coexist (version 1.0)

fast.flexsim: faster multiple species simulation, but just a bit, not fast actually

Description

similar to flexsim() function

Usage

fast.flexsim(scale = 2, island,dispersalscale = 51, allee = 1, T = 1000, initp,parcombination, spnum = 2, sourcetype = "constant", type, path = NULL)

Arguments

scale

controlling the parameter's low value=parameter original value/scale; while parameter's high value=parameter original value*scale. Specifically used in the spatial types.

island

number of patches in the simulation

dispersalscale

the scale value for reducing the dispersal rates between the neighbouring patches, allowing the simulation slow down or speed up.

allee

allee effect for the species, the minimum viable population in a local patch, default=1, indicating that if the population size in a patch for a species is less than 1, then the species was removed from that patch

T

simulation time/steps

parcombination

the matrix listing out all combinations of parameter setting

initp

initial population size for a species that will be released at the source patch at each time step

spnum

species number in the model

sourcetype

model configuration for the source supply methods, currently there are three types of source species supplying modes: "constant": constant population size of each species will be released at the source patch for each simulation step "flexible": random population size of each species will be released (all species should typically be assigned different values) based on a normal distribution with mean=initp, Variance=initp/2, at the source patch for each simulation step "cochange": random population size (but all species will be assigned a same value) at the source patch for each simulation step

type

a model configuration vector describing the spatial patterns of each of the parameters, for example, a vector like this, c("decrease","decrease","decrease","increase","increase","increase","mosaiclow") can be used as the input. There are 5 simple spatial types currently for the package: "decrease",the parameter will have one-time decreasing transition from source patch to other sink patches in the middle of the patches "increase",the parameter will have one-time increasing transition from source patch to other sink patches in the middle of the patches "constant",the parameter will keep in a constant value across the patches during the simulation "mosaiclow",the parameter will switch from a low value to a high value one-by-one from the source patch to sink patches, so source patch the parameter for the species there will be assigned a low value "mosaichigh",the parameter will switch from a high value to a low value one-by-one from the source patch to sink patches,so source patch the parameter for the species there will be assigned a high value

path

local disk file name or folder to save the simulated outputs

References

Chen YH (2012) coexist: an R package for performing species coexistence modeling and analysis under asymmetric dispersal and fluctuating source-sink dynamics. http://code.google.com/p/coexist.

See Also

flexsim, sim.coarse

Examples

##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.

## The function is currently defined as
function (scale = 2, island,dispersalscale = 51, allee = 1, T = 1000, 
    initp, parcombination, spnum = 2, sourcetype = "constant", 
    type, path = NULL) 
{
    parnum = spnum * 3
    parcomb <- parcombination
    outcome <- list()
    length(outcome) <- dim(parcomb)[1]
    resource <- matrix(0, ncol = island, nrow = spnum)
    grow <- resource
    comp <- resource
    spvector <- resource
    for (i in 1:spnum) {
        resource[i, ] <- parsetting(island, initp, scale, type[i])
    }
    outcomefile <- filename.check(path)
    comnum <- dim(parcomb)[1]
    colnames(parcomb) <- c(paste("r", c(1:spnum), sep = ""), 
        paste("dis", c(1:spnum), sep = ""), paste("com", c(1:spnum), 
            sep = ""))
    dismat <- list()
    length(dismat) <- spnum
    for (each in 1:comnum) {
        typenum <- spnum
        for (i in 1:spnum) {
            grow[i, ] <- parsetting(island, rate = parcomb[each, 
                i], scale, type[typenum + i])
        }
        typenum <- spnum + typenum
        for (i in 1:spnum) {
            dismat[[i]] <- dispvar(island, rate = parcomb[each, 
                i + spnum]/dispersalscale, scale, type[typenum + 
                i])
        }
        typenum <- spnum + typenum
        for (i in 1:spnum) {
            comp[i, ] <- parsetting(island, rate = parcomb[each, 
                i + spnum * 2], scale, type[typenum + i])
        }
        for (i in 1:spnum) {
            spvector[i, ] <- spabundance(island, 1000)
        }
        for (j in 1:T) {
            spvector <- flex.competition(spvector, resource, 
                grow, comp, allee)
            spvector <- flex.dispersal(spvector, dismat, allee, 
                type = sourcetype)
        }
        outcome[[each]] <- spvector
        write.table(outcome[[each]], file = outcomefile, sep = "\t", 
            append = TRUE)
        outcome[[each]] <- list(abund = outcome[[each]], pars = parcomb[each, 
            ])
    }
    return(outcome)
  }