COMP3260Data Science Foundations

Overview

A chart is an argument. Done well it makes a pattern immediately visible; done badly it hides, distorts or invents one. This unit teaches the principles of honest visual encoding, then works through the canonical chart types and the situations each is suited to.

You will build each chart type in Matplotlib and Seaborn, and you will learn to spot the misleading conventions — truncated axes, misleading colour scales, inappropriate aggregation — that undermine otherwise sound analysis.

Learning outcomes

By the end of this unit you will be able to

  • Apply the principles of effective visual encoding.
  • Choose the chart type appropriate to a question and data type.
  • Build bar, line, scatter, histogram, box and heatmap charts.
  • Design a coherent dashboard for a given audience.
  • Recognise and avoid misleading visualisation practices.

Lesson sequence

Key concepts

Visual encoding
Mapping a data value to a visual property such as position, length, area or colour.
Chart type
The visual form chosen to represent data, determined by the number and type of variables and the question being asked.
Bin width
The width of the intervals into which continuous data is grouped for a histogram, which strongly affects its apparent shape.
Misleading chart
A visualisation that creates an inaccurate impression through truncated axes, inappropriate aggregation or similar distortions.

Assessment alignment

Directly assessed in Assignment 5

Assignment 5 (Data Visualisation Dashboard) requires you to design and justify a professional dashboard. The design principles and chart choices in this unit are the substance of that assignment.

👤