Commit Graph

1351 Commits

Author SHA1 Message Date
Simon Ebner
df033d92ef Improve performance of decimalspace.py
decimalspace.py is heavily used in the hyperoptimization. The following
benchmark code runs an optimization which is taken from optimizing a
real strategy (wtc).
The optimized version takes on my machine approx. 11/12s compared to the
original 32s. Results are equivalent in both cases.

```
import freqtrade.optimize.space
import numpy as np
import skopt
import timeit

def init():
    Decimal = freqtrade.optimize.space.decimalspace.SKDecimal
    Integer = skopt.space.space.Integer
    dimensions = [Decimal(low=-1.0,
        high=1.0,
        decimals=4,
        prior='uniform',
        transform='identity')] * 20

    return skopt.Optimizer(
        dimensions,
        base_estimator="ET",
        acq_optimizer="auto",
        n_initial_points=5,
        acq_optimizer_kwargs={'n_jobs': 96},
        random_state=0,
        model_queue_size=10,
    )

def test():
    opt = init()
    actual = opt.ask(n_points=2)
    expected = [[
        0.7515, -0.4723, -0.6941, -0.7988, 0.0448, 0.8605, -0.108, 0.5399,
        0.763, -0.2948, 0.8345, -0.7683, 0.7077, -0.2478, -0.333, 0.8575,
        0.6108, 0.4514, 0.5982, 0.3506
    ], [
        0.5563, 0.7386, -0.6407, 0.9073, -0.5211, -0.8167, -0.3771,
        -0.0318, 0.2861, 0.1176, 0.0943, -0.6077, -0.9317, -0.5372,
        -0.4934, -0.3637, -0.8035, -0.8627, -0.5399, 0.6036
    ]]

    absdiff = np.max(np.abs(np.asarray(expected) - np.asarray(actual)))
    assert absdiff < 1e-5

def time():
    opt = init()
    print('dt', timeit.timeit("opt.ask(n_points=20)", globals=locals()))

if __name__ == "__main__":
    test()
    time()
```
2021-10-24 18:14:24 +02:00
Matthias
22dd2ca003 Fix mypy type errors 2021-10-24 15:18:29 +02:00
Matthias
17432b2823 Improve some stylings 2021-10-24 09:15:05 +02:00
Matthias
dffb4c5d53 Merge branch 'develop' into pr/TreborNamor/5607 2021-10-24 08:55:10 +02:00
Matthias
78724e304e Merge branch 'develop' into pr/theluxaz/5710 2021-10-21 17:46:39 +02:00
theluxaz
0e085298e9 Fixed test failures. 2021-10-21 17:25:38 +03:00
Matthias
96f99699e0 Merge pull request #4606 from rextea/add_days_breakdown_to_backtesting_summary
Add days breakdown table to backtesting
2021-10-21 13:56:30 +02:00
Matthias
e458c9867a Styling fixes 2021-10-21 07:45:15 +02:00
Matthias
7b5346b984 Add test for breakdown-stats 2021-10-21 07:11:39 +02:00
Matthias
fa028c2134 Support day/week/month breakdowns 2021-10-21 06:58:40 +02:00
Matthias
7197f4ce77 Don't show daily % profit (it's wrong) 2021-10-20 20:01:31 +02:00
Matthias
de5497c766 backtest_days cannot be below 1 2021-10-20 19:39:37 +02:00
theluxaz
905f3a1a50 Removed exit_tag from Trade objects. 2021-10-20 17:58:50 +03:00
Matthias
5454460227 Revert initial_points to 30
closes #5760
2021-10-20 07:46:15 +02:00
theluxaz
1fdc4425dd Changed exit_tag to be represented as sell_reason 2021-10-20 01:26:15 +03:00
GluTbl
00406ea7d5 Update backtesting.py
Support for custom entry-prices and exit-prices during backtesting.
2021-10-19 17:15:45 +05:30
theluxaz
5ecdd1d112 Merge branch 'develop' into freqtrade-development 2021-10-19 00:00:15 +03:00
theluxaz
69a59cdf37 Fixed flake 8, changed sell_tag to exit_tag and fixed telegram functions 2021-10-18 23:56:41 +03:00
Matthias
7d8cd736b8 Support days-breakdown also for hyperopt results 2021-10-17 16:49:39 +02:00
Matthias
47bba331c1 Merge branch 'develop' into pr/rextea/4606 2021-10-17 16:29:31 +02:00
Matthias
0e7d903a6f Merge pull request #5644 from slyons/develop
Add ability to ignore unparameterized spaces
2021-10-14 08:07:07 +02:00
theluxaz
0bb7ea10ab Fixed minor header for backtesting 2021-10-14 01:34:30 +03:00
theluxaz
ed39b8dab0 fixed profit total calculation 2021-10-14 01:18:16 +03:00
Matthias
aed919a05f Simplify "no-space-configured" error handling by moving it to hyperopt_auto 2021-10-13 19:54:35 +02:00
theluxaz
b151cf032b Merge branch 'develop' of https://github.com/theluxaz/freqtrade into main
# Conflicts:
#	freqtrade/freqtradebot.py
#	freqtrade/optimize/backtesting.py
2021-10-13 02:01:26 +03:00
theluxaz
80b71790bc Added some bigfixes for sell_tag 2021-10-13 01:22:53 +03:00
theluxaz
b898f86364 Added sell_tag and buy/sell telegram performance functions 2021-10-13 00:02:28 +03:00
sid
30bc96cf3f simplify expression 2021-10-09 06:36:23 +05:30
sid
46c320513a use profit_abs 2021-10-07 08:07:07 +05:30
sid
6ba46b38bd fix formatting 2021-10-06 13:46:05 +05:30
sid
c0d01dbc26 add max_drawdown loss 2021-10-06 13:24:27 +05:30
Scott Lyons
df45f467c6 Adding ability to ignore unparameterized spaces 2021-09-30 01:11:02 -07:00
Robert Roman
ca973c05d1 Merge branch 'freqtrade:develop' into develop 2021-09-28 10:16:36 -05:00
Robert Roman
626a40252d resolved mypy error
error: Signature of "hyperopt_loss_function" incompatible with supertype "IHyperOptLoss"
2021-09-27 17:33:29 -05:00
Robert Roman
c3414c3b78 resolved mypy error
error: Signature of "hyperopt_loss_function" incompatible with supertype "IHyperOptLoss"
2021-09-27 17:32:49 -05:00
Matthias
5726886b06 Reduce backtest-noise from "pandas slice" warning 2021-09-27 20:52:19 +02:00
Robert Roman
bdca3e2343 Merge branch 'freqtrade:develop' into develop 2021-09-26 15:37:09 -05:00
Matthias
6319c104fe Fix unreliable backtest-result when using webserver mode 2021-09-26 15:07:48 +02:00
Robert Roman
24baad7884 Add Calmar Ratio Daily
This hyper opt loss calculates the daily Calmar ratio.
2021-09-25 16:28:36 -05:00
Robert Roman
3b99c84b0a resolved the total profit issue
I resolved the total profit issue and locally ran flak8 and isort
2021-09-23 21:31:33 -05:00
Robert Roman
c6b684603c removed trade_count inside if statement
i removed trade_count inside if statement. Even though it helps overfitting, It is not useful when running hyperopt on small datasets.
2021-09-22 09:21:43 -05:00
Robert Roman
b946f8e7f1 I sorted imports with isort 2021-09-22 09:18:17 -05:00
Robert Roman
3834bb86ff updated line 42
I removed the minus sign on max drawdown.
2021-09-21 20:25:17 -05:00
Robert Roman
3845d55186 a new hyperopt loss created that uses calmar ratio
This is a new hyperopt loss file that uses the Calmar Ratio.

Calmar Ratio = average annual rate of return / maximum drawdown
2021-09-21 20:04:23 -05:00
Rokas Kupstys
5dc78a0c66 [SQUASH] Get rid of _initialize() and fix informatives for dynamic pairlists. 2021-09-18 10:48:53 +03:00
Rokas Kupstys
dfa61b7ad2 [SQUASH] Fix informatives for each pair not being created because dataprovider was not available.
Fix not being able to have informative dataframe of a pair in whitelist.
2021-09-18 10:48:53 +03:00
Matthias
853c3a4433 Merge pull request #5587 from raph92/patch-3
Update prepare_trials_columns() return type
2021-09-18 08:08:18 +02:00
raphael
4b2c1a9b8e Remove trailing whitespace 2021-09-17 14:39:15 -04:00
raphael
e715f2a253 Update formatting
Line 302 was too long
2021-09-17 14:23:26 -04:00
raphael
9525a5b96c Add type to "trials" parameter 2021-09-17 14:10:37 -04:00