Dear R-help users, I would like to produce a plot of ordered stacked coloured square where each square would represent a level of a categorical variable CAT (aprox. 150 levels), the size of the square would represent the frequency FREQ (count) of that level, the color would represent a proportion of failure FAILURE (0 to 1) and the squares would be organized by size from bottom left to top right. It would be an extra bonus if one could produce one such plot for each level of a higher order categorical variable GROUP with say 6 levels. All in one plot similar to panels in lattice but without the gaps between panels. Any ideas. I played around with mosaicplot and rectangle but without a satisfactory result. example data GROUP CAT FREQ FAILURE 1 a 100 0.05 1 b 20 0.1 And so forth. Thanks a million Osman Sent from my iPad
Stacked squares plots
2 messages · Osman Al-Radi, Jim Lemon
On Tue, 4 Nov 2014 01:54:45 AM Osman Al-Radi wrote:
Dear R-help users, I would like to produce a plot of ordered stacked coloured square
where each
square would represent a level of a categorical variable CAT (aprox.
150
levels), the size of the square would represent the frequency FREQ
(count)
of that level, the color would represent a proportion of failure
FAILURE (0
to 1) and the squares would be organized by size from bottom left
to top
right. It would be an extra bonus if one could produce one such plot for
each level
of a higher order categorical variable GROUP with say 6 levels. All in
one
plot similar to panels in lattice but without the gaps between panels. Any ideas. I played around with mosaicplot and rectangle but
without a
satisfactory result. example data GROUP CAT FREQ FAILURE 1 a 100 0.05 1 b 20 0.1 And so forth.
Hi Osman, The function color2D.matplot (plotrix) has a Hinton diagram option that might do what you want. The squares will be distributed on a plot in the same positions as the matrix that is passed as the "x" argument. The sizes of the squares will be proportional to "x" and the colors can be anything you pass as "cellcolors". Let me know if you have trouble working it out. Jim