SAS Optimization, and SAS Simulation Studio

turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

10-24-2008 09:30 AM

When I run the program, I get multiple warning messages and error messages in running the program, these messages appears many times. I cannot figure them out and have no idea how to fix them. More important, several estimated parameters are near to bounded values, which are not supposed to be these values. In other word, I do not get what I hope to get.

I cannot find a good way to "put" print out the intermediate output in order to find out the problems.

Moreover, I have no idea how to improve the part of the code in proc nlp.

The main idea for the program is to maximize a likelihood function to estimate a set of parameters ( in my model there are 10 parameters). Each time I firstly set a different set of initial values for the parameters, then run the program. Therefore I have a series of likelihood value given by the program, and I choose the set of estimated parameters which lead to the largest likelihood value.

I list below several problems in applying “proc nlp” to maximize the likelihood function.

WARNING: In a total of 1 calls an error occurred during execution of the program statements. NLP

attempted to recover by using a shorter step size.

NOTE: The above message was for the following by-group:

group=2

NOTE: Sparse storage of linear constraints: 220 < 400 Bytes.

NOTE: Initial point was changed to be feasible for boundary and linear constraints.

ERROR: Execution Errors for _OBS_= 42 :

ERROR: There are references to missing variables when the program code is executed for _OBS_= 42

WARNING: Your program statements cannot be executed completely.

WARNING: In a total of 2 calls an error occurred during execution of the program statements. NLP

attempted to recover by using a shorter step size.

NOTE: The above message was for the following by-group:

group=3

NOTE: Sparse storage of linear constraints: 220 < 400 Bytes.

NOTE: Initial point was changed to be feasible for boundary and linear constraints.

WARNING: Your program statements cannot be executed completely.

I am eager to learn any of your suggestions or comments.

By the way, do you think "proc optmodel" will work better for me than "proc nlp"?

my email: jinziyong2007@hotmail.com

I cannot find a good way to "put" print out the intermediate output in order to find out the problems.

Moreover, I have no idea how to improve the part of the code in proc nlp.

The main idea for the program is to maximize a likelihood function to estimate a set of parameters ( in my model there are 10 parameters). Each time I firstly set a different set of initial values for the parameters, then run the program. Therefore I have a series of likelihood value given by the program, and I choose the set of estimated parameters which lead to the largest likelihood value.

I list below several problems in applying “proc nlp” to maximize the likelihood function.

WARNING: In a total of 1 calls an error occurred during execution of the program statements. NLP

attempted to recover by using a shorter step size.

NOTE: The above message was for the following by-group:

group=2

NOTE: Sparse storage of linear constraints: 220 < 400 Bytes.

NOTE: Initial point was changed to be feasible for boundary and linear constraints.

ERROR: Execution Errors for _OBS_= 42 :

ERROR: There are references to missing variables when the program code is executed for _OBS_= 42

WARNING: Your program statements cannot be executed completely.

WARNING: In a total of 2 calls an error occurred during execution of the program statements. NLP

attempted to recover by using a shorter step size.

NOTE: The above message was for the following by-group:

group=3

NOTE: Sparse storage of linear constraints: 220 < 400 Bytes.

NOTE: Initial point was changed to be feasible for boundary and linear constraints.

WARNING: Your program statements cannot be executed completely.

I am eager to learn any of your suggestions or comments.

By the way, do you think "proc optmodel" will work better for me than "proc nlp"?

my email: jinziyong2007@hotmail.com

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

10-28-2008 11:14 AM

James,

I suggest that you send the details of your PROC NLP code and the results you are seeing to our Technical Support group so that we can help you to determine how best to use PROC NLP (or perhaps PROC OPTMODEL) to meet your requirements.

Contact SAS Technical Support by going to http://support.sas.com/ and clicking on "Submit a Problem" under the "SUPPORT" heading on the left-hand side of the window.

Ed Hughes

Product Manager, SAS/OR

I suggest that you send the details of your PROC NLP code and the results you are seeing to our Technical Support group so that we can help you to determine how best to use PROC NLP (or perhaps PROC OPTMODEL) to meet your requirements.

Contact SAS Technical Support by going to http://support.sas.com/ and clicking on "Submit a Problem" under the "SUPPORT" heading on the left-hand side of the window.

Ed Hughes

Product Manager, SAS/OR