lme4 problem: model defining and effect estimation ------ question from new bird to R community from SAS community

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20110415/271ae959/attachment.pl>

genetic_evaluation <- read.table(textConnection("
sire dam adg
1  1  2.24
1  1  1.85
1  2  2.05
1  2  2.41

....

4  2  1.86
4  2  1.79
5  1  2.82
5  1  2.64
5  2  2.58
5  2  2.56"), header = TRUE)

# my R practice codes
require (lme4)
lmer(adg ~ 1 + (1|sire) + (1|dam/sire), data=genetic_evaluation)

You're missing the equivalent of the SAS class statement. Grouping variables need to be factors:

genetic_evaluation<-transform(genetic_evaluation,
 sire=factor(sire),
 dam=factor(dam))

Also, you probably don't want a random main effect of dam, so

lmer(adg ~ 1 + (1|sire) + (1|dam:sire), data=genetic_evaluation)

or even

lmer(adg ~ 1 + (1|sire/dam), data=genetic_evaluation)

Peter Dalgaard
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com

I was trying to fit a mixed model in animal experiment but stuck at simple
point. The following similar example is from SAS mixed model pp 212.


genetic_evaluation <- read.table(textConnection("
sire dam adg
1  1  2.24
1  1  1.85
...
5  2  2.58
5  2  2.56"), header = TRUE)

# my R practice codes
require (lme4)
 lmer(adg ~ 1 + (1|sire) + (1|dam/sire), data=genetic_evaluation)

 ****error message********************************************88
 Error: length(f1) == length(f2) is not TRUE
In addition: Warning messages:

Thanks for providing a self-contained example. The error message is really a
bit confusing (anybody around who understands what lme thinks here?), but
the solution is simple. Just make sure that dam and sire are factors:

genetic_evaluation$dam = as.factor(genetic_evaluation$dam)
genetic_evaluation$sire = as.factor(genetic_evaluation$sire)

I recommend the excellent course notes in the MCMCglmm package as a
complementary reading.

Dieter




--
View this message in context: http://r.789695.n4.nabble.com/lme4-problem-model-defining-and-effect-estimation-question-from-new-bird-to-R-community-from-SAS-comy-tp3453530p3453690.html
Sent from the R help mailing list archive at Nabble.com.

Thanks for providing a self-contained example. The error message is really a
bit confusing (anybody around who understands what lme thinks here?), 

It's not thinking at all, it's just evaluating a:b (so usual model formula interpretation is partially disabled? Hm, well, multiplying the two continuous variables wouldn't make sense either, I suppose.) and what you see is a variant of

(1:2):(1:2)

[1] 1
Warning messages:
1: In (1:2):(1:2) :
  numerical expression has 2 elements: only the first used
2: In (1:2):(1:2) :
  numerical expression has 2 elements: only the first used

Peter Dalgaard
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com

An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20110416/6389d33e/attachment.pl>