0
I’m running this model on R:
modelogeral<-lmer(lognplanta~Clas_renda*sexo*idade*ocupacao*escolaridade*Clas_npessoas*(1|comunidade),data=bruno)
The following message appears:
Fixed-Effect model Matrix is rank deficient so dropping 668 Columns / coefficients Boundary (singular) fit: see ? isSingular
What does that mean?
As the message says, see
help('isSingular')for the meaning of unique Fits. The theme is difficult and I believe it is best to start with the documentation and follow the links on this page. This question is more appropriate for the Cross Validated the Stack Exchange community for Statistics questions, the OS and the OS in Portuguese are for code questions.– Rui Barradas
Thank you for your attention.
– Bruno
@Bruno, your model specification is wrong. Also, you are putting an interaction for all variables. Make:
modelogeral<-lmer(lognplanta~Clas_renda+sexo+idade+ocupacao+escolaridade+Clas_npessoas + (1|comunidade),data=bruno)and your model should run. After that, you can place the interactions (*) in its fixed effects. but never multiply the random part too, that is *(1|community). The fact that the matrix is singular means that it does not admit inverse (does not admit solution). In practical terms, it has many parameters estimating at the same time.– Guilherme Parreira