Date: Sun, 6 Sep 2009 03:57:16 0500
ReplyTo: OR Stats <stats112@GMAIL.COM>
Sender: "SAS(r) Discussion" <SASL@LISTSERV.UGA.EDU>
From: OR Stats <stats112@GMAIL.COM>
Subject: Re: Simulate t variables (gttir)
InReplyTo: <6eca73440909060131m70c3539aw4e2734114265dc77@mail.gmail.com>
ContentType: text/plain; charset=ISO88591
Included among the current unknown is the algorithm SAS uses to generate
random student's t.
data_null: this may help you better understand the purpose of the
questioning..... 1. tinv is not a closed form function and needs to be
approximated 2. student's t sometimes has a second (even a 3rd or 4th)
distnl parameter so that it is not a straight shoot application of SAS's t
r.v. generator 3. as before, implementing fastest algorithm for generating
sufficient sample (is it 10,000 copies that is optimal, or 5K or something
else?) of t makes a difference when we are running thru many scenarios w.
CPU limitations and what not and in general more convincing to audience that
is not just on this listserve.
SAS's documentation on how it does the two would be helpful sometimes.
On Sun, Sep 6, 2009 at 3:31 AM, OR Stats <stats112@gmail.com> wrote:
> Proof is in the pudding, right? W.o. documentation on the algorithm used
> by SAS' tinv, in some situations we cannot accept 'well, i'm sure SAS used
> the fastest algorithm.' It is not convincing for every audience, including
> member of this forum. If you have link to documentation please provide.
> The lack of knowing is fine too. But blind faith emails as below would
> defeat the purpose of a listserve.
>
>
> On Sat, Sep 5, 2009 at 4:47 PM, Dale McLerran <stringplayer_2@yahoo.com>wrote:
>
>> Data _null_,
>>
>> Harder, harder! Let's see those stripes turn crimson.
>> You are not flagellating yourself sufficiently. We
>> need more blood!
>>
>> Oh, this is the wrong forum for this post. Better go
>> now!
>>
>> Dale
>>
>> 
>> Dale McLerran
>> Fred Hutchinson Cancer Research Center
>> mailto: dmclerra@NO_SPAMfhcrc.org
>> Ph: (206) 6672926
>> Fax: (206) 6675977
>> 
>>
>>
>>  On Sat, 9/5/09, Data _null_; <iebupdte@GMAIL.COM> wrote:
>>
>> > From: Data _null_; <iebupdte@GMAIL.COM>
>> > Subject: Re: Simulate t variables (gttir)
>> > To: SASL@LISTSERV.UGA.EDU
>> > Date: Saturday, September 5, 2009, 3:33 AM
>> > Well DATA _NULL_, once again I see
>> > you have restated the obvious. It
>> > seems that the OP and the other knowledgeable helpful
>> > poster know all
>> > about the RAND function.
>> >
>> > On 9/4/09, Data _null_; <iebupdte@gmail.com>
>> > wrote:
>> > > Would RAND work?
>> > >
>> > > RAND Function
>> > > Generates random numbers from a specified
>> > distribution
>> > >
>> > > 230 data _null_;
>> > > 231 do rep=1 to 10;
>> > > 232 x=rand('T',4);
>> > > 233 put x=;;
>> > > 234 end;
>> > > 235 run;
>> > >
>> > > x=0.0484676897
>> > > x=0.3030507911
>> > > x=0.294910323
>> > > x=1.4615056019
>> > > x=0.15208295
>> > > x=0.6815017601
>> > > x=0.18924352
>> > > x=1.1911568
>> > > x=0.252726361
>> > > x=0.1148610137
>> > >
>> > >
>> > > On 9/4/09, OR Stats <stats112@gmail.com>
>> > wrote:
>> > > > Hello Listserve:
>> > > >
>> > > > Has anyone implemented the fast algorithm for
>> > generating deviates from the
>> > > > student's t family in SAS? The author wrote
>> > the GNU library
>> > > >
>> > > > gttir()
>> > > >
>> http://books.google.com/books?id=O0YoPJNWZbcC&pg=PA313&lpg=PA313&dq=GTTIR+MONAHAN&source=bl&ots=qmxzHzN8tz&sig=5JgsqjmH0f3QaufwU65kbjK8i4&hl=en&ei=OJhSsCmFtDfnAec6tiqBQ&sa=X&oi=book_result&ct=result&resnum=3#v=onepage&q=GTTIR%20MONAHAN&f=false
>> > > >
>> > > > The algorithm is suppose to generate the
>> > variables to use for
>> > > > AcceptanceRejection for goodness of fit test for
>> > student's t. My question
>> > > > is, for those of you who have coded this same
>> > algorithm in SAS, how many
>> > > > random # (from 0 to 1) did you use?
>> > > >
>> > >
>> >
>>
>
>
