[Lme4-authors] LME4 and Lmer model

The BY statement in SAS has nothing to do with BLUP, it simply runs an
analysis separately for each group defined by BY. The empirical BLUP
predicts the values of the random effect for each group taking into
consideration all the data.

There is a command in SAS for extracting the random effect predictions which
I would hope gives the same results as ranef in R.

Ken

On 30 September 2014 07:19, Zhanyou Xu <zxuiowa at iastate.edu> wrote:

 Dear Sir/Madam,

We are running lmer to calculate Best Linear Unbiased prediction (BLUP),
and we have 10 groups of data, each group has groupID and 42 individual.
We want to get BLUP values for each individual material within each
group. In SAS, there is a BY statement and we can get BLUP value for
each individuals within a group. We are wondering whether there is a
similar R code that we can calculate the BLUPs for each group similar
like BY statement in SAS? Below is the R code we pare testing for your
reference. Thank you very much in advance!

Zhanyou



---------- Forwarded message ----------
From: Ben Bolker <bbolker at gmail.com>
Date: Mon, Sep 29, 2014 at 2:05 PM
Subject: Re: [Lme4-authors] LME4 and Lmer model
To: zhanyou.xu at syngenta.com,
"lme4-authors at lists.r-forge.r-project.org"
<lme4-authors at lists.r-forge.r-project.org>,
marcia.almeida_de_macedo at syngenta.com, zxuiowa at iastate.edu,
lme4-authors at r-forge.wu-wien.ac.at


On 14-09-29 11:30 AM, zhanyou.xu at syngenta.com wrote:

Dear Ben and all LME4 authors,

We are running lmer to calculate Best Linear Unbiased prediction (BLUP),
and we have 10 groups of data, each group has groupID and 42 individual.
We want to get BLUP values for each individual material within each
group. In SAS, there is a BY statement and we can get BLUP value for
each individuals within a group. We are wondering whether there is a
similar R code that we can calculate the BLUPs for each group similar
like by statement in SAS? Below is the R code we pare testing for your
reference. Thank you very much in advance!

    Take a look at ranef() and coef().  If you have further questions,
could you please send them to r-sig-mixed-models at r-project.org ?

  I can believe that "%in%" works as a nesting indicator, but it is more
typical/I am more familiar with REPNO:HOSEID ...  (I'm also curious why
you need to ignore the greater-than-1-level check ...)

  thanks
   Ben Bolker

Zhanyou





Milkvarcomp =  lmer(YGSMN~ (1|ANIMALID) + (1|HOSEID) +
(1|REPNO%in%HOSEID) + (1|MINRNG%in%HOSEID) +

                (1|MINROW%in%HOSEID)+
(1|ANIMALID:HOSEID),control=lmerControl(check.nlev.gtr.1="ignore"),
data=Stage3Data)


------------------------------------------------------------------------
/This message may contain confidential information. If you are not the
designated recipient, please notify the sender immediately, and delete
the original and any copies. Any use of the message by you is
prohibited./

--

Ken Beath
Lecturer
Statistics Department
MACQUARIE UNIVERSITY NSW 2109, Australia

Phone: +61 (0)2 9850 8516

Building E4A, room 526
http://stat.mq.edu.au/our_staff/staff_-_alphabetical/staff/beath,_ken/

CRICOS Provider No 00002J
This message is intended for the addressee named and m...{{dropped:8}}