Share:

Data Visualization Workshop

Date:

June 26 to 30. MORNING: 9 to 12h

Instructor

Pere-Pau Vázquez.

Pere-Pau Vázquez is an associate professor in the Computer Science Department at the Universitat Politècnica de Catalunya in Barcelona. He is a member of the Research Center for Visualization, Virtual Reality and Graphics Interaction (ViRVIG). He has been working in the areas of Scientific Visualization and Computer Graphics for more than 20 years. His interests are the visualization of biomedical models, the development of novel visual analysis techniques for large data sets, the interaction of Virtual Reality environments, and the applications of Information Theory to Computer Graphics and Visualization. After graduating in Computer Science (1999), he obtained a PhD in Software (2003) at Universitat Politècnica de Catalunya. He joined Universitat Politècnica de Catalunya in 2002 as a lecturer, and has held an associate professor position since 2006.

Language

English

Description

Visual depictions are fundamental for understanding complex data. Complex, large datasets can often be better understood through visualizations. Nowadays, we deal with many large datasets that are difficult to analyze using other techniques such as statistics or deep learning. Moreover, visualization is also a powerful tool to communicate insights obtained from data analytics.
Data visualization is a process that has many steps: data cleaning, abstraction, design, and implementation.
In this course, we focus on the design and implementation parts by providing theoretical guidance on the visualization process, and the practical knowledge on how to create interactive visualizations using a simple yet powerful Python library called Altair.

Course goals

The goals of the course are threefold:

  • Providing the theoretical background on what visualizations are, how can they be made effective, and the aspects to consider.
  • Get some practice in critically analyzing the effectiveness of visualization designs.
  • Acquiring practical knowledge on the implementation of multi-view, interactive visualizations using a Python library.

All three goals are intertwined: we will describe visualization techniques and the perceptual aspects to consider in order to make them effective, and then we will implement such techniques using Python. The design should be analyzed and refined iteratively to reach a good result in terms of the tasks, the audience, and the data analyzed. Furthermore, we will introduce the concepts of interaction and describe how this can be achieved to facilitate exploratory analysis of data.

At the end of the course, students should be able to reflect on which kinds of visualizations are more useful for certain tasks and data, and implement, analyze, and redesign multiple view interactive visualizations.To gain proficiency in coding python, understand basic types

Course contents

The sessions will mix theoretical concepts with implementations. The theoretical background that will guide the course will consist of the following themes:

  • Basic concepts: definition, pipeline, history...
  • Visualization techniques: common techniques (bar charts, line charts, scatterplots...) and advanced techniques (sankey diagrams, ridgeplots...).
  • The role of perception in visualization (preattentive variables, use of color, color palettes, fonts...).
  • Making visualizations effective (communicating messages, things to avoid, guidelines...). 
  • Advanced concepts (multiple views, layouts, linked views, filtering and aggregation...).

 Prerequisites

The course will have a strong practical component. Therefore, students will be required to have a laptop. The implementation will be carried out using a Python notebook, preferably in Google Colab (a Google account will be needed).
Since the course will be very practical, with many examples suggested by the instructor to be implemented during the sessions, it is suitable that students participate in presence in the sessions.
Minimal programming knowledge is also required. Although Python is used, not extensive knowledge of the language is necessary.

Targeted at

Anybody who is interested in understanding the theoretical concepts behind visualization as a field, and wants to learn how to create interactive visualizations.
Although most of the students are typically master students, people that work with data in any area are also welcome.

Evaluation

The students will be assigned a practical work that will be started during the last sessions of the course and to be finished afterwards.

Software requirements

No particular software is needed. The course can be carried out using Google Colab, so only a browser (maybe preferably Google Chrome) will be necessary.