Making comparisons using graphical data analysis

Comparisons may be

• specific

• general s

• at different levels

library(ggplot2)
library(gridExtra)

data(bank, package="gclus")

bank2 <- within(bank, st <- ifelse(Status==0,"genuine","forgery"))

c1 <- ggplot(bank2,aes(x=Diagonal)) +
  geom_histogram(binwidth=0.2) + facet_grid(st~.)

c2 <- ggplot(bank2, aes(x=Right)) + 
  geom_histogram(binwidth=0.1) +facet_grid(st~.)

grid.arrange(c1, c2, ncol=2)

t.test(Right~ st,
       data = bank2,var.equal = T)

## 
##  Two Sample t-test
## 
## data:  Right by st
## t = 10.196, df = 198, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  0.3815181 0.5644819
## sample estimates:
## mean in group forgery mean in group genuine 
##               130.193               129.720

t.test(Diagonal ~ st,
       data = bank2, var.equal = T)

## 
##  Two Sample t-test
## 
## data:  Diagonal by st
## t = -28.915, df = 198, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -2.207971 -1.926029
## sample estimates:
## mean in group forgery mean in group genuine 
##               139.450               141.517

t-tests confirm that the differences in means between the two groups are highly significant.

The graphics show that the distribution patterns differ.