Question

I would like to performs some statistical analysis in a definite zone of a very big table created with hist2d function of R. Is there any elegant way to cut a definite zone of the 2-d histogram and to put it in a table with R? thanx

Answer 1

I’m not entirely clear on what you mean by “cut a definite zone”, but as per the documentation on hist2d, the function returns the counts for each cell in a matrix. So you can easily extract the specific cells you want by subsetting:

y <- rnorm(2000, sd=1)
x <- rnorm(2000, sd=4)
   # separate scales for each axis, this looks circular
tmp <- gplots:::hist2d(x,y)

str(tmp$counts)
dim(tmp$counts)
tmp$counts[1:10,1:10]

So just take the appropriate subset of tmp$counts.