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
- /
- Analytics
- /
- Stat Procs
- /
- PROC MIXED with RANDOM fails to ESTIMATE Random Ef...

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
- Permalink
- Email to a Friend
- Report Inappropriate Content

03-16-2017 06:03 PM

Can anyone provide me with some insight on why the random statement has an estimate of 0?

I have multiple data sets with the same issue (although the code runs just fine on others). The number of replicates ranges from 4 to 6 between the sets (most have 5) with 5 or more levels of treatment.

I am running a proc mixed model and using the random statement to control for replicate (and also to see if there is an effect).

This is the output I get for the random section:

Covariance Parameter Estimates Cov ParmEstimate StandardError Z Value Pr > Z

rep 0 . . .

Residual 0.000250 0.000072 3.460 .0003

Are there any tips or tricks that would allow me to get an estimate in this section?

Any advice would be appreciated.

Accepted Solutions

Solution

03-22-2017
06:08 AM

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

Posted in reply to ijhgo_uynrtpwiavnckml

03-16-2017 08:26 PM

Is the MIXED procedure also telling you that the estimated G matrix is not positive definite?

Most likely, this result implies that there is little variation among *rep*s for this response variable. The procedure has set the estimate to zero (hence, the SE is missing) and continued on its merry way. See Section III in this paper

http://support.sas.com/resources/papers/proceedings12/332-2012.pdf

In the documentation

under Parameter Constraints, you'll find:

"For some data sets the final estimate of a parameter might equal one of its boundary constraints. This is usually not a cause for concern, but it might lead you to consider a different model. For instance, a variance component estimate can equal zero; in this case, you might want to drop the corresponding random effect from the model. However, be aware that changing the model in this fashion can affect degrees-of-freedom calculations.

If that doesn't seem to be the reason, then you'll want to post your code and an example dataset so that people here have more to work with.

All Replies

Solution

03-22-2017
06:08 AM

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

Posted in reply to ijhgo_uynrtpwiavnckml

03-16-2017 08:26 PM

Is the MIXED procedure also telling you that the estimated G matrix is not positive definite?

Most likely, this result implies that there is little variation among *rep*s for this response variable. The procedure has set the estimate to zero (hence, the SE is missing) and continued on its merry way. See Section III in this paper

http://support.sas.com/resources/papers/proceedings12/332-2012.pdf

In the documentation

under Parameter Constraints, you'll find:

"For some data sets the final estimate of a parameter might equal one of its boundary constraints. This is usually not a cause for concern, but it might lead you to consider a different model. For instance, a variance component estimate can equal zero; in this case, you might want to drop the corresponding random effect from the model. However, be aware that changing the model in this fashion can affect degrees-of-freedom calculations.

If that doesn't seem to be the reason, then you'll want to post your code and an example dataset so that people here have more to work with.

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

Posted in reply to ijhgo_uynrtpwiavnckml

03-16-2017 10:51 PM

If you could use PROC GLIMMIX , try : random int time / type=chol