Working with Arrow

The @frictionless-ts/table package provides efficient support for loading and saving data in Apache Arrow format. It uses Polars DataFrames for high-performance columnar data processing.

Installation

npm install @frictionless-ts/table

Basic Usage

Tip

You can use loadTable and saveTable from frictionless-ts instead of @frictionless-ts/table to load and save ARROW files if the framework can infer that files are in the arrow/feather format.

Loading Data

import { loadArrowTable } from "@frictionless-ts/table"

// Load from local file
const table = await loadArrowTable({ path: "data.arrow" })

// Load multiple files (concatenated)
const table = await loadArrowTable({
  path: ["file1.arrow", "file2.arrow"]
})

Saving Data

import { saveArrowTable } from "@frictionless-ts/table"

// Save as Arrow format
await saveArrowTable(table, { path: "output.arrow" })