Compare and rank

Chernoff faces

Encodes several variables using facial features such as eye size and mouth curvature.

Plot
Multivariate comparison
Pattern recognition

What it answers

Which records have visually different profiles?

Shape: Faces with varying features
What it shows: Salient multivariate differences.

Example

The question

Which stores look alike on four measures at once?

Six stores drawn as faces (Chernoff faces)

Each face is one store: face width shows sales, smile satisfaction, eye size staff turnover, eyebrow slant returns. Malmö and Växjö look alike (wide, smiling); Kalmar has the narrowest, saddest face.

Illustrative data. It belongs to no company.

When to use it

Use it when

Exploratory demonstrations with a small number of variables.

Avoid it when

Precise comparisons or decisions vulnerable to facial-expression bias.

How to read it

Use the feature-to-variable key and compare one facial feature at a time.

Note

Features have unequal perceptual prominence. Define their mappings.

The data you need

Layout: Wide observation table
Fields: ObservationID: text; V1...Vp: numeric columns
One row is: One observation with several measured numeric variables. What is the grain?
Rules: Align variables by observation. Decide scaling, missing-data handling and which columns are features.
Optional: Optional Group. PCA, clustering and correlations require calculations before rendering their derived results.
Preparation: Select variables and map each consistently to bounded facial features. Keep the mapping key explicit.
Units: State the unit of every coordinate/variable. A unitless transformed or standardised scale must be identified; no common denominator is assumed.

A wide table: one row per item, with a separate column for each measure or period.

Worked example

Map quality to mouth curvature and speed to eye size. A higher quality score changes the mouth by the stated rule; it does not measure happiness.

ID	V1	V2	V3
A	10	30	5
B	12	20	8
C	18	25	6

All charts in the library