Time Series and Timedelta in Pandas

Introduction

In Pandas, Time series and Timedeltas are two very important data structures to work with time-related data. Time series data holds immense importance across diverse fields, from finance to healthcare. At the same time, the concept of timedelta plays an important role in managing time durations.

In this article, we will be looking into how to use Pandas in order to work with time series data and timedelta objects.

What is the Time Series?

A time series is a sequence of data points that occur in sequential order over a given period of time. The data points in the series could be numeric, categorical, or mixed. Time series data helps us to find a pattern or trends over a period of time.

Time series found its importance in problems such as the price of a stock, the number of website visitors, or the weather conditions.

Creating a Time series index

DatetimeIndex is used to create a timestamp index for time series data. The pd.date_range() function generates a sequence of timestamps based on a specified frequency.

See the below code for example:

Python

Python

import pandas as pd

# Create a timestamp index using date_range

date_rng = pd.date_range(start='2023-01-01', end='2023-01-10', freq='D')

print(date_rng)

You can also try this code with Online Python Compiler

Run Code

Output:

The above code generates a timestamp index from Nov 1, 2023, to Nov 19, 2023, with a frequency of one day ('D'). The resulting index is then printed.

Basic Operations in Time Series

Let us perform some basic DataFrame operations on time series data, such as data selection, filtering, and aggregation.

Before proceeding with the operations, let us first create a time series data by using the following command:

Python

Python

import pandas as pd

# Create a timestamp index

date_rng = pd.date_range(start='2023-01-01', end='2023-01-10', freq='D')


# Create a DataFrame with random data

data = {'value': [10, 15, 20, 25, 30, 35, 40, 45, 50, 55]}

df = pd.DataFrame(data, index=date_rng)

You can also try this code with Online Python Compiler

Run Code

In the above command, pd.date_range() function is used to generate a range of dates. Then we create a DataFrame named df with a column 'value', which contains corresponding random values for each date in the range.

Selecting Data for a Specific Date

In this section, we will show, through the code, how to select data for a specific date. For this purpose, we will set selected_date to '2023-01-05' and use .loc[selected_date] to access the row corresponding to that date.

Python

Python

# Selecting data for a specific date

selected_date = '2023-01-05'

selected_data = df.loc[selected_date]

print("Data for", selected_date, ":\n", selected_data)

You can also try this code with Online Python Compiler

Run Code

Output:

The output displays the data for the selected date.

Filtering Data Within a Date Range

Let us see how to filter data within a specific date range. We will be using slicing with the DataFrame, and we will retrieve rows that fall within the range from '2023-01-03' to '2023-01-08'.

Python

Python

# Filtering data within a date range

date_range_filter = df['2023-01-03':'2023-01-08']

print("\nData within date range:\n", date_range_filter)

You can also try this code with Online Python Compiler

Run Code

Output:

The output shows the data that falls within this date range.

Resampling Data

Let us see the resampling of the data to a lower frequency. Using the below code, we will resample data to weekly frequency by using .resample('W') and will calculate the mean for each week.

Python

Python

# Resampling data to weekly frequency and calculating mean

weekly_resampled = df.resample('W').mean()

print("\nWeekly resampled data:\n", weekly_resampled)

You can also try this code with Online Python Compiler

Run Code

Output:

The output displays the weekly resampled data along with the mean values.

Aggregating Data for the Entire Period

In this section, we will perform aggregation operations on the data for the entire time period. We will calculate the total sum and average value of the 'value' column using .sum() and .mean() functions.

Python

Python

# Aggregating data for the entire period

total_sum = df['value'].sum()

average_value = df['value'].mean()

print("\nTotal sum of values:", total_sum)

print("Average value:", average_value)

You can also try this code with Online Python Compiler

Run Code

Output:

The output displays the total sum and average value of the 'value' column across the entire dataset.

What is Timedelta?

A timedelta is nothing but refers to a duration or can, say, the difference between two times or dates. Timedelta is used to represent a period of time, such as one day, four hours, or five seconds and they can be positive or negative.

Timedeltas are immutable, which means that they cannot be changed once they are created.

Syntax

The syntax for creating pandas.timedelta object is:

class pandas.Timedelta(value=<object object>, unit=None, **kwargs)

It contains three parameters, namely value, unit and kwargs. Let us understand them one by one.

Parameters

Let us look at each parameter:

Value: It can be a number, a string, or another timedelta object, and It is mandatory to pass it
Unit: it represents units of duration and can be days, seconds, hours etc
**kwargs: used to specify additional properties of the timedelta object, such as tzinfo, which is the time zone of the timedelta object

Example

Let us see an example of how to use pandas.timedelta class:

In the code below, we import the panda's library and numpy for necessary operations, then create a timedelta object in which td2 represents 2 days, 5 hours, and 13 minutes.

Python

Python

#importing necessary libraries

import pandas as pd

import numpy as np



td2 = pd.Timedelta(days=2, hours = 5, minutes = 13)

print(td2)

You can also try this code with Online Python Compiler

Run Code

Output:

In the above code, we saw how to initialise the Timedelta object and print the output.

Timedeltas Arithmetic Operations

Arithmetic operations allow us to manipulate time durations, calculate past dates, and compare time spans. Arithmetic operations like addition and subtraction can be done using timedelta and datetime objects.

Let us see an example to understand it better; we will be performing addition and subtraction:

Python

Python

from datetime import datetime, timedelta


# Define a base date

base_date = datetime(2023, 8, 1)



# Define a timedelta

delta = timedelta(days=10)


# Adding timedelta to the base date

new_date = base_date + delta

print("New date:", new_date)


# Subtracting timedelta from the base date

previous_date = base_date - delta

print("Previous date:", previous_date)

You can also try this code with Online Python Compiler

Run Code

Output:

In the above example we define:

A base date is August 1, 2023
A timedelta of 10 days is defined
Adding the timedelta to the base date results in a new date that is 10 days into the future
Subtracting the timedelta from the base date gives a previous date that is 10 days in the past

Timedelta.seconds Property

The Timedelta.seconds property is used to return the total number of seconds within a timedelta object. With the help of this property, we can extract the seconds component from a timedelta duration.

It also does not need any special parameters, and its return type is in the form of seconds.

Consider the example below:

Python

Python

from datetime import timedelta

import pandas as pd


# Creating timedelta objects

delta1 = timedelta(days=5, hours=3, minutes=30)

delta2 = pd.Timedelta('2 days 4 hours 15 minutes 30 seconds')



# Accessing the seconds component using Timedelta.seconds

seconds1 = delta1.seconds

seconds2 = delta2.seconds


print("Seconds in delta1:", seconds1)

print("Seconds in delta2:", seconds2)

You can also try this code with Online Python Compiler

Run Code

In the above example, delta1 represents a timedelta of 5 days and 3 hours, and delta2 represents a timedelta of 2 days, 4 hours, 15 minutes, and 30 seconds.

By using the seconds property, we extract the total number of seconds from each timedelta and print them in the output.

Output:

Frequently Asked Questions

Is Time-based indexing important in time series data?

Yes, Time-based indexing is important in time series data because it allows easy retrieval of data within specific intervals of time and makes analysis effective.

How to create a timedelta object in Pandas?

To create a timedelta object, we can use the timedelta constructor or by parsing strings with a specified time duration format.

How is missing data handled during resampling of time series data?

Pandas provides us with various options to handle missing data during resampling. We have methods such as aggregation functions like mean or forward filling to maintain data integrity.

Conclusion

Congratulations, you did a fantastic job!!. This article is a comprehensive guide to Time Series and Timedelta in Pandas. We have discussed four basic operations of time series and covered timedelta with working examples. At last, some frequently asked questions are discussed.

Here are some more related articles:

Check out The Interview Guide for Product Based Companies and some famous Interview Problems from Top Companies, like Amazon, Adobe, Google, etc., on Coding Ninjas Studio.
Also, check out some of the Guided Paths on topics such as Data Structure and Algorithms, Competitive Programming, Operating Systems, Computer Networks, DBMS, System Design, etc., as well as some Contests, Test Series, Interview Bundles, and some Interview Experiences curated by top Industry Experts only on Coding Ninjas Studio.

We hope you liked this article.
"Have fun coding!”