用stata做空间计量回归时，做SEM模型结果是，固定效应可以出来结果，做随机效应时会出现以下提示，如下图，可以用什么方法解决？Iteration 0:   Log-likelihood =  -13315.62  Iteration 1:   Log-likelihood = -13224.632  
Iteration 2:   Log-likelihood = -13221.844  (backed up)
Iteration 3:   Log-likelihood = -13220.497  (backed up)
Iteration 4:   Log-likelihood = -13220.166  (backed up)
Iteration 5:   Log-likelihood = -13220.001  (backed up)
Iteration 6:   Log-likelihood =  -13219.99  (backed up)
Iteration 7:   Log-likelihood = -13219.988  (backed up)
Iteration 8:   Log-likelihood = -13219.987  (backed up)
numerical derivatives are approximate
nearby values are missing
Iteration 9:   Log-likelihood = -13219.986  (backed up)
numerical derivatives are approximate
nearby values are missing
Iteration 10:  Log-likelihood = -13219.986  (backed up)
numerical derivatives are approximate
nearby values are missing
Iteration 11:  Log-likelihood = -13219.986  (backed up)
numerical derivatives are approximate
nearby values are missing
Hessian is not negative semidefinite
r(430);
做其他的都可以出来，就是做SEM随机效应的时候出不来结果，求大神帮忙啊