profile-methods
Likelihood profiles
Compute likelihood profiles for a fitted model
- Keywords
- methods
Usage
proffun(fitted, which = 1:p, maxsteps = 100,
alpha = 0.01, zmax = sqrt(qchisq(1 - alpha/2, p)),
del = zmax/5, trace = FALSE, skiperrs=TRUE,
std.err,
tol.newmin = 0.001, debug=FALSE,
prof.lower, prof.upper,
skip.hessian = TRUE,
continuation = c("none","naive","linear"),
try_harder=FALSE, …)
# S4 method for mle2
profile(fitted, …)
Arguments
- fitted
A fitted maximum likelihood model of class “mle2”
- which
a numeric or character vector describing which parameters to profile (default is to profile all parameters)
- maxsteps
maximum number of steps to take looking for an upper value of the negative log-likelihood
- alpha
maximum (two-sided) likelihood ratio test confidence level to find
- zmax
maximum value of signed square root of deviance difference to find (default value corresponds to a 2-tailed chi-squared test at level alpha)
- del
step size for profiling
- trace
(logical) produce tracing output?
- skiperrs
(logical) ignore errors produced during profiling?
- std.err
Optional numeric vector of standard errors, for cases when the Hessian is badly behaved. Will be replicated if necessary, and NA values will be replaced by the corresponding values from the fit summary
- tol.newmin
tolerance for diagnosing a new minimum below the minimum deviance estimated in initial fit is found
- debug
(logical) debugging output?
- prof.lower
optional vector of lower bounds for profiles
- prof.upper
optional vector of upper bounds for profiles
- continuation
use continuation method to set starting values?
"none"
sets starting values to best fit;"naive"
sets starting values to those of previous profiling fit;"linear"
(not yet implemented) would use linear extrapolation from the previous two profiling fits- skip.hessian
skip hessian (defunct?)
- try_harder
(logical) ignore
NA
and flat spots in the profile, try to continue anyway?- …
additional arguments (not used)
Details
proffun
is the guts of the profile method, exposed
so that other packages can use it directly.
See the vignette (vignette("mle2",package="bbmle")
)
for more technical details of how profiling is done.