橘子树下

学习园地

0%

【Quantaxis】QAStrategy回测

发表于 更新于 Valine：

回测中使用的模块

  • QAStrategy

示例

Quantaxis 版本为 1.10.19。仍以 MACD_JCSC 策略进行回测。

创建 Strategy 类回测

在原 QAStrategy 回测文档中，多使用在 on_bar 函数中处理每一条 bar 数据，生成指标或回测信号。如下例代码所示的回测。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
from QAStrategy.qastockbase import QAStrategyStockBase
import QUANTAXIS as QA
from Risk_util import QA_Risk_Abstract
from collections import deque
import pandas as pd
import datetime
from rich import print


class MACD_JCSC(QAStrategyStockBase):
    def on_bar(self, bar):
        res = self.macd_jcsc()
        print(res)
        print(res.iloc[-1]) 

        if res.iloc[-1].CROSS_JC == 1:
            print("[%s] %s 出现买入信号！！！" % bar.name)
            if self.get_positions(bar.name[1]).volume_long == 0:
                self.send_order("BUY", "OPEN", code=bar.name[1], price=bar.close, volume=1000)
                print("交易：以%.2f价格买入！！！" % bar.close)
        elif res.iloc[-1].CROSS_SC == 1:
            print("[%s] %s 出现卖出信号！！！" % bar.name)
            if self.get_positions(bar.name[1]).volume_long > 0:
                self.send_order("SELL", "CLOSE",code=bar.name[1], price=bar.close, volume=1000)
                print("交易：以%.2f价格卖出！！！" % bar.close)

    def macd_jcsc(self, SHORT=12, LONG=26, M=9):
        """
        1.DIF向上突破DEA，买入信号参考。
        2.DIF向下跌破DEA，卖出信号参考。
        """
        CLOSE = self.market_data.close
        DIFF = QA.EMA(CLOSE, SHORT) - QA.EMA(CLOSE, LONG)
        DEA = QA.EMA(DIFF, M)
        MACD = 2*(DIFF-DEA)

        CROSS_JC = QA.CROSS(DIFF, DEA)
        CROSS_SC = QA.CROSS(DEA, DIFF)
        ZERO = 0
        return pd.DataFrame({'DIFF': DIFF, 'DEA': DEA, 'MACD': MACD, 'CROSS_JC': CROSS_JC, 'CROSS_SC': CROSS_SC, 'ZERO': ZERO})
    
st1=datetime.datetime.now()
print('开始回测 -- {}'.format(datetime.datetime.now()))
code_list = ['000001']
#code_list = QA.QA_fetch_stock_list_adv().code.tolist()

s = MACD_JCSC(
    code        = code_list,
    frequence   = 'day',
    start       = '2019-01-01',
    end         = '2024-06-27',
    strategy_id = 's-without_unit_init',
)


#s.debug()  //不保存进数据库
#s._debug_sim()   //模拟盘测试
s.run_backtest()
r = QA_Risk_Abstract(QA.QA_Risk(s.acc))
print('[bold magenta][风险指标][/bold magenta]')
print(r.message)
p = QA.QA_Performance(s.acc)
print('[bold magenta][绩效指标][/bold magenta]')
print(p.base_message(p.pnl))
print('回测耗时 -- {}'.format(datetime.datetime.now()-st1))

但对于像MACD这类指标，会使用前N天（比如26天）的数据进行计算，在生成market_data时，该数据在回测时即提取的历史数据源，在模拟/实盘时为行情数据。根据它的定义 

1
2
3
4
5
6
7
@property
def market_data(self):

    if self.running_mode == 'sim':
        return self._market_data
    elif self.running_mode == 'backtest':
        return pd.concat(self._market_data[-100:], axis=1, sort=False).T
会使用最后100条记录生成 DataFrame。所以在计算MACD时，指标会随着数据的更新发生一些变化。

  • 但数据不足100条时的计算情况
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
                        DIFF       DEA      MACD  CROSS_JC  CROSS_SC  ZERO
2019-01-02 000001       NaN       NaN       NaN         0         0     0
2019-01-03 000001       NaN       NaN       NaN         0         0     0
2019-01-04 000001       NaN       NaN       NaN         0         0     0
2019-01-07 000001       NaN       NaN       NaN         0         0     0
2019-01-08 000001       NaN       NaN       NaN         0         0     0
2019-01-09 000001       NaN       NaN       NaN         0         0     0
2019-01-10 000001       NaN       NaN       NaN         0         0     0
2019-01-11 000001       NaN       NaN       NaN         0         0     0
2019-01-14 000001       NaN       NaN       NaN         0         0     0
2019-01-15 000001       NaN       NaN       NaN         0         0     0
2019-01-16 000001       NaN       NaN       NaN         0         0     0
2019-01-17 000001       NaN       NaN       NaN         0         0     0
2019-01-18 000001       NaN       NaN       NaN         0         0     0
2019-01-21 000001       NaN       NaN       NaN         0         0     0
2019-01-22 000001       NaN       NaN       NaN         0         0     0
2019-01-23 000001       NaN       NaN       NaN         0         0     0
2019-01-24 000001       NaN       NaN       NaN         0         0     0
2019-01-25 000001       NaN       NaN       NaN         0         0     0
2019-01-28 000001       NaN       NaN       NaN         0         0     0
2019-01-29 000001       NaN       NaN       NaN         0         0     0
2019-01-30 000001       NaN       NaN       NaN         0         0     0
2019-01-31 000001       NaN       NaN       NaN         0         0     0
2019-02-01 000001       NaN       NaN       NaN         0         0     0
2019-02-11 000001       NaN       NaN       NaN         0         0     0
2019-02-12 000001  0.201445       NaN       NaN         0         0     0
2019-02-13 000001  0.213348       NaN       NaN         0         0     0
2019-02-14 000001  0.212171       NaN       NaN         0         0     0
2019-02-15 000001  0.190256       NaN       NaN         0         0     0
2019-02-18 000001  0.196467       NaN       NaN         0         0     0
2019-02-19 000001  0.193391       NaN       NaN         0         0     0
2019-02-20 000001  0.197590       NaN       NaN         0         0     0
2019-02-21 000001  0.195407  0.197895 -0.004975         0         0     0
2019-02-22 000001  0.202954  0.199063  0.007781         1         0     0
  • 数据大于100条后的计算情况
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
                       DIFF       DEA      MACD  CROSS_JC  CROSS_SC  ZERO
2019-01-07 000001       NaN       NaN       NaN         0         0     0
2019-01-08 000001       NaN       NaN       NaN         0         0     0
2019-01-09 000001       NaN       NaN       NaN         0         0     0
2019-01-10 000001       NaN       NaN       NaN         0         0     0
2019-01-11 000001       NaN       NaN       NaN         0         0     0
2019-01-14 000001       NaN       NaN       NaN         0         0     0
2019-01-15 000001       NaN       NaN       NaN         0         0     0
2019-01-16 000001       NaN       NaN       NaN         0         0     0
2019-01-17 000001       NaN       NaN       NaN         0         0     0
2019-01-18 000001       NaN       NaN       NaN         0         0     0
2019-01-21 000001       NaN       NaN       NaN         0         0     0
2019-01-22 000001       NaN       NaN       NaN         0         0     0
2019-01-23 000001       NaN       NaN       NaN         0         0     0
2019-01-24 000001       NaN       NaN       NaN         0         0     0
2019-01-25 000001       NaN       NaN       NaN         0         0     0
2019-01-28 000001       NaN       NaN       NaN         0         0     0
2019-01-29 000001       NaN       NaN       NaN         0         0     0
2019-01-30 000001       NaN       NaN       NaN         0         0     0
2019-01-31 000001       NaN       NaN       NaN         0         0     0
2019-02-01 000001       NaN       NaN       NaN         0         0     0
2019-02-11 000001       NaN       NaN       NaN         0         0     0
2019-02-12 000001       NaN       NaN       NaN         0         0     0
2019-02-13 000001       NaN       NaN       NaN         0         0     0
2019-02-14 000001       NaN       NaN       NaN         0         0     0
2019-02-15 000001  0.157022       NaN       NaN         0         0     0
2019-02-18 000001  0.164416       NaN       NaN         0         0     0
2019-02-19 000001  0.162747       NaN       NaN         0         0     0
2019-02-20 000001  0.168194       NaN       NaN         0         0     0
2019-02-21 000001  0.167374       NaN       NaN         0         0     0
2019-02-22 000001  0.176082       NaN       NaN         0         0     0
2019-02-25 000001  0.244633       NaN       NaN         0         0     0
2019-02-26 000001  0.273117  0.207987  0.130260         0         0     0
2019-02-27 000001  0.305015  0.230401  0.149227         0         0     0
2019-02-28 000001  0.323862  0.251342  0.145041         0         0     0
2019-03-01 000001  0.360723  0.275274  0.170898         0         0     0
2019-03-04 000001  0.400339  0.302133  0.196413         0         0     0
2019-03-05 000001  0.431294  0.329467  0.203652         0         0     0
2019-03-06 000001  0.451868  0.355074  0.193588         0         0     0
2019-03-07 000001  0.440428  0.372767  0.135322         0         0     0
2019-03-08 000001  0.397392  0.377835  0.039115         0         0     0
2019-03-11 000001  0.360469  0.374282 -0.027626         0         1     0
2019-03-12 000001  0.330084  0.365280 -0.070391         0         0     0
2019-03-13 000001  0.303178  0.352678 -0.098999         0         0     0
2019-03-14 000001  0.282642  0.338507 -0.111731         0         0     0
2019-03-15 000001  0.268008  0.324276 -0.112537         0         0     0
2019-03-18 000001  0.280965  0.315550 -0.069168         0         0     0
2019-03-19 000001  0.279837  0.308365 -0.057055         0         0     0
2019-03-20 000001  0.273061  0.301270 -0.056420         0         0     0
2019-03-21 000001  0.260586  0.293103 -0.065034         0         0     0
2019-03-22 000001  0.241083  0.282667 -0.083168         0         0     0
2019-03-25 000001  0.190592  0.264207 -0.147231         0         0     0
2019-03-26 000001  0.148189  0.240959 -0.185540         0         0     0
2019-03-27 000001  0.132268  0.219187 -0.173838         0         0     0
2019-03-28 000001  0.107424  0.196807 -0.178765         0         0     0
2019-03-29 000001  0.127513  0.182934 -0.110842         0         0     0
2019-04-01 000001  0.166284  0.179601 -0.026636         0         0     0
2019-04-02 000001  0.207012  0.185087  0.043850         1         0     0
2019-04-03 000001  0.242006  0.196477  0.091059         0         0     0
2019-04-04 000001  0.295303  0.216250  0.158106         0         0     0
2019-04-08 000001  0.340510  0.241110  0.198800         0         0     0
2019-04-09 000001  0.361793  0.265253  0.193080         0         0     0
2019-04-10 000001  0.368870  0.285981  0.165778         0         0     0
2019-04-11 000001  0.357212  0.300229  0.113965         0         0     0
2019-04-12 000001  0.335793  0.307343  0.056900         0         0     0
2019-04-15 000001  0.333666  0.312608  0.042116         0         0     0
2019-04-16 000001  0.389150  0.327918  0.122465         0         0     0
2019-04-17 000001  0.412419  0.344819  0.135200         0         0     0
2019-04-18 000001  0.425260  0.360908  0.128703         0         0     0
2019-04-19 000001  0.457213  0.380170  0.154085         0         0     0
2019-04-22 000001  0.437246  0.391586  0.091321         0         0     0
2019-04-23 000001  0.411128  0.395494  0.031268         0         0     0
2019-04-24 000001  0.411374  0.398670  0.025408         0         0     0
2019-04-25 000001  0.385593  0.396055 -0.020923         0         1     0
2019-04-26 000001  0.337650  0.384374 -0.093448         0         0     0
2019-04-29 000001  0.317533  0.371005 -0.106945         0         0     0
2019-04-30 000001  0.280976  0.352999 -0.144047         0         0     0
2019-05-06 000001  0.181789  0.318757 -0.273936         0         0     0
2019-05-07 000001  0.107507  0.276507 -0.338000         0         0     0
2019-05-08 000001  0.024025  0.226010 -0.403970         0         0     0
2019-05-09 000001 -0.071904  0.166427 -0.476663         0         0     0
2019-05-10 000001 -0.110485  0.111045 -0.443058         0         0     0
2019-05-13 000001 -0.165585  0.055719 -0.442606         0         0     0
2019-05-14 000001 -0.193798  0.005815 -0.399226         0         0     0
2019-05-15 000001 -0.184113 -0.032170 -0.303885         0         0     0
2019-05-16 000001 -0.179243 -0.061585 -0.235315         0         0     0
2019-05-17 000001 -0.201593 -0.089587 -0.224013         0         0     0
2019-05-20 000001 -0.220935 -0.115856 -0.210157         0         0     0
2019-05-21 000001 -0.221183 -0.136922 -0.168523         0         0     0
2019-05-22 000001 -0.229868 -0.155511 -0.148714         0         0     0
2019-05-23 000001 -0.241623 -0.172733 -0.137780         0         0     0
2019-05-24 000001 -0.243949 -0.186976 -0.113945         0         0     0
2019-05-27 000001 -0.241615 -0.197904 -0.087421         0         0     0
2019-05-28 000001 -0.228771 -0.204077 -0.049387         0         0     0
2019-05-29 000001 -0.222297 -0.207721 -0.029151         0         0     0
2019-05-30 000001 -0.227083 -0.211594 -0.030979         0         0     0
2019-05-31 000001 -0.230999 -0.215475 -0.031049         0         0     0
2019-06-03 000001 -0.250713 -0.222522 -0.056381         0         0     0
2019-06-04 000001 -0.266744 -0.231367 -0.070754         0         0     0
2019-06-05 000001 -0.268000 -0.238693 -0.058613         0         0     0
2019-06-06 000001 -0.269373 -0.244829 -0.049088         0         0     0

而且每次调用on_bar，都需要对指标进行一次计算。 所以考虑在QAStragety类自带的user_init中先统一根据所有数据计算完并生成信号后,再在on_bar中判断信号出现与否,再完成下单动作。修改代码如下：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
from QAStrategy.qastockbase import QAStrategyStockBase
import QUANTAXIS as QA
from Risk_util import QA_Risk_Abstract
from collections import deque
import pandas as pd
import datetime
from rich import print


class MACD_JCSC(QAStrategyStockBase):
    def user_init(self):
        data = QA.QA_fetch_stock_day_adv(
            self.code, self.start, self.end).to_qfq()
        self.signal = data.add_func(self.macd_jcsc)

    def on_bar(self, bar):
        cursor_date = bar.name[0].date().strftime("%Y-%m-%d")
        cursor_code = bar.name[1]

        if self.signal.loc[cursor_date].loc[cursor_code].CROSS_JC == 1:
            print("[%s] %s 出现买入信号！！！" % bar.name)
            if self.get_positions(bar.name[1]).volume_long == 0:
                self.send_order("BUY", "OPEN", code=bar.name[1], price=bar.close, volume=1000)
                print("交易：以%.2f价格买入！！！" % bar.close)
        elif self.signal.loc[cursor_date].loc[cursor_code].CROSS_SC == 1:
            print("[%s] %s 出现卖出信号！！！" % bar.name)
            if self.get_positions(bar.name[1]).volume_long > 0:
                self.send_order("SELL", "CLOSE",code=bar.name[1], price=bar.close, volume=1000)
                print("交易：以%.2f价格卖出！！！" % bar.close)

    def macd_jcsc(self, dataframe, SHORT=12, LONG=26, M=9):
        """
        1.DIF向上突破DEA，买入信号参考。
        2.DIF向下跌破DEA，卖出信号参考。
        """
        CLOSE = dataframe.close
        DIFF = QA.EMA(CLOSE, SHORT) - QA.EMA(CLOSE, LONG)
        DEA = QA.EMA(DIFF, M)
        MACD = 2*(DIFF-DEA)

        CROSS_JC = QA.CROSS(DIFF, DEA)
        CROSS_SC = QA.CROSS(DEA, DIFF)
        ZERO = 0
        return pd.DataFrame({'DIFF': DIFF, 'DEA': DEA, 'MACD': MACD, 'CROSS_JC': CROSS_JC, 'CROSS_SC': CROSS_SC, 'ZERO': ZERO})

st1=datetime.datetime.now()
print('开始回测 -- {}'.format(datetime.datetime.now()))
code_list = ['000001']
#code_list = QA.QA_fetch_stock_list_adv().code.tolist()

s = MACD_JCSC(
    code        = code_list,
    frequence   = 'day',
    start       = '2019-01-01',
    end         = '2024-06-27',
    strategy_id = 's-with_unit_init',
)


#s.debug()  //不保存进数据库
#s._debug_sim()   //模拟盘测试
s.run_backtest()
r = QA_Risk_Abstract(QA.QA_Risk(s.acc))
print('[bold magenta][风险指标][/bold magenta]')
print(r.message)
p = QA.QA_Performance(s.acc)
print('[bold magenta][绩效指标][/bold magenta]')
print(p.base_message(p.pnl))
print('回测耗时 -- {}'.format(datetime.datetime.now()-st1))

修改前完成单只股票5年半的回测需要 33.95秒, 而修改后仅需 25.71秒。

回测结果

回测的结果可以直接在终端上可见，也可以能过 Web 方式(http://127.0.0.1:81/) 查询。

风险分析

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
{
    'account_cookie': 's-with_unit_init',
    'annualize_return': 0.0,
    'profit': 0.0,
    'max_dropback': 0.01,
    'time_gap': 1330,
    'volatility': 0.43,
    'benchmark_code': '000300',
    'bm_annualizereturn': 0.03,
    'bm_profit': 0.16,
    'beta': 0.01,
    'alpha': -0.0,
    'sharpe': -0.12,
    'sortino': -395.75,
    'last_assets': '1002633.56',
    'total_tax': -644.7,
    'total_commission': -481.4,
    'profit_money': 2633.56,
    'ir': 0.0,
    'month_profit': {
        '2019-01-31 00:00:00': 0.0,
        '2019-02-28 00:00:00': 703.1234393180348,
        '2019-03-31 00:00:00': -50.181729689240456,
        '2019-04-30 00:00:00': 642.743004972348,
        '2019-05-31 00:00:00': 0.0,
        '2019-06-30 00:00:00': 1368.4163899420528,
        '2019-07-31 00:00:00': -353.5741839206312,
        '2019-08-31 00:00:00': -773.1137744144071,
        '2019-09-30 00:00:00': 68.32695359084755,
        '2019-10-31 00:00:00': 705.2476993447635,
        '2019-11-30 00:00:00': 0.0,
        '2019-12-31 00:00:00': 684.6625807939563,
        '2020-01-31 00:00:00': 250.99576606589835,
        '2020-02-29 00:00:00': -764.5018294820329,
        '2020-03-31 00:00:00': -639.1336695251521,
        '2020-04-30 00:00:00': 986.47938771802,
        '2020-05-31 00:00:00': -627.6430917717516,
        '2020-06-30 00:00:00': -518.7461710819043,
        '2020-07-31 00:00:00': 607.8399459136417,
        '2020-08-31 00:00:00': 220.76756709371693,
        '2020-09-30 00:00:00': 17.090446144808084,
        '2020-10-31 00:00:00': 1495.516639258014,
        '2020-11-30 00:00:00': 711.2246544788359,
        '2020-12-31 00:00:00': -293.4999780562939,
        '2021-01-31 00:00:00': 2069.263243380934,
        '2021-02-28 00:00:00': 621.7155406653183,
        '2021-03-31 00:00:00': 456.697427464067,
        '2021-04-30 00:00:00': 229.10599464632105,
        '2021-05-31 00:00:00': -48.13997128745541,
        '2021-06-30 00:00:00': -277.31763577868696,
        '2021-07-31 00:00:00': 0.0,
        '2021-08-31 00:00:00': -626.0138344422448,
        '2021-09-30 00:00:00': -1153.6153345350176,
        '2021-10-31 00:00:00': 84.48616837000009,
        '2021-11-30 00:00:00': -1622.0463032894768,
        '2021-12-31 00:00:00': -97.31165447854437,
        '2022-01-31 00:00:00': -2169.600942116813,
        '2022-02-28 00:00:00': -588.2864658574108,
        '2022-03-31 00:00:00': 701.9407301229658,
        '2022-04-30 00:00:00': -492.56538836390246,
        '2022-05-31 00:00:00': -407.6765322362771,
        '2022-06-30 00:00:00': 728.7865806339541,
        '2022-07-31 00:00:00': -519.0264483509818,
        '2022-08-31 00:00:00': 359.32954159996007,
        '2022-09-30 00:00:00': -307.7851737852907,
        '2022-10-31 00:00:00': 0.0,
        '2022-11-30 00:00:00': 2007.9207173399627,
        '2022-12-31 00:00:00': 46.82587729266379,
        '2023-01-31 00:00:00': 586.5987326084869,
        '2023-02-28 00:00:00': 0.0,
        '2023-03-31 00:00:00': -523.4222982611973,
        '2023-04-30 00:00:00': -376.4062199871987,
        '2023-05-31 00:00:00': -239.97391489637084,
        '2023-06-30 00:00:00': -623.9189663539873,
        '2023-07-31 00:00:00': 769.7435180819593,
        '2023-08-31 00:00:00': -622.6199320594314,
        '2023-09-30 00:00:00': -305.4354766883189,
        '2023-10-31 00:00:00': -183.30302026984282,
        '2023-11-30 00:00:00': -627.8779737370787,
        '2023-12-31 00:00:00': 200.35370358801447,
        '2024-01-31 00:00:00': 65.3398147780681,
        '2024-02-29 00:00:00': 1054.7712957019685,
        '2024-03-31 00:00:00': -257.3330304133706,
        '2024-04-30 00:00:00': -14.334259253926575,
        '2024-05-31 00:00:00': 292.6508590914309,
        '2024-06-30 00:00:00': 0.0
    }
}

绩效分析

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
{
    'total_profit': 17887.74,
    'total_loss': -14128.09,
    'total_pnl': 1.27,
    'trading_amounts': 48,
    'profit_amounts': 20,
    'loss_amounts': 28,
    'even_amounts': 0,
    'profit_precentage': 0.42,
    'loss_precentage': 0.58,
    'even_precentage': 0.0,
    'average_profit': 894.39,
    'average_loss': -504.57,
    'average_pnl': 1.77,
    'max_profit': 3782.37,
    'max_loss': -1342.29,
    'max_pnl': 2.82,
    'netprofio_maxloss_ratio': 2.8,
    'continue_profit_amount': 4,
    'continue_loss_amount': 9,
    'average_holdgap': '20 days 00:00:00',
    'average_profitholdgap': '32 days 15:36:00',
    'average_losssholdgap': '10 days 23:08:34.285714285'
}

回测中遇到的问题

回测使用的 pandas 版本为 2.2以上，比 Quantaxis 发布时使用的版本要高得多，所以在回测时会遇到大量用法未来会舍弃的告警。如：

①FutureWarning: Series.__getitem__ treating keys as positions is deprecated. In a future version, integer keys will always be treated as labels (consistent with DataFrame behavior). To access a value by position, use ser.iloc[pos]
②FutureWarning: 'M' is deprecated and will be removed in a future version, please use 'ME' instead.
③Removed deprecated Series.iteritems(), DataFrame.iteritems() and HDFStore.iteritems() use obj.items instead (GH 45321)
④FutureWarning: Series.fillna with 'method' is deprecated and will raise in a future version. Use obj.ffill() or obj.bfill() instead.
⑤FutureWarning: The 'axis' keyword in DataFrame.groupby is deprecated and will be removed in a future version.

可以有两种办法解决。

  1. 屏蔽告警 

    1
    2
    import warnings
    warnings.filterwarnings('ignore')

  2. 根据告警中的提示修改源代码以符合未来版本的要求。 我在所修改的部分，都加注了详细的文档注解，方便以后回溯。 具体改动不再详例。

参考

  1. 投资没有银弹 -- （五连阳回测代码勘误与更快的回测实现)
  2. QUANTAXIS探索（七）量化策略实现类QAStrategy的回测模式