I have problem with the running time of R when using lmer to fit a linear mixed effect model. Is there any way to reduce the time ? Because if I have 50 subjects, it took me