Resist.boot | R Documentation |
This is a 'pseudo' bootstrap procedure that will subsample a specified proportion of the sample locations/individuals, and will refit the MLPE model using the previously optimized resistance surface.
Resist.boot(mod.names,
dist.mat,
n.parameters,
sample.prop,
iters,
obs,
rank.method,
genetic.mat,
keep = NULL,
n.cores = NULL)
mod.names |
A vector of the model names to be assessed |
dist.mat |
A list containing all distance matrices from optimized resistance surfaces |
n.parameters |
A vector the length of mod.names, specifying the number of parameters in each model |
sample.prop |
Proportion of observations to be sampled each iteration (Default = 0.75) |
iters |
Number of bootstrap iterations to be conducted |
obs |
Total number of observations (populations or individuals) in your original analysis |
rank.method |
What metric should be used to rank models during bootstrap analysis? c('AIC', 'AICc', 'R2', 'LL', 'RMSE'). Default = 'AIC' |
genetic.mat |
Genetic distance matrix without row or column names. |
keep |
Optional vector of pairwise observations to keep (1) or omit (0) |
n.cores |
Number of cores to use when running bootstrap in parallel. Default 2 less than total available. |
This is a 'pseudo' bootstrap procedure that subsamples distance and genetic matrices, refits the MLPE model for each surface. AICc is calculated based on the number of parameters specified. Ranking of models during the bootstrap analysis is based on the specified rank.method
, which defaults to 'AICc'. The objective of this procedure is to identify the surface that is top ranked across all bootstrap iterations.
A data frame reporting the average model weight, average rank, number of times a model was the top model in the set, and the frequency a model was best.
Bill Peterman <Peterman.73@osu.edu>
## Not run:
## *** TO BE COMPLETED *** ##
## End (Not run)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.