OpenMx 0.3.0 (READ ME)

OpenMx 0.3.0 is a milestone along the path to the 1.0 release. See this post for the full set of features that will be included in the 1.0 release. The most significant change to the user interface is the improved mxConstraint() function. Instead of writing:

mxConstraint("A", "=", "B")

You should write:

mxConstraint(A == B)

New Workshops page

We have added a Workshops link to the top menu bar of the OpenMx website where workshops using OpenMx are listed. Links to registration pages (when available) are provided so that you can find a workshop that is convenient for you.

If you are putting on an OpenMx workshop or know of a workshop that is not yet listed, we would be happy to list it on our workshops page. Just post a message on one of the forums and the OpenMx team will add your workshop to the list.

genepi library: move into OpenMx, or keep on googlecode?

I put a version of the helper functions up on googlecode back at the workshop, renaming the functions with the prefix "genEpi_" so they are easier to pick out in a script.

As these are things that were very popular with users, can we figure out which (all?) of these to move into OpenMx and document (happy to help)? And how they will be named if they are part of the OpenMx distro itself: Perhaps mxl for library (misreadable as upper case i) or mxh for "Helper" or just genEpi_ for the sake of the guinea pigs :-)

1. mxhLabeler()
2. mxh_Labeler()
3. genEpi_Labeler()

citation("OpenMx")

citation("OpenMx") currently throws a warning at the end:

citation("OpenMx")
To cite package ‘OpenMx’ in publications use:

Steven Boker, Michael Neale, Hermine Maes, Paras Metah, Sarah Kenny, Timothy Bates, Ryne Estabrook, Jeffrey Spies, Timothy Brick
and Michael Spiegel (). OpenMx: The OpenMx Statistical Modeling Package. R package version 0.2.10-1172.
http://openmx.psyc.virginia.edu

A BibTeX entry for LaTeX users is

@Manual{,
title = {OpenMx: The OpenMx Statistical Modeling Package},

Weirdness with and without []: assignment of an object of class "numeric" is not valid for slot "values" in an object of class

The following indicates that assignment of numeric values to an @values part of an mxMatrix only works when the [] are there. When they are not there (which used to work fine) we get an error:

> DirectionofCausationBcausesA$ACE.B@values
[,1] [,2]
[1,] 0 0
[2,] 0 0
> DirectionofCausationBcausesA$ACE.B@values[]
[,1] [,2]
[1,] 0 0
[2,] 0 0
>
> DirectionofCausationBcausesA$ACE.B@values<-0
Error in checkSlotAssignment(object, name, value) :

GeneticEpi_Matrix.rst refers to undefined matrices

This portion of the user guide refers to the matrices 'expMeanMZ' and 'expMeanDZ' in several places. However these matrices don't exist. See http://openmx.psyc.virginia.edu/docs/OpenMx/latest/GeneticEpi_Matrix.html. Make sure to svn update before making changes, as I've made some other corrections to the file.

OpenMx 0.2.9 includes standard errors!

This is primarily a bug fix release. However we are now calculating both a Hessian matrix and standard error estimates at the conclusion of optimization and making those results available in the summary() function.

Release 0.2.9-1147

  • bug fix for ordinal FIML with columns that are not threshold columns
  • bug fix for detection of algebraic cycles when multiple objective functions are present
  • test cases included for standard error calculation
  • enabled standard error calculation by default