如何透视timeindex的值

Question

如何通过保持日期和小时数据的原样来透视分钟数据？

输入

,value
2024-07-07 03:00:00,-0.1
2024-07-07 03:15:00,-0.2
2024-07-07 03:30:00,-0.3
2024-07-07 04:00:00,-0.3
2024-07-07 04:15:00,-0.4
2024-07-07 04:30:00,-0.5

输出

,0,15,30,
2024-07-07 03:00:00,-0.1,-0.2,-0.3
2024-07-07 04:00:00,-0.3,-0.4,-0.5

Answer 1

这样如何：

out = (    
    df
    .assign(hour=df.index.floor('h'), minute=df.index.minute)
    .pivot_table(index='hour', columns='minute', values='value')
)

注意:首先要确保你的索引是一个DatetimeIndex。以您的数据为例，以下是构建正确的原始df的快速方法

from io import StringIO

txt = """,value
2024-07-07 03:00:00,-0.1
2024-07-07 03:15:00,-0.2
2024-07-07 03:30:00,-0.3
2024-07-07 04:00:00,-0.3
2024-07-07 04:15:00,-0.4
2024-07-07 04:30:00,-0.5
"""

df = pd.read_csv(StringIO(txt), parse_dates=[0], index_col=0)