the coding of factor f as a covariate.
The function is defined for f being a simple factor, Trait
The lin(f) function does not centre or scale the variable.
Motivation: Sometimes you may wish to fit a covariate as a random factor as well.
If the coding is say 1 ... n,
then you should define the field as a factor in the field definition
and use the lin() function to include it as a covariate in the model.
Do not centre the field in this case.
If the covariate values are irregular, you would leave the field as a covariate
and use the fac() function to derive a factor version.
Return to start