Open cytwill opened 4 years ago
I had the same issue, and I found this: https://github.com/rickecon/StructEst_W17/issues/26
Thank you, I have got the answer!
Thanks, @jesuspachecov and @cytwill. The approach (I don't know why it is this way) is to use the results.hess_inv.todense()
method at the end of the hess_inv
call.
My optimization method can not give an explicit inverse hessian variance-covariance matrix, my current method return the hess_inv as:
<4x4 LbfgsInvHessProduct with dtype=float64> Can anyone help with this?