fitNLSbouts: Fit mixture of Poisson Processes to Log Frequency data via...

In diveMove: Dive Analysis and Calibration

Fit mixture of Poisson Processes to Log Frequency data via Non-linear Least Squares regression

Description

Methods for modelling a mixture of 2 or 3 random Poisson processes to histogram-like data of log frequency vs interval mid points. This follows Sibly et al. (1990) method.

Usage

## S4 method for signature 'data.frame'
fitNLSbouts(obj, start, maxiter, ...)

## S4 method for signature 'Bouts'
fitNLSbouts(obj, start, maxiter, ...)

Arguments

obj	Object of class Bouts, or data.frame with named components lnfreq (log frequencies) and corresponding x (mid points of histogram bins).
start, maxiter	Arguments passed to nls.
...	Optional arguments passed to nls.

Value

nls object resulting from fitting this model to data.

Methods (by class)

• data.frame: Fit NLS model on data.frame

• Bouts: Fit NLS model on Bouts object

Author(s)

Sebastian P. Luque [e-mail]

References

Sibly, R.; Nott, H. and Fletcher, D. (1990) Splitting behaviour into bouts Animal Behaviour 39, 63-69.

See Also

fitMLEbouts for a better approach; boutfreqs; boutinit

Examples

## Run example to retrieve random samples for two- and three-process
## Poisson mixtures with known parameters as 'Bouts' objects
## ('xbouts2', and 'xbouts3'), as well as starting values from
## broken-stick model ('startval2' and 'startval3')
utils::example("boutinit", package="diveMove", ask=FALSE)

## 2-process
bout2.fit <- fitNLSbouts(xbouts2, start=startval2, maxiter=500)
summary(bout2.fit)
bec(bout2.fit)

## 3-process
## The problem requires using bound constraints, which is available
## via the 'port' algorithm
l_bnds <- c(100, 1e-3, 100, 1e-3, 100, 1e-6)
u_bnds <- c(5e4, 1, 5e4, 1, 5e4, 1)
bout3.fit <- fitNLSbouts(xbouts3, start=startval3, maxiter=500,
                         lower=l_bnds, upper=u_bnds, algorithm="port")
plotBouts(bout3.fit, xbouts3)

diveMove documentation built on Nov. 10, 2022, 5:11 p.m.