def gekko_search(**parameters): parallel = settings['parallel'] num_rounds = settings['num_rounds'] # remake CS & HS variability; candleSize = settings['candleSize'] historySize = settings['historySize'] if parallel: p = Pool(mp.cpu_count()) param_list = list([ (Strategy, parameters), ] * num_rounds) scores = p.starmap(evaluate_random, param_list) p.close() p.join() else: scores = [ evaluate_random(Strategy, parameters) for n in range(num_rounds) ] print(scores) series = pd.Series(scores) mean = series.mean() stats.append([series.count( ), mean, series.std(), series.min()] + [series.quantile(x) for x in percentiles] + [series.max()]) all_val.append(mean) write_evolution_logs(len(all_val), stats[-1]) return mean
def gekko_search(**parameters): parallel = settings['parallel'] num_rounds = settings['num_rounds'] # remake CS & HS variability; candleSize= settings['candleSize'] historySize= settings['historySize'] if parallel: p = Pool(mp.cpu_count()) param_list = list([(Strategy, parameters),] * num_rounds) scores = p.starmap(Evaluate, param_list) p.close() p.join() else: scores = [Evaluate(Strategy, parameters) for n in range(num_rounds)] series = pd.Series(scores) mean = series.mean() stats.append([series.count(), mean, series.std(), series.min()] + [series.quantile(x) for x in percentiles] + [series.max()]) all_val.append(mean) write_evolution_logs(len(all_val), stats[-1]) return mean