黄金矿工

1
# Generate sample price
2
price_idx = pd.date_range('2018-01-01 12:00:00', periods=48, freq='H')
3
np.random.seed(42)
4
price = pd.Series(np.random.uniform(size=price_idx.shape), index=price_idx)
5
print(price)
6
print(price.shape)
7

8
2018-01-03 02:00:00    0.684233
9
2018-01-03 03:00:00    0.440152
10
2018-01-03 04:00:00    0.122038
11
2018-01-03 05:00:00    0.495177
12
2018-01-03 06:00:00    0.034389
13
2018-01-03 07:00:00    0.909320
14
2018-01-03 08:00:00    0.258780
15
2018-01-03 09:00:00    0.662522
16
2018-01-03 10:00:00    0.311711
17
2018-01-03 11:00:00    0.520068
18
Freq: H, dtype: float64
19
(48,)

1
# Sessions must be equal - fill missing dates
2
# Fill on first date before 12:00 and on last date after 11:00
3
first_date = price.index[0].date()
4
last_date = price.index[-1].date()+timedelta(days=1)
5
filled_idx = pd.date_range(first_date, last_date, freq='H')
6
filled_price = price.reindex(filled_idx) #等于在原数据基础上扩展新增了部分数据
7
print(filled_price)
8

9
2018-01-01 00:00:00   NaN
10
2018-01-01 01:00:00   NaN
11
2018-01-01 02:00:00   NaN
12
2018-01-01 03:00:00   NaN
13
2018-01-01 04:00:00   NaN
14
                       ..
15
2018-01-03 20:00:00   NaN
16
2018-01-03 21:00:00   NaN
17
2018-01-03 22:00:00   NaN
18
2018-01-03 23:00:00   NaN
19
2018-01-04 00:00:00   NaN
20
Freq: H, Length: 73, dtype: float64 #数据长度从48变成73

1
# Remove dates that are outside of trading sessions
2
session_price_idx = filled_price.between_time('9:00', '17:00', include_end=False).index
3
session_price = filled_price.loc[session_price_idx]
4
print(session_price)

1
# Select first and last ticks of each trading session and split price into ranges between those ticks
2
start_idxs = session_price.index[session_price.index.hour == 9]
3
end_idxs = session_price.index[session_price.index.hour == 16]
4
price_per_session, _ = session_price.vbt(freq='1H').range_split(start_idxs=start_idxs, end_idxs=end_idxs)
5
print(price_per_session)
6

7
split_idx         0         1         2
8
0               NaN  0.139494  0.662522
9
1               NaN  0.292145  0.311711
10
2               NaN  0.366362  0.520068
11
3          0.374540  0.456070       NaN
12
4          0.950714  0.785176       NaN
13
5          0.731994  0.199674       NaN
14
6          0.598658  0.514234       NaN
15
7          0.156019  0.592415       NaN

1
session_price.vbt(freq='1H').range_split(start_idxs=start_idxs, end_idxs=end_idxs,plot=True)

1
# Run your strategy (here using random signals)
2
entries, exits = pd.DataFrame.vbt.signals.generate_random_both(price_per_session.shape, n=2, seed=42)
3
pf = vbt.Portfolio.from_signals(price_per_session, entries, exits, freq='1H')
4
print(pf.total_return())