根据另一个

Question

我有一个eurusd 5分钟数据的数据框，我有另一个非农场工资单数据框。我想创建一个新的数据框，其中包含NFP日期的所有eurusd 5-min数据行。见下文：

#NFP Days (Dataframe #1)
nfp_dates = forex_calendar.loc[forex_calendar['Event'] == '(United States) Nonfarm Payrolls']

    ＃将csv文件导入Pandas数据帧并转换为Pandas datetime并设置为index（Dataframe＃2）

eurusd_ask = pd.read_csv('EURUSD_Candlestick_5_m_ASK_01.01.2012-05.08.2017.csv')
eurusd_ask.index = pd.to_datetime(eurusd_ask.pop('Gmt time'))

所需的数据帧（Dataframe＃3） 我想要一个名为＆＃34; eurusd_ask_nfp＆＃34;的新数据帧。在NFP日期（即2012-01-06,2012-02-03,2012-03-09等），这将包含来自eurusd_ask的所有5分钟数据（整天）行

Open    High    Low Close   Volume
Gmt time                    
2012-01-06 00:00:00 1.23627 1.23627 1.23532 1.23553 860.47
2012-01-06 00:05:00 1.23553 1.23553 1.23509 1.23532 698.33
2012-01-06 00:10:00 1.23533 1.23535 1.23518 1.23533 338.22
2012-01-06 00:15:00 1.23533 1.23535 1.23511 1.23518 480.64
2012-01-06 00:20:00 1.23518 1.23525 1.23493 1.23508 551.73
2012-01-06 00:25:00 1.23509 1.23537 1.23501 1.23512 349.78
2012-01-06 00:30:00 1.23510 1.23520 1.23471 1.23481 659.66
... ... ... ... ... ...                 
2012-02-03 00:00:00 1.33271 1.33271 1.33210 1.33225 870.59
2012-02-03 00:05:00 1.33225 1.33248 1.33212 1.33227 943.04
2012-02-03 00:10:00 1.33226 1.33226 1.33189 1.33212 659.45
2012-02-03 00:15:00 1.33213 1.33272 1.33212 1.33268 788.32
2012-02-03 00:20:00 1.33265 1.33333 1.33264 1.33315 953.20
2012-02-03 00:25:00 1.33315 1.33322 1.33282 1.33286 678.37
2012-02-03 00:30:00 1.33286 1.33291 1.33260 1.33260 442.39
2012-02-03 00:35:00 1.33261 1.33293 1.33250 1.33268 621.53
2012-02-03 00:40:00 1.33268 1.33292 1.33257 1.33288 511.00
2012-02-03 00:45:00 1.33288 1.33300 1.33240 1.33240 554.76
2012-02-03 00:50:00 1.33240 1.33240 1.33191 1.33225 839.80
... ... ... ... ... ... 
2012-03-09 00:00:00 1.25739 1.25748 1.25710 1.25711 439.46
2012-03-09 00:05:00 1.25712 1.25732 1.25709 1.25715 388.29
2012-03-09 00:10:00 1.25715 1.25733 1.25711 1.25726 291.65
2012-03-09 00:15:00 1.25729 1.25731 1.25685 1.25692 626.85
2012-03-09 00:20:00 1.25692 1.25696 1.25662 1.25663 394.68
2012-03-09 00:25:00 1.25662 1.25677 1.25648 1.25652 302.73
2012-03-09 00:30:00 1.25652 1.25665 1.25645 1.25646 457.76
2012-03-09 00:35:00 1.25646 1.25659 1.25629 1.25633 423.00
2012-03-09 00:40:00 1.25633 1.25649 1.25632 1.25641 173.08
2012-03-09 00:45:00 1.25642 1.25671 1.25630 1.25665 263.99
... ... ... ... ... ... 
etc., etc.

Answer 1

对此进行了尝试，但您应该可以使用Index.isin查询日期匹配的所有记录。

idx = nfp_dates.index
eurusd_ask_nfp = eurusd_ask[eurusd_ask.index.isin(idx)]

Answer 2

我能够解决这个问题，以防万一有人想要答案，你走了：

nfp_dates = forex_calendar.loc[forex_calendar['Event'] == '(United States) Nonfarm Payrolls']
nfp_dates = nfp_dates.reset_index()
nfp_dates = nfp_dates['Date'].dt.date
nfp_5min = eurusd_ask[np.isin(eurusd_ask.index.date, nfp_dates)]

这返回了我想要的内容

Gmt time            Open    High    Low     Close   Volume              
2012-06-01 00:00:00 1.27925 1.27977 1.27925 1.27965 919.94
2012-06-01 00:05:00 1.27966 1.27971 1.27939 1.27958 668.37
2012-06-01 00:10:00 1.27958 1.27984 1.27929 1.27930 732.95
2012-06-01 00:15:00 1.27930 1.27956 1.27913 1.27921 732.04
2012-06-01 00:20:00 1.27922 1.27960 1.27918 1.27942 552.55
2012-06-01 00:25:00 1.27940 1.27940 1.27920 1.27928 384.76
2012-06-01 00:30:00 1.27929 1.27983 1.27929 1.27957 682.91
2012-06-01 00:35:00 1.27957 1.27963 1.27921 1.27927 572.12
2012-06-01 00:40:00 1.27928 1.27938 1.27889 1.27901 794.94
2012-06-01 00:45:00 1.27901 1.27901 1.27878 1.27887 535.03