Working with Parquet in Python

Updated May 16, 2026

Apache Parquet file handling with high-performance columnar data processing and compression.

Installation

pip install fairspec

Getting Started

The Parquet plugin provides:

load_parquet_table - Load Parquet files into tables
save_parquet_table - Save tables to Parquet files
ParquetPlugin - Plugin for framework integration

For example:

from fairspec import load_parquet_table, Resource

table = load_parquet_table(Resource(data="table.parquet"))
# Efficient columnar format with compression

Basic Usage

Loading Parquet Files

from fairspec import load_parquet_table, Resource

# Load from local file
table = load_parquet_table(Resource(data="data.parquet"))

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

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

Saving Parquet Files

from fairspec import save_parquet_table

# Save with default options
save_parquet_table(table, path="output.parquet")

Advanced Features

Remote File Loading

from fairspec import load_parquet_table, Resource

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

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

Search