我觉得是用的MLE估计

谢谢您，我找到出处了：“2.1 Estimation
Various methods of estimating spatial panel models have been proposed. Broadly, they
fall into two categories: i) GMM estimators, and ii) Quasi-Maximum Likelihood (QML)
estimators. All the models that can be estimated using xsmle fall into the second
category. A synopsis guide with all estimable models and their features is reported in
Table 1.3 The gain from programming gradients is large, so v1 evaluators are used for all
but one of the specifications. The exception is the random-effects SEM, whose likelihood
function involves a transformation using the Cholesky factors of a rather complicated
matrix containing the parameters to be estimated, so that the matrix differentiation is
extremely messy”
所以xsmle进行SAR、SEM、SAC、SDM模型估计时采用的全都是准极大似然方法，动态的SAR模型和动态SDM模型使用的Yu等（2008）的偏差修正QMLE方法，以上内容出自xsmle函数作者：Federico Belotti （2017）

哇，学习了，谢谢您！我想请教一下你会做动态的SDM模型吗？

可参考Federico Belotti 2017文章来做，里面介绍很详细

请问楼主   xsmle的sdm模型能检验空间收敛性吗