R: Compute table of information criteria and auxiliary info
ICtab
R Documentation
Compute table of information criteria and auxiliary info
Description
Computes information criteria for a series of models, optionally
giving information about weights, differences between ICs, etc.
Usage
ICtab(..., type=c("AIC","BIC","AICc","qAIC","qAICc"),
weights = FALSE, delta = TRUE, base = FALSE,
logLik=FALSE, sort = TRUE,
nobs=NULL, dispersion = 1, mnames, k = 2)
AICtab(...,mnames)
BICtab(...,mnames)
AICctab(...,mnames)
## S3 method for class 'ICtab'
print(x,...,min.weight)
Arguments
...
a list of (logLik or?) mle objects; in the case of
AICtab etc., could also include other arguments to ICtab
type
specify information criterion to use
base
(logical) include base IC (and log-likelihood) values?
weights
(logical) compute IC weights?
logLik
(logical) include log-likelihoods in the table?
delta
(logical) compute differences among ICs (and log-likelihoods)?
sort
(logical) sort ICs in increasing order?
nobs
(integer) number of observations: required for
type="BIC" or type="AICc" unless objects have
a nobs method
dispersion
overdispersion estimate, for computing qAIC:
required for type="qAIC" or type="qAICc" unless
objects have a "dispersion" attribute
mnames
names for table rows: defaults to names of objects passed
k
penalty term (largely unused: left at default of 2)
x
an ICtab object
min.weight
minimum weight for exact reporting (smaller values
will be reported as "<[min.weight]")
Value
A data frame containing:
IC
information criterion
df
degrees of freedom/number of parameters
dIC
difference in IC from minimum-IC model
weights
exp(-dIC/2)/sum(exp(-dIC/2))
Note
(1) The print method uses sensible defaults; all ICs are rounded
to the nearest 0.1, and IC weights are printed using
format.pval to print an inequality for
values <0.001. (2) The computation of degrees of freedom/number of
parameters (e.g., whether
variance parameters are included in the total) varies enormously
between packages. As long as the df computations
for a given set of models is consistent, differences
don't matter, but one needs to be careful with log likelihoods
and models taken from different packages. If necessary
one can change the degrees of freedom manually by
saying attr(obj,"df") <- df.new, where df.new
is the desired number of parameters.
(3) Defaults have changed to sort=TRUE, base=FALSE,
delta=TRUE, to match my conviction that it rarely makes
sense to report the overall values of information criteria