investing-algorithm-framework 4.0.0__tar.gz → 4.1.0__tar.gz

{investing_algorithm_framework-4.0.0 → investing_algorithm_framework-4.1.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: investing-algorithm-framework
-Version: 4.0.0
+Version: 4.1.0
 Summary: A framework for creating trading bots
 Author: MDUYN
 Requires-Python: >=3.10
@@ -278,7 +278,7 @@ app.add_portfolio_configuration(
 We are continuously working on improving the performance of the framework. If
 you have any suggestions, please let us know.
 
-## Download
+## How to install
 
 You can download the framework with pypi.
 
@@ -286,6 +286,27 @@ You can download the framework with pypi.
 pip install investing-algorithm-framework
 ```
 
+## Installation for local development
+
+The framework is built with poetry. To install the framework for local development, you can run the following commands:
+
+> Make sure you have poetry installed. If you don't have poetry installed, you can find installation instructions [here](https://python-poetry.org/docs/#installation)
+
+```bash
+git clone http
+cd investing-algorithm-framework
+poetry install
+```
+
+### Running tests
+
+To run the tests, you can run the following command:
+
+```bash
+# In the root of the project
+python -m unittest discover -s tests
+```
+
 ## Disclaimer
 
 If you use this framework for your investments, do not risk money

{investing_algorithm_framework-4.0.0 → investing_algorithm_framework-4.1.0}/README.md

@@ -241,7 +241,7 @@ app.add_portfolio_configuration(
 We are continuously working on improving the performance of the framework. If
 you have any suggestions, please let us know.
 
-## Download
+## How to install
 
 You can download the framework with pypi.
 
@@ -249,6 +249,27 @@ You can download the framework with pypi.
 pip install investing-algorithm-framework
 ```
 
+## Installation for local development
+
+The framework is built with poetry. To install the framework for local development, you can run the following commands:
+
+> Make sure you have poetry installed. If you don't have poetry installed, you can find installation instructions [here](https://python-poetry.org/docs/#installation)
+
+```bash
+git clone http
+cd investing-algorithm-framework
+poetry install
+```
+
+### Running tests
+
+To run the tests, you can run the following command:
+
+```bash
+# In the root of the project
+python -m unittest discover -s tests
+```
+
 ## Disclaimer
 
 If you use this framework for your investments, do not risk money

{investing_algorithm_framework-4.0.0 → investing_algorithm_framework-4.1.0}/investing_algorithm_framework/__init__.py

@@ -11,7 +11,8 @@ from investing_algorithm_framework.domain import ApiException, \
     pretty_print_backtest_reports_evaluation, load_backtest_reports, \
     RESERVED_BALANCES, APP_MODE, AppMode, DATETIME_FORMAT, \
     load_backtest_report, BacktestDateRange, convert_polars_to_pandas, \
-    DateRange, get_backtest_report, DEFAULT_LOGGING_CONFIG
+    DateRange, get_backtest_report, DEFAULT_LOGGING_CONFIG, \
+    BacktestReport
 from investing_algorithm_framework.infrastructure import \
     CCXTOrderBookMarketDataSource, CCXTOHLCVMarketDataSource, \
     CCXTTickerMarketDataSource, CSVOHLCVMarketDataSource, \
@@ -70,7 +71,6 @@ __all__ = [
     "BacktestDateRange",
     "convert_polars_to_pandas",
     "DateRange",
-    "get_rsi",
     "get_peaks",
     "is_uptrend",
     "is_downtrend",
@@ -89,5 +89,6 @@ __all__ = [
     "is_divergence",
     "get_backtest_report",
     "AzureBlobStorageStateHandler",
-    "DEFAULT_LOGGING_CONFIG"
+    "DEFAULT_LOGGING_CONFIG",
+    "BacktestReport"
 ]

{investing_algorithm_framework-4.0.0 → investing_algorithm_framework-4.1.0}/investing_algorithm_framework/app/algorithm.py

@@ -163,15 +163,6 @@ class Algorithm:
     def data_sources(self):
         return self._data_sources
 
-    @property
-    def identifier(self):
-        """
-        Function to get a config instance. This allows users when
-        having access to the algorithm instance also to read the
-        configs of the app.
-        """
-        return self.configuration_service.config
-
     @property
     def config(self):
         """
@@ -179,7 +170,7 @@ class Algorithm:
         having access to the algorithm instance also to read the
         configs of the app.
         """
-        return self.configuration_service.config
+        return self.configuration_service.get_config()
 
     @property
     def description(self):
@@ -489,6 +480,16 @@ class Algorithm:
 
         return self.portfolio_service.find({{"market": market}})
 
+    def get_portfolios(self):
+        """
+        Function to get all portfolios of the algorithm. This function
+        will return all portfolios of the algorithm.
+
+        Returns:
+            List[Portfolio]: A list of all portfolios of the algorithm
+        """
+        return self.portfolio_service.get_all()
+
     def get_unallocated(self, market=None) -> float:
         """
         Function to get the unallocated balance of the portfolio. This
@@ -834,7 +835,7 @@ class Algorithm:
         return (position.amount * ticker["bid"] / total) * 100
 
     def get_position_percentage_of_portfolio_by_net_size(
-            self, symbol, market=None, identifier=None
+        self, symbol, market=None, identifier=None
     ) -> float:
         """
         Returns the percentage of the portfolio that is allocated to a

{investing_algorithm_framework-4.0.0 → investing_algorithm_framework-4.1.0}/investing_algorithm_framework/app/app.py

@@ -518,17 +518,6 @@ class App:
             trade_service=self.container.trade_service(),
         )
 
-        # Create all portfolios
-        portfolio_configuration_service = self.container \
-            .portfolio_configuration_service()
-        portfolio_configurations = portfolio_configuration_service.get_all()
-        portfolio_service = self.container.portfolio_service()
-
-        for portfolio_configuration in portfolio_configurations:
-            portfolio_service.create_portfolio_from_configuration(
-                portfolio_configuration
-            )
-
     def run(
         self,
         payload: dict = None,
@@ -774,20 +763,24 @@ class App:
 
     def run_backtest(
         self,
-        algorithm,
         backtest_date_range: BacktestDateRange,
+        initial_amount=None,
         pending_order_check_interval=None,
-        output_directory=None
+        output_directory=None,
+        algorithm: Algorithm = None
     ) -> BacktestReport:
         """
         Run a backtest for an algorithm. This method should be called when
         running a backtest.
 
         Parameters:
-            algorithm: The algorithm to run a backtest for (instance of
-                Algorithm)
             backtest_date_range: The date range to run the backtest for
                 (instance of BacktestDateRange)
+            initial_amount: The initial amount to start the backtest with.
+                This will be the amount of trading currency that the backtest
+                portfolio will start with.
+            algorithm: The algorithm to run a backtest for (instance of
+                Algorithm)
             pending_order_check_interval: str - pending_order_check_interval:
               The interval at which to check pending orders (e.g. 1h, 1d, 1w)
             output_directory: str - The directory to
@@ -796,8 +789,11 @@ class App:
         Returns:
             Instance of BacktestReport
         """
-        logger.info("Initializing backtest")
-        self.algorithm = algorithm
+        if algorithm is not None:
+            self.algorithm = algorithm
+
+        if self.algorithm is None:
+            raise OperationalException("No algorithm registered")
 
         self._initialize_app_for_backtest(
             backtest_date_range=backtest_date_range,
@@ -807,18 +803,20 @@ class App:
         self._initialize_algorithm_for_backtest(
             algorithm=self.algorithm
         )
+
         backtest_service = self.container.backtest_service()
-        configuration_service = self.container.configuration_service()
-        config = configuration_service.get_config()
-        backtest_service.resource_directory = config[RESOURCE_DIRECTORY]
 
         # Run the backtest with the backtest_service and collect the report
         report = backtest_service.run_backtest(
-            algorithm=self.algorithm, backtest_date_range=backtest_date_range
+            algorithm=self.algorithm,
+            initial_amount=initial_amount,
+            backtest_date_range=backtest_date_range
         )
         backtest_report_writer_service = self.container \
             .backtest_report_writer_service()
 
+        config = self.container.configuration_service().get_config()
+
         if output_directory is None:
             output_directory = os.path.join(
                 config[RESOURCE_DIRECTORY], "backtest_reports"

{investing_algorithm_framework-4.0.0 → investing_algorithm_framework-4.1.0}/investing_algorithm_framework/dependency_container.py

@@ -126,10 +126,11 @@ class DependencyContainer(containers.DeclarativeContainer):
         BacktestService,
         configuration_service=configuration_service,
         order_service=order_service,
-        portfolio_repository=portfolio_repository,
+        portfolio_service=portfolio_service,
         performance_service=performance_service,
         position_repository=position_repository,
         market_data_source_service=market_data_source_service,
+        portfolio_configuration_service=portfolio_configuration_service,
     )
     backtest_report_writer_service = providers.Factory(
         BacktestReportWriterService,

{investing_algorithm_framework-4.0.0 → investing_algorithm_framework-4.1.0}/investing_algorithm_framework/domain/models/backtesting/backtest_report.py

@@ -9,7 +9,6 @@ from investing_algorithm_framework.domain.metrics import \
 from investing_algorithm_framework.domain.models \
     .backtesting.backtest_date_range import BacktestDateRange
 from investing_algorithm_framework.domain.models.base_model import BaseModel
-from investing_algorithm_framework.domain.models.time_unit import TimeUnit
 from investing_algorithm_framework.domain.models.position import Position
 from investing_algorithm_framework.domain.models.trade import Trade
 from investing_algorithm_framework.domain.models.order import Order
@@ -358,15 +357,7 @@ class BacktestReport(BaseModel):
         self._time_unit = value
 
     def get_runs_per_day(self):
-
-        if self.time_unit is None:
-            return 0
-        elif TimeUnit.SECOND.equals(self.time_unit):
-            return 86400 / self.interval
-        elif TimeUnit.MINUTE.equals(self.time_unit):
-            return 1440 / self.interval
-        else:
-            return 24 / self.interval
+        return self.number_of_runs / self.number_of_days
 
     @property
     def backtest_start_date(self):
@@ -525,6 +516,9 @@ class BacktestReport(BaseModel):
     def get_trading_symbol(self) -> str:
         return self.trading_symbol
 
+    def get_initial_unallocated(self) -> float:
+        return self.initial_unallocated
+
     def add_symbol(self, symbol):
 
         if symbol not in self.symbols:

{investing_algorithm_framework-4.0.0 → investing_algorithm_framework-4.1.0}/investing_algorithm_framework/domain/models/portfolio/portfolio.py

@@ -2,6 +2,37 @@ from investing_algorithm_framework.domain.models.base_model import BaseModel
 
 
 class Portfolio(BaseModel):
+    """
+    Portfolio base class.
+
+    A portfolio is a collection of positions that are managed by an algorithm.
+
+    Attributes:
+    * identifier: str, unique identifier of the portfolio
+    * trading_symbol: str, trading symbol of the portfolio
+    * unallocated: float, the size of the trading symbol that is not
+      allocated. For example, if the trading symbol is USDT and the unallocated
+      is 1000, it means that the portfolio has 1000 USDT that is not
+        allocated to any position.
+    * net_size: float, net size of the portfolio is the initial balance of the
+        portfolio plus the all the net gains of the trades. The
+    * realized: float, the realized gain of the portfolio is the sum of all the
+        realized gains of the trades.
+    * total_revenue: float, the total revenue of the portfolio is the sum
+        of all the orders (price * size)
+    * total_cost: float, the total cost of the portfolio is the sum of all the
+        costs of the trades (price * size (for buy orders)
+        or -price * size (for sell orders))
+    * total_net_gain: float, the total net gain of the portfolio is the sum of
+        all the net gains of the trades
+    * total_trade_volume: float, the total trade volume of the
+        portfolio is the sum of all the sizes of the trades
+    * market: str, the market of the portfolio (e.g. BITVAVO, BINANCE)
+    * created_at: datetime, the datetime when the portfolio was created
+    * updated_at: datetime, the datetime when the portfolio was last updated
+    * initialized: bool, whether the portfolio is initialized or not
+    * initial_balance: float, the initial balance of the portfolio
+    """
 
     def __init__(
         self,

{investing_algorithm_framework-4.0.0 → investing_algorithm_framework-4.1.0}/investing_algorithm_framework/domain/models/portfolio/portfolio_configuration.py

@@ -6,6 +6,23 @@ from investing_algorithm_framework.domain.models.base_model import BaseModel
 
 
 class PortfolioConfiguration(BaseModel):
+    """
+    This class represents a portfolio configuration. It is used to
+    configure the portfolio that the user wants to create.
+
+    The portfolio configuration will have the following attributes:
+    - market: The market where the portfolio will be created
+    - trading_symbol: The trading symbol of the portfolio
+    - track_from: The date from which the portfolio will be tracked
+    - identifier: The identifier of the portfolio
+    - initial_balance: The initial balance of the portfolio
+
+    For backtesting, a portfolio configuration is used to create a
+    portfolio that will be used to simulate the trading of the algorithm. if
+    the user does not provide an initial balance, the portfolio will be created
+    with a balance of according to the initial balanace of
+        the PortfolioConfiguration class.
+    """
 
     def __init__(
         self,
@@ -22,7 +39,9 @@ class PortfolioConfiguration(BaseModel):
         self._initial_balance = initial_balance
 
         if self.identifier is None:
-            self._identifier = market.lower()
+            self._identifier = market.upper()
+        else:
+            self._identifier = identifier.upper()
 
         if track_from:
             self._track_from = parse(track_from)
@@ -35,8 +54,8 @@ class PortfolioConfiguration(BaseModel):
     @property
     def market(self):
 
-        if hasattr(self._market, "lower"):
-            return self._market.lower()
+        if hasattr(self._market, "upper"):
+            return self._market.upper()
 
         return self._market
 

{investing_algorithm_framework-4.0.0 → investing_algorithm_framework-4.1.0}/investing_algorithm_framework/infrastructure/repositories/portfolio_repository.py

@@ -20,7 +20,7 @@ class SQLPortfolioRepository(Repository):
             query = query.filter_by(market=market_query_param.upper())
 
         if identifier_query_param:
-            query = query.filter_by(identifier=identifier_query_param.lower())
+            query = query.filter_by(identifier=identifier_query_param.upper())
 
         if position_query_param:
             position = db.query(SQLPosition)\

{investing_algorithm_framework-4.0.0 → investing_algorithm_framework-4.1.0}/investing_algorithm_framework/infrastructure/repositories/repository.py

@@ -71,6 +71,7 @@ class Repository(ABC):
             try:
                 delete_object = self.get(object_id)
                 db.delete(delete_object)
+                db.commit()
                 return delete_object
             except SQLAlchemyError as e:
                 logger.error(e)

{investing_algorithm_framework-4.0.0 → investing_algorithm_framework-4.1.0}/investing_algorithm_framework/infrastructure/services/performance_service/performance_service.py

@@ -193,6 +193,10 @@ class PerformanceService:
         return: The total net gain percentage of the backtest
         """
         portfolio = self.portfolio_repository.find({"id": portfolio_id})
+
+        if portfolio.total_net_gain == 0:
+            return 0
+
         return portfolio.total_net_gain \
             / backtest_profile.initial_unallocated * 100
 

{investing_algorithm_framework-4.0.0 → investing_algorithm_framework-4.1.0}/investing_algorithm_framework/services/backtesting/backtest_service.py

@@ -1,11 +1,11 @@
 from datetime import datetime, timedelta
-
 import re
 import os
 import json
 import pandas as pd
 from dateutil import parser
 from tqdm import tqdm
+import logging
 
 from investing_algorithm_framework.domain import BacktestReport, \
     BACKTESTING_INDEX_DATETIME, TimeUnit, BacktestPosition, \
@@ -15,6 +15,7 @@ from investing_algorithm_framework.services.market_data_source_service import \
     MarketDataSourceService
 
 
+logger = logging.getLogger(__name__)
 BACKTEST_REPORT_FILE_NAME_PATTERN = (
     r"^report_\w+_backtest-start-date_\d{4}-\d{2}-\d{2}:\d{2}:\d{2}_"
     r"backtest-end-date_\d{4}-\d{2}-\d{2}:\d{2}:\d{2}_"
@@ -28,17 +29,18 @@ class BacktestService:
     """
 
     def __init__(
-            self,
-            market_data_source_service: MarketDataSourceService,
-            order_service,
-            portfolio_repository,
-            position_repository,
-            performance_service,
-            configuration_service
+        self,
+        market_data_source_service: MarketDataSourceService,
+        order_service,
+        portfolio_service,
+        position_repository,
+        performance_service,
+        configuration_service,
+        portfolio_configuration_service
     ):
         self._resource_directory = None
         self._order_service = order_service
-        self._portfolio_repository = portfolio_repository
+        self._portfolio_service = portfolio_service
         self._data_index = {
             TradingDataType.OHLCV: {},
             TradingDataType.TICKER: {}
@@ -49,6 +51,7 @@ class BacktestService:
             = market_data_source_service
         self._backtest_market_data_sources = []
         self._configuration_service = configuration_service
+        self._portfolio_configuration_service = portfolio_configuration_service
 
     @property
     def resource_directory(self):
@@ -59,7 +62,10 @@ class BacktestService:
         self._resource_directory = resource_directory
 
     def run_backtest(
-        self, algorithm, backtest_date_range: BacktestDateRange
+        self,
+        algorithm,
+        backtest_date_range: BacktestDateRange,
+        initial_amount=None
     ) -> BacktestReport:
         """
         Run a backtest for the given algorithm. This function will run
@@ -73,15 +79,40 @@ class BacktestService:
 
         At the end of the run all traces
 
-        Parameters:
+        Args:
             algorithm: The algorithm to run the backtest for
             backtest_date_range: The backtest date range
+            initial_amount: The initial amount of the backtest portfolio
 
         Returns:
             BacktestReport - The backtest report
         """
+        logging.info(
+            f"Running backtest for algorithm with name {algorithm.name}"
+        )
+
+        # Create backtest portfolio
+        portfolio_configurations = \
+            self._portfolio_configuration_service.get_all()
+
+        for portfolio_configuration in portfolio_configurations:
+
+            if self._portfolio_service.exists(
+                {"identifier": portfolio_configuration.identifier}
+            ):
+                # Delete existing portfolio
+                portfolio = self._portfolio_service.find(
+                    {"identifier": portfolio_configuration.identifier}
+                )
+                self._portfolio_service.delete(portfolio.id)
+
+            # Check if the portfolio configuration has a initial balance
+            self._portfolio_service.create_portfolio_from_configuration(
+                portfolio_configuration, initial_amount=initial_amount
+            )
+
         strategy_profiles = []
-        portfolios = self._portfolio_repository.get_all()
+        portfolios = self._portfolio_service.get_all()
         initial_unallocated = 0
 
         for portfolio in portfolios:
@@ -114,10 +145,23 @@ class BacktestService:
                 strategy=algorithm.get_strategy(strategy_profile.strategy_id),
                 index_date=index_date,
             )
-        return self.create_backtest_report(
+
+        report = self.create_backtest_report(
             algorithm, len(schedule), backtest_date_range, initial_unallocated
         )
 
+        # Cleanup backtest portfolio
+        portfolio_configurations = \
+            self._portfolio_configuration_service.get_all()
+
+        for portfolio_configuration in portfolio_configurations:
+            portfolio = self._portfolio_service.find(
+                {"identifier": portfolio_configuration.identifier}
+            )
+            self._portfolio_service.delete(portfolio.id)
+
+        return report
+
     def run_backtests(
         self, algorithms, backtest_date_range: BacktestDateRange
     ):
@@ -126,9 +170,9 @@ class BacktestService:
         backtests for the given algorithms and return a list of backtest
         reports.
 
-        Parameters
-            - algorithms: The algorithms to run the backtests for
-            - backtest_date_range: The backtest date range of the backtests
+        Args:
+            algorithms: The algorithms to run the backtests for
+            backtest_date_range: The backtest date range of the backtests
 
         Returns:
             List - A list of backtest reports
@@ -146,6 +190,9 @@ class BacktestService:
         return backtest_reports
 
     def run_backtest_for_profile(self, algorithm, strategy, index_date):
+        self._configuration_service.add_value(
+            BACKTESTING_INDEX_DATETIME, index_date
+        )
         algorithm.config[BACKTESTING_INDEX_DATETIME] = index_date
         market_data = {}
 
@@ -173,7 +220,7 @@ class BacktestService:
         calculate when the strategies should run based on the given start
         and end date. The schedule will be stored in a pandas DataFrame.
 
-        Parameters:
+        Args:
             strategies: The strategies to generate the schedule for
             start_date: The start date of the schedule
             end_date: The end date of the schedule
@@ -207,6 +254,7 @@ class BacktestService:
                     raise ValueError(f"Unsupported time unit: {time_unit}")
 
         schedule_df = pd.DataFrame(data)
+
         if schedule_df.empty:
             raise OperationalException(
                 "Could not generate schedule "
@@ -253,7 +301,7 @@ class BacktestService:
             BacktestReport: The backtest report instance of BacktestReport
         """
 
-        for portfolio in self._portfolio_repository.get_all():
+        for portfolio in self._portfolio_service.get_all():
             ids = [strategy.strategy_id for strategy in algorithm.strategies]
 
             # Check if strategy_id is None
@@ -440,7 +488,6 @@ class BacktestService:
         if a ticker market data source is registered for the symbol and market.
         """
         symbols = self._configuration_service.config[SYMBOLS]
-        print(symbols)
 
         if symbols is not None:
 

{investing_algorithm_framework-4.0.0 → investing_algorithm_framework-4.1.0}/investing_algorithm_framework/services/market_data_source_service/backtest_market_data_source_service.py

@@ -61,7 +61,6 @@ class BacktestMarketDataSourceService(MarketDataSourceService):
                     .get_config()[BACKTESTING_START_DATE],
                     backtest_end_date=configuration_service
                     .get_config()[BACKTESTING_END_DATE],
-                    market_credential_service=self._market_credential_service
                 )
 
     def get_data(self, identifier):