熊猫按条件切成拼块，然后选择

时间：2018-11-03 03:01:18

标签： pandas

我有一个报价数据帧，其中的 delta 列标识了“ true”中的行。我想在“假”低点之间选择“真”中的那些，并在“低”点中找到最低的，而将那些不是。最后删除“ false”中的所有行。

		symbol	open	high	low	close	adjusted	volume	delta
date								
2017-11-06	TOU	23.70	25.09	23.70	25.07	24.7563	999400	False
2017-11-07	TOU	25.10	25.25	24.73	24.77	24.4600	546500	True
2017-11-08	TOU	24.75	25.16	24.41	24.90	24.5884	450000	True
2017-11-09	TOU	25.36	27.26	25.30	26.83	26.4942	2347500	False
2017-11-10	TOU	26.70	27.01	26.45	26.81	26.4745	903600	False
2017-11-13	TOU	26.76	26.85	26.10	26.40	26.0696	733200	False
2017-11-14	TOU	26.30	26.41	25.37	25.48	25.1611	619300	False
2017-11-15	TOU	25.22	25.27	24.72	24.74	24.4304	525800	False
2017-11-16	TOU	24.69	24.90	24.33	24.34	24.0354	516000	True
2017-11-17	TOU	24.67	24.86	23.98	24.00	23.6997	1233100	True
2017-11-20	TOU	24.01	24.03	23.68	23.70	23.4034	977800	True
2017-11-21	TOU	23.86	23.98	23.35	23.46	23.1664	544300	True
...	...	...	...	...	...	...	...	...
2018-09-21	TOU	21.11	21.30	20.91	20.99	20.9900	1235800	True
2018-09-24	TOU	21.19	21.72	21.19	21.66	21.6600	995800	False
2018-09-25	TOU	21.83	21.83	21.45	21.45	21.4500	574100	False
2018-09-26	TOU	21.38	21.65	20.88	20.97	20.9700	791600	True
2018-09-27	TOU	21.36	22.69	21.23	22.67	22.6700	1192500	False
2018-09-28	TOU	22.58	23.27	22.29	22.74	22.7400	1376300	False
2018-10-01	TOU	23.15	23.86	22.75	23.01	23.0100	1137200	False
2018-10-02	TOU	23.05	23.05	22.51	22.59	22.5900	801600	False
2018-10-03	TOU	22.65	23.59	22.43	23.52	23.5200	1391100	False
2018-10-04	TOU	23.35	23.35	22.39	22.47	22.4700	1272900	False
2018-10-05	TOU	22.62	22.66	22.19	22.62	22.6200	668300	False
2018-10-09	TOU	22.70	23.44	22.53	23.41	23.4100	832800	False
2018-10-10	TOU	23.38	23.38	22.27	22.30	22.3000	1435300	False
2018-10-11	TOU	21.84	22.08	21.16	21.28	21.2800	1127700	False
2018-10-12	TOU	21.78	21.80	21.12	21.18	21.1800	887300	True
2018-10-15	TOU	21.32	21.42	20.58	20.68	20.6800	852300	True
2018-10-16	TOU	20.80	20.80	20.34	20.44	20.4400	1115200	True
2018-10-17	TOU	20.38	20.48	20.03	20.09	20.0900	700900	True
2018-10-18	TOU	20.00	20.01	19.32	19.50	19.5000	1188600	True
2018-10-19	TOU	19.59	20.15	19.57	19.94	19.9400	1321600	True
2018-10-22	TOU	19.96	20.08	19.73	19.80	19.8000	828200	True

通过以下代码大致解决了该问题：

tdf = tdf.reset_index()
temp = tdf.iloc[0:0]
final = tdf.iloc[0:0]
for index, row in tdf.iterrows():
    if(row.delta == False):
        temp =  temp[temp.low ==  temp.low.min()]
        final = final.append(temp)
        temp = temp.iloc[0:0]
    elif(row.delta == True):
        temp = temp.append(row)
        if(row.date == tdf.iloc[-1].date):
            temp =  temp[temp.low ==  temp.low.min()]
            final = final.append(temp)
            temp = temp.iloc[0:0]

df final的结果：

enter image description here

如果我错了，请纠正我！

1 个答案:

答案 0 :(得分：0)

这是我要进行的操作：

#creating the dataframe

import pandas as pd

from io import StringIO

s = '''date symbol  open    high    low close   adjusted    volume  delta
2017-11-06  TOU 23.70   25.09   23.70   25.07   24.7563 999400  False
2017-11-07  TOU 25.10   25.25   24.73   24.77   24.4600 546500  True
2017-11-08  TOU 24.75   25.16   24.41   24.90   24.5884 450000  True
2017-11-09  TOU 25.36   27.26   25.30   26.83   26.4942 2347500 False
2017-11-10  TOU 26.70   27.01   26.45   26.81   26.4745 903600  False
2017-11-13  TOU 26.76   26.85   26.10   26.40   26.0696 733200  False
2017-11-14  TOU 26.30   26.41   25.37   25.48   25.1611 619300  False
2017-11-15  TOU 25.22   25.27   24.72   24.74   24.4304 525800  False
2017-11-16  TOU 24.69   24.90   24.33   24.34   24.0354 516000  True
2017-11-17  TOU 24.67   24.86   23.98   24.00   23.6997 1233100 True
2017-11-20  TOU 24.01   24.03   23.68   23.70   23.4034 977800  True
2017-11-21  TOU 23.86   23.98   23.35   23.46   23.1664 544300  True
2018-09-21  TOU 21.11   21.30   20.91   20.99   20.9900 1235800 True
2018-09-24  TOU 21.19   21.72   21.19   21.66   21.6600 995800  False
2018-09-25  TOU 21.83   21.83   21.45   21.45   21.4500 574100  False
2018-09-26  TOU 21.38   21.65   20.88   20.97   20.9700 791600  True
2018-09-27  TOU 21.36   22.69   21.23   22.67   22.6700 1192500 False
2018-09-28  TOU 22.58   23.27   22.29   22.74   22.7400 1376300 False
2018-10-01  TOU 23.15   23.86   22.75   23.01   23.0100 1137200 False
2018-10-02  TOU 23.05   23.05   22.51   22.59   22.5900 801600  False
2018-10-03  TOU 22.65   23.59   22.43   23.52   23.5200 1391100 False
2018-10-04  TOU 23.35   23.35   22.39   22.47   22.4700 1272900 False
2018-10-05  TOU 22.62   22.66   22.19   22.62   22.6200 668300  False
2018-10-09  TOU 22.70   23.44   22.53   23.41   23.4100 832800  False
2018-10-10  TOU 23.38   23.38   22.27   22.30   22.3000 1435300 False
2018-10-11  TOU 21.84   22.08   21.16   21.28   21.2800 1127700 False
2018-10-12  TOU 21.78   21.80   21.12   21.18   21.1800 887300  True
2018-10-15  TOU 21.32   21.42   20.58   20.68   20.6800 852300  True
2018-10-16  TOU 20.80   20.80   20.34   20.44   20.4400 1115200 True
2018-10-17  TOU 20.38   20.48   20.03   20.09   20.0900 700900  True
2018-10-18  TOU 20.00   20.01   19.32   19.50   19.5000 1188600 True
2018-10-19  TOU 19.59   20.15   19.57   19.94   19.9400 1321600 True
2018-10-22  TOU 19.96   20.08   19.73   19.80   19.8000 828200  True'''

df = pd.read_csv(StringIO(s), sep='\t', index_col=0)

限制增量列中第一个False和最后一个True之间的数据，因为我的False必须夹在df1 = df[df[df.delta==False].index.min():df[df.delta==False].index.max()] 值之间。

delta

将df1['deltaInt'] = df1.delta.astype('int') df1.deltaInt = 1 - df1.deltaInt 列转换为整数并翻转其值。其原因将在下面变得很清楚：

df1['cumSum'] = df1.deltaInt.cumsum()

添加累计总和列； 

df1 = df1.iloc[1:,:]

删除第一行； 

df2 = df1.groupby('cumSum').apply(lambda x:x[x.low==x.low.min()])

通过累计进行汇总，并选择低列中的最小值； 

df2 = df2[df2.delta]
print(df2)

仅选择delta为true的行，我们可以得到所需的结果； 

     getPosts(){
    return this.http.get('https://jsonplaceholder.typicode.com/posts').map(res =>res.json());
  }

输出：

enter image description here

当然，您可以删除两个额外的列。希望这会有所帮助，这不是经过优化的代码，而是我想到的第一件事。

相关问题
最新问题