[R] MLE for lambda of Poisson distribution using fitdistr

[Peter Ehlers]

Ankush Bhargava wrote:
> Thank you everyone for responding.
> 
> David,
> 3.75 in my example was equivalent to the mean of the values, which i thought was too much a coincidence...
> What do you think the significance of "(0.03343)". What is this value?
> 
> Kjetil,
> Are you saying that mean(x) is same as the MLE  for the poisson lambda?

Yes, the mle for lambda is mean(x) [easy to derive from the
likelihood function] and the estimated sd for the mle is
sqrt(mean(x)/length(x)) which is 0.03343 in your case.

Cheers,

  -Peter Ehlers

> 
> Thanks again!
> Ankush
> 
> 
> 
> 
> ________________________________
> From: Peter Ehlers <ehlers at ucalgary.ca>
> To: Kjetil Halvorsen <kjetilbrinchmannhalvorsen at gmail.com>
> Cc: R-help at stat.math.ethz.ch; ankush2kn at yahoo.com
> Sent: Tue, October 27, 2009 10:15:31 AM
> Subject: Re: [R] MLE for lambda of Poisson distribution using fitdistr
> 
> 
> 
> Kjetil Halvorsen wrote:
>> What is wrong with using
>> mean(x)
>> to get the MLE of the poisson lambda?
>>
> and
> 
>   mean(x)/length(x)
> 
> to get its estimated variance.
> 
>   -Peter Ehlers
> 
>> Kjetil
>>
>> On Tue, Oct 27, 2009 at 9:17 AM, David Winsemius <dwinsemius at comcast.net> wrote:
>>> On Oct 26, 2009, at 11:25 PM, ankush2kn at yahoo.com wrote:
>>>
>>>> Hi,
>>>>
>>>> I am using the fitdistr of MASS to get the MLE for the lambda of a Poisson
>>>> distribution.
>>>> When i run the fitdistr command, i get an output that looks like -
>>>>
>>>>    lambda
>>>>    3.750000
>>>>    (0.03343)
>>>>
>>>> Couple of questions -
>>>> 1. is the MLE 0.03343 for the lambda of the given distribution then?
>>>> 2. How would I calculate the variance of the MLE for the lambda?
>>> It would be more typical statistical usage to have output of the form:
>>>  estimate ( se(estimate) ) .... so I was expecting 3.75 to be the estimate.
>>> Looking at the help page and running str(.) on the fitdistr object of the
>>> first example confirms my expectations. Why did you think the help page was
>>> suggesting otherwise?
>>>
>>> --
>>>
>>> David Winsemius, MD
>>> Heritage Laboratories
>>> West Hartford, CT
>>>
>>> ______________________________________________
>>> 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.
>>>
>> ______________________________________________
>> 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.
>