[R] Extract estimate of error variance from glm() object

[Rui Barradas]

Às 23:29 de 24/12/2024, Bert Gunter escreveu:
> ... but do note:
> glm(lot1 ~ log(u), data = clotting, family = gaussian)
> 
> is a plain old *linear model*, which is of course a specific type of
> glm, but not one that requires the machinery of glm() to fit.  That
> is, the above is exactly the same as:
> 
> lm(lot1 ~ log(u), data = clotting)
> 
> and gives exactly the same sigma() !
> 
> (and I would therefore hazard the guess that the poster may
> misunderstand what a glm actually is, though of course I may be wrong
> about this).
> 
> Cheers,
> Bert
> 
> On Tue, Dec 24, 2024 at 5:45 AM Christofer Bogaso
> <bogaso.christofer using gmail.com> wrote:
>>
>> Hi,
>>
>> I have below GLM fit
>>
>> clotting <- data.frame(
>>      u = c(5,10,15,20,30,40,60,80,100),
>>      lot1 = c(118,58,42,35,27,25,21,19,18),
>>      lot2 = c(69,35,26,21,18,16,13,12,12))
>> summary(glm(lot1 ~ log(u), data = clotting, family = gaussian))
>>
>> Is there any direct function to extract estimate of Error standard deviation?
>>
>> ______________________________________________
>> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide https://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
> 
> ______________________________________________
> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide https://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

Hello,

In case of doubt, program a residual standard error function and compare 
the results.



rse <- function(object)  {
   rss <- resid(object)^2
   sum(rss / object[["df.residual"]]) |> sqrt()
}

clotting <- data.frame(
   u = c(5,10,15,20,30,40,60,80,100),
   lot1 = c(118,58,42,35,27,25,21,19,18),
   lot2 = c(69,35,26,21,18,16,13,12,12))

fit1 <- glm(lot1 ~ log(u), data = clotting, family = gaussian)
fit2 <- lm(lot1 ~ log(u), data = clotting)

# all of the results below are identically equal
sigma(fit1)
rse(fit1)

sigma(fit2)
rse(fit2)



Hope this helps,

Rui Barradas


-- 
Este e-mail foi analisado pelo software antivírus AVG para verificar a presença de vírus.
www.avg.com