R help - Mar 2013 - funtion equivalent of jitter to move figures on device

hello all,
I'm overlaying numerous scatter plots onto a map (done in PBSmodelling). In
this case I'm placing each plot by setting par(fig) to the centroid of map
polygons. The location/mapping part is not so important. There are cases of
small overlaps for some plots (ie figures) so I'm keen to write or find a
function that moves my small scatter plots so they don't overlap. A little
like jitter, but not random in behaviour, it needs to move away from the plots
it's overlapping.

thanks to all
Michael
_______________________________________________________
Michael Folkes
Salmon Stock Assessment
Canadian Dept. of Fisheries & Oceans     
Pacific Biological Station


	[[alternative HTML version deleted]]

On 2013-03-12 08:04, Folkes, Michael wrote:
> hello all,
> I'm overlaying numerous scatter plots onto a map (done in
PBSmodelling). In this case I'm placing each plot by setting par(fig) to the
centroid of map polygons. The location/mapping part is not so important. There
are cases of small overlaps for some plots (ie figures) so I'm keen to write
or find a function that moves my small scatter plots so they don't overlap.
A little like jitter, but not random in behaviour, it needs to move away from
the plots it's overlapping.
>
> thanks to all
> Michael
> _______________________________________________________
> Michael Folkes
> Salmon Stock Assessment
> Canadian Dept. of Fisheries & Oceans
> Pacific Biological Station
>
I would investigate the subplot() function in Greg Snow's
TeachingDemos package, possibly in conjunction with his
updateusr() function.

For interactive placement of figures, it should be possible to
use locator() to place each plot.

Peter Ehlers

In addition to the other suggestions that you have received you may want to
look at the spread.labs function in the TeachingDemos package and the
spread.labels function in the plotrix package.  These only spread in 1
dimension, but may be able to do what you want.  The thigmaphobe.labels
function in plotrix may be of help as well.  If none of these functions
does what you want then looking at the code may help with some ideas.

Preventing overlap in 2 dimensions automatically while keeping things close
to a certain location is not a simple task.  I have written functions that
do well for a particular dataset, but then fail miserably on a different
one.  If there is only a little overlap and that can be solved by moving
just one subplot then an algorithm that may work is to start with the
smallest polygon and put its plot on the map and store the coordinates that
it covers, then try the next smallest polygon and if it overlaps move it a
little in the direction away from the center of the one it overlaps with
until they don't overlap, keep going in the order smallest to largest
(since larger polygons should have more flexibility in where you can move
things).

Another option that sometimes works (and sometimes fails with strange
results) is to create a function that for a give set of "centers" will
calculate the degree of overlap and the distance from the centers to the
desired location and returns a weighted sum of these values.  You then use
that function with the optim function it can give you location for the
centers of the subplots.  Getting the weighting correct is difficult,
weight the overlap too much and it will spread the plots out way too much,
don't give it enough weight and it will still leave you with some overlap.


On Tue, Mar 12, 2013 at 9:04 AM, Folkes, Michael <
Michael.Folkes@dfo-mpo.gc.ca> wrote:
> hello all,
> I'm overlaying numerous scatter plots onto a map (done in
PBSmodelling).
> In this case I'm placing each plot by setting par(fig) to the centroid
of
> map polygons. The location/mapping part is not so important. There are
> cases of small overlaps for some plots (ie figures) so I'm keen to
write or
> find a function that moves my small scatter plots so they don't
overlap. A
> little like jitter, but not random in behaviour, it needs to move away from
> the plots it's overlapping.
>
> thanks to all
> Michael
> _______________________________________________________
> Michael Folkes
> Salmon Stock Assessment
> Canadian Dept. of Fisheries & Oceans
> Pacific Biological Station
>
>
>         [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>


-- 
Gregory (Greg) L. Snow Ph.D.
538280@gmail.com

	[[alternative HTML version deleted]]