Merge pull request #6563 from italodamato/opt-ask-force-new-points
Optimize only new points
This commit is contained in:
Matthias
GitHub
@@ -10,7 +10,7 @@ import warnings
|
||||
from datetime import datetime, timezone
|
||||
from math import ceil
|
||||
from pathlib import Path
|
||||
from typing import Any, Dict, List, Optional
|
||||
from typing import Any, Dict, List, Optional, Tuple
|
||||
|
||||
import progressbar
|
||||
import rapidjson
|
||||
@@ -409,6 +409,51 @@ class Hyperopt:
|
||||
# Store non-trimmed data - will be trimmed after signal generation.
|
||||
dump(preprocessed, self.data_pickle_file)
|
||||
|
||||
def get_asked_points(self, n_points: int) -> Tuple[List[List[Any]], List[bool]]:
|
||||
'''
|
||||
Enforce points returned from `self.opt.ask` have not been already evaluated
|
||||
|
||||
Steps:
|
||||
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 length truncated at `n_points`
|
||||
'''
|
||||
def unique_list(a_list):
|
||||
new_list = []
|
||||
for item in a_list:
|
||||
if item not in new_list:
|
||||
new_list.append(item)
|
||||
return new_list
|
||||
i = 0
|
||||
asked_non_tried: List[List[Any]] = []
|
||||
is_random: List[bool] = []
|
||||
while i < 5 and len(asked_non_tried) < n_points:
|
||||
if i < 3:
|
||||
self.opt.cache_ = {}
|
||||
asked = unique_list(self.opt.ask(n_points=n_points * 5))
|
||||
is_random = [False for _ in range(len(asked))]
|
||||
else:
|
||||
asked = unique_list(self.opt.space.rvs(n_samples=n_points * 5))
|
||||
is_random = [True for _ in range(len(asked))]
|
||||
is_random += [rand for x, rand in zip(asked, is_random)
|
||||
if x not in self.opt.Xi
|
||||
and x not in asked_non_tried]
|
||||
asked_non_tried += [x for x in asked
|
||||
if x not in self.opt.Xi
|
||||
and x not in asked_non_tried]
|
||||
i += 1
|
||||
|
||||
if asked_non_tried:
|
||||
return (
|
||||
asked_non_tried[:min(len(asked_non_tried), n_points)],
|
||||
is_random[:min(len(asked_non_tried), n_points)]
|
||||
)
|
||||
else:
|
||||
return self.opt.ask(n_points=n_points), [False for _ in range(n_points)]
|
||||
|
||||
def start(self) -> None:
|
||||
self.random_state = self._set_random_state(self.config.get('hyperopt_random_state', None))
|
||||
logger.info(f"Using optimizer random state: {self.random_state}")
|
||||
@@ -473,7 +518,7 @@ class Hyperopt:
|
||||
n_rest = (i + 1) * jobs - self.total_epochs
|
||||
current_jobs = jobs - n_rest if n_rest > 0 else jobs
|
||||
|
||||
asked = self.opt.ask(n_points=current_jobs)
|
||||
asked, is_random = self.get_asked_points(n_points=current_jobs)
|
||||
f_val = self.run_optimizer_parallel(parallel, asked, i)
|
||||
self.opt.tell(asked, [v['loss'] for v in f_val])
|
||||
|
||||
@@ -492,6 +537,7 @@ class Hyperopt:
|
||||
# evaluations can take different time. Here they are aligned in the
|
||||
# order they will be shown to the user.
|
||||
val['is_best'] = is_best
|
||||
val['is_random'] = is_random[j]
|
||||
self.print_results(val)
|
||||
|
||||
if is_best:
|
||||
|
||||
@@ -310,6 +310,8 @@ class HyperoptTools():
|
||||
if not has_drawdown:
|
||||
# Ensure compatibility with older versions of hyperopt results
|
||||
trials['results_metrics.max_drawdown_account'] = None
|
||||
if 'is_random' not in trials.columns:
|
||||
trials['is_random'] = False
|
||||
|
||||
# New mode, using backtest result for metrics
|
||||
trials['results_metrics.winsdrawslosses'] = trials.apply(
|
||||
@@ -322,12 +324,12 @@ class HyperoptTools():
|
||||
'results_metrics.profit_total', 'results_metrics.holding_avg',
|
||||
'results_metrics.max_drawdown',
|
||||
'results_metrics.max_drawdown_account', 'results_metrics.max_drawdown_abs',
|
||||
'loss', 'is_initial_point', 'is_best']]
|
||||
'loss', 'is_initial_point', 'is_random', 'is_best']]
|
||||
|
||||
trials.columns = [
|
||||
'Best', 'Epoch', 'Trades', ' Win Draw Loss', 'Avg profit',
|
||||
'Total profit', 'Profit', 'Avg duration', 'max_drawdown', 'max_drawdown_account',
|
||||
'max_drawdown_abs', 'Objective', 'is_initial_point', 'is_best'
|
||||
'max_drawdown_abs', 'Objective', 'is_initial_point', 'is_random', 'is_best'
|
||||
]
|
||||
|
||||
return trials
|
||||
@@ -349,9 +351,11 @@ class HyperoptTools():
|
||||
trials = HyperoptTools.prepare_trials_columns(trials, has_account_drawdown)
|
||||
|
||||
trials['is_profit'] = False
|
||||
trials.loc[trials['is_initial_point'], 'Best'] = '* '
|
||||
trials.loc[trials['is_initial_point'] | trials['is_random'], 'Best'] = '* '
|
||||
trials.loc[trials['is_best'], 'Best'] = 'Best'
|
||||
trials.loc[trials['is_initial_point'] & trials['is_best'], 'Best'] = '* Best'
|
||||
trials.loc[
|
||||
(trials['is_initial_point'] | trials['is_random']) & trials['is_best'],
|
||||
'Best'] = '* Best'
|
||||
trials.loc[trials['Total profit'] > 0, 'is_profit'] = True
|
||||
trials['Trades'] = trials['Trades'].astype(str)
|
||||
# perc_multi = 1 if legacy_mode else 100
|
||||
@@ -407,7 +411,7 @@ class HyperoptTools():
|
||||
trials.iat[i, j] = "{}{}{}".format(Style.BRIGHT,
|
||||
str(trials.loc[i][j]), Style.RESET_ALL)
|
||||
|
||||
trials = trials.drop(columns=['is_initial_point', 'is_best', 'is_profit'])
|
||||
trials = trials.drop(columns=['is_initial_point', 'is_best', 'is_profit', 'is_random'])
|
||||
if remove_header > 0:
|
||||
table = tabulate.tabulate(
|
||||
trials.to_dict(orient='list'), tablefmt='orgtbl',
|
||||
|
||||
Reference in New Issue
Block a user