Convert Timedelta to Total Seconds

Convert timedelta to total seconds

Use timedelta.total_seconds().

>>> import datetime
>>> datetime.timedelta(seconds=24*60*60).total_seconds()
86400.0

convert timedelta to seconds

You should be able to use the total_seconds method. However, you'd need to access that method via the datetime accessor dt.

>>> df['duration'].dt.total_seconds()

However, if series Duration are strings/object- you should do use pd.to_timedelta

>>> pd.to_timedelta(df['duration']).dt.total_seconds()

Python timedelta seconds vs total_seconds

seconds is the number of seconds within a day, which is in [0, 86399]. total_seconds is the entire timedelta converted to seconds, and can be any value, for example 604800.0 for one week, or 0.1 for 100 milliseconds.

Convert pandas timedelta to be displayed in seconds

Use the total_seconds() method on the dt accessor:

foo['DURATION_NEW'].dt.total_seconds()

Obtaining total seconds from a datetime.time object

You can combine the time with a reference date to get a datetime object. If you then subtract that reference date, you get a timedelta object, from which you can take the total_seconds:

from datetime import datetime, time

t = time(2,0,0)

ts = (datetime.combine(datetime.min, t) - datetime.min).total_seconds()

print(ts)
# 7200.0

---

With pandas, I'd use the string representation of the time object column (Series) and convert it to timedelta datatype - which then allows you to use the dt accessor to get the total seconds:

import pandas as pd

df = pd.DataFrame({'time': [time(2,0,0)]})

df['totalseconds'] = pd.to_timedelta(df['time'].astype(str)).dt.total_seconds()

# df['totalseconds']
# 0    7200.0
# Name: totalseconds, dtype: float64

Getting the total seconds from a time_delta column python

When you wrote ck = pd.to_timedelta(data['col1'], unit='ns') your ck no longer have original col1, as you write "to create another column" I assume you want to add new columns, then you could use below code:

After your data = pd.read_csv(...:

data['td'] = pd.to_timedelta(data['col1'], unit='ns')
data['secs'] = data['td'].dt.seconds

will give your something like:

      col1              td           secs
0   7291413234      00:00:07.291413    7
1   3245345223423   00:54:05.345223   3245

Tested on Jupyter notebook with:

import pandas as pd
import io

temp2=u"""col1
7291413234
3245345223423
"""
data = pd.read_csv(io.StringIO(temp2),sep='\t' , engine='python')
data['td'] = pd.to_timedelta(data['col1'], unit='ns')
data['secs'] = data['td'].dt.seconds
data

Note:
https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.Series.dt.seconds.html#pandas-series-dt-seconds

Number of seconds (>= 0 and less than 1 day) for each element.

ADDED after comment: total_seconds will return float:

data['secs2'] = data['td'].dt.total_seconds()
data
        col1          td            secs    secs2
0   7291413234      00:00:07.291413 7       7.291413
1   3245345223423   00:54:05.345223 3245    3245.345223

How to extract seconds from datetime.timedelta

Just change

date = [(datetime.datetime.strptime(i, format_string)-now) for i in date]

To

date = [(datetime.datetime.strptime(i, format_string)-now).seconds for i in date]date = [(datetime.datetime.strptime(i, format_string)-now) for i in date]

convert timedelta hh:mm to seconds

Just append :00 to the end of the string to convert it in the format hh:mm:ss.

Ex: 08:00 could be rewritten as 08:00:00

Quick way to fix this:

d['A'] = [x + ':00' for x in d['A']]

And then, you can run the following:

secs = (pd.to_timedelta(d['A']).dt.total_seconds())

---

Related Topics