Installation

PardoX ships as a pre-compiled native library. No Rust compiler or C toolchain is required on the end-user machine. The Rust core (libpardox.so / .dylib / .dll) is bundled inside the package for each supported platform.

Python SDK

via pip

pip install pardox

System Requirements

OS	Architecture	Status
Linux	x86_64	✅ Stable
Windows	x86_64	✅ Stable
macOS	ARM64 (M1/M2/M3)	✅ Stable
macOS	x86_64 (Intel)	✅ Stable

!!! note “Python version” Python 3.8 or higher is required. No Rust compiler is needed — binaries are pre-packaged for all platforms.

!!! note “No database drivers needed” PardoX connects to PostgreSQL, MySQL, SQL Server, and MongoDB entirely through the Rust core. You do not need to install psycopg2, pymysql, pyodbc, or pymongo.

Verify Installation

import pardox as px
print(px.__version__)    # e.g. "0.3.4"

df = px.read_csv("my_data.csv")
print(df.shape)

Node.js SDK

The Node.js SDK uses koffi for FFI bindings and ships the same pre-compiled .so / .dylib / .dll.

npm i @pardox/pardox

Usage

const { DataFrame, read_csv } = require('@pardox/pardox');

const df = px.read_csv('my_data.csv');
console.log(df.shape);

Requirements

Runtime	Minimum version
Node.js	18 LTS or higher

PHP SDK

The PHP SDK uses the built-in FFI extension (PHP 7.4+).

composer require betoalien/pardox-php

Requirements

Runtime	Minimum version
PHP	8.1 or higher
FFI extension	Must be enabled in `php.ini`

Enable the FFI extension:

; php.ini
extension=ffi
ffi.enable=true

Usage

<?php
require_once 'vendor/autoload.php';

use PardoX\DataFrame;

$df = DataFrame::read_csv('my_data.csv');
echo $df->shape[0] . " rows\n";

Docker (Databases for Testing)

A docker-compose.yml is provided to spin up all supported databases locally:

docker-compose up -d

This starts:

Container	Database	Host port
`pardox_validation_db`	PostgreSQL 16	5434
`pardox_mysql`	MySQL 8.0	3306
`pardox_sqlserver`	SQL Server 2022	1433

!!! tip “MongoDB” Add a MongoDB service to the compose file with mongo:7.0 if needed for your environment.