How to Convert Timedelta to Int in Pandas

This tutorial will discuss converting the timedelta to int using the dt attribute in Pandas.

Convert the timedelta to int Using the dt Attribute in Pandas

To convert the timedelta to an integer value, we can use the pandas library’s dt attribute. The dt attribute allows us to extract components of the timedelta. For example, we can extract the year, month, day, minutes, or seconds using the dt attribute. To do that, we need to write the component’s name after the dt attribute. To show all the components of the timedelta variable, we can use the components attribute. For example, let’s create a time series using the pandas Series attribute and show its components using the components attribute.

import pandas as pd

time_series = pd.Series(pd.timedelta_range(start="1 days", end="10 days", freq="1500T"))
time_series.dt.components

Output:

As you can see, the components attribute showed all the components of the time series. The timedelta_range() attribute is used in the above code to create the time series. We can define the starting and ending point and the frequency at which the time will change inside the timedelta_range() attribute. We can extract any of these components using the name of that component. For example, let’s extract the days component from the above time series. See the code below.

import pandas as pd

time_series = pd.Series(pd.timedelta_range(start="1 days", end="10 days", freq="1500T"))
time_series.dt.days

Output:

0    1
1    2
2    3
3    4
4    5
5    6
6    7
7    8
8    9
dtype: int64

You can extract any component you want from the above time series. We can also convert the timedelta to an integer by dividing it by the timedelta of one day or by extracting its integer part using the astype() attribute. For example, let’s create a timedelta object and convert it into an integer to get the value of the day using NumPy. See the code below.

import numpy as np

x = np.timedelta64(2058311000000000, "ns")
day = x.astype("timedelta64[D]")
days.astype(int)

Output:

23.0

The timedelta is actually in the int64 data type, and we can extract our desired component by converting it into an int using the astype() attribute. We can also convert the timedelta to hours or seconds or any other component using the same method. To do that, we need to change D in the third line of code to h for hours and s for seconds, and so on.