1. Data Analysis Using R: 3. Graphical Analyses
Tuan V. Nguyen
Garvan Institute of Medical Research,
Sydney, Australia

2. Overview
Data
Barchart
Historgram
Stripchart
Boxplot
Scatter plot

3. Data
Body composition data measured by dual energy X-ray absorptiometry
43 men and women, aged between 11 and 28
Variable names: id
age
sex
dur
weight
height
lm (lean mass)
pclm (percent lean mass)
fm (fat mass)
pcfm (percent fat mass)
bmc (bone mineral contents)

4. Reading data into R setwd(“c:/works/stats”)
bc <- read.table(“comp.txt”, header=T)
attach(bc)
names(bc)
[1] "id" "age" "sex" "dur" "weight" "height" "lm" "pclm"
[9] "fm" "pcfm" "bmc"

5. View data bc id age sex dur weight height lm pclm fm pcfm bmc
... M M M M

6. Counting: barplot freq <- table(sex) barplot(freq)
barplot(freq, horiz=T, main="Sex distribution")

7. Counting by group : barplot
agegroup <- cut(age, 3)
agesex <- table(sex, agegroup)
barplot(agesex)

8. Counting by group : barplot
agegroup <- cut(age, 3)
agesex <- table(sex, agegroup)
barplot(agesex, xlab="Age group")
barplot(agesex, beside=T, xlab="Age group")

9. Distribution of data: Histogram
par(mfrow=c(2,2))
hist(age)
hist(age, breaks=20)
hist(age, breaks=40)
hist(age, breaks=50)

10. Distribution of data: Histogram
par(mfrow=c(2,2))
hist(age)
hist(weight)
hist(lm)
hist(fm)

11. Distribution of data: plot(density)
hist(lm, main="Distribution of lean mass")
plot(density(lm), main="Distribution of lean mass")

12. Normal distribution? qqnorm
qqnorm(lm)

13. Contiunity of data: stripchart
stripchart(lm, xlab=“Lean mass; kg") ?

14. Summary of continuous data: boxplot
boxplot(fm)
boxplot(lm) LM
Min. 1st Qu. Median Mean 3rd Qu. Max. FM

15. Summary of data by group: boxplot
Lean mass by sex
Fat mass by sex
boxplot(fm ~ sex)
boxplot(lm ~ sex)

16. Analysis of association: scatter plot
plot(lm ~ age)
plot(lm ~ age, pch=16)

17. Analysis of association: scatter plot
line <- lm(lm ~ age)
plot(lm ~ age, pch=16)
abline(line)

18. Analysis of association by group: scatter plot
plot(lm ~ age, pch=ifelse(sex=="M", "M", "F"), xlab="Age", ylab="Kg")

19. Analysis of multiple associations
data <- data.frame(age, weight, lm, fm, bmc)
pairs(data)

20. Analysis of multiple associations – more fancy graph
matrix.cor <- function(x, y, digits=2, prefix="", cex.cor){ usr <- par("usr"); on.exit(par(usr)) par(usr = c(0, 1, 0, 1)) r <- abs(cor(x, y)) txt <- format(c(r, ), digits=digits)[1] txt <- paste(prefix, txt, sep="") if(missing(cex.cor)) cex <- 0.8/strwidth(txt) test <- cor.test(x,y) # borrowed from printCoefmat Signif <- symnum(test$p.value, corr = FALSE, na = FALSE, cutpoints = c(0, 0.001, 0.01, 0.05, 0.1, 1), symbols = c("***", "**", "*", ".", " ")) text(0.5, 0.5, txt, cex = cex * r) text(.8, .8, Signif, cex=cex, col=2)}
pairs(data,lower.panel=panel.smooth, upper.panel=matrix.cor)

21. Results

22. Summary R is a very powerful package for graphical analysis
First step in data analysis: graphical analysis
Look for
Distributions
Differences
Associations