Um outro exemplo, dessa vez usando sciplot::bargraph.CI...

### <code r>
dURL <- 'http://www.ualberta.ca/~ahamann/teaching/renr480/Lab5d.zip'
download.file(dURL, basename(dURL), mode='wb')
unzip(basename(dURL),'lentils.csv')
dat <- read.table('lentils.csv', head=T, sep=',')
attach(dat)

require(sciplot)
bargraph.CI(x.factor=FARM, group=VARIETY, response=YIELD, legend=T, 
 xlab="Location", ylab="Yield (kg/ha)", ylim=c(0,800), 
 cex.names=1, cex.lab=1, 
 x.leg=6, y.leg=700, cex.leg=1, leg.lab=c("Var A","Var B","Var C"), 
 col=grey.colors(3), 
 uc=T, lc=T, err.width=0.1, err.col="black", err.lty=1) 
### </code>

Fonte/Source: http://www.ualberta.ca/~ahamann/teaching/renr480/Lab5D.pdf

Éder Comunello <comunello.eder@gmail.com>
Dourados, MS - [22 16.5'S, 54 49'W]