diff --git a/src/.keep b/src/.keep deleted file mode 100644 index e69de29bb2d1d6434b8b29ae775ad8c2e48c5391..0000000000000000000000000000000000000000 diff --git a/src/FixedRatioStrategyFill_V2.py b/src/FixedRatioStrategyFill_V2.py deleted file mode 100644 index 929aa08b1c43c9b91d68711621bb3a7ffd47fcde..0000000000000000000000000000000000000000 --- a/src/FixedRatioStrategyFill_V2.py +++ /dev/null @@ -1,104 +0,0 @@ -from re import X -import numpy as np -import pandas as pd - -from core.model.strategy_config import PosStrategyConfig - - -def calc_ratio(equity_dfs: list[pd.DataFrame], stg_conf: PosStrategyConfig) -> pd.DataFrame: - """ - 计算选币仓位结果，只接受两个参数，1. 资金曲线们，2. 策略配置 - :param equity_dfs: 资金曲线，是一个df的列表，里面包含配置中要求计算好的因子 - :param stg_conf: 策略配置 - :return: 返回仓位比 - """ - # 取出对应参数 - cap_ratios = stg_conf.params['cap_ratios'] # 默认资金分配 - pos_limit = stg_conf.params['pos_limit'] - monitor_strategy = pos_limit['monitor_strategy'] # 监控的策略 - - # 获取各种情况下的资金分配方案 - alt_ratios_0 = pos_limit.get('alt_ratios_0', cap_ratios) # 空仓时的资金分配 - alt_ratios_1 = pos_limit.get('alt_ratios_1', cap_ratios) # 选1个币时的资金分配 - alt_ratios_2 = pos_limit.get('alt_ratios_2', cap_ratios) # 选2个币时的资金分配 - alt_ratios_3 = pos_limit.get('alt_ratios_3', cap_ratios) # 选3个币时的资金分配 - alt_ratios_4 = pos_limit.get('alt_ratios_4', cap_ratios) # 选4个币时的资金分配 - alt_ratios_5plus = pos_limit.get('alt_ratios_5plus', cap_ratios) # 选5个及以上币时的资金分配 - - monitor_df = None - monitor_index = None - - print(f"策略列表长度: {len(stg_conf.strategy_cfg_list)}") - for i, cfg in enumerate(stg_conf.strategy_cfg_list): - print(f"策略{i}: {cfg.name}") - - # 找到需要监控的策略 - for index, cfg in enumerate(stg_conf.strategy_cfg_list): - if cfg.name.endswith(monitor_strategy): - equity_df = pd.read_csv(cfg.get_result_folder() / '资金曲线.csv', encoding='utf-8-sig', parse_dates=['candle_begin_time']) - monitor_df = equity_df - monitor_index = index - break - - if monitor_df is None: - print(f"未找到监控策略: {monitor_strategy}，使用默认资金分配...") - - # 判断资金占比的参数数量是否与资金曲线的数量一致 - all_ratios = [cap_ratios, alt_ratios_0, alt_ratios_1, alt_ratios_2, alt_ratios_3, alt_ratios_4, alt_ratios_5plus] - for ratios in all_ratios: - if len(ratios) != len(equity_dfs): - print(f'资金比例数量与资金曲线数量不同，请检查配置: {ratios}') - exit() - - # 构建时间轴 - begin_times = equity_dfs[0]['candle_begin_time'] - - # 创建一个全 0 的 NDArray - cap_weights = np.zeros([len(begin_times), len(equity_dfs)]) - - # 分配默认权重 - for i in range(len(cap_ratios)): - cap_weights[:, i] = cap_ratios[i] - - # 根据监控策略的选币数量动态调整资金分配 - if monitor_df is not None: - # 创建选币数量掩码 - long_num = monitor_df['symbol_long_num'].values - short_num = monitor_df['symbol_short_num'].values - - # 确保长度匹配 - if len(long_num) != len(begin_times): - print('警告：监控策略的时间轴与其他策略不一致，可能导致结果不准确') - # 创建DataFrame时需要转换index类型 - return pd.DataFrame(cap_weights, columns=range(cap_weights.shape[1]), index=begin_times) - - # 计算总选币数量 - total_num = np.add(long_num, short_num) # 使用numpy的add函数 - - # 创建不同选币数量的掩码 - mask_0 = (total_num == 0) # 空仓 - mask_1 = (total_num == 1) # 选1个币 - mask_2 = (total_num == 2) # 选2个币 - mask_3 = (total_num == 3) # 选3个币 - mask_4 = (total_num == 4) # 选4个币 - mask_5plus = (total_num >= 5) # 选5个及以上币 - - # 根据不同选币数量应用不同的资金分配方案 - for i in range(len(equity_dfs)): - # 空仓情况 - cap_weights[mask_0, i] = alt_ratios_0[i] - # 选1个币情况 - cap_weights[mask_1, i] = alt_ratios_1[i] - # 选2个币情况 - cap_weights[mask_2, i] = alt_ratios_2[i] - # 选3个币情况 - cap_weights[mask_3, i] = alt_ratios_3[i] - # 选4个币情况 - cap_weights[mask_4, i] = alt_ratios_4[i] - # 选5个及以上币情况 - cap_weights[mask_5plus, i] = alt_ratios_5plus[i] - - # 将结果转换回 DataFrame，注意index可能需要转换 - cap_weights = pd.DataFrame(cap_weights, columns=range(cap_weights.shape[1]), index=begin_times) - - return cap_weights \ No newline at end of file