Single-Machine Analytics Engine

Parallel Dataframes. Purposeful UX. Zero-Nonsense Workflows.

FrameX combines Pandas-like tables, NumPy-compatible arrays, and Arrow-native memory so teams can process medium-to-large datasets without jumping straight to a cluster.

  • Arrow-firstColumnar internals
  • Eager + LazyChoose execution style
  • NDArray InteropNumPy protocol support
pip install pyframe-xpy
import framex as fx

Arrow-Native Core

Tables are backed by Arrow for efficient columnar operations, cleaner interoperability, and better memory behavior.

DataFrame + NDArray

Use `DataFrame`, `Series`, and chunked `NDArray` in one workflow, with NumPy protocol support included.

Eager and Lazy Modes

Keep day-to-day work intuitive with eager execution, then switch to lazy pipelines when transformations become complex.

Real-World File IO

Ship data across parquet, ORC, SQLite, JSON, CSV, fixed-width text, and export-ready HTML/XML with one `read_file` / `write_file` surface.

Guides That Ship

Full docs now include onboarding, practical tutorials, use cases, architecture details, and API reference material.

Learning Path