Dear R fans, MASS::mvrnorm with empirical=T is very useful to reproduce practice data of published papers. Now I want to reproduce a binomial variable with multiple normal variables. The target paper gave a big corr table with sd(s) and Means including a GENDER variable. Is there a point-biserial corr version of mvrnorm? lixiaoxu at fudan.edu.cn / xiaoxu.lxxm.com