ALDEx3 Mixed-Effects Models20 days ago
Motivation | The model | The profiled REML objective | After the fixed effects and the residual scale are profiled out, the REMLcriterion depends only on (\theta_s). Following lme4, define[A_{\theta} = I_q + \Lambda_{\theta}^T Z^T Z \Lambda_{\theta} = L_{\theta}L_{\theta}^T,\qquadR_{ZX,\theta} = L_{\theta}^{-1}\Lambda_{\theta}^T Z^T X,\qquadM_{\theta} = X^T X - R_{ZX,\theta}^T R_{ZX,\theta},]where (L_\theta) is the lower Cholesky factor of (A_\theta). The matrix(M_\theta) equals (X^T V_\theta^{-1} X) for the relative marginal covariance(V_\theta = Z\Sigma_\theta Z^T + I_N); it is the generalized least-squares(GLS) information for (\beta). The response enters through the penalizedweighted residual sum of squares,[\mathrm{PWRSS}_s(\theta) | With (d = N - p), the profiled REML objective for draw (s) is, up to aconstant,[\ell_s(\theta) | What BLMM approximates | Standard errors and degrees of freedom | Stopping and boundary checks | Worked example | Contrasts | Example comparison: lme4 versus blmm | Runtime scaling | Practical guidance | Other inference options | Assumptions, validation and limitations
ALDEx3 1.7.0Justin Silverman ALDEx3-mixed-effects.Rmd