Prev: Sample size estimation for Prescott's test
Next: Trouble Connecting to SAS from Java using TelnetConnectClient via Spawner
From: Dale McLerran on 10 Apr 2007 15:32
--- Shawn Haskell <shawn.haskell(a)TTU.EDU> wrote:

> On Apr 10, 12:49 am, davidlcass...(a)MSN.COM (David L Cassell) wrote:
> >
> > You know, it's so rare that someone starts out with a scientific
> model
> > that leads to a logical investigation of a ZINB model. Thanks. I
> needed
> > that. :-)
> >
> > The place to look is in the SAS-L archives. Look up Dale
> McLerran's
> > tomes on PROC NLMIXED for ZIP and ZINB models. Oh, and when
> > you do, don't look for his name in the 'author' box, because he
> uses
> > the 'from' name stringplayer2 .
> >
> > HTH,
> > David
> > --
> > David L. Cassell
> > mathematical statistician
> > Design Pathways
> > 3115 NW Norwood Pl.
> > Corvallis OR 97330
> >
> > _________________________________________________________________
> > Exercise your brain! Try
>
Flexicon.http://games.msn.com/en/flexicon/default.htm?icid=flexicon_hmemailtag...-
> Hide quoted text -
> >
> > - Show quoted text -
>
> David, thanks for the tip, I'll check it out.

Just one hint that will help you to more successfully locate my
posts: look for my posts under the pseudonym "stringplayer_2".

Dale

---------------------------------------
Dale McLerran
Fred Hutchinson Cancer Research Center
mailto: dmclerra(a)NO_SPAMfhcrc.org
Ph: (206) 667-2926
Fax: (206) 667-5977
---------------------------------------



____________________________________________________________________________________
The fish are biting.
Get more visitors on your site using Yahoo! Search Marketing.
http://searchmarketing.yahoo.com/arp/sponsoredsearch_v2.php
From: "H. Wang" on 11 Apr 2007 09:49
Just to add one paper -
"Negative Binomial Loglinear mixed models" by Booth et al. in Statistical
modeling 2003. The authors provided SAS code (using proc NLMIXED). One can
also look into MLwiN.


Hongjie
From: Dave Fournier on 11 Apr 2007 10:43
There is also a discussion of the Booth et al. paper and
a comparison of their SAS results with AD Model Builder's
random effects module at

http://www.otter-rsch.com/admbre/examples/nbmm/nbmm.html

as well as free ZINB MM software you can use with R

http://www.otter-rsch.com/admbre/examples/glmmadmb/glmmADMB.html

which was built with AD Model Builder to illustrate the power and
flexibility of the ADMB software
From: Dale McLerran on 11 Apr 2007 20:26
Dave,

I tried your ADMB executable and I also fit the negative binomial
mixed model in NLMIXED. Given my parameterization of the NBMM,
I had no trouble obtaining convergence from starting values which
were all zero for the mean model, 1 for the variance of the random
intercept, 1/e for the variance of the random slope, and 0 for the
covariance between the two random effects. I obtained a value for
-LL=624.43, smaller than the value returned by the ADMB executable
which returned -LL=624.55. I tried other starting values as well
and obtained results that were quite consistent, unlike what Booth
et al. observed. The estimate of the random slope effect and the
estimate of the covariance were both very close to zero, and no
standard error was produced for the covariance. Since the
random slope is close to zero variance, then there can be
no covariance, so a missing value for the standard error of the
covariance is not unreasonable.

I will state that ADMB returned results on my machine in about 45
seconds vs 9 minutes, 11 seconds for NLMIXED. A lot of the time
required by the NLMIXED procedure was spent on the last few
iterations. I think that I could get the time for SAS down to
something not too much greater than what I observed for ADMB.
The ADMB results are for compiled code that may have been optimized
for the problem, so I don't know that time comparisons are completely
meaningful. But, I suspect that ADMB may be faster than the SAS
NLMIXED procedure.

One thing that I am not clear about is your parameterization of
the random effect covariance structure. The file nbmm.par returns
estimates in tmpL of

-0.00729744 -0.763550 -7.92662

What is the interpretation of these parameters? Since they are
all negative, I know that none can be random effect variance
estimates. I suspect that the last two are log(V(int)) and
log(V(slope)), but I really can't tell. That would leave the
first parameter in tmpL to be some function of the covariance,
but what function?


Dale McLerran


--- Dave Fournier <otter(a)OTTER-RSCH.COM> wrote:

> There is also a discussion of the Booth et al. paper and
> a comparison of their SAS results with AD Model Builder's
> random effects module at
>
> http://www.otter-rsch.com/admbre/examples/nbmm/nbmm.html
>
> as well as free ZINB MM software you can use with R
>
> http://www.otter-rsch.com/admbre/examples/glmmadmb/glmmADMB.html
>
> which was built with AD Model Builder to illustrate the power and
> flexibility of the ADMB software
>


---------------------------------------
Dale McLerran
Fred Hutchinson Cancer Research Center
mailto: dmclerra(a)NO_SPAMfhcrc.org
Ph: (206) 667-2926
Fax: (206) 667-5977
---------------------------------------



____________________________________________________________________________________
Now that's room service! Choose from over 150,000 hotels
in 45,000 destinations on Yahoo! Travel to find your fit.
http://farechase.yahoo.com/promo-generic-14795097
From: David L Cassell on 11 Apr 2007 22:21
shawn.haskell(a)TTU.EDU replied:
>
>On Apr 10, 12:49 am, davidlcass...(a)MSN.COM (David L Cassell) wrote:
> > shawn.hask...(a)TTU.EDU wrote:
> >
> > >On Apr 9, 12:53 am, davidlcass...(a)MSN.COM (David L Cassell) wrote:
> > > > o...(a)yahoo.com wrote:
> >
> > > > >Does any one konow any referance on bayesian multilevel
>zero-inflated
> > > > >negative binom?
> >
> > > > >all the best,
> >
> > > > >Oslo
> >
> > > > Are your data really suited to such an analysis?
> >
> > > > Please write back to SAS-L first and explain what your data are, and
> > > > where they come from, and how they were collected, and what you
> > > > want to do with them, and what the project goals are. Then people
> > > > here can give more helpful suggestions.
> >
> > > > HTH,
> > > > David
> > > > --
> > > > David L. Cassell
> > > > mathematical statistician
> > > > Design Pathways
> > > > 3115 NW Norwood Pl.
> > > > Corvallis OR 97330
> >
> > > > _________________________________________________________________
> > > > Download Messenger. Join the i'm Initiative. Help make a difference
> > >today.http://im.live.com/messenger/im/home/?source=TAGHM_APR07
> >
> > >Don't mean to hijack your thread, but I was actually wondering about
> > >this myself the other day. Zero-inflated Poisson exists, does zero-
> > >inflated NegBin? If so, how so in SAS? i have used PROC GENMOD for
> > >negbin GLM with offset before.
> >
> > >In wildlife studies we often examine habitat selection - i.e.,
> > >resource selection functions that are often modeled with logistic
> > >regression as 'used' vs either 'unused' or 'available'. Those last 2
> > >words carry a lot of baggage that we need not get into. Anyhow,
> > >sometimes the study area is broken into small plots using GIS or some
> > >other means. With new satellite and GPS technologies for tracking
> > >animals - you could end up with some used plots having 10, 20, 50, etc
> > >animal locations. A situation where the variance becomes much larger
> > >than the mean b/c you typically have many many more plots with zero
> > >uses - hence, a zero-inflated negative binomial? Also, you can use
> > >mixed models to deal with dependency of observations in order to
> > >separate individualistic effects vs population level effects. thanks.
> >
> > >Shawn H
> >
> > You know, it's so rare that someone starts out with a scientific model
> > that leads to a logical investigation of a ZINB model. Thanks. I
>needed
> > that. :-)
> >
> > The place to look is in the SAS-L archives. Look up Dale McLerran's
> > tomes on PROC NLMIXED for ZIP and ZINB models. Oh, and when
> > you do, don't look for his name in the 'author' box, because he uses
> > the 'from' name stringplayer2 .
> >
> > HTH,
> > David
> > --
> > David L. Cassell
> > mathematical statistician
> > Design Pathways
> > 3115 NW Norwood Pl.
> > Corvallis OR 97330
> >
> > _________________________________________________________________
> > Exercise your brain! Try
>Flexicon.http://games.msn.com/en/flexicon/default.htm?icid=flexicon_hmemailtag...-
>Hide quoted text -
> >
> > - Show quoted text -
>
>David, thanks for the tip, I'll check it out. The situation I gave
>was pretty particular b/c it was on my mind at the time. However, a
>ZINB model may actually be fairly common for a lot of wildife
>surveys. The reason is that many wildlife species are gregarious but
>not terribly common. Say you want to model factors affecting duck use
>of ponds. You visit 30 ponds and find no ducks, but when you do find
>a pond with ducks there might be 50, 500, or 1000. You might
>standardize the model by using pond area as an offset to get density
>but that is another matter. Anyhow, thanks for the help.
>
>Shawn Haskell
>Texas Tech Univ
>PhD candidate, Wildlife Science

What I *meant* but apparently did not convey is that you have thought
out the process and come up with a rational model.. which is more than
a lot of people do, so you deserve some sort of pat on the back.

It was supposed to be a compliment, since a lot of people asking for ZINB
and ZIP models do *not* have fundamental constructs leading logically
to such a model, but have heard of the model and are trying to shoehorn
their data into it whether or not it's a size 6 shoe and a size 12 foot.

David
--
David L. Cassell
mathematical statistician
Design Pathways
3115 NW Norwood Pl.
Corvallis OR 97330

_________________________________________________________________
Can�t afford to quit your job? � Earn your AS, BS, or MS degree online in 1
year.
http://www.classesusa.com/clickcount.cfm?id=866145&goto=http%3A%2F%2Fwww.classesusa.com%2Ffeaturedschools%2Fonlinedegreesmp%2Fform-dyn1.html%3Fsplovr%3D866143
First  |  Prev  |  Next  |  Last
Pages: 1 2 3 4
Prev: Sample size estimation for Prescott's test
Next: Trouble Connecting to SAS from Java using TelnetConnectClient via Spawner