[R] How to adaptively sample a simulated population of individuals

Krishna Pacifici pacificik at warnell.uga.edu
Sat May 31 17:38:30 CEST 2008


Hello,
I am looking for advice on how to adaptively sample from a simulated population.

I have simulated a population of individuals in a matrix and need to do the following:

1) select a random sample from that population
2) apply a function to that sample to determine if I detect any individuals at each of the referenced site locations [i,j] in the matrix

3) then if I detect any individuals after that first function (adaptive condition y[i,j] > 0) I need to sample the 4 adjacent plots in that matrix [i,j+1]; [i,j-1]; [i+1,j]; [i-1,j]

4) apply that function to those adjacent plots to see if I detect any individuals 
5) continue in that fashion to eventually develop a network

I have already simulated the population with spatial dependence (used a gamma-poisson mixture), so that in each [i,j] site there is a number of individuals such that the the adjacent sites (contained in a 4x3 sub-matrix) are distributed via the same lambda from a poisson distribution, but each 4x3 sub-matrix has a unique lambda describing the distribution.  I have then been able to sample from that population using the sampling package and have been able to apply the function to determine if I detect any individuals.  

The real problem is how to adapt the sampling protocol to the adjacent plots in the matrix given I have initially detected an individual.  I am not sure what the best approach is to do this.  I have thought that because I have the population simulated in a matrix I already have spatially referenced information about the population through the [i,j] locations in the matrix, but am not sure if this is enough spatially referenced information and if it is how do I apply the function to the adjacent sites.  An alternative would be to assign spatial coordinates to the sites so that they are referenced or put it into a data.frame that would keep the locations and the numbers of individuals at each of those locations.  I am not sure if this is possible ( I am guessing it is) and if so how to do it.

Any advice on how to do this or where to start would be greatly appreciated.  

Thank you very much,
Krishna 

Krishna Pacifici
Ph.D. Student
Georgia Cooperative Fish and Wildlife Unit
Warnell School of Forestry and Natural Resources
University of Georgia
Athens, GA 30602
(706) 248-9968



More information about the R-help mailing list