被解释变量二值
用logit模型回归，Iteration 0:   log likelihood =  -9.667766  
Iteration 1:   log likelihood = -9.0704415  
Iteration 2:   log likelihood = -7.7032835  
Iteration 3:   log likelihood = -7.3805801  
Iteration 4:   log likelihood = -6.9349477  
Iteration 5:   log likelihood = -6.5999437  
Iteration 6:   log likelihood = -6.1167926  
Iteration 7:   log likelihood = -5.7196697  
Iteration 8:   log likelihood =  -5.204502  
Iteration 9:   log likelihood = -4.9805788  
Iteration 10:  log likelihood = -4.9329421  (not concave)
Iteration 11:  log likelihood = -4.9319949  (not concave)
Iteration 12:  log likelihood = -4.9316364  (not concave)
Iteration 13:  log likelihood = -4.9315606  (not concave)
Iteration 14:  log likelihood = -4.9315392  (not concave)这种错误原因是什么？
probit 模型也是一样，求大神给指点