Administration

At developers meeting on 8/5 we discussed the following:

• Changes in the front and back end to the make RAM Objectives work with ML and FIML. The proposed changes will have RAM populate things like the covariance matrix and latent variables. These changes will make it easier to add LISREL Objective Functions to OpenMx.
• We discussed a parallel strategy for FIML:

At developers meeting on 7/29 we discussed the following:

• The bug in OpenMx 1.1 Beta release where an ordinal non-positive definite covariance matrix goes into an infinite loop has been fixed.
• The 'connect' slot which is replacing the deprecated 'all' slot in MxPath is ready for testing. It will be checked in over the weekend for testing. For more information on this please see the forum post here.

At developers meeting on 7/22 we discussed the following:

• There continues to be a known bug in the OpenMx 1.1 Beta release. When running FIML with an ordinal non-positive definite covariance matrix OpenMx goes into an infinite loop instead of reporting an error to the user. This issue has been moved very high on the OpenMx 1.1 priorities list.

At developers meeting on 7/15 we discussed the following:

• There is a known bug in the OpenMx 1.1 Beta release. When running FIML with an ordinal non-positive definite covariance matrix OpenMx goes into an infinite loop instead of reporting an error to the user.

At developers meeting on 7/8 we discussed the following list of upcoming development changes for OpenMx 1.2. OpenMx 1.1. is currently being beta tested and has been feature frozen.

• Changes to the back-end of the OpenMx to enforce consistency and make the addition of new model types more straight forward.
• The inclusion of multi-level models. Currently we have an interface design that the group is pleased with but we are still soliciting some additional feedback.
• The implementation of weighted least squares to offer performance improvements.

Here are the minutes for today's developers meeting.

1. Add three test conditions for summary output. See forum discussion.
   • calculated hessian is non-positive-definite
   • calculated hessian values are zero
   • calculated hessian values are very far from zero
2. Rename omx*** functions to imx**** functions. See forum discussion.

It looks like the R 2.12.0 binary has been released for windows. I've made some quick adjustments such that it should be possible to download a binary build of OpenMx for 32-bit windows with R 2.12.0. Our team is working on the 64-bit windows, this will be very cool when cross-compilation is working. We need to compile libnpsol for 64-bit windows so have patience.

Here are the minutes from our first meeting after the OpenMx 1.0 release. We made a list of objectives for the short term. These tasks involve code cleanup and rewriting some components of the library.

Objectives:

• rewrite mxEval. Handle definition variables.
• correctly transform cbind(), rbind(), and t() into matrix objects behind the scenes
• error handling in the backend
• pass algebra dependency graph to the backend
• mxRowObjective - complete NA reduction algebra

In OpenMx 1.0, it will be possible to use square brackets inside MxMatrix labels only when the row and column arguments of the square brackets are numeric literal values. Specifically, if you wish to use a definition variable in either the row or column argument, then you must place that reference in an algebra and then reference the algebra from the MxMatrix label.

The OpenMx developers team has started looking at the most effective optimizations we can implement prior to the version 1.0 release this summer. One large candidate is the sorting of raw data, and then using the information learned from the sorting process to speed up full information maximum likelihood (FIML) estimation. We should write a forum post devoted entirely to the sorting process, and the various transformation we can perform after we have sorted the data, but that forum post will appear in the future.