Solves a univariate mixed model of form \(y=X\beta+M_1u_1+M_2u_2+e\) using the Expectation-Maximization algorithm.
Arguments
- Y
a matrix with n rows and 1 column
- X
a matrix with n rows and x columns
- M1
a matrix with n rows and m1 columns
- M2
a matrix with n rows and m2 columns
- Vu1
initial guess for variance of the first marker effects
- Vu2
initial guess for variance of the second marker effects
- Ve
initial guess for error variance
- tol
tolerance for declaring convergence
- maxIter
maximum iteration for attempting convergence
- useEM
should EM algorithm be used. If false, no estimation of variance components is performed. The initial values are treated as true.