Merge branch 'develop' into feat/short

freqtrade/optimize/hyperopt_loss_calmar.py

@@ -0,0 +1,64 @@
+"""
+CalmarHyperOptLoss
+
+This module defines the alternative HyperOptLoss class which can be used for
+Hyperoptimization.
+"""
+from datetime import datetime
+from math import sqrt as msqrt
+from typing import Any, Dict
+
+from pandas import DataFrame
+
+from freqtrade.data.btanalysis import calculate_max_drawdown
+from freqtrade.optimize.hyperopt import IHyperOptLoss
+
+
+class CalmarHyperOptLoss(IHyperOptLoss):
+    """
+    Defines the loss function for hyperopt.
+
+    This implementation uses the Calmar Ratio calculation.
+    """
+
+    @staticmethod
+    def hyperopt_loss_function(
+        results: DataFrame,
+        trade_count: int,
+        min_date: datetime,
+        max_date: datetime,
+        config: Dict,
+        processed: Dict[str, DataFrame],
+        backtest_stats: Dict[str, Any],
+        *args,
+        **kwargs
+    ) -> float:
+        """
+        Objective function, returns smaller number for more optimal results.
+
+        Uses Calmar Ratio calculation.
+        """
+        total_profit = backtest_stats["profit_total"]
+        days_period = (max_date - min_date).days
+
+        # adding slippage of 0.1% per trade
+        total_profit = total_profit - 0.0005
+        expected_returns_mean = total_profit.sum() / days_period * 100
+
+        # calculate max drawdown
+        try:
+            _, _, _, high_val, low_val = calculate_max_drawdown(
+                results, value_col="profit_abs"
+            )
+            max_drawdown = (high_val - low_val) / high_val
+        except ValueError:
+            max_drawdown = 0
+
+        if max_drawdown != 0:
+            calmar_ratio = expected_returns_mean / max_drawdown * msqrt(365)
+        else:
+            # Define high (negative) calmar ratio to be clear that this is NOT optimal.
+            calmar_ratio = -20.0
+
+        # print(expected_returns_mean, max_drawdown, calmar_ratio)
+        return -calmar_ratio

freqtrade/optimize/hyperopt_tools.py

@@ -1,4 +1,3 @@
-
 import io
 import logging
 from copy import deepcopy
@@ -64,10 +63,11 @@ class HyperoptTools():
             'export_time': datetime.now(timezone.utc),
         }
         logger.info(f"Dumping parameters to {filename}")
-        rapidjson.dump(final_params, filename.open('w'), indent=2,
-                       default=hyperopt_serializer,
-                       number_mode=rapidjson.NM_NATIVE | rapidjson.NM_NAN
-                       )
+        with filename.open('w') as f:
+            rapidjson.dump(final_params, f, indent=2,
+                           default=hyperopt_serializer,
+                           number_mode=rapidjson.NM_NATIVE | rapidjson.NM_NAN
+                           )
 
     @staticmethod
     def try_export_params(config: Dict[str, Any], strategy_name: str, params: Dict):

freqtrade/optimize/space/decimalspace.py

@@ -7,11 +7,15 @@ class SKDecimal(Integer):
     def __init__(self, low, high, decimals=3, prior="uniform", base=10, transform=None,
                  name=None, dtype=np.int64):
         self.decimals = decimals
-        _low = int(low * pow(10, self.decimals))
-        _high = int(high * pow(10, self.decimals))
+
+        self.pow_dot_one = pow(0.1, self.decimals)
+        self.pow_ten = pow(10, self.decimals)
+
+        _low = int(low * self.pow_ten)
+        _high = int(high * self.pow_ten)
         # trunc to precision to avoid points out of space
-        self.low_orig = round(_low * pow(0.1, self.decimals), self.decimals)
-        self.high_orig = round(_high * pow(0.1, self.decimals), self.decimals)
+        self.low_orig = round(_low * self.pow_dot_one, self.decimals)
+        self.high_orig = round(_high * self.pow_dot_one, self.decimals)
 
         super().__init__(_low, _high, prior, base, transform, name, dtype)
 
@@ -25,9 +29,9 @@ class SKDecimal(Integer):
         return self.low_orig <= point <= self.high_orig
 
     def transform(self, Xt):
-        aa = [int(x * pow(10, self.decimals)) for x in Xt]
-        return super().transform(aa)
+        return super().transform([int(v * self.pow_ten) for v in Xt])
 
     def inverse_transform(self, Xt):
         res = super().inverse_transform(Xt)
-        return [round(x * pow(0.1, self.decimals), self.decimals) for x in res]
+        # equivalent to [round(x * pow(0.1, self.decimals), self.decimals) for x in res]
+        return [int(v) / self.pow_ten for v in res]