Developer's Meeting Notes 2/17/12

No replies
rgore's picture
Offline
Joined: 01/27/2011

At developers meeting on 2/17 we discussed the following:

  • There is now a GUI for OpenMx installation. The GUI addresses issues with the version of gcc that is provided in Mac OS 10.7.
  • More LISREL code is checked in and LISREL objective functions appears close to working.
  • The list of features being considered for an OpenMx. The goal for the GUI is to quickly engage new OpenMx users and encourage them to learn and modify OpenMx source code. Specifically, the following features were discussed:
    • The inclusion of coefficients and variables in a publication quality output of the model.
    • Different views of a OpenMx model: (1) OpenMx source code, (2) modifiable graphical representation, (3) connector view.
      • Model paths with good curvature.
      • Provide the same features for RAM and LISREL models.
  • The following list of new features for OpenMx which were previously discussed on 2/3/12 were further flushed out.
    • OpenMx GUI - Timo von Oertzen is leading the development of this with several other of his colleagues. He will be meeting with them in several weeks to begin development.
    • Weighted Least Squares Minimization - Ryne Eastbrook and Tim Brick are leading the development on this feature. The are investigating how to adapt some of the group's previous work to implement WLS in a new (and more documented) way.
    • Smart gradient calculations that improve efficiency - Tim Brick has made substantial progress on and will continue to finish it in upcoming weeks.
    • Creating a versions of solnp in C so that OpenMx can be distributed as a CRAN package - Ross Gore is rewriting the Rsolnp package (written in R) in C.
    • Multilevel modeling - Michael Spiegel is exploring the viability of an interface specced out by the group several months ago.
    • Improved Methods to Handle Data in OpenMx in a variety of contexts
      • Multilevel OpenMx Data
      • Likelihood data that goes across different machines
      • Heterogeneous Data
      • Swift R Integration
      • Relational Database methods (sql-like) to query data
    • Bayesian Objective Function - no immediate action, there maybe some progress on this starting in two or three weeks.