mostly, using the default options is good enough, that performs nearest match based on probit or you can use logit option.  The program generates _weight, _id, and _nk (_n1 for 1 to 1), and will be useful to identify the matched samples.

同求啊 楼主你解决问题了吗？我现在也是这个问题 找配对样本  头大死了 想了整整几天了 看help文件的optmatch2 ,没法完全理解呢