The varlist1 (varlist2) syntax is of special use to programmers.  It verifies that the data are sorted by
    varlist1 varlist2 and then performs a by as if only varlist1 were specified.  For instance,

        by pid (time): generate growth = (bp - bp[_n-1])/bp

    performs the generate by values of pid but first verifies that the data are sorted by pid and time within pid.


在论坛见到很多次by varlist1 (varlist2) 和 by varlist1 varlist2，之前有混用过，今天查了下帮助文件，发现应该是有区别的，括号的意思是，括号里的变量参与排序但不作为分组变量参与计算。例子如上。

另外，
by和bys的区别在于前者没有排序，要求数据是已经排好序的，后者命令可以排序。

Syntax

        by varlist: stata_cmd

        bysort varlist: stata_cmd

    The above diagrams show by and bysort as they are typically used.  The full syntax of the commands is

        by varlist1 [(varlist2)] [, sort rc0]:  stata_cmd

        bysort varlist1 [(varlist2)] [, rc0]:  stata_cmd

例如by rep78, sort:  tabulate foreign等价于bysort rep78: tabulate foreign