用stata做空間計(jì)量回歸時(shí),做SEM模型結(jié)果是,固定效應(yīng)可以出來結(jié)果,做隨機(jī)效應(yīng)時(shí)會(huì)出現(xiàn)以下提示,如下圖,可以用什么方法解決?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);
做其他的都可以出來,就是做SEM隨機(jī)效應(yīng)的時(shí)候出不來結(jié)果,求大神幫忙啊