Python 熊猫：在熊猫数据框中填写缺失的日期

如何填充“日期”列，以便在检测到日期时将该日期添加到下面的行中，直到它看到新日期开始添加该日期

可复制示例：

输入：

                Date                                           Headline
0   Mar-20-21 04:03AM  Apple CEO Cook, executives on tentative list o...
1             03:43AM  Apple CEO Cook, execs on tentative list of wit...
2   Mar-19-21 10:19PM  Dow Jones Futures: Why This Market Rally Is So...
3             06:13PM  Zuckerberg: Apples Privacy Move Could Spur Mor...
4             05:45PM  Apple (AAPL) Dips More Than Broader Markets: W...
5             04:17PM  Facebook Stock Jumps As Zuckerberg Changes Tun...
6             04:03PM  Best Dow Jones Stocks To Buy And Watch In Marc...
7             01:02PM  The Nasdaq's on the Rise Friday, and These 2 S...

期望输出：

                 Date                                           Headline
0   Mar-20-21 04:03AM  Apple CEO Cook, executives on tentative list o...
1   Mar-20-21 03:43AM  Apple CEO Cook, execs on tentative list of wit...
2   Mar-19-21 10:19PM  Dow Jones Futures: Why This Market Rally Is So...
3   Mar-19-21 06:13PM  Zuckerberg: Apples Privacy Move Could Spur Mor...
4   Mar-19-21 05:45PM  Apple (AAPL) Dips More Than Broader Markets: W...
5   Mar-19-21 04:17PM  Facebook Stock Jumps As Zuckerberg Changes Tun...
6   Mar-19-21 04:03PM  Best Dow Jones Stocks To Buy And Watch In Marc...
7   Mar-19-21 01:02PM  The Nasdaq's on the Rise Friday, and These 2 S...

尝试：

df['Time'] = [x[-7:] for x in df['Date']]
df['Date'] = [x[:-7] for x in df['Date']]
# Some code that fills the date
# Then convert to datetime

在使用

ffill（）

之前，您需要拆分两列以获得正确的时间，并且只填写日期部分。要使用

ffill（）

，您需要将空格替换为

np.nan

。然后将这些列放回一起，并将该操作包装在

pd.to_datetime

中，以获得正确的

dtype

最后，您可以删除时间列

# Imports
import numpy as np
import pandas as pd

# Split the column
df[['Date','Time']] = df['Date'].str.split(' ',expand=True)

# Replace space with nan and use ffill()
df['Date'] = df['Date'].replace(r'^\s*$', np.nan, regex=True).ffill()

# Put the columns back and convert to datetime
df['Date'] =  pd.to_datetime(df['Date'] + ' ' + df['Time'])

# Drop the time column
del(df['Time'])

你会回来的：

df
                 Date                                           Headline
0 2021-03-20 04:03:00  Apple CEO Cook, executives on tentative list o...
1 2021-03-20 03:43:00  Apple CEO Cook, execs on tentative list of wit...
2 2021-03-19 22:19:00  Dow Jones Futures: Why This Market Rally Is So...
3 2021-03-19 18:13:00  Zuckerberg: Apples Privacy Move Could Spur Mor...
4 2021-03-19 17:45:00  Apple (AAPL) Dips More Than Broader Markets: W...
5 2021-03-19 16:17:00  Facebook Stock Jumps As Zuckerberg Changes Tun...
6 2021-03-19 16:03:00  Best Dow Jones Stocks To Buy And Watch In Marc...
7 2021-03-19 13:02:00  The Nasdaq's on the Rise Friday, and These 2 S...

编辑 如果您希望您的“日期”与您期望的结果完全一致，即此格式的“Mar-20-21”，请不要将其包装在

pd.to_datetime（）

中，并将其作为

对象保存：

df['Date'] =  df['Date'] + ' ' + df['Time']

df
                Date                                           Headline
0  Mar-20-21 04:03AM  Apple CEO Cook, executives on tentative list o...
1  Mar-20-21 03:43AM  Apple CEO Cook, execs on tentative list of wit...
2  Mar-19-21 10:19PM  Dow Jones Futures: Why This Market Rally Is So...
3  Mar-19-21 06:13PM  Zuckerberg: Apples Privacy Move Could Spur Mor...
4  Mar-19-21 05:45PM  Apple (AAPL) Dips More Than Broader Markets: W...
5  Mar-19-21 04:17PM  Facebook Stock Jumps As Zuckerberg Changes Tun...
6  Mar-19-21 04:03PM  Best Dow Jones Stocks To Buy And Watch In Marc...
7  Mar-19-21 01:02PM  The Nasdaq's on the Rise Friday, and These 2 S...

解决方案在没有以下行的情况下工作：
df[['Date'，'Time']=df['Date'].str.split（''，expand=True）
。这句台词让我觉得纳特