Togaware DATA MINING
Desktop Survival Guide
by Graham Williams
Google

Pie Chart

A http://en.wikipedia.org/wiki/Pie_chartpie chart partitions a circle into proportions related to some data. The R function pie is used to produce a pie chart. A pie chart can be used to again display the proportion of entities spread across some partitioning of the dataset.

Pie charts are a perennial favourite even though common wisdom suggests avoiding them. The human eye is not well suited to differentiating angular variations and a bar chart provides a better alternative. However, many people still enjoy the look of a pie chart

In our example, using the wine dataset, the data is partitioned on categorical variable Type. The default plot produced by pie will produce quite a respectable looking pie chart. We add in to the basic plot the percentage of entities in each category, including this with the labels of the pie chart. This helps in communicating the distribution of the data over Type.

[width=0.7]rplot-wine-pie



load("wine.Rdata")
attach(wine)
percent <- round(summary(Type) * 100 / nrow(wine))
labels <- sprintf("%s (%d%%)", levels(Type), percent)
pie(summary(Type), lab=labels)

http://rattle.togaware.com/code/rplot-wine-pie.R

Copyright © 2004-2006 [email protected]
Support further development through the purchase of the PDF version of the book.
Brought to you by Togaware.