Elizabeth E. Holmes. 2004. Beyond theory to application and evaluation: diffusion approximations for population viability analysis. Ecological Applications 14:1272–1292.
Matlab code to run a stochastic matrix model and generate diagnostic plots of parameterization and risk metric performance.
Ecological Archives A014-023-S2.
Elizabeth E. Holmes
National Marine Fisheries Service
Northwest Fisheries Science Center
2725 Montlake Blvd. East
Seattle, WA 98112-2097
U.S.A.
Email: eli.holmes@noaa.gov
Matlab code:
gensimests.m
marshmat.m
paramperf.m
SimpleRun.mAll files are in ASCII text.
The file SimpleRun.m calls marshmat.m to specify the stochastic model. It then calls gensimests.m to make 1000 simulated time series, estimate the diffusion approximation parameters using the ML, running sum, Heyde-Cohen, Kalman and slope methods, and saves the results to a file. Finally, SimpleRun.m calls paramperf.m which makes diagnostic plots of the different parameterization methods.
The file marshmat.m specifies the matrix model, the level of stochasticity for each matrix element, and what segment of the population is censused.
The file gensimests.m runs the model specified in marshmat.m to create simulated time series. To each time series, it adds either none or one of the three levels of sampling error. From this time series, it then estimates diffusion approximation parameters via ML, running sum, Heyde-Cohen, Kalman or slope methods. It saves the results in a data file.
The file paramperf.m takes the data file created by gensimests.m and makes diagnostic plots of the percentage error in estimation of m, s2, l, and the probability of 90% decline.