added daily sharpe ratio hyperopt loss method, ty @djacky (#2826)

* more consistent backtesting tables and labels

* added rounding to Tot Profit % on Sell Reasosn table to be consistent with other percentiles on table.

* added daily sharpe ratio hyperopt loss method, ty @djacky

* removed commented code

* removed unused profit_abs

* added proper slippage to each trade

* replaced use of old value total_profit

* Align quotes in same area

* added daily sharpe ratio test and modified hyperopt_loss_sharpe_daily

* fixed some more line alignments

* updated docs to include SharpeHyperOptLossDaily

* Update dockerfile to 3.8.1

* Run tests against 3.8

* added daily sharpe ratio hyperopt loss method, ty @djacky

* removed commented code

* removed unused profit_abs

* added proper slippage to each trade

* replaced use of old value total_profit

* added daily sharpe ratio test and modified hyperopt_loss_sharpe_daily

* updated docs to include SharpeHyperOptLossDaily

* docs fixes

* missed one fix

* fixed standard deviation line

* fixed to bracket notation

* fixed to bracket notation

* fixed syntax error

* better readability, kept np.sqrt(365) which results in  annualized sharpe ratio

* fixed method arguments indentation

* updated commented out debug print line

* renamed after slippage profit_percent so it wont affect _calculate_results_metrics()

* Reworked to fill leading and trailing days

* No need for np; make flake happy

* Fix risk free rate

Co-authored-by: Matthias <xmatthias@outlook.com>
Co-authored-by: hroff-1902 <47309513+hroff-1902@users.noreply.github.com>
This commit is contained in:
Yazeed Al Oyoun 2020-02-06 06:49:08 +01:00 committed by GitHub
parent b5ee4f17cb
commit 9639ffb140
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
5 changed files with 110 additions and 21 deletions

View File

@ -337,8 +337,8 @@ optional arguments:
generate completely different results, since the
target for optimization is different. Built-in
Hyperopt-loss-functions are: DefaultHyperOptLoss,
OnlyProfitHyperOptLoss, SharpeHyperOptLoss (default:
`DefaultHyperOptLoss`).
OnlyProfitHyperOptLoss, SharpeHyperOptLoss,
SharpeHyperOptLossDaily (default: `DefaultHyperOptLoss`).
Common arguments:
-v, --verbose Verbose mode (-vv for more, -vvv to get all messages).

View File

@ -57,7 +57,7 @@ Rarely you may also need to override:
!!! Tip "Quickly optimize ROI, stoploss and trailing stoploss"
You can quickly optimize the spaces `roi`, `stoploss` and `trailing` without changing anything (i.e. without creation of a "complete" Hyperopt class with dimensions, parameters, triggers and guards, as described in this document) from the default hyperopt template by relying on your strategy to do most of the calculations.
``` python
```python
# Have a working strategy at hand.
freqtrade new-hyperopt --hyperopt EmptyHyperopt
@ -75,8 +75,8 @@ Copy the file `user_data/hyperopts/sample_hyperopt.py` into `user_data/hyperopts
There are two places you need to change in your hyperopt file to add a new buy hyperopt for testing:
- Inside `indicator_space()` - the parameters hyperopt shall be optimizing.
- Inside `populate_buy_trend()` - applying the parameters.
* Inside `indicator_space()` - the parameters hyperopt shall be optimizing.
* Inside `populate_buy_trend()` - applying the parameters.
There you have two different types of indicators: 1. `guards` and 2. `triggers`.
@ -141,7 +141,7 @@ one we call `trigger` and use it to decide which buy trigger we want to use.
So let's write the buy strategy using these values:
``` python
```python
def populate_buy_trend(dataframe: DataFrame) -> DataFrame:
conditions = []
# GUARDS AND TRENDS
@ -192,6 +192,7 @@ Currently, the following loss functions are builtin:
* `DefaultHyperOptLoss` (default legacy Freqtrade hyperoptimization loss function)
* `OnlyProfitHyperOptLoss` (which takes only amount of profit into consideration)
* `SharpeHyperOptLoss` (optimizes Sharpe Ratio calculated on the trade returns)
* `SharpeHyperOptLossDaily` (optimizes Sharpe Ratio calculated on daily trade returns)
Creation of a custom loss function is covered in the [Advanced Hyperopt](advanced-hyperopt.md) part of the documentation.
@ -323,7 +324,7 @@ method, what those values match to.
So for example you had `rsi-value: 29.0` so we would look at `rsi`-block, that translates to the following code block:
``` python
```python
(dataframe['rsi'] < 29.0)
```
@ -372,6 +373,7 @@ In order to use this best ROI table found by Hyperopt in backtesting and for liv
118: 0
}
```
As stated in the comment, you can also use it as the value of the `minimal_roi` setting in the configuration file.
#### Default ROI Search Space
@ -379,7 +381,7 @@ As stated in the comment, you can also use it as the value of the `minimal_roi`
If you are optimizing ROI, Freqtrade creates the 'roi' optimization hyperspace for you -- it's the hyperspace of components for the ROI tables. By default, each ROI table generated by the Freqtrade consists of 4 rows (steps). Hyperopt implements adaptive ranges for ROI tables with ranges for values in the ROI steps that depend on the ticker_interval used. By default the values vary in the following ranges (for some of the most used ticker intervals, values are rounded to 5 digits after the decimal point):
| # step | 1m | | 5m | | 1h | | 1d | |
|---|---|---|---|---|---|---|---|---|
| ------ | ------ | ----------------- | -------- | ----------- | ---------- | ----------------- | ------------ | ----------------- |
| 1 | 0 | 0.01161...0.11992 | 0 | 0.03...0.31 | 0 | 0.06883...0.71124 | 0 | 0.12178...1.25835 |
| 2 | 2...8 | 0.00774...0.04255 | 10...40 | 0.02...0.11 | 120...480 | 0.04589...0.25238 | 2880...11520 | 0.08118...0.44651 |
| 3 | 4...20 | 0.00387...0.01547 | 20...100 | 0.01...0.04 | 240...1200 | 0.02294...0.09177 | 5760...28800 | 0.04059...0.16237 |
@ -416,6 +418,7 @@ In order to use this best stoploss value found by Hyperopt in backtesting and fo
# This attribute will be overridden if the config file contains "stoploss"
stoploss = -0.27996
```
As stated in the comment, you can also use it as the value of the `stoploss` setting in the configuration file.
#### Default Stoploss Search Space
@ -452,6 +455,7 @@ In order to use these best trailing stop parameters found by Hyperopt in backtes
trailing_stop_positive_offset = 0.06038
trailing_only_offset_is_reached = True
```
As stated in the comment, you can also use it as the values of the corresponding settings in the configuration file.
#### Default Trailing Stop Search Space

View File

@ -256,7 +256,7 @@ AVAILABLE_CLI_OPTIONS = {
help='Specify the class name of the hyperopt loss function class (IHyperOptLoss). '
'Different functions can generate completely different results, '
'since the target for optimization is different. Built-in Hyperopt-loss-functions are: '
'DefaultHyperOptLoss, OnlyProfitHyperOptLoss, SharpeHyperOptLoss.'
'DefaultHyperOptLoss, OnlyProfitHyperOptLoss, SharpeHyperOptLoss, SharpeHyperOptLossDaily.'
'(default: `%(default)s`).',
metavar='NAME',
default=constants.DEFAULT_HYPEROPT_LOSS,

View File

@ -0,0 +1,61 @@
"""
SharpeHyperOptLossDaily
This module defines the alternative HyperOptLoss class which can be used for
Hyperoptimization.
"""
import math
from datetime import datetime
from pandas import DataFrame, date_range
from freqtrade.optimize.hyperopt import IHyperOptLoss
class SharpeHyperOptLossDaily(IHyperOptLoss):
"""
Defines the loss function for hyperopt.
This implementation uses the Sharpe Ratio calculation.
"""
@staticmethod
def hyperopt_loss_function(results: DataFrame, trade_count: int,
min_date: datetime, max_date: datetime,
*args, **kwargs) -> float:
"""
Objective function, returns smaller number for more optimal results.
Uses Sharpe Ratio calculation.
"""
resample_freq = '1D'
slippage_per_trade_ratio = 0.0005
days_in_year = 365
annual_risk_free_rate = 0.0
risk_free_rate = annual_risk_free_rate / days_in_year
# apply slippage per trade to profit_percent
results.loc[:, 'profit_percent_after_slippage'] = \
results['profit_percent'] - slippage_per_trade_ratio
# create the index within the min_date and end max_date
t_index = date_range(start=min_date, end=max_date, freq=resample_freq)
sum_daily = (
results.resample(resample_freq, on='close_time').agg(
{"profit_percent_after_slippage": sum}).reindex(t_index).fillna(0)
)
total_profit = sum_daily["profit_percent_after_slippage"] - risk_free_rate
expected_returns_mean = total_profit.mean()
up_stdev = total_profit.std()
if (up_stdev != 0.):
sharp_ratio = expected_returns_mean / up_stdev * math.sqrt(days_in_year)
else:
# Define high (negative) sharpe ratio to be clear that this is NOT optimal.
sharp_ratio = -20.
# print(t_index, sum_daily, total_profit)
# print(risk_free_rate, expected_returns_mean, up_stdev, sharp_ratio)
return -sharp_ratio

View File

@ -42,7 +42,13 @@ def hyperopt_results():
'profit_percent': [-0.1, 0.2, 0.3],
'profit_abs': [-0.2, 0.4, 0.6],
'trade_duration': [10, 30, 10],
'sell_reason': [SellType.STOP_LOSS, SellType.ROI, SellType.ROI]
'sell_reason': [SellType.STOP_LOSS, SellType.ROI, SellType.ROI],
'close_time':
[
datetime(2019, 1, 1, 9, 26, 3, 478039),
datetime(2019, 2, 1, 9, 26, 3, 478039),
datetime(2019, 3, 1, 9, 26, 3, 478039)
]
}
)
@ -336,6 +342,24 @@ def test_sharpe_loss_prefers_higher_profits(default_conf, hyperopt_results) -> N
assert under > correct
def test_sharpe_loss_daily_prefers_higher_profits(default_conf, hyperopt_results) -> None:
results_over = hyperopt_results.copy()
results_over['profit_percent'] = hyperopt_results['profit_percent'] * 2
results_under = hyperopt_results.copy()
results_under['profit_percent'] = hyperopt_results['profit_percent'] / 2
default_conf.update({'hyperopt_loss': 'SharpeHyperOptLossDaily'})
hl = HyperOptLossResolver.load_hyperoptloss(default_conf)
correct = hl.hyperopt_loss_function(hyperopt_results, len(hyperopt_results),
datetime(2019, 1, 1), datetime(2019, 5, 1))
over = hl.hyperopt_loss_function(results_over, len(hyperopt_results),
datetime(2019, 1, 1), datetime(2019, 5, 1))
under = hl.hyperopt_loss_function(results_under, len(hyperopt_results),
datetime(2019, 1, 1), datetime(2019, 5, 1))
assert over < correct
assert under > correct
def test_onlyprofit_loss_prefers_higher_profits(default_conf, hyperopt_results) -> None:
results_over = hyperopt_results.copy()
results_over['profit_percent'] = hyperopt_results['profit_percent'] * 2