Hi all,
I have a huge dataset of values and I have precalculated outside of R the
frequency/count distribution (as for example counts returned by hist
function or the output of table function). For example,
>x<-hist(c(1,2,2,2,1,3), breaks=0:3, plot=F)
> x
$breaks
[1] 0 1 2 3
$counts
[1] 2 3 1
Is there any function to calculate boxplot based on the counts?
I guess I could "expand" the count vector to a values set and run
boxplot
but the problem is that I have a lot of data (300,000^2). Another way could
be to calculate median and quartiles myself and custom draw the boxplot but
seems complicated.
Do you happen to know some more automated way, maybe a function in R?
Many thanks for the help.
Best regards,
Ioannis
[[alternative HTML version deleted]]