<p>正如@anky所建议的,您可以在df上设置日期,因为这是一个较小的数据帧。然后合并数据</p>
<pre><code>import pandas as pd
df1 = pd.DataFrame({'date':['2021-01-08 07:52:18','2021-01-08 08:53:34',
'2021-01-09 07:56:54','2021-01-09 09:52:17',
'2021-01-12 07:55:58'],
'count':[1,10,12,13,5]})
df2 = pd.DataFrame({'date':pd.date_range('2021-01-08 07:52:00',periods=9000,freq='T'),
'count':[0]*9000})
print (df2)
df1['date'] = pd.to_datetime(df1['date'])
df1['date_str'] = df1['date'].dt.floor('T')
df2 = df2.merge(df1[['date_str','count']],left_on='date',right_on='date_str', how='left')
df2.drop(columns=['date_str','count_x'],inplace=True)
df2.rename(columns={'count_y':'count'},inplace=True)
print (df2)
</code></pre>
<p>其输出将为:</p>
<p>df1:</p>
<pre><code> date count date_str
0 2021-01-08 07:52:18 1 2021-01-08 07:52:00
1 2021-01-08 08:53:34 10 2021-01-08 08:53:00
2 2021-01-09 07:56:54 12 2021-01-09 07:56:00
3 2021-01-09 09:52:17 13 2021-01-09 09:52:00
4 2021-01-12 07:55:58 5 2021-01-12 07:55:00
</code></pre>
<p>df2:与df1合并后</p>
<pre><code> date count
0 2021-01-08 07:52:00 1.0
1 2021-01-08 07:53:00 NaN
2 2021-01-08 07:54:00 NaN
3 2021-01-08 07:55:00 NaN
4 2021-01-08 07:56:00 NaN
... ... ...
8995 2021-01-14 13:47:00 NaN
8996 2021-01-14 13:48:00 NaN
8997 2021-01-14 13:49:00 NaN
8998 2021-01-14 13:50:00 NaN
8999 2021-01-14 13:51:00 NaN
</code></pre>