Antonio Della Fortuna
464cb4761c
Fixed max_open_trades update from hyperopt
...
Fixed max_open_trades update from hyperopt + removed max_open_trades as a param to backtesting + refactoring
2023-01-08 12:39:39 +01:00
Antonio Della Fortuna
f2fa476dc6
max_open_trades should be an integer
...
Max open trades will be always an integer in the strategy (-1 for infinity), but in the config -1 will be parsed as infinity
2023-01-04 16:09:27 +01:00
Antonio Della Fortuna
5fd85368a9
Added support for max_open_trades hyperopting
2023-01-04 10:34:44 +01:00
Matthias
c3d4fb9f1b
Simplify backtest calling interface
2022-10-18 06:39:55 +02:00
Matthias
6252ae466e
Convert position_stacking to attribute of backtest
2022-10-17 06:57:26 +02:00
Matthias
a6296be2f5
Update market_change datatype
2022-10-04 10:27:04 +00:00
Matthias
016e438468
Calculate market-change in hyperopt
...
closes #7532
2022-10-04 08:37:07 +00:00
Matthias
a1a62681bf
add fixes for random test failures
2022-09-26 08:11:00 +00:00
Matthias
667853c504
Use Alias to type config objects
2022-09-18 13:20:55 +02:00
Matthias
982c0315fa
Rename variable
2022-09-11 19:31:11 +02:00
Matthias
72d197a99d
Run first epoch in non-parallel mode
...
this allows dataprovider to load it's cache.
closes #7384
2022-09-11 17:51:07 +02:00
Matthias
78cd46ecd5
hyperopt Remove unnecessary arguments
2022-09-11 14:59:39 +02:00
Matthias
32e13d65c3
Refactor hyperopt to extract evaluate_result
2022-09-11 14:59:39 +02:00
Matthias
a48923c0e4
Extract widget colorization to separate function
2022-09-11 14:59:39 +02:00
Matthias
4dec19de9f
Add comment to explain why we use the non-trimmed DF
2022-09-06 06:52:50 +02:00
Italo
9f5642fd97
fix hyperopt df preprocessing
2022-09-05 18:12:19 +02:00
Matthias
bc359675a2
Add --analyze-per-epoch - moving populate_analysis to the epoch process
2022-08-19 15:19:43 +02:00
Matthias
09f8904545
Extract analysis to separate method
2022-08-19 15:12:55 +02:00
Matthias
0b0e7eaf96
Mutex TTL Cache accesses which can be accessed by multiple threads
...
Apparently, cachetools is (intentionally) not threadsafe
when using the Caches directly.
It's therefore recommended to wrap these with an explicit lock to avoid
problems.
source: https://github.com/tkem/cachetools/issues/245
closes #7215
2022-08-16 19:48:21 +02:00
Matthias
40e2da10f3
Add hypeorpt cloudpickle magic
...
closes #7078
2022-07-16 11:49:33 +02:00
Matthias
8bf0bf10c5
Merge branch 'develop' into pr/SmartManoj/6859
2022-06-23 20:43:35 +02:00
Italo
7142394121
remove random_state condition
...
otherwise the random sample always draws the same set of points
2022-06-10 09:46:45 +01:00
Italo
dce9fdd0e4
don't overwrite is_random
...
this should fix issue #6746
2022-06-09 20:06:23 +01:00
Matthias
e3beaae8be
update hyperopt typing
2022-05-22 19:32:32 +02:00
மனோஜ்குமார் பழனிச்சாமி
2b61aa282a
Removed None in dict.get()
...
https://stackoverflow.com/a/12631641
Extra Changes: freqtrade\freqtradebot.py:70
freqtrade\plugins\pairlistmanager.py:31
2022-05-18 03:41:10 +05:30
Matthias
1e83589641
Fix hyperopt
2022-04-28 06:59:03 +02:00
Matthias
f2912f8815
Improve mypy runs
2022-04-23 11:31:12 +02:00
Matthias
5a90d5ece6
Fix docstring quotes
2022-04-23 09:44:04 +02:00
Italo
340c0ea391
update is_random before asked_non_tried
...
is_random depends on asked_non_tried and needs to be updated first
2022-04-14 14:15:11 +01:00
Italo
35cea6dcfa
fix unique_list
2022-04-13 09:36:46 +01:00
Italo
fa298d6f1c
fix unique_list logic
2022-04-12 23:57:40 +01:00
Matthias
54633e90a7
Merge branch 'develop' into pr/italodamato/6563
2022-04-10 09:16:31 +02:00
Italo
1559692e47
Update hyperopt.py
...
remove duplicates from list of asked points
2022-04-08 11:44:42 +01:00
Matthias
4897731030
use_sell_signal -> use_exit_signal
2022-04-05 20:10:20 +02:00
Italo
9f171193ef
Revert "Merge branch 'plot_hyperopt_stats' into opt-ask-force-new-points"
...
This reverts commit 4eb9cc6e8b
, reversing
changes made to a3b401a762
.
2022-03-30 09:39:07 +01:00
Matthias
b91b7b4464
Fix hyperopt assigning sell_signal to wrong field
2022-03-30 07:16:48 +02:00
Italo
4eb9cc6e8b
Merge branch 'plot_hyperopt_stats' into opt-ask-force-new-points
2022-03-30 00:30:33 +01:00
Italo
a3b401a762
highlight random points in hyperopt results table
2022-03-30 00:29:14 +01:00
Italo
229b0b037e
reduce search loops
2022-03-29 19:33:35 +01:00
Matthias
bcf326a035
Initial steps to change bid/ask pricing to enter/exit
2022-03-27 18:03:49 +02:00
Italo
b5a346a46d
Update hyperopt.py
2022-03-22 11:01:38 +00:00
Italo
2733aa33b6
Update hyperopt.py
2022-03-22 00:28:11 +00:00
Italo
37a43019d6
fix
...
- clear cache before calling `ask`
- avoid errors in case asked_non_tried has less than n_points elements
2022-03-21 11:36:53 +00:00
Italo
fca93d8dfe
Update hyperopt.py
2022-03-20 16:12:06 +00:00
Italo
f8a674f24d
make robust in case all points have been tried
2022-03-20 16:08:38 +00:00
Italo
23f1a1904b
more compact
2022-03-20 16:06:41 +00:00
Italo
0fd269e4f0
typo
2022-03-20 16:03:07 +00:00
Italo
e16bb1b34e
Optimize only new points
...
Enforce points returned from `self.opt.ask` have not been already evaluated
2022-03-20 16:02:03 +00:00
Italo
d796ce0935
Update hyperopt.py
...
1. Try to get points using `self.opt.ask` first
2. Discard the points that have already been evaluated
3. Retry using `self.opt.ask` up to 3 times
4. If still some points are missing in respect to `n_points`, random sample some points
5. Repeat until at least `n_points` points in the `asked_non_tried` list
6. Return a list with legth truncated at `n_points`
2022-03-20 15:41:14 +00:00
Italo
d2a5448305
Update hyperopt.py
2022-03-11 17:38:32 +00:00