Personal notes
Note that I make many mistakes in converting numerical values and time series data in the dataframe. Conversion that converts the following data (seconds) to time
> df['Time']
0        30620.5
1        39226.0
2        81318.5
3         3438.0
4        11108.5
          ...   
40916     3544.0
40917    37506.0
40918    23386.0
40919    37884.5
40920    24607.5
Name: Time, Length: 40921, dtype: float64
> df['Time']=pd.to_timedelta(df['Time'], unit='S') + pd.to_datetime('1900-01-01')
> df[Time]
0       1900-01-01 08:30:20.500
1       1900-01-01 10:53:46.000
2       1900-01-01 22:35:18.500
3       1900-01-01 00:57:18.000
4       1900-01-01 03:05:08.500
                  ...          
40916   1900-01-01 00:59:04.000
40917   1900-01-01 10:25:06.000
40918   1900-01-01 06:29:46.000
40919   1900-01-01 10:31:24.500
40920   1900-01-01 06:50:07.500
Name: Time, Length: 40921, dtype: datetime64[ns]
unit str, optional Denotes the unit of the arg for numeric arg. Defaults to "ns".
Possible values
Changed in version 1.1.0: Must not be specified when arg context strings and errors="raise".
Recommended Posts