adelorenzo/stable
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Add hyperopt-list and hyperopt-show commands

Browse Source
This commit is contained in:
hroff-1902 2019-11-26 15:01:42 +03:00
parent cab748588c
commit 8e7512161a
5 changed files with 353 additions and 131 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
freqtrade/configuration/arguments.py
View File

@ -49,8 +49,14 @@ ARGS_PLOT_DATAFRAME = ["pairs", "indicators1", "indicators2", "plot_limit",
ARGS_PLOT_PROFIT = ["pairs", "timerange", "export", "exportfilename", "db_url",
"trade_source", "ticker_interval"]
ARGS_HYPEROPT_LIST = ["hyperopt_list_best", "hyperopt_list_profitable", "print_colorized",
"print_json", "hyperopt_list_no_details"]
ARGS_HYPEROPT_SHOW = ["hyperopt_list_best", "hyperopt_list_profitable", "hyperopt_show_index",
"print_json", "hyperopt_show_no_header"]
NO_CONF_REQURIED = ["download-data", "list-timeframes", "list-markets", "list-pairs",
"plot-dataframe", "plot-profit"]
"hyperopt_list", "hyperopt_show", "plot-dataframe", "plot-profit"]
NO_CONF_ALLOWED = ["create-userdir", "list-exchanges"]
@ -116,6 +122,7 @@ class Arguments:
from freqtrade.optimize import start_backtesting, start_hyperopt, start_edge
from freqtrade.utils import (start_create_userdir, start_download_data,
start_hyperopt_list, start_hyperopt_show,
start_list_exchanges, start_list_markets,
start_list_timeframes, start_trading)
from freqtrade.plot.plot_utils import start_plot_dataframe, start_plot_profit
@ -220,3 +227,21 @@ class Arguments:
)
plot_profit_cmd.set_defaults(func=start_plot_profit)
self._build_args(optionlist=ARGS_PLOT_PROFIT, parser=plot_profit_cmd)
# Add hyperopt-list subcommand
hyperopt_list_cmd = subparsers.add_parser(
'hyperopt-list',
help='List Hyperopt results',
parents=[_common_parser],
)
hyperopt_list_cmd.set_defaults(func=start_hyperopt_list)
self._build_args(optionlist=ARGS_HYPEROPT_LIST, parser=hyperopt_list_cmd)
# Add hyperopt-show subcommand
hyperopt_show_cmd = subparsers.add_parser(
'hyperopt-show',
help='Show details of Hyperopt results',
parents=[_common_parser],
)
hyperopt_show_cmd.set_defaults(func=start_hyperopt_show)
self._build_args(optionlist=ARGS_HYPEROPT_SHOW, parser=hyperopt_show_cmd)

39
freqtrade/configuration/cli_options.py
View File

@ -18,6 +18,18 @@ def check_int_positive(value: str) -> int:
return uint
def check_int_nonzero(value: str) -> int:
try:
uint = int(value)
if uint == 0:
raise ValueError
except ValueError:
raise argparse.ArgumentTypeError(
f"{value} is invalid for this parameter, should be a non-zero integer value"
)
return uint
class Arg:
# Optional CLI arguments
def __init__(self, *args, **kwargs):
@ -364,4 +376,31 @@ AVAILABLE_CLI_OPTIONS = {
choices=["DB", "file"],
default="file",
),
# hyperopt-list, hyperopt-show
"hyperopt_list_profitable": Arg(
'--profitable',
help='Select only profitable epochs.',
action='store_true',
),
"hyperopt_list_best": Arg(
'--best',
help='Select only best epochs.',
action='store_true',
),
"hyperopt_list_no_details": Arg(
'--no-details',
help='Do not print best epoch details.',
action='store_true',
),
"hyperopt_show_index": Arg(
'-n', '--index',
help='Specify the index of the epoch to print details for.',
type=check_int_nonzero,
metavar='INT',
),
"hyperopt_show_no_header": Arg(
'--no-header',
help='Do not print epoch details header.',
action='store_true',
),
}

15
freqtrade/configuration/configuration.py
View File

@ -308,6 +308,21 @@ class Configuration:
self._args_to_config(config, argname='hyperopt_loss',
logstring='Using Hyperopt loss class name: {}')
self._args_to_config(config, argname='hyperopt_show_index',
logstring='Parameter -n/--index detected: {}')
self._args_to_config(config, argname='hyperopt_list_best',
logstring='Parameter --best detected: {}')
self._args_to_config(config, argname='hyperopt_list_profitable',
logstring='Parameter --profitable detected: {}')
self._args_to_config(config, argname='hyperopt_list_no_details',
logstring='Parameter --no-details detected: {}')
self._args_to_config(config, argname='hyperopt_show_no_header',
logstring='Parameter --no-header detected: {}')
def _process_plot_options(self, config: Dict[str, Any]) -> None:
self._args_to_config(config, argname='pairs',

300
freqtrade/optimize/hyperopt.py
View File

@ -74,11 +74,11 @@ class Hyperopt:
else:
logger.info("Continuing on previous hyperopt results.")
self.num_trials_saved = 0
# Previous evaluations
self.trials: List = []
self.num_trials_saved = 0
# Populate functions here (hasattr is slow so should not be run during "regular" operations)
if hasattr(self.custom_hyperopt, 'populate_indicators'):
self.backtesting.strategy.advise_indicators = \
@ -104,6 +104,10 @@ class Hyperopt:
self.config['ask_strategy'] = {}
self.config['ask_strategy']['use_sell_signal'] = True
self.print_all = self.config.get('print_all', False)
self.print_colorized = self.config.get('print_colorized', False)
self.print_json = self.config.get('print_json', False)
@staticmethod
def get_lock_filename(config) -> str:
@ -119,20 +123,18 @@ class Hyperopt:
logger.info(f"Removing `{p}`.")
p.unlink()
def get_args(self, params):
def _get_params_dict(self, raw_params: List[Any]) -> Dict:
dimensions = self.dimensions
dimensions: List[Dimension] = self.dimensions
# Ensure the number of dimensions match
# the number of parameters in the list x.
if len(params) != len(dimensions):
raise ValueError('Mismatch in number of search-space dimensions. '
f'len(dimensions)=={len(dimensions)} and len(x)=={len(params)}')
# the number of parameters in the list.
if len(raw_params) != len(dimensions):
raise ValueError('Mismatch in number of search-space dimensions.')
# Create a dict where the keys are the names of the dimensions
# and the values are taken from the list of parameters x.
arg_dict = {dim.name: value for dim, value in zip(dimensions, params)}
return arg_dict
# Return a dict where the keys are the names of the dimensions
# and the values are taken from the list of parameters.
return {d.name: v for d, v in zip(dimensions, raw_params)}
def save_trials(self, final: bool = False) -> None:
"""
@ -147,106 +149,126 @@ class Hyperopt:
logger.info(f"{num_trials} {plural(num_trials, 'epoch')} "
f"saved to '{self.trials_file}'.")
def read_trials(self) -> List:
@staticmethod
def _read_trials(trials_file) -> List:
"""
Read hyperopt trials file
"""
logger.info("Reading Trials from '%s'", self.trials_file)
trials = load(self.trials_file)
self.trials_file.unlink()
logger.info("Reading Trials from '%s'", trials_file)
trials = load(trials_file)
return trials
def log_trials_result(self) -> None:
def _get_params_details(self, params: Dict) -> Dict:
"""
Display Best hyperopt result
Return the params for each space
"""
# This is printed when Ctrl+C is pressed quickly, before first epochs have
# a chance to be evaluated.
if not self.trials:
print("No epochs evaluated yet, no best result.")
return
result: Dict = {}
results = sorted(self.trials, key=itemgetter('loss'))
best_result = results[0]
params = best_result['params']
log_str = self.format_results_logstring(best_result)
print(f"\nBest result:\n\n{log_str}\n")
if self.config.get('print_json'):
result_dict: Dict = {}
if self.has_space('buy') or self.has_space('sell'):
result_dict['params'] = {}
if self.has_space('buy'):
result_dict['params'].update({p.name: params.get(p.name)
for p in self.hyperopt_space('buy')})
result['buy'] = {p.name: params.get(p.name) for p in self.hyperopt_space('buy')}
if self.has_space('sell'):
result_dict['params'].update({p.name: params.get(p.name)
for p in self.hyperopt_space('sell')})
result['sell'] = {p.name: params.get(p.name) for p in self.hyperopt_space('sell')}
if self.has_space('roi'):
result['roi'] = self.custom_hyperopt.generate_roi_table(params)
if self.has_space('stoploss'):
result['stoploss'] = params.get('stoploss')
return result
@staticmethod # noqa: C901
def print_epoch_details(results, total_epochs, print_json: bool,
no_header: bool = False, header_str: str = None) -> None:
"""
Display details of the hyperopt result
"""
params = results['params_details']
# Default header string
if header_str is None:
header_str = "Best result"
if not no_header:
explanation_str = Hyperopt._format_explanation_string(results, total_epochs)
print(f"\n{header_str}:\n\n{explanation_str}\n")
if print_json:
result_dict: Dict = {}
result_params_dict: Dict = {}
if 'buy' in params:
result_params_dict.update(params['buy'])
if 'sell' in params:
result_params_dict.update(params['sell'])
if result_params_dict:
result_dict['params'] = result_params_dict
if 'roi' in params:
# Convert keys in min_roi dict to strings because
# rapidjson cannot dump dicts with integer keys...
# OrderedDict is used to keep the numeric order of the items
# in the dict.
result_dict['minimal_roi'] = OrderedDict(
(str(k), v) for k, v in self.custom_hyperopt.generate_roi_table(params).items()
(str(k), v) for k, v in params['roi'].items()
)
if self.has_space('stoploss'):
result_dict['stoploss'] = params.get('stoploss')
if 'stoploss' in params:
result_dict['stoploss'] = params['stoploss']
print(rapidjson.dumps(result_dict, default=str, number_mode=rapidjson.NM_NATIVE))
else:
if self.has_space('buy'):
if 'buy' in params:
print('Buy hyperspace params:')
pprint({p.name: params.get(p.name) for p in self.hyperopt_space('buy')},
indent=4)
if self.has_space('sell'):
pprint(params['buy'], indent=4)
if 'sell' in params:
print('Sell hyperspace params:')
pprint({p.name: params.get(p.name) for p in self.hyperopt_space('sell')},
indent=4)
if self.has_space('roi'):
pprint(params['sell'], indent=4)
if 'roi' in params:
print("ROI table:")
# Round printed values to 5 digits after the decimal point
pprint(round_dict(self.custom_hyperopt.generate_roi_table(params), 5), indent=4)
if self.has_space('stoploss'):
pprint(round_dict(params['roi'], 5), indent=4)
if 'stoploss' in params:
# Also round to 5 digits after the decimal point
print(f"Stoploss: {round(params.get('stoploss'), 5)}")
print(f"Stoploss: {round(params['stoploss'], 5)}")
def is_best(self, results) -> bool:
return results['loss'] < self.current_best_loss
@staticmethod
def is_best_loss(results, current_best_loss) -> bool:
return results['loss'] < current_best_loss
def log_results(self, results) -> None:
def print_results(self, results) -> None:
"""
Log results if it is better than any previous evaluation
"""
print_all = self.config.get('print_all', False)
is_best_loss = self.is_best(results)
if not print_all:
is_best = results['is_best']
if not self.print_all:
# Print '\n' after each 100th epoch to separate dots from the log messages.
# Otherwise output is messy on a terminal.
print('.', end='' if results['current_epoch'] % 100 != 0 else None) # type: ignore
sys.stdout.flush()
if print_all or is_best_loss:
if is_best_loss:
self.current_best_loss = results['loss']
log_str = self.format_results_logstring(results)
# Colorize output
if self.config.get('print_colorized', False):
if results['total_profit'] > 0:
log_str = Fore.GREEN + log_str
if print_all and is_best_loss:
log_str = Style.BRIGHT + log_str
if print_all:
print(log_str)
else:
print(f'\n{log_str}')
if self.print_all or is_best:
if not self.print_all:
# Separate the results explanation string from dots
print("\n")
self.print_results_explanation(results, self.total_epochs, self.print_all,
self.print_colorized)
def format_results_logstring(self, results) -> str:
current = results['current_epoch']
total = self.total_epochs
res = results['results_explanation']
loss = results['loss']
log_str = f'{current:5d}/{total}: {res} Objective: {loss:.5f}'
log_str = f'*{log_str}' if results['is_initial_point'] else f' {log_str}'
return log_str
@staticmethod
def print_results_explanation(results, total_epochs, highlight_best: bool,
print_colorized: bool) -> None:
"""
Log results explanation string
"""
explanation_str = Hyperopt._format_explanation_string(results, total_epochs)
# Colorize output
if print_colorized:
if results['total_profit'] > 0:
explanation_str = Fore.GREEN + explanation_str
if highlight_best and results['is_best']:
explanation_str = Style.BRIGHT + explanation_str
print(explanation_str)
@staticmethod
def _format_explanation_string(results, total_epochs) -> str:
return (("*" if results['is_initial_point'] else " ") +
f"{results['current_epoch']:5d}/{total_epochs}: " +
f"{results['results_explanation']} " +
f"Objective: {results['loss']:.5f}")
def has_space(self, space: str) -> bool:
"""
@ -276,33 +298,34 @@ class Hyperopt:
spaces += self.custom_hyperopt.stoploss_space()
return spaces
def generate_optimizer(self, _params: Dict, iteration=None) -> Dict:
def generate_optimizer(self, raw_params: List[Any], iteration=None) -> Dict:
"""
Used Optimize function. Called once per epoch to optimize whatever is configured.
Keep this function as optimized as possible!
"""
params = self.get_args(_params)
params_dict = self._get_params_dict(raw_params)
params_details = self._get_params_details(params_dict)
if self.has_space('roi'):
self.backtesting.strategy.minimal_roi = \
self.custom_hyperopt.generate_roi_table(params)
self.custom_hyperopt.generate_roi_table(params_dict)
if self.has_space('buy'):
self.backtesting.strategy.advise_buy = \
self.custom_hyperopt.buy_strategy_generator(params)
self.custom_hyperopt.buy_strategy_generator(params_dict)
if self.has_space('sell'):
self.backtesting.strategy.advise_sell = \
self.custom_hyperopt.sell_strategy_generator(params)
self.custom_hyperopt.sell_strategy_generator(params_dict)
if self.has_space('stoploss'):
self.backtesting.strategy.stoploss = params['stoploss']
self.backtesting.strategy.stoploss = params_dict['stoploss']
processed = load(self.tickerdata_pickle)
min_date, max_date = get_timeframe(processed)
results = self.backtesting.backtest(
backtesting_results = self.backtesting.backtest(
{
'stake_amount': self.config['stake_amount'],
'processed': processed,
@ -312,58 +335,58 @@ class Hyperopt:
'end_date': max_date,
}
)
results_explanation = self.format_results(results)
results_metrics = self._calculate_results_metrics(backtesting_results)
results_explanation = self._format_results_explanation_string(results_metrics)
trade_count = len(results.index)
total_profit = results.profit_abs.sum()
trade_count = results_metrics['trade_count']
total_profit = results_metrics['total_profit']
# If this evaluation contains too short amount of trades to be
# interesting -- consider it as 'bad' (assigned max. loss value)
# in order to cast this hyperspace point away from optimization
# path. We do not want to optimize 'hodl' strategies.
if trade_count < self.config['hyperopt_min_trades']:
return {
'loss': MAX_LOSS,
'params': params,
'results_explanation': results_explanation,
'total_profit': total_profit,
}
loss = self.calculate_loss(results=results, trade_count=trade_count,
loss: float = MAX_LOSS
if trade_count >= self.config['hyperopt_min_trades']:
loss = self.calculate_loss(results=backtesting_results, trade_count=trade_count,
min_date=min_date.datetime, max_date=max_date.datetime)
return {
'loss': loss,
'params': params,
'params_dict': params_dict,
'params_details': params_details,
'results_metrics': results_metrics,
'results_explanation': results_explanation,
'total_profit': total_profit,
}
def format_results(self, results: DataFrame) -> str:
def _calculate_results_metrics(self, backtesting_results: DataFrame) -> Dict:
return {
'trade_count': len(backtesting_results.index),
'avg_profit': backtesting_results.profit_percent.mean() * 100.0,
'total_profit': backtesting_results.profit_abs.sum(),
'profit': backtesting_results.profit_percent.sum() * 100.0,
'duration': backtesting_results.trade_duration.mean(),
}
def _format_results_explanation_string(self, results_metrics: Dict) -> str:
"""
Return the formatted results explanation in a string
"""
trades = len(results.index)
avg_profit = results.profit_percent.mean() * 100.0
total_profit = results.profit_abs.sum()
stake_cur = self.config['stake_currency']
profit = results.profit_percent.sum() * 100.0
duration = results.trade_duration.mean()
return (f'{trades:6d} trades. Avg profit {avg_profit: 5.2f}%. '
f'Total profit {total_profit: 11.8f} {stake_cur} '
f'({profit: 7.2f}\N{GREEK CAPITAL LETTER SIGMA}%). '
f'Avg duration {duration:5.1f} mins.'
return (f"{results_metrics['trade_count']:6d} trades. "
f"Avg profit {results_metrics['avg_profit']: 6.2f}%. "
f"Total profit {results_metrics['total_profit']: 11.8f} {stake_cur} "
f"({results_metrics['profit']: 7.2f}\N{GREEK CAPITAL LETTER SIGMA}%). "
f"Avg duration {results_metrics['duration']:5.1f} mins."
).encode(locale.getpreferredencoding(), 'replace').decode('utf-8')
def get_optimizer(self, dimensions, cpu_count) -> Optimizer:
def get_optimizer(self, dimensions: List[Dimension], cpu_count) -> Optimizer:
return Optimizer(
dimensions,
base_estimator="ET",
acq_optimizer="auto",
n_initial_points=INITIAL_POINTS,
acq_optimizer_kwargs={'n_jobs': cpu_count},
random_state=self.config.get('hyperopt_random_state', None)
random_state=self.config.get('hyperopt_random_state', None),
)
def fix_optimizer_models_list(self):
@ -387,14 +410,16 @@ class Hyperopt:
return parallel(delayed(
wrap_non_picklable_objects(self.generate_optimizer))(v, i) for v in asked)
def load_previous_results(self):
""" read trials file if we have one """
if self.trials_file.is_file() and self.trials_file.stat().st_size > 0:
self.trials = self.read_trials()
logger.info(
'Loaded %d previous evaluations from disk.',
len(self.trials)
)
@staticmethod
def load_previous_results(trials_file) -> List:
"""
Load data for epochs from the file if we have one
"""
trials: List = []
if trials_file.is_file() and trials_file.stat().st_size > 0:
trials = Hyperopt._read_trials(trials_file)
logger.info(f"Loaded {len(trials)} previous evaluations from disk.")
return trials
def start(self) -> None:
data, timerange = self.backtesting.load_bt_data()
@ -415,17 +440,17 @@ class Hyperopt:
# We don't need exchange instance anymore while running hyperopt
self.backtesting.exchange = None # type: ignore
self.load_previous_results()
self.trials = self.load_previous_results(self.trials_file)
cpus = cpu_count()
logger.info(f"Found {cpus} CPU cores. Let's make them scream!")
config_jobs = self.config.get('hyperopt_jobs', -1)
logger.info(f'Number of parallel jobs set as: {config_jobs}')
self.dimensions = self.hyperopt_space()
self.dimensions: List[Dimension] = self.hyperopt_space()
self.opt = self.get_optimizer(self.dimensions, config_jobs)
if self.config.get('print_colorized', False):
if self.print_colorized:
colorama_init(autoreset=True)
try:
@ -439,19 +464,38 @@ class Hyperopt:
self.opt.tell(asked, [v['loss'] for v in f_val])
self.fix_optimizer_models_list()
for j in range(jobs):
# Use human-friendly index here (starting from 1)
# Use human-friendly indexes here (starting from 1)
current = i * jobs + j + 1
val = f_val[j]
val['current_epoch'] = current
val['is_initial_point'] = current <= INITIAL_POINTS
logger.debug(f"Optimizer epoch evaluated: {val}")
is_best = self.is_best(val)
self.log_results(val)
is_best = self.is_best_loss(val, self.current_best_loss)
# This value is assigned here and not in the optimization method
# to keep proper order in the list of results. That's because
# evaluations can take different time. Here they are aligned in the
# order they will be shown to the user.
val['is_best'] = is_best
self.print_results(val)
if is_best:
self.current_best_loss = val['loss']
self.trials.append(val)
# Save results after each best epoch and every 100 epochs
if is_best or current % 100 == 0:
self.save_trials()
except KeyboardInterrupt:
print('User interrupted..')
self.save_trials(final=True)
self.log_trials_result()
if self.trials:
sorted_trials = sorted(self.trials, key=itemgetter('loss'))
results = sorted_trials[0]
self.print_epoch_details(results, self.total_epochs, self.print_json)
else:
# This is printed when Ctrl+C is pressed quickly, before first epochs have
# a chance to be evaluated.
print("No epochs evaluated yet, no best result.")

99
freqtrade/utils.py
View File

@ -1,12 +1,14 @@
import logging
import sys
from collections import OrderedDict
from operator import itemgetter
from pathlib import Path
from typing import Any, Dict, List
import arrow
import csv
import rapidjson
from colorama import init as colorama_init
from tabulate import tabulate
from freqtrade import OperationalException
@ -18,6 +20,7 @@ from freqtrade.data.history import (convert_trades_to_ohlcv,
from freqtrade.exchange import (available_exchanges, ccxt_exchanges, market_is_active,
symbol_is_pair)
from freqtrade.misc import plural
from freqtrade.optimize.hyperopt import Hyperopt
from freqtrade.resolvers import ExchangeResolver
from freqtrade.state import RunMode
@ -236,3 +239,99 @@ def start_list_markets(args: Dict[str, Any], pairs_only: bool = False) -> None:
args.get('list_pairs_print_json', False) or
args.get('print_csv', False)):
print(f"{summary_str}.")
def start_hyperopt_list(args: Dict[str, Any]) -> None:
"""
"""
config = setup_utils_configuration(args, RunMode.UTIL_NO_EXCHANGE)
only_best = config.get('hyperopt_list_best', False)
only_profitable = config.get('hyperopt_list_profitable', False)
print_colorized = config.get('print_colorized', False)
print_json = config.get('print_json', False)
no_details = config.get('hyperopt_list_no_details', False)
no_header = False
trials_file = (config['user_data_dir'] /
'hyperopt_results' / 'hyperopt_results.pickle')
# Previous evaluations
trials = Hyperopt.load_previous_results(trials_file)
total_epochs = len(trials)
trials = _hyperopt_filter_trials(trials, only_best, only_profitable)
# TODO: fetch the interval for epochs to print from the cli option
epoch_start, epoch_stop = 0, None
if print_colorized:
colorama_init(autoreset=True)
try:
# Human-friendly indexes used here (starting from 1)
for val in trials[epoch_start:epoch_stop]:
Hyperopt.print_results_explanation(val, total_epochs, not only_best, print_colorized)
except KeyboardInterrupt:
print('User interrupted..')
if trials and not no_details:
sorted_trials = sorted(trials, key=itemgetter('loss'))
results = sorted_trials[0]
Hyperopt.print_epoch_details(results, total_epochs, print_json, no_header)
def start_hyperopt_show(args: Dict[str, Any]) -> None:
"""
"""
config = setup_utils_configuration(args, RunMode.UTIL_NO_EXCHANGE)
only_best = config.get('hyperopt_list_best', False)
only_profitable = config.get('hyperopt_list_profitable', False)
no_header = config.get('hyperopt_show_no_header', False)
trials_file = (config['user_data_dir'] /
'hyperopt_results' / 'hyperopt_results.pickle')
# Previous evaluations
trials = Hyperopt.load_previous_results(trials_file)
total_epochs = len(trials)
trials = _hyperopt_filter_trials(trials, only_best, only_profitable)
n = config.get('hyperopt_show_index', -1)
if n > total_epochs:
raise OperationalException(
f"The index of the epoch to show should be less than {total_epochs + 1}.")
if n < -total_epochs:
raise OperationalException(
f"The index of the epoch to showshould be greater than {-total_epochs - 1}.")
# Translate epoch index from human-readable format to pythonic
if n > 0:
n -= 1
print_json = config.get('print_json', False)
if trials:
val = trials[n]
Hyperopt.print_epoch_details(val, total_epochs, print_json, no_header,
header_str="Epoch details")
def _hyperopt_filter_trials(trials: List, only_best: bool, only_profitable: bool) -> List:
"""
Filter our items from the list of hyperopt results
"""
if only_best:
trials = [x for x in trials if x['is_best']]
if only_profitable:
trials = [x for x in trials if x['results_metrics']['profit'] > 0]
logger.info(f"{len(trials)} " +
("best " if only_best else "") +
("profitable " if only_profitable else "") +
"epochs found.")
return trials