Take the 2-minute tour ×
Geographic Information Systems Stack Exchange is a question and answer site for cartographers, geographers and GIS professionals. It's 100% free, no registration required.

Im working on marking behaviour in group living animals and im interested in how marking behaviour is affected by certain characteristics of the neighbouring groups. I have plotted the territories of each group from 95% density isopleths created using a-LoCoH. The marks of each individual from a group are plotted as a separate points layer. Marking behaviour does not appear to be random, and they do seem to cluster around boundaries with neighbouring groups. What I would like to investigate whether individuals mark more near some groups than near others, for example, do individuals mark less near groups that are closely related? is it possible to show that points are clustering near some features but not near others?

Any help that can be provided would be greatly appreciated,

Dave Humphries

##### update

For demonstration, below is an image of four territories (A,B,C, and D) and the marking locations of two individuals from group A. Groups A and B are closely related. Groups C and D are unrelated to group A.

enter image description here

share|improve this question
1  
Are you able to supply sample data? –  om_henners Jan 3 '13 at 2:12
    
Sorry, i'm new to stack exchange - and GIS for that matter, what is the standard format to supply sample data? ArcMap document? –  Dave Humphries Jan 3 '13 at 2:35
1  
Sample data should be published in a open GIS format readable by many GIS systems. –  markusN Jan 3 '13 at 7:41
    
Some of the files involved are shapefiles and as i understand it there is no way currently of sharing these types of files on SE. I am happy to supply examples via e-mail at: david.humphries@mq.edu.au –  Dave Humphries Jan 6 '13 at 4:21
    
Another question while I'm thinking about this (back off holidays) - are the a-LoCoH boundaries produced from the same marking dataset, or a different dataset altogether? Also, what kind of software are you using (and do you have access to)? –  om_henners Jan 7 '13 at 2:48

1 Answer 1

I'd strongly recommend taking a look at R, in particular the spatstat package from CSRIO. There's an excellent manual/tutorial by Prof. Adrian Baddeley available which will take you through a host of different methods for point pattern analysis. Particulraly I'd take a look at the Part VII: Marked Point Patterns (specifically the Mark correlation function on page 202). Note marks here refer to data being classified by an attribute instead of having a numerical relationship.

There's also a chapter on reading data in (using rgdal, sp etc.) at the beginning of the document. If you're using ArcGIS I'd take a look at the Geospatial Modelling Environment as one way to link between data in R and ArcGIS, otherwise QGIS has manageR which will do the same.

Lastly @whuber (among many others) here on GIS SE has a huge amount of experinece in R, so ask questions as you find them or try and grab him in the chat.

share|improve this answer
    
Thanks very much for your suggestions. I'll try the spatstat package and see how i go. –  Dave Humphries Jan 10 '13 at 7:50

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.