This is a helper function for creating a model comparison
<data.frame>
primarily for use in the superspreading vignettes. It
is designed specifically for handling fitdistrplus::fitdist()
output and
not a generalised function. See bbmle::ICtab()
for a more general use
function to create information criteria tables.
Usage
ic_tbl(..., sort_by = c("AIC", "BIC", "none"))
Arguments
- ...
dots One or more model fit results from
fitdistrplus::fitdist()
.- sort_by
A
character
string specifying which information criterion to order the table by, either"AIC"
(default),"BIC"
, or"none"
(i.e. no ordering).
Examples
if (requireNamespace("fitdistrplus", quietly = TRUE)) {
cases <- rnbinom(n = 100, mu = 5, size = 0.7)
pois_fit <- fitdistrplus::fitdist(data = cases, distr = "pois")
geom_fit <- fitdistrplus::fitdist(data = cases, distr = "geom")
nbinom_fit <- fitdistrplus::fitdist(data = cases, distr = "nbinom")
ic_tbl(pois_fit, geom_fit, nbinom_fit)
}
#> distribution AIC DeltaAIC wAIC BIC DeltaBIC
#> 1 geom 546.6447 0.000000 6.265959e-01 549.2499 0.000000
#> 2 nbinom 547.6800 1.035281 3.734041e-01 552.8903 3.640451
#> 3 pois 849.8365 303.191794 9.114358e-67 852.4417 303.191794
#> wBIC
#> 1 8.605932e-01
#> 2 1.394068e-01
#> 3 1.251804e-66