Question

我正试图从一个月中获取最后一天，但现在只返回了一年。我需要获取所有年份的所有记录。有什么建议吗？

import pandas as pd
import numpy as np

df = pd.read_csv('PETR4_BOV_D_cor.csv', engine='c', skiprows=1, parse_dates=['date'], names=['ticker', 'date', 'trades', 'close', 'low', 'high', 'open', 'vol', 'qty', 'avg'])

df.date = pd.to_datetime(df.date)
df = df.set_index('date')

结果

            ticker  trades  close       low         high       open         vol         qty        avg
   date                                 
2015-05-29  PETR4   44895   11.577403   11.577403   11.999936   11.934209   901139500.0 72120400    11.732238
2015-06-01  PETR4   31861   11.614961   11.502286   11.877871   11.671299   489916746.0 39483500    11.650736
2015-06-02  PETR4   47249   12.056274   11.708858   12.159559   11.783975   582467511.0 45754100    11.953363
2015-06-03  PETR4   37454   12.046884   11.943598   12.300404   12.168949   629815443.0 48703400    12.142376
2015-06-05  PETR4   34917   11.793364   11.661910   11.999936   11.812143   452516624.0 36024200    11.794773
2016-12-23  PETR4   23100   13.370821   13.154859   13.474106   13.192418   309168316.0 21776900    13.330539
2016-12-26  PETR4   4840    13.539834   13.398989   13.568003   13.445938   82501537.0  5734300     13.509224
2016-12-27  PETR4   13617   13.530444   13.389600   13.661899   13.614951   215534672.0 14949200    13.537768
2016-12-28  PETR4   20265   13.877860   13.483496   13.906029   13.549223   277762881.0 18979900    13.741335
2016-12-29  PETR4   19721   13.962367   13.633730   13.971756   13.943587   266439891.0 18090600    13.829128
395 rows × 9 columns

分组依据

df.groupby(df.index.month).apply(pd.Series.tail, 1).reset_index(level=0, drop=True)

返回

    ticker  trades  close   low high    open    vol qty avg
date                                    
2016-01-29  PETR4   64685   4.544577    4.244109    4.563356    4.413122    4.398262e+08    93013900    4.439976
2016-02-29  PETR4   36334   4.826265    4.676031    4.910772    4.769928    4.312967e+08    84165000    4.811617
2016-03-31  PETR4   44127   7.840334    7.690100    8.103243    7.849723    5.834259e+08    69529900    7.878831
2016-04-29  PETR4   39767   9.605582    9.399011    9.849713    9.774596    5.482716e+08    53536700    9.615911
2016-05-31  PETR4   56676   7.549255    7.549255    8.046905    7.849723    4.804290e+08    58131400    7.760052
2016-06-30  PETR4   19998   8.845023    8.676010    8.910751    8.845023    4.090867e+08    43553100    8.819483
2016-07-29  PETR4   44681   11.145480   10.901350   11.248766   11.042195   7.142205e+08    60478800    11.088579
2016-08-31  PETR4   45622   12.065663   11.934209   12.413079   12.328573   7.848222e+08    60716500    12.137024
2016-09-30  PETR4   28869   12.741716   12.619651   12.929508   12.704157   5.284275e+08    38771900    12.797209
2016-10-31  PETR4   48694   16.610240   16.535123   17.060942   16.929487   7.480264e+08    42059100    16.699535
2016-11-30  PETR4   57759   15.023394   14.657199   15.201797   14.929498   1.316175e+09    82547300    14.971282
2016-12-29  PETR4   19721   13.962367   13.633730   13.971756   13.943587   2.664399e+08    18090600    13.829128

Answer 1

Tks @ Ch3steR和David。

我更改了类似于David示例的代码，并在新列中将月份和年份分开以对数据进行排序。

import pandas as pd
import numpy as np

df = pd.read_csv('PETR4_BOV_D_cor.csv', engine='c', skiprows=1, parse_dates=['date'], names=['ticker', 'date', 'trades', 'close', 'low', 'high', 'open', 'vol', 'qty', 'avg'])

df['year'] = pd.DatetimeIndex(df.date).year
df['month'] = pd.DatetimeIndex(df.date).month

df.date = pd.to_datetime(df.date)
df = df.set_index('date')

result = df.groupby([df.index.month, df.index.year]).apply(pd.Series.tail, 1).reset_index(level=0, drop=True)

sorted = result.sort_values(['year', 'month'])

结果

        ticker  trades  close   low high    open    vol qty avg year    month
date    date                                            
2015 2015-05-29 PETR4   44895   11.577403   11.577403   11.999936   11.934209   9.011395e+08    72120400    11.732238   2015    5
     2015-06-30 PETR4   41060   11.934209   11.877871   12.197118   12.112611   4.078387e+08    31922400    11.996086   2015    6
     2015-07-31 PETR4   38481   9.859102    9.690089    10.046895   9.840323    3.838792e+08    36465800    9.884548    2015    7
     2015-08-31 PETR4   64015   8.629062    7.999957    8.722958    8.178360    6.765446e+08    75379100    8.427373    2015    8
     2015-09-30 PETR4   77202   6.798086    6.544566    6.816865    6.807475    7.660222e+08    107245600   6.706725    2015    9
    ... ... ... ... ... ... ... ... ... ... ... ... ...
2020 2020-02-28 PETR4   109660  25.340000   24.620000   25.560000   25.160000   2.230362e+09    89095300    25.033400   2020    2
     2020-03-31 PETR4   169315  13.990000   13.600000   14.540000   13.600000   2.180450e+09    155314800   14.038900   2020    3
     2020-04-30 PETR4   80537   18.050000   17.700000   18.420000   17.980000   1.433551e+09    79395500    18.055800   2020    4
     2020-05-29 PETR4   85912   20.340000   19.300000   20.340000   19.550000   2.528730e+09    127224200   19.876200   2020    5
     2020-06-02 PETR4   99499   21.400000   20.600000   21.400000   20.750000   1.592479e+09    76091600    20.928500   2020    6

在熊猫DataFrame列表中查找月份和年份的最大日期

1 个答案: