VFrames

A high-performance DataFrame library for V, powered by DuckDB.

Overview

VFrames provides a Pandas-like interface for data manipulation in V, leveraging DuckDB's powerful SQL engine for fast in-memory analytics. Designed for developers who want the expressiveness of Pandas with the performance of a compiled language.

Features

• Pandas-like API - Familiar syntax for data scientists coming from Python
• DuckDB Backend - Leverages DuckDB's vectorized execution for exceptional performance
• Multiple Data Sources - Read CSV, JSON, and Parquet files with automatic type inference
• Immutable Design - All operations return new DataFrames, preventing accidental mutations
• Rich Functionality - Support for filtering, grouping, pivoting, joining, and more

Installation

# Install dependencies
v install https://github.com/rodabt/vduckdb
v install https://github.com/rodabt/vframes

Quick Start

import vframes

fn main() {
    // Initialize context (in-memory or persisted)
    mut ctx := vframes.init()
    
    // Load data from file
    df := ctx.read_auto('data.csv')!
    
    // Explore data
    df.head(5)
    df.shape()
    df.describe()
    
    ctx.close()
}

Requirements

• V (Vlang) compiler
• DuckDB library (LIBDUCKDB_DIR environment variable must be set)

Documentation

• Tutorial - Step-by-step guide with Pandas comparisons
• API Reference - Complete function list
• Examples - Real-world usage examples

Why VFrames?

Feature	VFrames	Pandas
Language	V (compiled)	Python (interpreted)
Performance	~10-100x faster	Baseline
Memory	Efficient (DuckDB)	RAM-intensive
Type Safety	Compile-time	Runtime

Contributing

Contributions are welcome! Please read our contributing guidelines before submitting PRs.

License

MIT License - see LICENSE for details.