References for covariance matrix in PROC NLP

[Camillus]

Hi,

My question in in the subject.

PROC NLP proposes several methods to compute covariance of the parameter estimates. Some of them are really uncommon (at least in econometrics). I would like obtain some references in particular for the methods M, B and U. Thank you

[sbxkoenk]

Hello @Camillus ,

PROC NLP is really a SAS legacy procedure (from SAS 9.3).

Such that people know what you are talking about , I put here a link to the (9.3) doc:

SAS/OR(R) 9.3 User’s Guide: Mathematical Programming Legacy Procedures
https://support.sas.com/documentation/cdl/en/ormplpug/64004/HTML/default/viewer.htm#ormplpug_nlp_sec...

The same six (6) types of covariance matrices are still there in PROC OPTMODEL

( The Nonlinear Programming Solver PROC OPTMODEL ).

Any reason why you stick to PROC NLP?
You can migrate your PROC NLP code to PROC OPTMODEL.

BR,

Koen

[Camillus]

Thank you.

I'm used to using NLP and I don't know OPTMODEL. I am finishing an article using NLP. So I'm not going to change now but I'll look into that next time.

[RobPratt]

When you are ready to migrate, you might find this example helpful:

PROC NLP: Rewriting NLP Models for PROC OPTMODEL - 9.3 (sas.com)

[sbxkoenk]

Hello,

With regard to your original question , see :

Example 7.5: Approximate Standard Errors

at page 648 -- 653 ✦ Chapter 7: The NLP Procedure

https://support.sas.com/documentation/onlinedoc/or/132/nlp.pdf

Koen