I have a panel data with fixed N = 197 countries. T is in years (15). My dependent variable varies between 0 and 1 (no observation at extremes). There is serial autocorrelation, heteroskedasticity, and cross-sectional dependence. Now, which command should I use to get correct standard errors for a fixed effects model? What is the command for random effects if it doesn't exist for fixed effects model?