Hmisc Dotplot
function can produce multi-panel dot charts with error bars while the Hmisc dotchart2
function is useful for showing auxiliary data (e.g., sample sizes) in the right margin.
library(Hmisc) Load(hayes) a <- apr pa <- factor(ifelse(a$peds==0, 'Adult', 'Pediatric')) dept <- factor(sample(letters[1:30], nrow(a), replace=TRUE)) rdept <- with(a, reorder(dept, oe_dead, mean)) Dotplot(rdept ~ oe_dead | pa, data=a) with(a, dotchart2(oe_dead, auxdata=total_obs, auxtitle='N')) ## add another variable with different symbol .... add=TRUE, pch=2, ... ## See also ~/doc/teaching/hes/statcomp/Notes/graphscourse.pdf ## ~/doc/Talks/ACDRS/analysis.r in Frank's files