Adapter Selection Guide¶

Overview¶

RustyBT provides multiple data adapters for fetching market data from various sources. This guide helps you choose the right adapter for your use case.

Quick Decision Tree¶

Need market data?
│
├─ Local CSV files? → CSV Adapter
│
├─ Cryptocurrency?
│  └─ 100+ exchanges? → CCXT Adapter
│
├─ US Stocks?
│  ├─ Free/testing? → YFinance Adapter
│  ├─ Paper trading? → Alpaca Adapter (paper mode)
│  ├─ Production/real-time? → Polygon or Alpaca (live mode)
│  └─ Global stocks? → AlphaVantage Adapter
│
├─ Forex?
│  ├─ Real-time? → Polygon Adapter
│  └─ Free tier okay? → AlphaVantage Adapter
│
└─ Options?
   └─ US options? → Polygon Adapter

Adapter Comparison Matrix¶

Adapter	Asset Classes	Cost	Rate Limits	Real-time	Best For
CSV	Any (custom)	Free	None	No	Custom data, backtesting
CCXT	Crypto	Free	Exchange-specific	Yes (some)	Crypto trading, multi-exchange
YFinance	Stocks, ETFs, indices, forex	Free	~2000/hr	No (15-min delay)	Development, research
Polygon	Stocks, options, forex, crypto	Free-$99/mo	5-100/min	Yes	Production US markets
Alpaca	US stocks	Free-Paid	200/min	Yes	US stock trading
AlphaVantage	Stocks, forex, crypto	Free-$50/mo	5-75/min	No	Global markets

Detailed Comparisons¶

For Cryptocurrency Data¶

CCXT Adapter¶

Pros: - ✅ Access to 100+ exchanges - ✅ Unified interface across all exchanges - ✅ Free for most exchanges - ✅ Real-time data from many exchanges - ✅ High-frequency intraday data (1m, 5m, 15m, etc.)

Cons: - ❌ Each exchange has different rate limits - ❌ Symbol formats vary by exchange - ❌ Some exchanges require API keys

Use When: - Trading or analyzing multiple cryptocurrencies - Need data from specific exchanges (Binance, Coinbase, Kraken, etc.) - Building crypto trading strategies

Example:

from rustybt.data.adapters.ccxt_adapter import CCXTAdapter

adapter = CCXTAdapter(exchange_id="binance")
data = await adapter.fetch(["BTC/USDT", "ETH/USDT"], start, end, "1h")

For US Stock Data¶

YFinance Adapter (Free)¶

Pros: - ✅ Completely free - ✅ No API key required - ✅ Good for US stocks and ETFs - ✅ 10+ years of historical data - ✅ Dividend and split data available

Cons: - ❌ 15-minute delayed data - ❌ Intraday data limited to 60 days - ❌ Rate limits (informal, ~2000/hr) - ❌ No real-time streaming

Use When: - Development and testing - Academic research - Backtesting strategies - Don't need real-time data

Example:

from rustybt.data.adapters.yfinance_adapter import YFinanceAdapter

adapter = YFinanceAdapter()
data = await adapter.fetch(["AAPL", "MSFT"], start, end, "1d")

Alpaca Adapter (Paper: Free, Live: Paid)¶

Pros: - ✅ Free paper trading feed (IEX) - ✅ Real-time data (live mode) - ✅ 200 requests/minute - ✅ Good for algorithmic trading - ✅ WebSocket support for live streaming

Cons: - ❌ US stocks only (no forex/options) - ❌ Live feed requires paid subscription - ❌ IEX feed (paper) has limited market coverage

Use When: - Building algorithmic trading systems - Need paper trading environment - Want real-time US stock data (live mode)

Example:

from rustybt.data.adapters.alpaca_adapter import AlpacaAdapter

# Paper trading (free)
adapter = AlpacaAdapter(is_paper=True)

# Live trading (paid)
adapter_live = AlpacaAdapter(is_paper=False)

Polygon Adapter (Free-$99/mo)¶

Pros: - ✅ Real-time data (all tiers) - ✅ US stocks, options, forex, crypto - ✅ High rate limits (developer tier: 100/min) - ✅ Professional-grade data quality - ✅ WebSocket support

Cons: - ❌ Free tier very limited (5 req/min) - ❌ Best features require paid tiers - ❌ Focused on US markets

Use When: - Production trading systems - Need real-time US market data - Trading options - Budget for professional data

Example:

from rustybt.data.adapters.polygon_adapter import PolygonAdapter

# US stocks (developer tier for production)
adapter = PolygonAdapter(tier="developer", asset_type="stocks")

# Options
adapter_options = PolygonAdapter(tier="developer", asset_type="options")

For Forex Data¶

Polygon Adapter¶

Pros: - ✅ Real-time forex data - ✅ High-frequency resolutions - ✅ Good rate limits (paid tiers)

Cons: - ❌ Requires paid tier for production - ❌ Limited to major currency pairs

Use When: - Need real-time forex data - Trading major currency pairs - Have budget for professional data

Example:

from rustybt.data.adapters.polygon_adapter import PolygonAdapter

adapter = PolygonAdapter(tier="developer", asset_type="forex")
data = await adapter.fetch_ohlcv("EURUSD", start, end, "1h")

AlphaVantage Adapter¶

Pros: - ✅ Free tier available - ✅ Global forex coverage - ✅ Simple API

Cons: - ❌ Very strict rate limits (free: 5 req/min) - ❌ Delayed data - ❌ No real-time streaming

Use When: - Backtesting forex strategies - Research and development - Don't need real-time data - Limited budget

Example:

from rustybt.data.adapters.alphavantage_adapter import AlphaVantageAdapter

adapter = AlphaVantageAdapter(tier="free", asset_type="forex")
data = await adapter.fetch_ohlcv("EUR/USD", start, end, "1d")

For Custom Data¶

CSV Adapter¶

Pros: - ✅ Import any CSV format - ✅ Flexible schema mapping - ✅ Auto-detection of delimiters and dates - ✅ No rate limits (local files) - ✅ Timezone conversion support - ✅ Missing data handling strategies

Cons: - ❌ Requires data preparation - ❌ No automatic updates - ❌ Not suitable for live trading

Use When: - Have proprietary data sources - Testing with custom datasets - Academic research with specific data - Data from vendors not supported by adapters

Example:

from rustybt.data.adapters.csv_adapter import CSVAdapter, CSVConfig, SchemaMapping

config = CSVConfig(
    file_path="data/custom_data.csv",
    schema_mapping=SchemaMapping(
        date_column="Date",
        open_column="Open",
        high_column="High",
        low_column="Low",
        close_column="Close",
        volume_column="Volume"
    )
)

adapter = CSVAdapter(config)

Use Case Recommendations¶

Backtesting Historical Strategies¶

Recommended: YFinance (free) or CSV (custom data)

Why: No need for real-time data, want to minimize costs, need long historical data.

Setup:

from rustybt.data.adapters.yfinance_adapter import YFinanceAdapter

adapter = YFinanceAdapter()

# Fetch 10 years of daily data
data = await adapter.fetch(
    symbols=["SPY"],
    start_date=pd.Timestamp("2014-01-01"),
    end_date=pd.Timestamp("2024-01-01"),
    resolution="1d"
)

Live Algorithmic Trading (US Stocks)¶

Recommended: Alpaca (live) or Polygon (developer tier)

Why: Need real-time data, WebSocket support, high rate limits.

Setup:

from rustybt.data.adapters.alpaca_adapter import AlpacaAdapter

# Real-time trading
adapter = AlpacaAdapter(is_paper=False)

Cryptocurrency Trading¶

Recommended: CCXT

Why: Access to 100+ exchanges, real-time data, high-frequency support.

Setup:

from rustybt.data.adapters.ccxt_adapter import CCXTAdapter

adapter = CCXTAdapter(exchange_id="binance")

# High-frequency data
data = await adapter.fetch(
    symbols=["BTC/USDT"],
    start_date=pd.Timestamp("2024-01-01"),
    end_date=pd.Timestamp("2024-01-02"),
    resolution="1m"
)

Research and Development¶

Recommended: YFinance

Why: Free, no API key required, sufficient for testing and development.

Setup:

from rustybt.data.adapters.yfinance_adapter import YFinanceAdapter

adapter = YFinanceAdapter()

Global Market Analysis¶

Recommended: AlphaVantage (premium) or Polygon

Why: Access to global markets, multiple asset classes, professional data.

Setup:

from rustybt.data.adapters.alphavantage_adapter import AlphaVantageAdapter

# Global stocks
adapter = AlphaVantageAdapter(tier="premium", asset_type="stocks")

# Can fetch US, UK, European, Asian stocks
data_us = await adapter.fetch_ohlcv("AAPL", start, end, "1d")
data_uk = await adapter.fetch_ohlcv("IBM.LON", start, end, "1d")

Cost Optimization Strategies¶

Development Phase¶

Use Free Adapters:
YFinance for US stocks
CCXT for crypto (most exchanges free)
CSV for custom data
Test with Small Date Ranges:
Minimize API calls during development
Cache downloaded data locally
Use Paper Trading Modes:
Alpaca paper trading (free IEX feed)
Polygon free tier for testing

Production Phase¶

Evaluate Data Needs:
Real-time vs. 15-minute delayed
Frequency of updates required
Number of symbols to track
Choose Cost-Effective Tier:
Alpaca: $0 (paper) to enterprise pricing (live)
Polygon: $0 (limited) to $99/mo (developer)
AlphaVantage: $0 to $49.99/mo
Cache Aggressively:
Store historical data locally
Only fetch recent updates
Use bundle system for efficient storage

Performance Considerations¶

Rate Limits¶

Adapter	Free Tier	Paid Tier	Daily Limits
YFinance	~2000/hr	N/A	Informal
CCXT	Exchange-specific	Exchange-specific	Varies
Polygon	5/min	100/min	None
Alpaca	200/min	200/min	None
AlphaVantage	5/min	75/min	500-1200/day

Tip: For high-frequency needs, use Polygon (developer) or Alpaca.

Data Freshness¶

Adapter	Delay	Live Streaming
YFinance	15 minutes	No
CCXT	Exchange-specific	Yes (some)
Polygon	Real-time	Yes
Alpaca	Real-time (live)	Yes
AlphaVantage	Delayed	No

Tip: For live trading, use Polygon, Alpaca, or CCXT.

Migration Paths¶

From YFinance to Production¶

Path: YFinance → Alpaca (paper) → Alpaca (live)

Steps: 1. Develop strategy with YFinance (free) 2. Test with Alpaca paper trading (free) 3. Deploy to Alpaca live (paid) when ready

From Free to Paid Tiers¶

Path: Free tier → Starter tier → Professional tier

When to Upgrade: - Hit rate limits regularly - Need real-time data - Scaling to more symbols - Moving to production

Adapter Selection Guide¶

Overview¶

Quick Decision Tree¶

Adapter Comparison Matrix¶

Detailed Comparisons¶

For Cryptocurrency Data¶

CCXT Adapter¶

For US Stock Data¶

YFinance Adapter (Free)¶

Alpaca Adapter (Paper: Free, Live: Paid)¶

Polygon Adapter (Free-$99/mo)¶

For Forex Data¶

Polygon Adapter¶

AlphaVantage Adapter¶

For Custom Data¶

CSV Adapter¶

Use Case Recommendations¶

Backtesting Historical Strategies¶

Live Algorithmic Trading (US Stocks)¶

Cryptocurrency Trading¶

Research and Development¶

Global Market Analysis¶

Cost Optimization Strategies¶

Development Phase¶

Production Phase¶

Performance Considerations¶

Rate Limits¶

Data Freshness¶

Migration Paths¶

From YFinance to Production¶

From Free to Paid Tiers¶

See Also¶