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

- Home
- /
- SAS Programming
- /
- Base SAS Programming
- /
- Product sum function OR similar

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

06-03-2009 12:31 PM

Two questions:

1. When calculating the probability values from the system of equations(parameter estimates/variable values) for a categorical model, I am seeing code that includes Y=B1+(Var1='xxx' * B2)+..., for example. How does SAS interpret that...meaning what does it do with Var1='xxx', evaluate it to 0 or1? I'd like to modify the code to dynamically(and therefore generically) calculate the Y value without explicit reference to each Var value. I was thinking about an array of variables/values and the parameter estimate array. Is that the most efficient option for large datasets?

2. The first question is somewhat contingent on the answer to this one...is it possible to create a sum of products when arrays are involved without employing IML or is that in fact what I should use...is it efficient for a large dataset.

Probabilities will then be used for forecasting.

Any assistance appreciated.

1. When calculating the probability values from the system of equations(parameter estimates/variable values) for a categorical model, I am seeing code that includes Y=B1+(Var1='xxx' * B2)+..., for example. How does SAS interpret that...meaning what does it do with Var1='xxx', evaluate it to 0 or1? I'd like to modify the code to dynamically(and therefore generically) calculate the Y value without explicit reference to each Var value. I was thinking about an array of variables/values and the parameter estimate array. Is that the most efficient option for large datasets?

2. The first question is somewhat contingent on the answer to this one...is it possible to create a sum of products when arrays are involved without employing IML or is that in fact what I should use...is it efficient for a large dataset.

Probabilities will then be used for forecasting.

Any assistance appreciated.

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

Posted in reply to deleted_user

06-03-2009 11:29 PM

>evaluate it to 0 or1?

Correct. This is a boolean.

1-How about[pre]

do i=1 to 20;

Y=B1+(Var{i}='xxx' * B2)+...,

end;[/pre]

2-

>is it possible to create a sum of products when arrays are involved without employing IML

Yes it is. Your question is too vague.

Correct. This is a boolean.

1-How about[pre]

do i=1 to 20;

Y=B1+(Var{i}='xxx' * B2)+...,

end;[/pre]

2-

>is it possible to create a sum of products when arrays are involved without employing IML

Yes it is. Your question is too vague.