optimal.params.sloss: Estimation of neutral community parameters using a two-stage...

In RobinHankin/untb: Ecological Drift under the UNTB

Estimation of neutral community parameters using a two-stage maximum-likelihood procedure

Description

Function optimal.params.sloss() returns maximum likelihood estimates of theta and m(k) using numerical optimization.

It differs from untb's optimal.params() function as it applies to a network of smaller community samples k instead of to a single large community sample.

Although there is a single, common theta for all communities, immigration estimates are provided for each local community k, sharing a same biogeographical background.

Usage

optimal.params.sloss(D, nbres = 100, ci = FALSE, cint = c(0.025, 0.975))

Arguments

D	Species counts over a network of community samples (species by sample table)
nbres	Number of resampling rounds for theta estimation
ci	Specifies whether bootstraps confidence intervals should be provided for estimates
cint	Bounds of confidence intervals, if ci = T

Value

theta	Mean theta estimate
I	The vector of estimated immigration numbers I(k)

Output of the bootstrap procedure, if ci = T:

thetaci	Confidence interval for theta
msampleci	Confidence intervals for m(k)
thetasamp	theta estimates provided by the resampling procedure
Iboot	Bootstrapped values of I(k)
mboot	Bootstrapped values of m(k)

Note

The function returns unhelpful output when run with the caruso dataset as in optimal.params.sloss(caruso). The reason for this behaviour is unknown.

Author(s)

Francois Munoz

References

Francois Munoz, Pierre Couteron, B. R. Ramesh, and Rampal S. Etienne 2007. “Estimating parameters of neutral communities: from one single large to several small samples”. Ecology 88(10):2482-2488

See Also

optimal.params, optimal.params.gst

Examples

data(ghats)
optimal.params.sloss(ghats)

RobinHankin/untb documentation built on Dec. 7, 2024, 6:30 p.m.