Check if dataframe has datetime data type

Author: MDUYN

investing_algorithm_framework/domain/models/trade/trade.py

@@ -209,12 +209,21 @@ def is_manual_stop_loss_trigger(
             # If dataframes are provided, we use the dataframe to calculate
             # the stop loss price
             if ohlcv_df is not None:
-                filtered_df = ohlcv_df.filter(
-                    pl.col('Datetime') >= self.opened_at.strftime(
-                        DATETIME_FORMAT
+                column_type = ohlcv_df['Datetime'].dtype
+
+                if isinstance(column_type, pl.Datetime):
+                    filtered_df = ohlcv_df.filter(
+                        pl.col('Datetime') >= self.opened_at
+                    )
+                else:
+                    filtered_df = ohlcv_df.filter(
+                        pl.col('Datetime') >= self.opened_at.strftime(
+                            DATETIME_FORMAT
+                        )
                     )
-                )
+
                 prices = filtered_df['Close'].to_numpy()
+
             highest_price = max(prices)
             stop_loss_price = highest_price * (1 - stop_loss_percentage / 100)
             return current_price <= stop_loss_price

investing_algorithm_framework/services/order_service/order_backtest_service.py

@@ -1,6 +1,8 @@
+import datetime
 import logging
 
 import pandas as pd
+import polars as pl
 
 from investing_algorithm_framework.domain import BACKTESTING_INDEX_DATETIME, \
     OrderStatus, BACKTESTING_PENDING_ORDER_CHECK_INTERVAL, \
@@ -160,9 +162,18 @@ def has_executed(self, order, ohlcv_data_frame):
         created_at = order.get_created_at()
         order_side = order.get_order_side()
         order_price = order.get_price()
+        column_type = ohlcv_data_frame['Datetime'].dtype
 
-        # Filter OHLCV data after the order creation time
-        ohlcv_data_after_order = ohlcv_data_frame.loc[created_at:]
+        if isinstance(column_type, pl.Datetime):
+            ohlcv_data_after_order = ohlcv_data_frame.filter(
+                pl.col('Datetime') >= created_at
+            )
+        else:
+            ohlcv_data_after_order = ohlcv_data_frame.filter(
+                pl.col('Datetime') >= created_at.strftime(
+                    self.configuration_service.config["DATETIME_FORMAT"]
+                )
+            )
 
         # Check if the order execution conditions are met
         if OrderSide.BUY.equals(order_side):

tests/infrastructure/market_data_sources/test_ccxt_ohlcv_backtest_market_data_source.py

@@ -3,7 +3,7 @@
 from unittest import TestCase
 
 from investing_algorithm_framework.domain import RESOURCE_DIRECTORY, \
-    BACKTEST_DATA_DIRECTORY_NAME
+    BACKTEST_DATA_DIRECTORY_NAME, DATETIME_FORMAT
 from investing_algorithm_framework.infrastructure import \
     CCXTOHLCVBacktestMarketDataSource
 
@@ -97,6 +97,9 @@ def test_right_columns(self):
             timeframe="15m",
             window_size=200
         )
+        data_source.config = {
+            "DATETIME_FORMAT": DATETIME_FORMAT
+        }
         data_source.prepare_data(
             config={
                 RESOURCE_DIRECTORY: self.resource_dir,
@@ -107,7 +110,11 @@ def test_right_columns(self):
         )
         self.assertEqual(200, data_source.window_size)
         self.assertEqual(csv_file_path, data_source._create_file_path())
-
+        df = data_source\
+            .get_data(datetime(year=2023, month=12, day=17, hour=0, minute=0))
+        self.assertEqual(
+            ["Datetime", "Open", "High", "Low", "Close", "Volume"], df.columns
+        )
 
 # def test_start_date(self):
     #     start_date = datetime(2023, 12, 1)
@@ -151,26 +158,7 @@ def test_right_columns(self):
     #     data_source.start_date = datetime(2023, 12, 25)
     #     data_source.end_date = datetime(2023, 12, 16)
     #     self.assertTrue(data_source.empty())
-    #
-    # def test_get_data(self):
-    #     file_name = "OHLCV_BTC-EUR_BINANCE_15m_2023-12-" \
-    #                 "01:00:00_2023-12-25:00:00.csv"
-    #     datasource = CSVOHLCVMarketDataSource(
-    #         csv_file_path=f"{self.resource_dir}/"
-    #                       "market_data_sources/"
-    #                       f"{file_name}"
-    #     )
-    #     number_of_runs = 0
-    #
-    #     while not datasource.empty():
-    #         data = datasource.get_data()
-    #         datasource.start_date = datasource.start_date + timedelta(days=1)
-    #         datasource.end_date = datasource.end_date + timedelta(days=1)
-    #         self.assertTrue(len(data) > 0)
-    #         number_of_runs += 1
-    #
-    #     self.assertTrue(number_of_runs > 0)
-    #
+
     # def test_get_identifier(self):
     #     file_name = "OHLCV_BTC-EUR_BINANCE_15m_2023-12-" \
     #                 "01:00:00_2023-12-25:00:00.csv"
@@ -214,3 +202,4 @@ def test_right_columns(self):
     #         timeframe="15m"
     #     )
     #     self.assertEqual("15m", datasource.get_timeframe())
+

tests/infrastructure/market_data_sources/test_csv_ohlcv_market_data_source.py

@@ -31,12 +31,17 @@ def setUp(self) -> None:
     def test_right_columns(self):
         file_name = "OHLCV_BTC-EUR_BINANCE" \
                     "_2h_2023-08-07:07:59_2023-12-02:00:00.csv"
-        CSVOHLCVMarketDataSource(
+        data_source = CSVOHLCVMarketDataSource(
             csv_file_path=f"{self.resource_dir}/"
                           "market_data_sources/"
                           f"{file_name}",
             window_size=10
         )
+        df = data_source \
+            .get_data(datetime(year=2023, month=12, day=17, hour=0, minute=0))
+        self.assertEqual(
+            ["Datetime", "Open", "High", "Low", "Close", "Volume"], df.columns
+        )
 
     def test_throw_exception_when_missing_column_names_columns(self):
         file_name = "OHLCV_BTC-EUR_BINANCE_2h_NO_COLUMNS_2023-" \
@@ -114,6 +119,7 @@ def test_get_data(self):
             datasource.start_date = datasource.start_date + timedelta(days=1)
             datasource.end_date = datasource.end_date + timedelta(days=1)
             self.assertTrue(len(data) > 0)
+            self.assertAlmostEqual(10, len(data), 2)
             self.assertTrue(isinstance(data, DataFrame))
             number_of_runs += 1
 
@@ -170,3 +176,6 @@ def test_get_timeframe(self):
             window_size=10,
         )
         self.assertEqual("15m", datasource.get_timeframe())
+
+    def test_get_data(self):
+        pass