pa = parameter_analysis('26e2c7b189e7e2cc49447adf71415cbd') pa.get_backtest_data(file_name = 'results.pkl', running_max = 10, benchmark_id = '000300.XSHG', start_date = '2015-01-01', end_date = '2016-12-31', frequency = 'day', initial_cash = '1000000', param_names = [], param_values = [] ) NameError Traceback (most recent call last) <ipython-input-1-4781ef98af74> in <module>() ----> 1 pa = parameter_analysis('26e2c7b189e7e2cc49447adf71415cbd') 2 pa.get_backtest_data(file_name = 'results.pkl', 3 running_max = 10, 4 benchmark_id = '000300.XSHG', 5 start_date = '2015-01-01', NameError: name 'parameter_analysis' is not defined 这是原因，是指样子使用这个的吗？我想看一个多因子策略10个分位的表现

结果一多中间就显示点点点...有什么办法处理么

@Sunx 还没有跑成功过，麻烦帮忙看下，谢谢 【运行中|已完成|待运行】: [0|0|10]. [0|10|0]. [9|1|0]. 【回测完成】总用时：26秒(即0.01小时)。 --------------------------------------------------------------------------- Exception Traceback (most recent call last) <ipython-input-20-311ba23468b3> in <module>() 8 initial_cash = '2000000', 9 param_names = ['factor', 'quantile'], ---> 10 param_values = [['BP'], tuple(zip(range(0,100,10), range(10,101,10)))] 11 ) 12 <ipython-input-5-83b03aa5034a> in get_backtest_data(self, algorithm_id, benchmark_id, file_name, running_max, start_date, end_date, frequency, initial_cash, param_names, param_values) 234 ) 235 # 回测结果指标中加入 log 收益率和超额收益率等指标 --> 236 self.organize_backtest_results(benchmark_id) 237 # 生成 dict 保存所有结果。 238 results = {'returns':self.returns, <ipython-input-5-83b03aa5034a> in organize_backtest_results(self, benchmark_id) 180 else: 181 # 基准使用新加入的基准回测结果 --> 182 self.benchmark_returns = [x['returns'] for x in get_backtest(benchmark_id).get_results()] 183 # 回测日期为结果中记录的第一项对应的日期 184 self.dates = [x['time'] for x in self.results[0]] /opt/conda/envs/python2/lib/python2.7/site-packages/jqresearch/base.py in get_results(self) 123 如果没有收益则返回一个空的 list. 124 """ --> 125 return self.__request_json("results") 126 127 def get_positions(self): /opt/conda/envs/python2/lib/python2.7/site-packages/jqresearch/base.py in __request_json(self, info_type) 86 } 87 ---> 88 return request_json("/get_backtest", params) 89 90 def get_params(self): /opt/conda/envs/python2/lib/python2.7/site-packages/jqresearch/base.py in request_json(path, data) 56 57 def request_json(path, data={}): ---> 58 return json.loads(request_data(path, data)) 59 60 /opt/conda/envs/python2/lib/python2.7/site-packages/jqresearch/base.py in request_data(path, data) 51 body = resp.content.decode('utf-8-sig') 52 if resp.status_code != requests.codes.ok: ---> 53 raise Exception(body) 54 return body 55 Exception: 无效的回测ID, ID应该从回测详情页的 url 中获取, 或者是 create_backtest 的返回值 我的代码地址： https://www.joinquant.net/research?target=research&url=/user/28852065939/notebooks/%E5%8F%82%E6%95%B0%E4%BC%98%E5%8C%96.ipynb 我的引用代码是这样的： #2 设定回测策略 id pa = parameter_analysis('4d72f79e11b98a84699e49be58f24128') #3 运行回测 pa.get_backtest_data(file_name = 'results.pkl', running_max = 10, benchmark_id = '', start_date = '2017-8-01', end_date = '2017-12-01', frequency = 'day', initial_cash = '2000000', param_names = ['factor', 'quantile'], param_values = [['BP'], tuple(zip(range(0,100,10), range(10,101,10)))]

@代数量化 那个有两个原因，一个是运行占的内存超限，进程被杀死了。另一个是有时候网络不稳，中间断了一下。

@maxking 看看是不是没有修改原是的策略id编号，因为对应的是我自己的策略，所以你调用时默认那个找不到

@AlfredH4869 可以查看api文档，那个东西就是通过研究调用回测得一个函数

@mouetteo 最多只能一次运行10个回测，你如果手动停止研究模块运行，但是回测那里还没运行完就会出现这个问题，你可以手动结束对应的回测进程就好了

@Sunx 谢谢！解决了

@alan-kulolo 先把前边一大堆运行一次，然后在跑后边的，系统就认识那个函数了

@天道量化 没看明白，首先创造一个回测，然后复制会侧的id对下面这个初始化。说是无效id，是不是那个id没复制对？ ``` pa = parameter_analysis('4d72f79e11b98a84699e49be58f24128') ```

好

@周小泉 他的代码有bug 修改一下第112行，把fail的backtest打出来, 然后把id塞进去，自己看看错哪里， https://www.joinquant.com//algorithm/backtest/detail?backtestId= ``` bt = get_backtest(running_backtests[key]) print(bt.backtest_id) ``` 另外，代码没有处理fail之后，已经跑了的test的问题，导致to_run计数错误，在 ``` for key in failed: running_backtests.pop(key) ``` 改成 ``` for key in failed: finished_backtests[key] = running_backtests.pop(key) ```

@JoinQuant量化课堂 大神求帮忙！！ 我用多回测模板，输入回测信息如下： ``` pa.get_backtest_data(file_name='%s.pkl'%out_file_name,#结果保存，名称最好改变 running_max=10, start_date='2005-01-01',#开始时间 end_date='2018-03-31', #结束时间 frequency='day', #回测频度 initial_cash='10000000', #初始资金 param_names=['group_id'], #调试的参数名称 param_values=[list(range(1,11))]) #对应的参数值列表 ``` 但是会报错，错误为： TypeError Traceback (most recent call last) <ipython-input-54-2637b8f371bd> in <module>() 7 initial_cash='10000000', #初始资金 8 param_names=['group_id'], #调试的参数名称 ----> 9 param_values=[list(range(1,11))]) #对应的参数值列表 <ipython-input-51-a43d9abb4b4f> in get_backtest_data(self, algorithm_id, benchmark_id, file_name, running_max, start_date, end_date, frequency, initial_cash, param_names, param_values) 237 initial_cash=initial_cash, 238 param_names=param_names, --> 239 param_values=param_values 240 ) 241 # 回测结果指标中加入 log 收益率和超额收益率等指标 <ipython-input-51-a43d9abb4b4f> in run_backtest(self, algorithm_id, running_max, start_date, end_date, frequency, initial_cash, param_names, param_values) 101 extras = params, 102 # 在回测中把改参数的结果起一个名字，包含了所有涉及的变量参数值 --> 103 name = str(params) 104 ) 105 # 记录运行中 i 回测的回测 id TypeError: 1 not serializable 但是我用自己编写的循环建立多回测中，却可以提交多回测，不会报错，代码如下： ``` #创建回测 algorithm_id='0e1fec8acb20bce09f7e6d01ad0dd429' initial_positions=[] group_id=list(range(1,11)) bt_id_list=[] for i in range(10): params = { "algorithm_id": algorithm_id, "start_date": "2005-01-01", "end_date": "2018-03-31", "frequency": "day", "initial_cash": "10000000", "initial_positions": initial_positions, "extras": {'group_id':group_id[i]} } bt_id_list.append(create_backtest(**params)) ``` @JoinQuant量化课堂 求帮忙debug一下，谢谢！感激不尽！

昨天开始出现这个错误，有些参数组能正常运行，很多时候运行到最后就报这个错误，麻烦大神帮忙看看 【已完成|运行中|待运行】: [0|0|216]. [0|8|208]. ……………………………………………………[182|8|26]. [182|8|26]. [183|7|26]. KeyError Traceback (most recent call last) <ipython-input-22-4bc0e22cca0f> in <module>() 6 initial_cash = '10000000', 7 param_names = ['N','R_period', 'HOLD'], ----> 8 param_values = [range(8,20,2),range(45,71,5),range(2,13,2)] 9 ) <ipython-input-20-d81d57d08a84> in get_backtest_data(self, algorithm_id, benchmark_id, file_name, running_max, start_date, end_date, frequency, initial_cash, param_names, param_values) 231 initial_cash=initial_cash, 232 param_names=param_names, --> 233 param_values=param_values 234 ) 235 # 回测结果指标中加入 log 收益率和超额收益率等指标 <ipython-input-20-d81d57d08a84> in run_backtest(self, algorithm_id, running_max, start_date, end_date, frequency, initial_cash, param_names, param_values) 87 for i in range(pointer, pointer+to_run): 88 # 备选的参数排列组合的 df 中第 i 行变成 dict，每个 key 为列名字，value 为 df 中对应的值 ---> 89 params = to_run_df.ix[i].to_dict() 90 # 记录策略回测结果的 id，调整参数 extras 使用 params 的内容 91 backtest = create_backtest(algorithm_id = algorithm_id, /opt/conda/envs/python2/lib/python2.7/site-packages/pandas/core/indexing.pyc in __getitem__(self, key) 68 return self._getitem_tuple(key) 69 else: ---> 70 return self._getitem_axis(key, axis=0) 71 72 def _get_label(self, label, axis=0): /opt/conda/envs/python2/lib/python2.7/site-packages/pandas/core/indexing.pyc in _getitem_axis(self, key, axis) 927 return self._get_loc(key, axis=axis) 928 --> 929 return self._get_label(key, axis=axis) 930 931 def _getitem_iterable(self, key, axis=0): /opt/conda/envs/python2/lib/python2.7/site-packages/pandas/core/indexing.pyc in _get_label(self, label, axis) 84 raise IndexingError('no slices here, handle elsewhere') 85 ---> 86 return self.obj._xs(label, axis=axis) 87 88 def _get_loc(self, key, axis=0): /opt/conda/envs/python2/lib/python2.7/site-packages/pandas/core/generic.pyc in xs(self, key, axis, level, copy, drop_level) 1471 drop_level=drop_level) 1472 else: -> 1473 loc = self.index.get_loc(key) 1474 1475 if isinstance(loc, np.ndarray): /opt/conda/envs/python2/lib/python2.7/site-packages/pandas/core/index.pyc in get_loc(self, key, method) 1570 """ 1571 if method is None: -> 1572 return self._engine.get_loc(_values_from_object(key)) 1573 1574 indexer = self.get_indexer([key], method=method) pandas/index.pyx in pandas.index.IndexEngine.get_loc (pandas/index.c:3824)() pandas/index.pyx in pandas.index.IndexEngine.get_loc (pandas/index.c:3704)() pandas/hashtable.pyx in pandas.hashtable.Int64HashTable.get_item (pandas/hashtable.c:7224)() pandas/hashtable.pyx in pandas.hashtable.Int64HashTable.get_item (pandas/hashtable.c:7162)() KeyError: 216

# 因order函数调整为顺序调整，为防止先行调仓股票由于后行调仓股票占金额过大不能一次调整到位，这里运行两次以解决这个问题 for stock_to_buy in holding_list[start_q * total_number / 100 : end_q * total_number / 100]: order_target_value(stock_to_buy, every_stock) for stock_to_buy in holding_list[start_q * total_number / 100 : end_q * total_number / 100]: order_target_value(stock_to_buy, every_stock) 请问这里是什么原因运行两次，没看明白

换了自己的策略，总通不过，报错，该怎么改呀，谢谢了！ @JoinQuant量化课堂 TypeError Traceback (most recent call last) in () 13 14 # 定义类'参数分析' ---> 15 class parameter_analysis('126926f701399e7c63fa1e6aee1d40d6'): 16 17 # 定义函数中不同的变量 TypeError: Error when calling the metaclass bases str() takes at most 1 argument (3 given)

@wwwlll ---> 15 class parameter_analysis('126926f701399e7c63fa1e6aee1d40d6'):-------不是这样使用的。您在这里面传入您的id #2 设定回测策略 id pa = parameter_analysis('bce2e5c55b3b631f91985c9bf113414f')

@alanx # 因order函数调整为顺序调整，为防止先行调仓股票由于后行调仓股票占金额过大不能一次调整到位，这里运行两次以解决这个问题

@同乐娃哈哈 param_values = [['BP'], tuple(zip(range(0,100,10), range(10,101,10)))] 您传入的参数数据类型不对。