I am having trouble deciding which statistical technique I should use in
S-PLUS.
The context...
I have 23 biome types (classes of vegetation) defined for approximately 7500
point locations around the world. I also have environmental variable data
(i.e. annual mean temperature, January mean temperature, July mean
temperature, annual mean precipitation, January mean precipitation, July
mean precipitation, elevation, continentality) for each of these points, as
well as other global points that have no associated biome data.
The problem...
I want to be able to determine which environmental variables determine biome
type, and the degree to which the environmental variable (or combination of
environmental variables) determine the biome type. It would be ideal to
define a range of environmental variable values and the probability of the
values occurring for each biome type. This way I can interpolate the biome
type for my other point locations that only contain the environmental
variable data.
Is there a procedure in S-PLUS that will help me with this?
The solution ?????
I would appreciate any advice/help.
Thank you,
Erin
|