Overview The DataManager class handles loading market data from various providers including sample data, CSV files, and external APIs like Alpha Vantage. It provides a unified interface for data access with built-in caching and catalog management.
Constructor from glowback import DataManager dm = DataManager()
Creates a new DataManager instance with an asynchronous runtime for data operations.
A new DataManager instance
The DataManager creates its own async runtime internally. You don’t need to worry about async/await in your Python code.
Methods load_data Load market data for a specific symbol and date range.
from glowback import DataManager, Symbol dm = DataManager() dm.add_sample_provider() # Add data source first symbol = Symbol( "AAPL" , "NASDAQ" , "equity" ) bars = dm.load_data( symbol = symbol, start_date = "2020-01-01T00:00:00Z" , end_date = "2023-12-31T23:59:59Z" , resolution = "day" ) print ( f "Loaded { len (bars) } bars" ) for bar in bars[: 5 ]: # First 5 bars print ( f " { bar.timestamp } : O= { bar.open } H= { bar.high } L= { bar.low } C= { bar.close } " )
The symbol to load data for
Start date in RFC3339 format (e.g., “2020-01-01T00:00:00Z”)
End date in RFC3339 format (e.g., “2023-12-31T23:59:59Z”)
Time resolution: “minute”/“1m”, “hour”/“1h”, or “day”/“1d”
List of Bar objects containing OHLCV data
add_sample_provider Add a sample data provider for testing and development.
dm = DataManager() dm.add_sample_provider() # Now you can load sample data symbol = Symbol( "AAPL" , "NASDAQ" , "equity" ) bars = dm.load_data( symbol = symbol, start_date = "2020-01-01T00:00:00Z" , end_date = "2023-12-31T23:59:59Z" , resolution = "day" )
The sample provider generates synthetic data useful for testing strategies without needing real market data.
add_csv_provider Add a CSV data provider to load data from local files.
dm = DataManager() dm.add_csv_provider( base_path = "./data" ) # CSV files should be named: {symbol}_{exchange}_{resolution}.csv # Example: AAPL_NASDAQ_day.csv symbol = Symbol( "AAPL" , "NASDAQ" , "equity" ) bars = dm.load_data( symbol = symbol, start_date = "2020-01-01T00:00:00Z" , end_date = "2023-12-31T23:59:59Z" , resolution = "day" )
Base directory path containing CSV files
CSV files should have the following columns:
timestamp, open, high, low, close, volume 2020-01-02T00:00:00Z, 100.25, 102.50, 99.75, 101.00, 1000000 2020-01-03T00:00:00Z, 101.00, 103.25, 100.50, 102.75, 1200000
add_alpha_vantage_provider Add an Alpha Vantage API provider for fetching real market data.
dm = DataManager() dm.add_alpha_vantage_provider( api_key = "YOUR_API_KEY" ) symbol = Symbol( "AAPL" , "NASDAQ" , "equity" ) bars = dm.load_data( symbol = symbol, start_date = "2020-01-01T00:00:00Z" , end_date = "2023-12-31T23:59:59Z" , resolution = "day" )
Alpha Vantage free tier has rate limits:
5 API calls per minute
500 API calls per day
Consider caching data locally or using a premium API key for production use. get_catalog_stats Get statistics about the data catalog.
dm = DataManager() dm.add_sample_provider() # Load some data first symbol = Symbol( "AAPL" , "NASDAQ" , "equity" ) dm.load_data( symbol = symbol, start_date = "2020-01-01T00:00:00Z" , end_date = "2023-12-31T23:59:59Z" , resolution = "day" ) # Get catalog stats stats = dm.get_catalog_stats() print ( f "Total symbols: { stats.total_symbols } " ) print ( f "Total records: { stats.total_records } " ) print ( f "Date range: { stats.date_range_start } to { stats.date_range_end } " )
Statistics object with the following attributes:
total_symbols (int): Number of unique symbols in catalogtotal_records (int): Total number of data recordsdate_range_start (str): Earliest date in catalog (RFC3339 format)date_range_end (str): Latest date in catalog (RFC3339 format) get_provider_count Get the number of configured data providers.
dm = DataManager() print ( f "Providers: { dm.get_provider_count() } " ) # 0 dm.add_sample_provider() print ( f "Providers: { dm.get_provider_count() } " ) # 1 dm.add_csv_provider( "./data" ) print ( f "Providers: { dm.get_provider_count() } " ) # 2
Number of configured data providers
Complete example Here’s a complete example using multiple data providers:
from glowback import DataManager, Symbol # Initialize DataManager dm = DataManager() # Add multiple data providers dm.add_sample_provider() # For testing dm.add_csv_provider( "./historical_data" ) # Local CSV files dm.add_alpha_vantage_provider( "YOUR_API_KEY" ) # External API print ( f "Configured { dm.get_provider_count() } data providers" ) # Load data for multiple symbols symbols = [ Symbol( "AAPL" , "NASDAQ" , "equity" ), Symbol( "MSFT" , "NASDAQ" , "equity" ), Symbol( "GOOGL" , "NASDAQ" , "equity" ), ] for symbol in symbols: try : bars = dm.load_data( symbol = symbol, start_date = "2020-01-01T00:00:00Z" , end_date = "2023-12-31T23:59:59Z" , resolution = "day" ) print ( f "Loaded { len (bars) } bars for { symbol.symbol } " ) # Display first and last bar if bars: first_bar = bars[ 0 ] last_bar = bars[ - 1 ] print ( f " First: { first_bar.timestamp } - Close: $ { first_bar.close } " ) print ( f " Last: { last_bar.timestamp } - Close: $ { last_bar.close } " ) except Exception as e: print ( f "Error loading { symbol.symbol } : { e } " ) # Display catalog statistics stats = dm.get_catalog_stats() print ( f " \n Catalog Stats:" ) print ( f " Symbols: { stats.total_symbols } " ) print ( f " Records: { stats.total_records :,} " ) print ( f " Range: { stats.date_range_start } to { stats.date_range_end } " )
Data provider priority When multiple providers are configured, GlowBack queries them in the order they were added. The first provider that returns data wins.
dm = DataManager() # Priority order: dm.add_csv_provider( "./data" ) # 1. Try local cache first dm.add_alpha_vantage_provider( "KEY" ) # 2. Then try API dm.add_sample_provider() # 3. Fall back to sample data # This will try CSV first, then API, then sample data bars = dm.load_data(symbol, start_date, end_date, resolution)
See also
Symbol class Learn about the Symbol type
Bar class Understand OHLCV bar data
Data providers Learn more about data sources
BacktestEngine Use loaded data in backtests