=pjoos64ym3CR7Hk75qq2dY0NN4OT0Sn3hiBSG_h44SE,406
-featcopilot/selection/importance.py,sha256=GzSHTwKIeHJt42mzb1x8ClGFnqU9_2MxV1kOce9SODg,5481
-featcopilot/selection/redundancy.py,sha256=bz47JQtEBKoz9eMIMB3YMOxH_1s7q1OsEUE_Ob3t0xg,5301
-featcopilot/selection/statistical.py,sha256=POuWuMoyggJt1bCML10I6C7gKZoUP7Yc4nMsKBwECYs,6184
-featcopilot/selection/unified.py,sha256=UPrTJ0svTjy6SnqQVeyiAGNjkgTRbXO_AzbucGZNUjc,6273
-featcopilot/transformers/__init__.py,sha256=pHHSivxNVyC9AmKzdA8UJS0B-b9V1B4eI1QgTP0Z8uw,234
-featcopilot/transformers/sklearn_compat.py,sha256=z5oA_y8YENQaGTTnR13u7CWhJpIhNmUoQP-DNVIgPeA,13536
-featcopilot/utils/__init__.py,sha256=PH-fC1WYJYevcMVI4CCoXIUTJlb5lPXwKeKNj56vs_k,198
-featcopilot/utils/cache.py,sha256=CYIVXQU9dpVdrLkemhjVAoCzXRMRzuQ-U36-q_8zDoc,6489
-featcopilot/utils/parallel.py,sha256=MjLAYWYgoeu02IoWNWEZ82sCcRDi5bICJVcSq7dXmN4,2874
-featcopilot-0.1.0.dist-info/METADATA,sha256=EI1MSGG2KkICOCRnLX1CI5-v91N40pmRS6oyAVgk1w4,6804
-featcopilot-0.1.0.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
-featcopilot-0.1.0.dist-info/top_level.txt,sha256=Lf6lB8VBDX71TBzSXW3qq44RO5hIUKQ6DMi_dL8G_Ak,12
-featcopilot-0.1.0.dist-info/RECORD,,