I've trying to perform a logistic mixed model introducing several covariables with GMMAT. To correct for cryptic relatedness, I've been using a matrix that is positive semi-definite (it contains 0 values).
I could adjust the null model and perform the scores test. However, when running the function glmm.wald(), I am getting this type of errors: "Error in h(simpleError(msg, call)) : error in evaluating the argument 'x' in selecting a method for function 'chol2inv': the leading minor of order 44 is not positive definite"
Is it because the kinship matrix is positive semi-definite? Should I use a definite one instead? Nonetheless, the Wald test is performed and I'm getting an output.
Hello,
I've trying to perform a logistic mixed model introducing several covariables with GMMAT. To correct for cryptic relatedness, I've been using a matrix that is positive semi-definite (it contains 0 values).
I could adjust the null model and perform the scores test. However, when running the function glmm.wald(), I am getting this type of errors: "Error in h(simpleError(msg, call)) : error in evaluating the argument 'x' in selecting a method for function 'chol2inv': the leading minor of order 44 is not positive definite"
Is it because the kinship matrix is positive semi-definite? Should I use a definite one instead? Nonetheless, the Wald test is performed and I'm getting an output.
Thank you in advance for your help!