2 • Why build data apps?

What are data apps?

A data app allows visual exploration of a dataset. A data app can range from being simple (perhaps just drawing a graph) to sophisticated (including interactive plots, user log-ins and profiles, connections to databases, and methods to transform data). A data app is often also a web app, if it is deployed to the internet so that others can use it through a web browser.

In any case, data apps aims can include:

• bridge technical aspects of working with data, making analysis more accessible,
• create visualisations for natural interpretation of data,
• promote exploration, prototyping, and hypothesis development,
• facilitate collaboration and the sharing of visualisations and datasets,
• educate users on what methods are being applied.

Central to all this is the interface, and how it is used as a communicative tool. Examples of use-cases of data-apps include

• dataset explorer
• data visualisation builder and customiser
• method demonstrator
• data science portfolio presenter
• machine learning model host
• real-time dataset status reporter
• data donation portal

Data apps in Python

Until recently, building data apps has required good knowledge of the core web programming languages - namely HTML, CSS and JavaScript. These languages are uncommon in research or education, because they are not designed for general computation or data-processing. As such, a gap has existed between those with general-purpose code skills (for example in Python, R, C++ etc) and those with web skills. To address this, libraries have been developed to bridge this gap between the “back end” (code that handles data management and computation) and the “front end” (the graphical interface (or the code that generates an interface)).

In this course we will build your skills and confidence, by bringing together Python tools for data science, data visualisation and data apps.