How to keep BIC, AIC and maxloglikelihood in variables when using the method maxLogL

Question

I am fitting some distribution with the method Maximizing the Likelihood. The method implemented in R "maxLogL" is an amazing tools that works great. Documentation:

• https://cran.r-project.org/web/packages/EstimationTools/vignettes/maxlogL.pdf
• https://www.rdocumentation.org/packages/EstimationTools/versions/1.2.1/topics/maxlogL

AIC and BIC are printed with the summary function but I want to keep those values in variables. Here a code that you can easily reproduce:

library(EstimationTools)

set.seed(10)
z <- rnorm(n = 1000, mean = 0.1, sd = 1)
fit1 <- maxlogL(x = z, dist = 'dnorm', start=c(0, 2), lower = 0, upper = 2)
a <-summary(fit1)

which prints the following:

Optimization routine: nlminb 
Standard Error calculation: Hessian from optim 

       AIC      BIC
  2824.494 2820.494

     Estimate  Std. Error
mean  0.011375     0.0313
sd    0.991346     0.0222

My question is:

1. How to keep AIC and BIC saved in a variable?
2. Is the max value of loglikelihood saved somewhere? I could use reverse engineering with BIC and AIC for getting it, but I would like to avoid that.

Thanks

PD: I particularly want to use the maxLogL function.

Answer 1

I apologize because I have not seen this post before. Since some previous versions, we enabled some generic methods, including those you mentioned. Recently we have released version 2.0.0. Let's take the example before:

library(EstimationTools)

set.seed(10)
z <- rnorm(n = 1000, mean = 0.1, sd = 1)
fit1 <- maxlogL(x = z, dist = 'dnorm', start=c(0, 2), lower = 0, upper = 2)
summary(fit1)

_______________________________________________________________
Optimization routine: nlminb 
Standard Error calculation: Hessian from optim 
_______________________________________________________________
       AIC      BIC
  2824.494 2834.309
_______________________________________________________________
     Estimate  Std. Error Z value Pr(>|z|)    
mean   0.11137    0.03135   3.553 0.000381 ***
sd     0.99135    0.02217  44.722  < 2e-16 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
_______________________________________________________________

For keeping AIC, BIC and Log-likelihood in a variable, just invoke the generic functions:

myAIC <- AIC(fit1)
myBIC <- BIC(fit1)
mylogL <- logLik(fit1)

Then, you can print any of them

>myAIC
[1] 2824.494
> myBIC
[1] 2834.309
> mylogL
'log Lik.' -1410.247 (df=2)

Hope this could be useful. Thanks on your interest in EstimationTools.

Best wishes.