PCA (Factor Analysis) with R spatial
you can have a look to http://pbil.univ-lyon1.fr/members/dray/articles/SD807.php which presents a spatially-constrained version of multivariate analysis. It is implemented in the multispati function of the ade4 package. Cheers.
Raphael Saldanha wrote:
Hi Brian, A few time ago, I had recommended these references to a new "spatial" R user: You need first take a good look at the spdep package documentation (here <http://cran.r-project.org/web/packages/spdep/index.html> and here <http://cran.r-project.org/web/packages/spdep/spdep.pdf>), and some good tutorials maded by the CISS <http://www.csiss.org/>, like this one<http://geodacenter.asu.edu/system/files/rex1.pdf>. Also, I recommed the Roger Bivand's book Applied Spatial Analysis<http://www.springer.com/public+health/epidemiology/book/978-0-387-78170-9> . A while ago, I had done some spatial analysis with PCA help. First, I had managed the spatial lag of the variables, and then, with detrended variables, started the PCA analysis. Please, let me known more about your research. 2010/4/6 brian cooper <brianadr at optusnet.com.au>
Dear List,
I am just learning R and associated spatial analysis.
I am attempting to compare the impact of location has on a PCA. To do this
I
will be using the same data set to conduct a PCA using SPSS. Then using the
same data set attached to LAT/LONG coordinates (Polygon Centroid) and
polygons, I will attempt to conduct a PCA using geo statistics.
The question I am asking is there any difference between aspatial and
spatial analysis approaches on the same data. The particular approach is
factor analysis.
The file I am using has approximately 640 cases and 23 variables one of
which is the geographical tag (in this case postcode), SPSS format. I have
a
polygon file for the postcodes, which is currently in MapInfo format.
In my preliminary analysis, I established using vertical mapper (a MapInfo
add-on) there was a difference in results.
Now in R spatial, I want to replicate the results. Is it possible to
include
the polygon as a variable in the calculation as opposed to the centroid of
the polygon?
1. How does one store the results especially the factor scores associated
with each row as new columns.
2. Is it better to have the variables attached to the polygon, or merge
files.
3. Which is the most effective modules to use.
Brian Cooper
[[alternative HTML version deleted]]
_______________________________________________ R-sig-Geo mailing list R-sig-Geo at stat.math.ethz.ch https://stat.ethz.ch/mailman/listinfo/r-sig-geo
St?phane DRAY (dray at biomserv.univ-lyon1.fr ) Laboratoire BBE-CNRS-UMR-5558, Univ. C. Bernard - Lyon I 43, Bd du 11 Novembre 1918, 69622 Villeurbanne Cedex, France Tel: 33 4 72 43 27 57 Fax: 33 4 72 43 13 88 http://pbil.univ-lyon1.fr/members/dray/