Data Visualisation and EDA

Data, aesthetics, geometries, scales, facets and coordinates as a compositional system.

A decision procedure from question and data type to chart form, and the charts to avoid.

Assumes: The Grammar of Graphics

Histograms and bin-width sensitivity, KDE, box plots, violin plots, ECDFs and Q–Q plots.

Assumes: Descriptive Statistics

Scatter plots, overplotting remedies, smoothers, residual plots and correlation heatmaps.

Assumes: Visualising Distributions

Small multiples, pair plots, parallel coordinates, and projecting high-dimensional data to two.

Assumes: Visualising Relationships

Line charts, seasonal plots, calendar heatmaps, choropleths and projection distortion.

Assumes: Visualising Relationships

Pre-attentive attributes, colour scales for sequential/diverging/categorical data, and accessible palettes.

Assumes: The Grammar of Graphics

Truncated axes, dual axes, area distortion, cherry-picked ranges, and how to audit a chart.

Assumes: Perceptual Principles

A repeatable exploration checklist from first load to modelling hypotheses, applied to a real dataset.

Assumes: Multivariate Visualisation · Missing Values and Outliers