[R] Log likelihood of Gamma distributions

[Xiaohui Chen]

By the scale of log-likelihood, I did not mean the scale parameter of
the gamma density...

Generally, as you get more and more data, the log-likelihood will get
more and more negative. Hence, what I mean by scale is how negative of
the values of loglik.

So the 10 values returned from your dgamma are the log-densities
evaluated for your data points, respectively.

The loglik for your samples is just the sum of those from all data
points, under the independency assumption.

X

Edward Wijaya 写道:
> Dear Xiaohui,
>
> Thanks.
>
>   
>> The scale of log-likelihood depends on the number of your data samples
>>     
> Can you explain what do you mean by this?
>
> For example if I have 10 data points. Should I use "scale=10" ?
> And how about "shape" parameters. What's the rule to choose its value?
>
> Hope to hear from you again.
>
> Regards,
> Edward
>
>
>
>   
>> Edward Wijaya 写道:
>>     
>>> Dear all,
>>>
>>> How can I compute the log likelihood of a gamma
>>> distributions of a vector.
>>>
>>> I tried the following. But it doesn't seem to work:
>>>
>>> samples<-c(6.1, 2.2, 14.9, 9.9, 24.6, 13.2)
>>> llgm <- dgamma(samples, scale=1, shape=2, log = TRUE)
>>>
>>> It gives
>>>
>>> [1]  -4.291711  -1.411543 -12.198639  -7.607465 -21.397254 -10.619783
>>>
>>> I expect it only returns "one" value instead of vector.
>>> What's wrong with my command above?
>>>
>>> - Edward
>>>
>>> ______________________________________________
>>> R-help at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-help
>>> PLEASE do read the posting guide
>>> http://www.R-project.org/posting-guide.html
>>> and provide commented, minimal, self-contained, reproducible code.
>>>
>>>
>>>       
>>     
>
>