Understanding the technical requirements for SAS/OR

[Satyakshma]

Hi All,

I need to write a code for optimisation problem, and as per my understanding we need either SAS/IML or SAS/OR component. None of them is a part of my current SAS setup. I have two questions regarding this:
1. Is it possible to add the SAS/OR component to our existing SAS setup.
2. Is it necessary to have SAS Stimulation Studio for SAS/OR.
3. Is there any other alternative to SAS/OR or SAS/IML to write code for optimisation problem.

Thanks,
Satyakshma Rawat

[PaigeMiller]

1. Your company or university has to pay money to SAS
2. NO
3. Unlikely using SAS, but since we don't really know what problems you are trying to solve, we can't be more specific

[Satyakshma]

Thank you for your response. Below is the link of the optimization problem statement which I posted some time back.
https://communities.sas.com/t5/SAS-Programming/Generate-an-index-value-used-in-calculating-Predicted...

Looking forward to your response on this as I have been struggling with resolution of this problem statement for quite sometime now.

Thanks,
Satyakshma Rawat

[PaigeMiller]

Regarding your link, I never (and many people never) download and open MS Office files, as they can be a security risk. So I can't see your data.

The details provided at that link are not sufficient for me to determine what other types of coding might work. Since you say you want to minimize the sum of squares of errors, what method is being used to do this? Can you be specific? SAS has this capability in some areas. 

[Satyakshma]

To minimize the sum square error, we calculate the predicted values using a index value (formula for calculation of this predicted value is NORM.S.DIST(index value + zscore))  , the index value in the formula can be any random value such that the sum square error is minimum between the actual values and predicted values. Hope I am able to explain, else let me know some other way if I can share the dummy data with you.

[PaigeMiller]

Sounds like some type of regression. Is it?

[Satyakshma]

regression can help in final output, but generating the index value such that when used in calculation gives minimum sum square error, is the challenge here.

[Ksharp]

1)Yes. But you need pay it .
2)No.
3)Try other third-side free software. Like :
Lingo
https://www.lindo.com/

MatLab

Also could try FREE sas software (SAS OnDemand for Academics).
https://welcome.oda.sas.com/login