Package to use for AICc (Akaike Information Criterion corrected) in r

Question

Does anyone know what package to use for AICc (Akaike Information Criterion corrected) in r? I am currently using the package 'MASS' and function 'step' to find the best AIC, but I would like to find the best AICc because my sample size is small.

Thank you in advance.

score 2 · Accepted Answer · answered Mar 06 '20 at 04:33

2

Package MuMIn wil calculate it with the function AICc.

https://www.rdocumentation.org/packages/MuMIn/versions/1.43.15/topics/AICc

https://cran.r-project.org/web/packages/MuMIn/index.html

answered Mar 06 '20 at 04:33

Scransom

3,175
3
31
51

Unfortunately, MuMIn has a hard-coded limit of 30 variables (https://stackoverflow.com/questions/59114813/error-in-mumin-pdredge-when-using-more-than-30-predictor-variables). Also, it doesn't let you do bi-directional stepwise selection-- the starting model is also the maximum model. – bokov Oct 01 '21 at 22:30

Gerhard · Answer 2 · 2020-03-06T08:38:19.767

-1

It is build into the R core stats package. No extra packages need to be installed.

https://www.rdocumentation.org/packages/stats/versions/3.6.2/topics/AIC

edited Mar 06 '20 at 08:38

answered Mar 06 '20 at 08:33

Gerhard

6,850
8
51
81

That's AIC not AICc, as the question is specifically asking for – Scransom Mar 09 '20 at 00:27
@Umaomamaomao Sorry I see I missed the *corrected* part – Gerhard Mar 09 '20 at 07:35

Package to use for AICc (Akaike Information Criterion corrected) in r

2 Answers2