Python Pandas 没有按要求设置日期格式

发布于11月16日

I have a column called Pausetime and another one called Worktime.
Both of them are strings and look like this: "02:28:49".
I want to convert them from strings to date objects; however, keep the format %H%M%S
I am using the following line:

# define the format of the time columns
time_format = "%H:%M:%S"

# convert the time columns
df["Pausetime"] = pd.to_datetime(df["Pausetime"], format=time_format)
df["Worktime"]  = pd.to_datetime(df["Worktime"],  format=time_format)

print(df["Worktime"])

但是，输出不是传递的time_格式

1   1900-01-01 02:28:49
2   1900-01-01 04:47:45
3   1900-01-01 04:49:09
4   1900-01-01 02:46:47
Name: Worktime, dtype: datetime64[ns]

Any Tips on how to achieve this?
I am new to Pandas and kinda lost in this issue.

The reason is I am calculating the time later on in Excel, and apparently, it only works if the cell is date formatted.

推荐答案

如果需要只有时间而没有日期的DateTime对象，这在Python中是不可能的.

您可以将值转换为时间:

df["Pausetime"] = pd.to_datetime(df["Pausetime"], format=time_format).dt.time
df["Worktime"]  = pd.to_datetime(df["Worktime"],  format=time_format).dt.time

print (df['Worktime'])
0    02:28:49
1    04:47:45
2    04:49:09
Name: Worktime, dtype: object

或时间增量，但格式不同:

df["Pausetime"] = pd.to_timedelta(df["Pausetime"])
df["Worktime"]  = pd.to_timedelta(df["Worktime"])

print (df['Worktime'])
0   0 days 02:28:49
1   0 days 04:47:45
2   0 days 04:49:09
Name: Worktime, dtype: timedelta64[ns]