ajrgodfrey/BrailleR

VI() descriptions for geom_histogram() give really inaccurate numbers of bars

boringNectarine0 opened this issue · 11 comments

boringNectarine0 commented

VI() for geom_histogram has really weird ideas about numbers of bars. I expect that when sorting by categories it is including any missing category from each bar as a completely missing bar.

eg.
geom_histogram(aes(x, fill=catsOneToSeven), bins=5)

VI() will say that there are 35 vertical bars if there are 7 cats, regardless of whether the colours are included in the bar, and the fact that the default has them stacked and will not leave gaps for missing fill cats.

ajrgodfrey commented
boringNectarine0 commented

l<-ggplot(iris) +
geom_histogram(aes(y=Sepal.Width, fill=Species),bins = 5)
l
VI(l)
In the VI() description it says "The chart is a bar chart containing 15 vertical bars."

ajrgodfrey commented
boringNectarine0 commented

If there was a bar chart with space for 10 bars, but 2 gaps (or 2 bars of 0 height) would we ideally have that said to be 8 bars?

ajrgodfrey commented
boringNectarine0 commented

Is bar height of 0 something worth having in VI() descriptions? By the way, would you have time to reply to my emails? Thanks.

ajrgodfrey commented
boringNectarine0 commented

Another example here is saying that graph is 8 bars

data <- data.frame(
quarter=c("Q1", "Q1", "Q2", "Q2", "Q3", "Q3", "Q4", "Q4"),
product=c("A", "B", "A", "B", "A", "B", "A", "B"),
profit=c(10, 11, 12, 11, 13, 15, 16, 18)
)
VI(ggplot(data, aes(x = quarter, y = profit)) +
geom_bar(stat = "identity")
)

because there is a small number of bars the points are described so it does make some sense but as there is no sorting by product A/B this is rather odd.

ajrgodfrey commented
boringNectarine0 commented
ajrgodfrey commented