Add backtest data retrieve functionality

Author: MDUYN

investing_algorithm_framework/app/app.py

@@ -1246,9 +1246,28 @@ def run_vector_backtest(
             show_progress (bool): Whether to show progress bars during
                 data source initialization. This is useful for long-running
                 initialization processes.
+            dynamic_position_sizing (bool): Whether to use dynamic position
+                sizing based on volatility or other factors. Defaults to False.
 
         Returns:
             Backtest: Instance of Backtest
+
+        Examples:
+            # Basic usage
+            backtest = app.run_vector_backtest(
+                strategy=my_strategy,
+                backtest_date_range=my_date_range,
+                initial_amount=1000
+            )
+
+            # With custom storage and checkpoints
+            backtest = app.run_vector_backtest(
+                strategy=my_strategy,
+                backtest_date_range=my_date_range,
+                initial_amount=1000,
+                use_checkpoints=True,
+                backtest_storage_directory="./backtest_results"
+            )
         """
         # Use registered strategy if none provided
         if strategy is None:
@@ -1479,6 +1498,107 @@ def run_backtests(
 
         return backtests
 
+    def get_backtest_data(
+        self,
+        strategy: TradingStrategy,
+        backtest_date_range: BacktestDateRange,
+        show_progress: bool = True,
+        fill_missing_data: bool = True,
+    ) -> Dict[str, Any]:
+        """
+        Get all data sources with their corresponding data for a given
+        strategy and backtest window.
+
+        This method retrieves the market data for all data sources defined
+        in the strategy, considering the warmup window for each data source.
+        The data is returned as a dictionary where keys are data source
+        identifiers and values are the corresponding DataFrames.
+
+        Args:
+            strategy (TradingStrategy): The strategy containing the data
+                sources to retrieve data for.
+            backtest_date_range (BacktestDateRange): The date range for
+                the backtest window.
+            show_progress (bool): Whether to show progress bars during
+                data retrieval. Defaults to True.
+            fill_missing_data (bool): If True, missing time series data
+                entries will be filled automatically. Defaults to True.
+
+        Returns:
+            Dict[str, Any]: A dictionary where keys are data source
+                identifiers (e.g., "BTC/EUR_ohlcv") and values are the
+                corresponding data (typically pandas DataFrames).
+
+        Example:
+            ```python
+            from investing_algorithm_framework import (
+                create_app, TradingStrategy, BacktestDateRange, DataSource
+            )
+            from datetime import datetime, timezone
+
+            class MyStrategy(TradingStrategy):
+                data_sources = [
+                    DataSource(
+                        identifier="btc_data",
+                        symbol="BTC/EUR",
+                        time_frame="1h",
+                        warmup_window=100,
+                        market="BITVAVO"
+                    )
+                ]
+                # ... strategy implementation
+
+            app = create_app()
+            app.add_strategy(MyStrategy)
+
+            backtest_range = BacktestDateRange(
+                start_date=datetime(2024, 1, 1, tzinfo=timezone.utc),
+                end_date=datetime(2024, 6, 1, tzinfo=timezone.utc)
+            )
+
+            # Get all data for the strategy
+            data = app.get_backtest_data(
+                strategy=MyStrategy(),
+                backtest_date_range=backtest_range
+            )
+
+            # Access data by identifier
+            btc_df = data["btc_data"]
+            ```
+
+        Raises:
+            OperationalException: If no data sources are defined in the
+                strategy or if data cannot be retrieved for a data source.
+        """
+        # Get data sources from the strategy
+        data_sources = strategy.data_sources
+
+        if data_sources is None or len(data_sources) == 0:
+            raise OperationalException(
+                "No data sources defined in the strategy. "
+                "Please define data sources to retrieve backtest data."
+            )
+
+        # Setup backtest data providers
+        self.initialize_data_sources_backtest(
+            data_sources=data_sources,
+            backtest_date_range=backtest_date_range,
+            show_progress=show_progress,
+            fill_missing_data=fill_missing_data,
+        )
+
+        # Get the data provider service
+        data_provider_service = self.container.data_provider_service()
+
+        # Retrieve vectorized backtest data for all data sources
+        data = data_provider_service.get_vectorized_backtest_data(
+            data_sources=data_sources,
+            start_date=backtest_date_range.start_date,
+            end_date=backtest_date_range.end_date,
+        )
+
+        return data
+
     def run_backtest(
         self,
         backtest_date_range: BacktestDateRange,
@@ -1545,10 +1665,10 @@ def run_backtest(
                 the backtest. This is useful for long-running backtests.
             market (str): The market to use for the backtest. This is used
                 to create a portfolio configuration if no portfolio
-                configuration is provided.
+                configuration is provided in the strategy.
             trading_symbol (str): The trading symbol to use for the backtest.
                 This is used to create a portfolio configuration if no
-                portfolio configuration is provided.
+                portfolio configuration is provided in the strategy.
             fill_missing_data (bool): If True (default), missing time series
                 data entries will be filled automatically before running the
                 backtest.

investing_algorithm_framework/domain/data_provider.py

@@ -32,9 +32,11 @@ class DataProvider(ABC):
         time_frame (Optional[str]): The time frame for the data. This is
             useful for data providers that support multiple time frames.
             Example: "1m", "5m", "1h", "1d"
-        window_size (Optional[int]): The window size for the data. This is
-            useful for data providers that support multiple window sizes.
-            Example: 100, 200, 500
+        warmup_window (Optional[int]): The warmup window size for the data.
+            This is useful for data providers that support multiple window
+            sizes. Example: 100, 200, 500
+        window_size (Optional[int]): Deprecated. Use warmup_window instead.
+            Will be removed in release 0.8.0.
         storage_path (Optional[str]): The path to the storage location
             for the data. This is useful for data providers that support
             saving data to a file
@@ -84,9 +86,12 @@ def __init__(
                 can be set later. This is useful for data providers
                 that support multiple time frames.
                 Example: "1m", "5m", "1h", "1d"
-            window_size (int): The window size for the data. This is optional
-                and can be set later. This is useful for data providers that
-                support multiple window sizes. Example: 100, 200, 500
+            warmup_window (int): The warmup window size for the data. This is
+                optional and can be set later. This is useful for data
+                providers that support multiple window sizes.
+                Example: 100, 200, 500
+            window_size (int): Deprecated. Use warmup_window instead.
+                Will be removed in release 0.8.0.
             storage_path (str): The path to the storage location for the data.
                 This is optional and can be set later. This is useful for data
                 providers that support saving data to a file.

investing_algorithm_framework/domain/models/data/data_source.py

@@ -1,3 +1,4 @@
+import warnings
 from dataclasses import dataclass
 from datetime import datetime, timezone, timedelta
 from typing import Union
@@ -16,15 +17,21 @@ class DataSource:
     DataSource(
         symbol="BTC/EUR",
         data_type="ohlcv",
-        window_size=200,
+        warmup_window=200,
         market="BITVAVO",
         identifier="BTC/EUR_ohlcv"
     )
+
+    .. deprecated::
+        The `window_size` parameter is deprecated and will be removed in
+        release 0.8.0. Please use `warmup_window` instead.
     """
     identifier: str = None
     data_provider_identifier: str = None
     data_type: Union[DataType, str] = None
     symbol: str = None
+    warmup_window: int = None
+    # Deprecated: use warmup_window instead. Will be removed in release 0.8.0
     window_size: int = None
     time_frame: Union[TimeFrame, str] = None
     market: str = None
@@ -36,6 +43,19 @@ class DataSource:
     save: bool = False
 
     def __post_init__(self):
+        # Handle backward compatibility for window_size -> warmup_window
+        # window_size is deprecated and will be removed in release 0.8.0
+        if self.window_size is not None:
+            warnings.warn(
+                "The 'window_size' parameter is deprecated and will be "
+                "removed in release 0.8.0. Please use 'warmup_window' instead.",
+                DeprecationWarning,
+                stacklevel=2
+            )
+            # If warmup_window is not set, use window_size value
+            if self.warmup_window is None:
+                object.__setattr__(self, 'warmup_window', self.window_size)
+
         # Convert data_type and time_frame to their respective enums if needed
         if isinstance(self.data_type, str):
             object.__setattr__(self, 'data_type',
@@ -107,8 +127,8 @@ def get_identifier(self):
             if self.market is not None:
                 identifier += f"_{self.market}"
 
-            if self.window_size is not None:
-                identifier += f"_{self.window_size}"
+            if self.warmup_window is not None:
+                identifier += f"_{self.warmup_window}"
 
             return identifier
 
@@ -133,7 +153,7 @@ def __repr__(self):
             f"DataSource(identifier={self.identifier}, "
             f"data_provider_identifier={self.data_provider_identifier}, "
             f"data_type={self.data_type}, symbol={self.symbol}, "
-            f"window_size={self.window_size}, time_frame={self.time_frame}, "
+            f"warmup_window={self.warmup_window}, time_frame={self.time_frame}, "
             f"market={self.market}, storage_path={self.storage_path}, "
             f"pandas={self.pandas}, date={self.date}, "
             f"start_date={self.start_date}, end_date={self.end_date}, "
@@ -152,10 +172,10 @@ def __eq__(self, other):
         """
         if DataType.OHLCV.equals(self.data_type):
 
-            if other.time_frame is None and other.window_size is None:
+            if other.time_frame is None and other.warmup_window is None:
                 return (self.data_type == other.data_type and
                         self.symbol == other.symbol)
-            elif self.time_frame is None and self.window_size is None:
+            elif self.time_frame is None and self.warmup_window is None:
                 return (self.data_type == other.data_type and
                         self.symbol == other.symbol)
 
@@ -167,7 +187,7 @@ def __eq__(self, other):
         elif DataType.CUSTOM.equals(self.data_type):
             return (self.data_type == other.data_type and
                     self.symbol == other.symbol and
-                    self.window_size == other.window_size and
+                    self.warmup_window == other.warmup_window and
                     self.time_frame == other.time_frame and
                     self.market == other.market)
 
@@ -180,11 +200,11 @@ def __eq__(self, other):
 
     def create_start_date_data(self, index_date: datetime) -> datetime:
 
-        if self.window_size is None or self.time_frame is None:
+        if self.warmup_window is None or self.time_frame is None:
             return index_date
 
         return index_date - \
-            (self.window_size * timedelta(
+            (self.warmup_window * timedelta(
                 minutes=self.time_frame.amount_of_minutes
             ))
 
@@ -216,7 +236,7 @@ def get_number_of_required_data_points(
         total_minutes = delta.total_seconds() / 60
         data_points = total_minutes / self.time_frame.amount_of_minutes
 
-        if self.window_size is not None:
-            data_points += self.window_size
+        if self.warmup_window is not None:
+            data_points += self.warmup_window
 
         return int(data_points)

investing_algorithm_framework/download_data.py

@@ -73,7 +73,8 @@ def download(
     data_type: Union[str, DataType] = DataType.OHLCV,
     start_date: Union[datetime, str] = None,
     end_date: Union[datetime, str] = None,
-    window_size: int = 200,
+    warmup_window: int = 200,
+    window_size: int = None,  # Deprecated: use warmup_window instead
     pandas: bool = True,
     save: bool = True,
     storage_path: Union[str, Path] = None,
@@ -91,20 +92,32 @@ def download(
             download (e.g., "ohlcv", "ticker").
         start_date (str): The start date for the data download.
         end_date (str): The end date for the data download.
-        window_size (int): The size of the data window.
+        warmup_window (int): The size of the warmup window.
+        window_size (int): Deprecated. Use warmup_window instead.
+            Will be removed in release 0.8.0.
         pandas (bool): Whether to return the data as a pandas DataFrame.
         save (bool): Whether to save the downloaded data.
         storage_path (str): The directory to save the downloaded data.
         save_to (str): The path to save the downloaded data. If provided,
             it overrides storage_path.
         time_frame (str): The time frame for the data download.
         date (str): The date for the data download.
-        window_size (int): The size of the data window.
-        pandas (bool): Whether to return the data as a pandas DataFrame.
 
     Returns:
         None
     """
+    import warnings
+
+    # Handle backward compatibility for window_size
+    if window_size is not None:
+        warnings.warn(
+            "The 'window_size' parameter is deprecated and will be "
+            "removed in release 0.8.0. Please use 'warmup_window' instead.",
+            DeprecationWarning,
+            stacklevel=2
+        )
+        if warmup_window == 200:  # Default value, use window_size
+            warmup_window = window_size
     configuration_service = ConfigurationService()
     market_credential_service = MarketCredentialService()
     data_provider_service = DataProviderService(
@@ -148,7 +161,7 @@ def download(
         date=date,
         start_date=start_date,
         end_date=end_date,
-        window_size=window_size,
+        warmup_window=warmup_window,
         pandas=pandas,
         save=save,
         storage_path=storage_path
@@ -172,7 +185,8 @@ def download_v2(
     data_type: Union[str, DataType] = DataType.OHLCV,
     start_date: Union[datetime, str] = None,
     end_date: Union[datetime, str] = None,
-    window_size: int = 200,
+    warmup_window: int = 200,
+    window_size: int = None,  # Deprecated: use warmup_window instead
     pandas: bool = True,
     save: bool = True,
     storage_path: Union[str, Path] = None,
@@ -191,14 +205,29 @@ def download_v2(
         start_date: Start date for data range
         end_date: End date for data range
         date: Specific date for data download
-        window_size: Size of the data window
+        warmup_window: Size of the warmup window
+        window_size: Deprecated. Use warmup_window instead.
+            Will be removed in release 0.8.0.
         pandas: Whether to return the data as a pandas DataFrame
         save: Whether to save the data to disk
         storage_path: Base directory for storing files
 
     Returns:
         DownloadResult with .data (DataFrame) and .path (Path or None)
     """
+    import warnings
+
+    # Handle backward compatibility for window_size
+    if window_size is not None:
+        warnings.warn(
+            "The 'window_size' parameter is deprecated and will be "
+            "removed in release 0.8.0. Please use 'warmup_window' instead.",
+            DeprecationWarning,
+            stacklevel=2
+        )
+        if warmup_window == 200:  # Default value, use window_size
+            warmup_window = window_size
+
     # Parse dates if they are strings
     parsed_start_date = start_date
     parsed_end_date = end_date
@@ -220,7 +249,7 @@ def download_v2(
         data_type=data_type,
         start_date=start_date,
         end_date=end_date,
-        window_size=window_size,
+        warmup_window=warmup_window,
         pandas=pandas,
         save=save,
         storage_path=storage_path,