[R-sig-ME] [Fwd: Errors in lme]
"Gerhard Göldner"
u28101104 at zoology.up.ac.za
Mon Oct 15 21:39:55 CEST 2012
Good day,
I am trying to do analysis on a dataset using mixed models because I have
samples from repeated individual animals. Please find attached my
dataset. The question I would like to pose is if and what are the effects
of the variable of distances and time on the proportional mass
change(propmc). I have included other factorial effects such as age sex
and phase as well as year, but I don't know if I am using them correctly
in the model.
I have also attached the codes that I use for the model and the error
messages I am getting. I have done some internet searches and some
suggest missing data as the culprit or fixed effects that are correlated.
Even if I exclude correlated fixed effects and only include fixed effects
which are not correlated with each other, I still get the same message
"error in MEEM - singularity in backsolve, level 0, block 1".
In addition, if I use for example a simpler model with only one fixed
effect, the model runs, but the output doesn't have p-values.. it only
says NaN. I also did a search on that and found that it means Not a
number, though I don't know why it is producing that. I noticed that the
DF is negative however and that if I use less variables, the DF increases
and the p value is calculated. Is there a statistical way of using
t-values, without using p-values, as someone told me using AIC and
p-values are two different methodologies or philosophies. In other words,
is it correct to still use the AIC values even if I don't get p-values and
just use the AIC values.
Can you please help me by making any suggestions about how to correct
this. The "missing" data in my dataset are not missing - the values are
actually 0!
Thank you for any help you can give me - if not, could you please refer me
elsewhere? I have your book "Mixed effects models in S and S-plus" 2000,
but would appreciate any help you can give me.
I would appreciate it very much.
Gerhard
---------------------------------------------------------------------
This message and attachments are subject to a disclaimer. Please refer to http://www.it.up.ac.za/documentation/governance/disclaimer/ for full details. / Hierdie boodskap en aanhangsels is aan 'n vrywaringsklousule onderhewig. Volledige besonderhede is by http://www.it.up.ac.za/documentation/governance/disclaimer/ beskikbaar.
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: errors in lme.txt
URL: <https://stat.ethz.ch/pipermail/r-sig-mixed-models/attachments/20121015/d0c67d42/attachment.txt>
More information about the R-sig-mixed-models
mailing list