[R] Using the digest and t distribution.

[David Winsemius]

On Jan 28, 2012, at 1:08 PM, Rohit Pandey wrote:

> Hello R community,
>
> I have two questions:
> The first might be one of the silliest ever posted here and I
> apologize if I've missed some thing very obvious. It relates to using
> this digest. When I subscribed to the forum, I had chosen the "digest"
> option that bundles all mails every day into a single digest.
>
> Now, I posted a question a while ago on the logistic regression
> function and got no reply. After a while, I realized that it was
> probably bundled up in one of the digests. So, looked through them and
> found my question in Vol107, issue 16. My question was number 34 in it
> and some one had posted a solution to it which appeared in item number
> 42. However, when I scrolled down to the bottom of the digest, it went
> only up to 24.

Those identifiers ( but no date and  no subject line)  are meaningless  
to those of us who read the postings with a mail-client. Just go to  
the Archives in the proper month and search for your name:

https://stat.ethz.ch/pipermail/r-help/
https://stat.ethz.ch/pipermail/r-help/2012-January/thread.html

> In any case, I've tried searching a a variety of other
> things and just can't find the solution posted to my question in the
> body (I couldn't even thank the solution provider). Can some one
> please tell me where  I can find it and if there is an easy way to
> access particular mails sent to you.
>
> My second question is related to the fitting the t distribution with
> the fitdistr function. I am actually trying to fit it to the standard
> data set in the Ecdat package. Here is my code:
>> library(ecdat)
>> data(SP500)
> #Fitting a normal distribution runs just fine.
>> fitdistr(SP500[[1]],"normal")
>       mean            sd
>  0.0004180994   0.0108610082
> (0.0002058797) (0.0001455789)
> #But when I try the t:
>> fitdistr(SP500[[1]],"t")
>        m              s              df
>  0.0005236326   0.0083937340   9.9999984297
> (0.0001721041) (         NaN) (         NaN)
> There were 17 warnings (use warnings() to see them)
>
> #I get these NaN and warnings.
> Am I doing some thing wrong here?
>
> Also, if some one relies, could you please mark any reply to my
> personal ID as well (rohitpandey576 at gmail.com)?

That is the recommended practice.

And you should stop copying replies to:  r-help-request at r-project.org.  
They just going to create confusing error message sent back to you.  
That address is for administrative requests to the mailserver software.


-- 
David Winsemius, MD
West Hartford, CT