You are here

Bivariate Cholesky model adjusted for age-Error message for ACE

2 posts / 0 new
Last post
Salutfate's picture
Joined: 11/22/2013 - 04:34
Bivariate Cholesky model adjusted for age-Error message for ACE

I am using a bivariate cholesky model for combined continuous and bivariate variables with Age effects on thresholds.

I am stuck with the error message in running model 'CholAce':A definition variable has been declared in model 'ACE' that does not contain a data set.

Could you tell me what the error means? Any help appreciated! Thanks.

AdminRobK's picture
Joined: 01/24/2014 - 12:15
I'm going to take a guess

I'm going to take a guess that 'ACE' is the name of your "top-level" model, which contains several submodels (probably four, for MZ males, DZ males, MZ females, and DZ females). I suspect that your top-level model does not contain an MxData object, but does contain at least one MxMatrix which has labels starting with "data" followed by a period (i.e., "data."). Try rebuilding your overall ACE model in a way that does not include those MxMatrix objects and see if the problem persists.

If I'm correct, then here's why you're getting an error. MxMatrix labels that start with "data." tell OpenMx to look at the raw MxData object to find the values for the MxMatrix elements with such labels. Specifically, OpenMx looks for the data column the name of which matches the part after the "data.". Such columns are called "definition variables," because they can be used to define a slightly different variation of the model to each row of the raw dataset. For instance, if an MxMatrix has an element labeled 'data.age', then OpenMx uses the first value of the data column labeled 'age' when fitting the model to the first row, and uses the second value from that column when fitting the model to the second row, and so on.