I have following dataframe in pandas
code start_time end_time
0 13:00:09 13:30:09
1 14:23:33 15:23:23
2 11:30:00 12:30:00
start_time and end_time are of type objects.
I want to get difference of these two columns in minutes. I am doing following in pandas
df['time_diff'] = pd.Timestamp(df['start_time']) - pd.Timestamp(df['end_time'])
Convert columns to datetimes by to_datetime or to timedeltas by to_timedelta, subtract by sub, convert output timedeltas to total_seconds and divide 60:
start_time = pd.to_datetime(df['start_time'].astype(str))
end_time = pd.to_datetime(df['end_time'].astype(str))
#another solution
#start_time = pd.to_timedelta(df['start_time'].astype(str))
#end_time = pd.to_timedelta(df['end_time'].astype(str))
df['time_diff'] = start_time.sub(end_time).dt.total_seconds().div(60)
print (df)
code start_time end_time time_diff
0 0 13:00:09 13:30:09 -30.000000
1 1 14:23:33 15:23:23 -59.833333
2 2 11:30:00 12:30:00 -60.000000
If swap end time with start time:
df['time_diff'] = end_time.sub(start_time).dt.total_seconds().div(60)
print (df)
code start_time end_time time_diff
0 0 13:00:09 13:30:09 30.000000
1 1 14:23:33 15:23:23 59.833333
2 2 11:30:00 12:30:00 60.000000
Sample:
import datetime
df = pd.DataFrame({'code': [0, 1, 2],
'start_time': [datetime.time(13, 0, 9), datetime.time(14, 23, 33),
datetime.time(11, 30)],
'end_time': [datetime.time(13, 30, 9), datetime.time(15, 23, 23),
datetime.time(12, 30)]})
print (df)
code start_time end_time
0 0 13:00:09 13:30:09
1 1 14:23:33 15:23:23
2 2 11:30:00 12:30:00
IDE from comment, thanks @Anton vBR:
start_time = pd.to_timedelta(df['start_time'].astype(str)).dt.total_seconds()
end_time = pd.to_timedelta(df['end_time'].astype(str)).dt.total_seconds()
df['time_diff'] = end_time.sub(start_time).div(60)
print (df)
code start_time end_time time_diff
0 0 13:00:09 13:30:09 30.000000
1 1 14:23:33 15:23:23 59.833333
2 2 11:30:00 12:30:00 60.000000
この記事はインターネットから収集されたものであり、転載の際にはソースを示してください。
侵害の場合は、連絡してください[email protected]
コメントを追加