Re: [R-sig-Geo] spatially clustered factors?

Roger Bivand Mon, 01 Jul 2013 04:13:40 -0700

On Mon, 1 Jul 2013, Chris Mcowen wrote:

Dear list, first let me apologise, i am an ecologist and this is my first
foray into spatial statistics.




I have, for a number of regions, the factor that best explains variation in
the yield (fisheries catch) obtained within that region, the factor has 4
levels - ChlA, SST, Effort or none.

See joincount.test for single comparisons, or joincount.multi for multiplecomparisons by join "colours" in spdep. You'll also need neighbour lists,explained in:


vignette("nb")

in the most recent version of that package.

Hope this helps,

Roger




What i am essentially wanting to know is,  is there a spatial pattern to the
data? (clumped, over-dispersed or random)



For example if region 1 is classed as Effort, is it more or less likely that
the adjoining region (region 2) will be of the same factor, or does it makes
no difference.



I have seen lots of packages that allow this to be done for numerical data
but have not found one for factors.



Please see below for an example of the data. Where latitude is the centroid
of the region and factor is the source of variation.



clusters <- structure(list(Latitude = c(4.8261, 26.127123, 3.409063,
-31.964573,
11.524328, 10.857926, 30.95066, 6.322733, 40.963009, -4.370738,
53.77057, -46.270909, 51.431808, 50.091992, -12.606217, 16.26404,
-27.513126, -40.976427, -40.491914, -35.169487, -16.739053, -22.651113,
30.445027, -27.701939, -17.154608, -10.147356, -29.117245, 57.472215,
45.538569, 65.31731, 75.308501, 54.063299, 45.194528, 41.079154,
-35.169487, 51.431808, 24.647845, 68.201714, 41.262656, 24.976371,
33.251908, 16.347456, 30.95066, -1.408364, 37.054319, 23.843305,
57.400753, 8.712258), Factor = structure(c(1L, 1L, 1L, 1L, 1L,
1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L,
2L, 2L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 4L, 4L,
4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L), .Label = c("ChlA",
"Effort", "SST", "Unidentified"), class = "factor")), .Names = c("Latitude",

"Factor"), class = "data.frame", row.names = c(NA, -48L))





Thanks,



Chris




        [[alternative HTML version deleted]]

_______________________________________________
R-sig-Geo mailing list
[email protected]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo


--
Roger Bivand
Department of Economics, NHH Norwegian School of Economics,
Helleveien 30, N-5045 Bergen, Norway.
voice: +47 55 95 93 55; fax +47 55 95 95 43
e-mail: [email protected]

_______________________________________________
R-sig-Geo mailing list
[email protected]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo

Re: [R-sig-Geo] spatially clustered factors?

Reply via email to