Working with Arrow in Python

Updated May 16, 2026

Apache Arrow IPC file handling with high-performance columnar data processing.

Installation

pip install fairspec

Getting Started

The Arrow plugin provides:

load_arrow_table - Load Arrow IPC files into tables
save_arrow_table - Save tables to Arrow IPC files
ArrowPlugin - Plugin for framework integration

For example:

from fairspec import load_arrow_table, Resource

table = load_arrow_table(Resource(data="table.arrow"))
# High-performance columnar format

Basic Usage

Loading Arrow Files

from fairspec import load_arrow_table, Resource

# Load from local file
table = load_arrow_table(Resource(data="data.arrow"))

# Load from remote URL
table = load_arrow_table(Resource(data="https://example.com/data.arrow"))

# Load multiple files (concatenated)
table = load_arrow_table(Resource(data=["file1.arrow", "file2.arrow"]))

Saving Arrow Files

from fairspec import save_arrow_table

# Save with default options
save_arrow_table(table, path="output.arrow")

Advanced Features

Remote File Loading

from fairspec import load_arrow_table, Resource

# Load from URL
table = load_arrow_table(Resource(data="https://example.com/data.arrow"))

# Load multiple remote files
table = load_arrow_table(Resource(data=[
    "https://api.example.com/data-2023.arrow",
    "https://api.example.com/data-2024.arrow",
]))

Search