update

2024.03.07_akshare_strategy/check_data.py

@@ -0,0 +1,14 @@
+import guan
+
+stock_symbols_60, stock_symbols_00, stock_symbols_30, stock_symbols_68, \
+stock_symbols_8_4,stock_symbols_others = guan.stock_symbols_classification()
+print(len(stock_symbols_60))
+print(len(stock_symbols_00))
+print()
+
+file_list_60 = guan.get_all_filenames_in_directory(directory='./all_to_2024_03_07/stock_data_60', \
+                                        file_format=None, show_root_path=0, sort=1, include_subdirectory=1)
+print(len(file_list_60))
+file_list_00 = guan.get_all_filenames_in_directory(directory='./all_to_2024_03_07/stock_data_00', \
+                                        file_format=None, show_root_path=0, sort=1, include_subdirectory=1)
+print(len(file_list_00))

2024.03.07_akshare_strategy/data_processing.py

@@ -0,0 +1,38 @@
+import guan
+import numpy as np
+import datetime
+
+file_list_60 = guan.get_all_filenames_in_directory(directory='./all_to_2024_03_07/stock_data_60', \
+                                        file_format=None, show_root_path=0, sort=1, include_subdirectory=1)
+file_list_00 = guan.get_all_filenames_in_directory(directory='./all_to_2024_03_07/stock_data_00', \
+                                        file_format=None, show_root_path=0, sort=1, include_subdirectory=1)
+
+
+len_stock_60 = len(file_list_60)
+guan.make_directory('./2018_01_01_to_2024_03_07/')
+guan.make_directory('./2018_01_01_to_2024_03_07/stock_data_60/')
+for i00 in range(len_stock_60):
+    stock_data = guan.load_data('./all_to_2024_03_07/stock_data_60/'+file_list_60[i00], file_format='')
+    num_data = len(stock_data)
+    new_stock_data = []
+    for i0 in range(num_data):
+        if stock_data[i0, 0]>datetime.date(2018, 1, 1):
+            new_stock_data.append(list(stock_data[i0, :]))
+    new_stock_data = np.array(new_stock_data)
+    guan.dump_data(data=new_stock_data, filename='./2018_01_01_to_2024_03_07/stock_data_60/'+file_list_60[i00], file_format='')
+
+print('数据处理结束1')
+
+len_stock_00 = len(file_list_00)
+guan.make_directory('./2018_01_01_to_2024_03_07/stock_data_00/')
+for i00 in range(len_stock_00):
+    stock_data = guan.load_data('./all_to_2024_03_07/stock_data_00/'+file_list_00[i00], file_format='')
+    num_data = len(stock_data)
+    new_stock_data = []
+    for i0 in range(num_data):
+        if stock_data[i0, 0]>datetime.date(2018, 1, 1):
+            new_stock_data.append(list(stock_data[i0, :]))
+    new_stock_data = np.array(new_stock_data)
+    guan.dump_data(data=new_stock_data, filename='./2018_01_01_to_2024_03_07/stock_data_00/'+file_list_00[i00], file_format='')
+
+print('数据处理结束2')

2024.03.07_akshare_strategy/download_data.py

@@ -0,0 +1,28 @@
+import guan
+stock_symbols_60, stock_symbols_00, stock_symbols_30, stock_symbols_68, \
+stock_symbols_8_4,stock_symbols_others = guan.stock_symbols_classification()
+
+import os
+guan.make_directory('./all_to_2024_03_07')
+
+guan.make_directory('./all_to_2024_03_07/stock_data_60')
+for stock_symbol in stock_symbols_60:
+    try:
+        if not os.path.exists('./all_to_2024_03_07/stock_data_60/'+stock_symbol+'.txt'):
+            title, stock_data = guan.history_data_of_one_stock(symbol=stock_symbol, \
+                                period='daily', start_date="19000101", end_date='21000101')
+            guan.dump_data(data=stock_data, filename='./all_to_2024_03_07/stock_data_60/'+stock_symbol, file_format='.txt')
+    except:
+        pass
+print('60下载完成')
+
+guan.make_directory('./all_to_2024_03_07/stock_data_00')
+for stock_symbol in stock_symbols_00:
+    try:
+        if not os.path.exists('./all_to_2024_03_07/stock_data_00/'+stock_symbol+'.txt'):
+            title, stock_data = guan.history_data_of_one_stock(symbol=stock_symbol, \
+                                period='daily', start_date="19000101", end_date='21000101')
+            guan.dump_data(data=stock_data, filename='./all_to_2024_03_07/stock_data_00/'+stock_symbol, file_format='.txt')
+    except:
+        pass
+print('00下载完成')

2024.03.07_akshare_strategy/strategy.py

@@ -0,0 +1,79 @@
+import guan
+file_list_60 = guan.get_all_filenames_in_directory(directory='./2018_01_01_to_2024_03_07/stock_data_60', \
+                                        file_format=None, show_root_path=0, sort=1, include_subdirectory=1)
+file_list_00 = guan.get_all_filenames_in_directory(directory='./2018_01_01_to_2024_03_07/stock_data_00', \
+                                        file_format=None, show_root_path=0, sort=1, include_subdirectory=1)
+# print(len(file_list_60))
+# print(len(file_list_00))
+# print()
+
+# 回测策略：当跌到成本时的 ratio 倍时，翻倍投入补仓。（这里只对00开头的股票进行回测）
+buy_limit = 100
+for buy_limit in [100000, 10, 4]: # 限制补仓次数
+    print('限制补仓次数：', buy_limit, '\n')
+    for ratio in [0.95, 0.9, 0.85, 0.8, 0.75, 0.7, 0.65]:
+        win_count = 0
+        lose_count = 0
+        win_count_relative = 0
+        lose_count_relative = 0
+        win_count_0 = 0
+        lose_count_0 = 0
+        print('跌到成本的某个倍数补仓：', ratio)
+        invest_money_array = []
+        current_money_array = []
+        buy_times_array = []
+        profit_array = []
+        profit_array_0 = []
+        for i00 in range(len(file_list_00)):
+            stock_data = guan.load_data('./2018_01_01_to_2024_03_07/stock_data_00/'+file_list_00[i00], file_format='')
+            stock_data = stock_data[::-1]
+            if len(stock_data) > 0:
+                invest_money = 1000  # 初始投入资金
+                current_money = 1000  # 当前所有资金
+                buy_price = stock_data[0, 2]  # 买入的价格（第一天买入）
+                cost_price = stock_data[0, 2]  # 当前持有的成本价（第一天买入）
+                cost_num = invest_money/cost_price   # 当前持有的股数（为了简化，这里不做取整处理）
+                len_date = len(stock_data[:, 2])
+                buy_times = 0
+                for i0 in range(len_date):
+                    if stock_data[i0, 2]/cost_price < ratio and buy_times < buy_limit:  # 当满足策略，且补仓次数小于buy_limit时，补仓
+                        current_price = stock_data[i0, 2]   # 当前的市场价
+                        current_money = invest_money + (invest_money/cost_price)*current_price  # 补仓后的所有资金
+                        buy_price = stock_data[i0, 2]  # 买入价为当前市场价
+                        cost_num = (invest_money/cost_price)+(invest_money/buy_price)  # 更新当前持有的股数
+                        cost_price = 2*invest_money/cost_num # 更新当前持有的成本价
+                        invest_money = invest_money*2  # 更新总投入资金
+                        buy_times += 1
+                    current_price = stock_data[i0, 2]   # 当前的市场价
+                    current_money = cost_num*current_price # 当前所有资金 
+                buy_times_array.append(buy_times)
+                invest_money_array.append(invest_money)
+                current_money_array.append(current_money)
+                profit_array.append(current_money/invest_money)
+                profit_array_0.append(current_price/stock_data[0, 2])
+                if current_money > invest_money:
+                    win_count += 1
+                else:
+                    lose_count += 1
+                if current_money/invest_money > current_price/stock_data[0, 2]:
+                    win_count_relative += 1
+                else:
+                    lose_count_relative += 1
+                if current_price > stock_data[0, 2]:
+                    win_count_0 += 1
+                else:
+                    lose_count_0 += 1
+        print('策略的平均加仓次数：', sum(buy_times_array)/len(buy_times_array))
+        print('策略的平均投入资金：', sum(invest_money_array)/len(invest_money_array))
+        print('策略的平均最终所有资金：', sum(invest_money_array)/len(invest_money_array))
+        print('策略的平均利润率（包含本金）：', sum(profit_array)/len(profit_array))
+        print('完成不操作的市场平均利润率（包含本金）：', sum(profit_array_0)/len(profit_array_0))
+        print('策略赢的次数：', win_count)
+        print('策略输的次数；', lose_count)
+        print('策略相对于市场赢的次数：', win_count_relative)
+        print('策略相对于市场输的次数；', lose_count_relative)
+        print('完全不操作市场赢的次数：', win_count_0)
+        print('完全不操作市场输的次数；', lose_count_0)
+        print()
+    print('-----')
+    print()