Pandas data frame with time column-CodePudding

I have df with column with time like this:

time
12:30
12:15
12:00

I need to have it in float type like this:

time
12.5
12.25
12.00

I have tried:

df.time = df.time.replace(":", ".", regex=True).astype(float)

but I have unsatisfactory result, like this:

time
12.3
12.15
12.00

Any idea please?

CodePudding user response：

One solution is to use pandas.to_timedelta on a modified version of the input string (to match the expected format), then convert to total_seconds and divide by 60:

df['time_float'] = pd.to_timedelta('00:' df['time']).dt.total_seconds()/60

output:

    time  time_float
0  12:30       12.50
1  12:15       12.25
2  12:00       12.00

Alternative using str.replace as you originally intended:

df['time'] = (df['time']
              .str.replace('(:)(\d )',
                           lambda x: f'.{100*int(x.group(2))//60}',
                           regex=True)
              .astype(float)
              )

CodePudding user response：

Use to_timedelta with append :00 by Series.add, then convert to seconds by Series.dt.total_seconds and last divide by 60:

df['time'] = pd.to_timedelta(df['time'].add(':00')).dt.total_seconds().div(3600)
print (df)
0  12.50
1  12.25
2  12.00