Data Provider

The Data Provider is the heart of market data access in PyEventBT. It abstracts the complexity of data sourcing, seamlessly switching between historical CSV files for backtesting and real-time MetaTrader 5 feeds for live trading.

Unified Interface

Write your strategy once. The Data Provider ensures your logic works identically whether you are running a backtest or trading live.

Key Features

Unified API

A single API for both historical and live data. No code changes required to go live.

Multi-Timeframe

Access data from multiple timeframes simultaneously (e.g., trade on H1 while checking trends on D1).

Auto-Management

Handles data loading, caching, and updates automatically, so you can focus on strategy logic.

Indicator Ready

Returns data structures perfect for computing technical indicators like ATR, RSI, or Moving Averages.

Accessing Data

In your strategy, the data provider is accessible via the modules object.

Get Latest Bars

Retrieve a sequence of historical bars, useful for calculating indicators or analyzing trends.

@strategy.custom_signal_engine(...)
def my_strategy(event: BarEvent, modules: Modules):
    # Get the last 50 bars for H1 timeframe
    bars = modules.DATA_PROVIDER.get_latest_bars(
        symbol=event.symbol,
        timeframe=StrategyTimeframes.ONE_HOUR,
        N=50
    )

    if bars is None or bars.height < 50:
        return [] # Not enough data yet

    # Access data columns
    closes = bars.select('close').to_numpy()
    highs = bars.select('high').to_numpy()

Get Single Bar

Retrieve just the most recent completed bar.

bar = modules.DATA_PROVIDER.get_latest_bar(
    symbol='EURUSD',
    timeframe=StrategyTimeframes.ONE_HOUR
)

if bar:
    print(f"Last close: {bar.close}")

Current Prices

Get the absolute latest Bid/Ask prices (Live trading context).

# Useful for precise entry/exit in live execution
bid = modules.DATA_PROVIDER.get_latest_bid(symbol)
ask = modules.DATA_PROVIDER.get_latest_ask(symbol)

spread = ask - bid

Common Use Cases

Here are practical examples of how to use the Data Provider in real trading scenarios.

1. Moving Average Crossover

1. Request Bars: Fetch enough history to calculate your longest moving average.
2. Calculate Indicators: Compute the Fast and Slow SMAs.
3. Generate Signal: Compare the values to determine entries.

def ma_crossover_logic(symbol, modules):
    # 1. Request Data (need at least 30 bars for the slow SMA)
    bars = modules.DATA_PROVIDER.get_latest_bars(
        symbol,
        StrategyTimeframes.ONE_HOUR,
        50
    )
    if bars is None or bars.height < 30: return

    # 2. Calculate Indicators
    closes = bars.select('close').to_numpy().flatten()
    sma_fast = closes[-10:].mean()  # 10-period
    sma_slow = closes[-30:].mean()  # 30-period

    # 3. Logic
    if sma_fast > sma_slow:
        # Bullish condition
        pass

2. Multi-Timeframe Analysis

Combine data from different timeframes to confirm trends.

# Entry Timeframe: 1 Hour
hourly_bars = modules.DATA_PROVIDER.get_latest_bars(
    symbol, StrategyTimeframes.ONE_HOUR, 20
)

# Trend Timeframe: 1 Day
daily_bars = modules.DATA_PROVIDER.get_latest_bars(
    symbol, StrategyTimeframes.ONE_DAY, 20
)

# Check if daily trend is UP before taking hourly buy signals
daily_closes = daily_bars.select('close').to_numpy()
is_uptrend = daily_closes[-1] > daily_closes[-5] # Simple trend check

Data Sources & Configuration

The Data Provider abstracts the source, but it’s helpful to know where data comes from.

Backtesting Mode

Reads from CSV files in your configured csv_dir. File structure is expected to be simple:

• Directorydata/

  • EURUSD.csv
  • GBPUSD.csv
  • SPX500.csv

CSV Format: datetime,open,high,low,close,volume

Live Mode

Connects directly to the MetaTrader 5 terminal.

• Real-time ticks: Updates bid/ask instantly.
• Bar generation: Aggregates ticks into bars automatically.

Best Practices

Check for None

Always check if the returned data is None or if the list is empty before performing calculations. At the start of a backtest, data might not be available yet.

• Performance: Request only the N bars you need for your calculation. Fetching 1000 bars when you only need 14 for an RSI calculation is inefficient.
• Safety: Use get_latest_bid/ask for execution logic to ensure you are trading at the most current price, rather than the close of the last bar.

Methods

Here is the complete list of methods available in the DATA_PROVIDER module.

Get Latest Bars

Returns the most recent N bars for a given symbol and timeframe.

def get_latest_bars(self, symbol: str, timeframe: str, N: int) -> Bars

Parameter	Type	Description
symbol	str	The instrument symbol (e.g., “EURUSD”).
timeframe	str	The timeframe (e.g., StrategyTimeframes.ONE_HOUR).
N	int	Number of bars to retrieve.

Returns: A Bars object containing the requested history.

Get Latest Bar

Returns the single latest completed bar.

def get_latest_bar(self, symbol: str, timeframe: str) -> Bar

Returns: A single Bar object representing the most recent completed period.

Get Latest Tick

Returns the most recent tick data (Real-time).

def get_latest_tick(self, symbol: str) -> dict

Returns: A dictionary matching the MT5 Tick structure (bid, ask, last, volume, time).

Get Latest Bid

Returns the absolute latest bid price.

def get_latest_bid(self, symbol: str) -> Decimal

Get Latest Ask

Returns the absolute latest ask price.

def get_latest_ask(self, symbol: str) -> Decimal

Get Latest Datetime

Returns the timestamp of the last data point.

def get_latest_datetime(self, symbol: str, timeframe: str) -> Timestamp