Skip to content

Latest commit

 

History

History
175 lines (123 loc) · 3.59 KB

dependencies.md

File metadata and controls

175 lines (123 loc) · 3.59 KB

Dependencies

IceFrame is designed to be lightweight with a core set of dependencies and optional extras for specific features.

Core Dependencies

  • pyiceberg: Core Iceberg client
  • polars: High-performance DataFrame library
  • pyarrow: Apache Arrow support
  • python-dotenv: Environment variable management

Optional Dependencies

CLI ([cli])

Required for the command-line interface.

  • typer: CLI application builder
  • rich: Terminal formatting

Install with:

pip install "iceframe[cli]"

Pydantic ([pydantic])

Required for Pydantic integration.

  • pydantic>=2.0.0: Data validation and settings management

Install with:

pip install "iceframe[pydantic]"

Ingestion ([ingestion])

Required for various data ingestion and format support.

  • pdf: PDF generation (fpdf2, markdown-it-py)
  • delta: Delta Lake support (deltalake)
  • lance: Lance support (pylance)
  • vortex: Vortex support (vortex-data)
  • excel: Excel support (fastexcel)
  • gsheets: Google Sheets support (gspread)
  • hudi: Hudi support (getdaft)
  • sql: SQL Database support (connectorx, sqlalchemy)
  • xml: XML support (lxml)
  • stats: Statistical file support (pyreadstat)
  • api: REST API support (requests)
  • hf: HuggingFace Datasets support (datasets)
  • html: HTML table scraping (lxml, html5lib, beautifulsoup4)
  • clipboard: Clipboard support (pyperclip)

Install with:

pip install "iceframe[ingestion]"

Or individually:

pip install "iceframe[sql]"
pip install "iceframe[xml]"
pip install "iceframe[stats]"
pip install "iceframe[api]"
pip install "iceframe[hf]"
pip install "iceframe[html]"
pip install "iceframe[clipboard]"

MCP Server ([mcp])

Required for running the Model Context Protocol server.

  • mcp: MCP server library

Install with:

pip install "iceframe[mcp]"

Notebook ([notebook])

Required for Jupyter Notebook integration.

  • ipython>=8.0.0: Interactive computing
  • ipywidgets>=8.0.0: Interactive HTML widgets

Install with:

pip install "iceframe[notebook]"

Cloud Storage ([aws], [gcs], [azure])

Required for accessing cloud storage backends.

  • s3fs: AWS S3 support
  • gcsfs: Google Cloud Storage support
  • adlfs: Azure Data Lake Storage support

Install with:

pip install "iceframe[aws]"
# or
pip install "iceframe[aws,gcs]"

Async Support

Async operations use Python's built-in asyncio library (no additional dependencies required).

from iceframe.async_ops import AsyncIceFrame

AI Agent

The AI agent requires LLM provider packages. Install with:

pip install "iceframe[agent]"

This includes:

  • openai>=1.0.0 - For GPT models
  • anthropic>=0.18.0 - For Claude models
  • google-generativeai>=0.3.0 - For Gemini models
  • rich>=13.0.0 - For CLI formatting

You only need the API key for the LLM provider you want to use:

# Choose one:
export OPENAI_API_KEY="your-key"
export ANTHROPIC_API_KEY="your-key"
export GOOGLE_API_KEY="your-key"

Scalability Features

pip install "iceframe[cache,streaming,monitoring]"

SQL Support ([datafusion])

Required for high-performance SQL execution.

  • datafusion>=35.0.0: Apache DataFusion
pip install "iceframe[datafusion]"

Distributed Processing ([distributed])

Required for distributed execution.

  • ray>=2.0.0: Ray
pip install "iceframe[distributed]"

Visualization ([viz])

Required for generating charts.

  • altair>=5.0.0: Declarative statistical visualization
pip install "iceframe[viz]"