4. Interactive Visualization of Data across Strata

Learning Objectives

By the end of this chapter, you will be able to:

• Create interactivity in scatter plots using altair
• Use zoom in and out, hover and tooltip, and select and highlight on scatter plots
• Create interactive bar plots and heatmaps
• Create dynamic links between different types of plots within a single rich interactive visualization

In this chapter, you will learn to create interactive visualizations for data stratified with respect to any categorical variable.

Introduction

In the previous chapters, we went through a variety of techniques for visualizing data effectively based on the type of features in the dataset and learned how to introduce interactivity in plots using the plotly library. The second section of this book, starting with this chapter, will guide you on building interactive visualizations with Python for a variety of contexts. An observation made in the previous chapter was that when it comes to introducing interactivity in certain types of Python plots, plotly can sometimes be verbose, and may involve a steep learning curve. Therefore, in this chapter, we'll introduce altair, a library designed especially for generating interactive plots. We will demonstrate how to create interactive visualizations with altair for data stratified with respect to any categorical variable. For illustration, we will use a publicly available dataset to generate scatter plots and bar plots with the features in the dataset and add a variety...

Interactive Scatter Plots

As you know by now, scatter plots are one of the most essential types of plots for presenting global patterns within a dataset. Naturally, it is important to know how to introduce interactivity in these plots. We will first look at the zoom and reset actions on plots. Before that, though, let's have a look at the dataset.

We can view the HPI dataset using the following code:

import pandas as pd
#Download the data from Github repo 
hpi_url = "https://raw.githubusercontent.com/TrainingByPackt/Interactive-Data-Visualization-with-Python/master/datasets/hpi_data_countries.tsv"
# Once downloaded, read it into a DataFrame using pandas
hpi_df = pd.read_csv(hpi_url, sep='\t')
hpi_df.head()

The output is as follows:

Figure 4.1: HPI dataset

Note that there are 5 numerical/quantitative features in this dataset: Life Expectancy (years), Wellbeing (0-10), Inequality of outcomes, Ecological Footprint...

Other Interactive Plots in altair

Now that we know how to add interactivity to scatter plots, let's learn how to introduce interactivity to two other important visualization types – bar plots and heatmaps. We also encourage you to read the official documentation and look at the official example gallery at https://altair-viz.github.io/gallery/index.html to explore altair so as to be aware of the wide variety of visualization types possible in it.

Exercise 33: Adding a Zoom-In and Zoom-Out Feature and Calculating the Mean on a Static Bar Plot

In this exercise, first, we will generate a simple (static) bar plot and then explore interactivity such as zooming in and out. Then, we'll use the same bar plot and find out the mean of the Happy Planet Index of each region. We'll use the altair library here and the Happy Planet Index dataset:

1. Import the altair module as alt:

   import altair as alt

2. Read from the dataset:

   hpi_url = "https://raw.githubusercontent...

Summary

In this chapter, we learned how to create visualizations that respond to the selection of specific strata in a dataset. For illustration purposes, we used the Happy Planet Index dataset of 140 countries, creating a variety of plots with stratification based on the different regions to which countries belonged. We generated scatter plots, bar plots, and heatmaps with interactive features such as zooming in and out, tool tipping, the selection of datapoints in a user-specified interval, and the selection of datapoints belonging to specific strata. We also generated more complex visualizations with multiple plots interlinked with each other that dynamically respond to user inputs. In the next chapter, we will learn how to create interactive visualizations of data across time.