Data Visualization

The intended audience consists of professionals, engineers, and researchers who want to learn the fundamentals and get hands-on experience with

• how to make useful visualizations and dashboards of data,
• how to help yourself and others to understand and analyse that data,
• how to use state-of-the-art tools and techniques.

To follow the course, you will need to bring a laptop. The course requires familiarity with programming (Python). Exercises will be hosted through Google Colab, for which you need a Google account. For Session 4 it is recommended to install Tableau Public on your laptop. Datasets will be provided.

The course consists of three modules of one day each (12-19h). Namely:

• Data storytelling
• Data visualization & interactive graphics
• Dashboards & structured data

Module 1 (5 November): Data storytelling

• Interactive Theory & Practice
  • Exploratory Data Analytics (EDA) with Python/Pandas/Spark and/or Spark SQL
    • Data tools, technologies, industry best practices for data processing
    • Data processing on a realistic & representative dataset
    • Handling missing observations & data imputation
    • Handling outliers & anomalies
    • Generative AI as an aid during this process
  • Basics of Data Storytelling
    • Overview of tools like matplotlib, Seaborn, plotly
    • Selecting the right visualization & structure
    • Gestalt Principles
    • Reporting DOs and DON'Ts
• Creating Your Own Data Story
  • Brainstorming about storyline & your own approach
  • Creating visuals & storyline
  • Presenting your data story
  • Feedback from colleagues

Lecturers: Quinten Rosseel & Michael Schwaenen

Module 2 (12 November): Data visualization, tools & libraries

In these workshop sessions, you will get an introduction to data visualization for data science, analysis, and storytelling. You will get hands-on experience working with popular Python libraries and standard tools on multiple datasets. The module is divided into the following five segments:

• Introduction to fundamentals of Data Visualization
• Introduction to Data Visualization using Matplotlib, Pandas, and NumPy in Python
• Advanced Data Visualization using Seaborn in Python
• Interactive Data Visualization using Plotly in Python
• Introduction to Data Reporting using Tableau
  

As a key takeaway from these sessions, along with the theoretical knowledge of data visualization, you will develop hands-on skills working with practical datasets. The code and the datasets will be provided to you and we will use Google Colab to practice the exercises from the first three segments. For the last segment, we recommend you install Tableau Public.

Lecturers: Katrien Verbert & Aditya Bhattacharya

Module 3 (19 November): Dashboards and structured data

Part 1: Dashboards & linked visualizations

This session gives a hands-on introduction to dashboards and linked graphics, using Plotly Dash:

• Page layout, dash components, custom components
• Recap interactive plots
• Input, output, callbacks
  • Make components respond to each other
• Examples
• Integration of server-side computations and algorithms
• Deployment
  

Part 2: Interactive graphics to explore complex data

This session covers creating visualizations of structured and complex data, and the integration of these into dashboards:

• Visualization of (large) time series and collections of time series
• (Choropleth) Maps and their complexities, terrain visualization, Sankey diagram/flow map
• Visualization using constructed axes: dimensionality reduction/representation learning, visualization of graphs/networks (layout algorithms, chord diagram), image databases

Lecturer: Jefrey Lijffijt