Crate neural_trader 

Neural Trader - Node.js FFI Bindings

This crate provides high-performance Node.js bindings using napi-rs for the Neural Trader Rust core. It exposes async APIs that map directly to JavaScript Promises.

Architecture

JavaScript (Node.js)
      ↓
napi-rs (auto-generated)
      ↓
This crate (manual wrappers)
      ↓
Rust core (nt-core, nt-strategies, etc.)

Features

• Zero-copy buffer operations for large datasets
• Async/await with automatic Promise conversion
• Type-safe error marshaling
• Resource management with Arc for shared state
• Thread-safe concurrent operations

Modules

• broker - Broker integrations (Alpaca, IBKR, CCXT, Oanda, Questrade, Lime)
• strategy - Trading strategy implementations
• neural - Neural network models for forecasting
• risk - Risk management (VaR, Kelly, drawdown)
• backtest - Backtesting engine
• market_data - Market data streaming and indicators
• portfolio - Portfolio management

Modules

backtest

Backtesting engine bindings for Node.js

broker

Broker integration bindings for Node.js

market_data

Market data streaming and fetching bindings for Node.js

neural

Neural model bindings for Node.js

portfolio

Portfolio management bindings for Node.js

risk

Risk management bindings for Node.js

strategy

Strategy bindings for Node.js

Structs

JsBar

JavaScript-compatible bar data

JsConfig

JavaScript-compatible configuration

JsOrder

JavaScript-compatible order data

JsPosition

JavaScript-compatible position data

JsSignal

JavaScript-compatible signal data

NeuralTrader

Main Neural Trader system instance

VersionInfo

Get version information

Functions

calculate_indicator

Calculate technical indicators

decode_bars_from_buffer

Decode bars from MessagePack buffer

encode_bars_to_buffer

Encode bars to MessagePack buffer for efficient transfer

fetch_market_data

Fetch historical market data

get_version_info

init_runtime

Initialize tokio runtime with custom thread count