In the matrix example on the front page, the mxAlgebra statement needs dimnames. It should be: mxAlgebra(A %*% L %*% t(A) + U, name="R", dimnames = list(names(demoOneFactor), names(demoOneFactor))). Also, the development team doesn't have access to move forum posts, or edit the home page of the wiki.

It seems that the diagonal of the $Hessian results returned are used as the Error Estimate in summary. Running the homepage model, pathic version we get:

name matrix row col parameter estimate error estimate
1 A 1 6 0.39715247 163.52372
2 A 2 6 0.50366153 149.57542
3 A 3 6 0.57724183 141.16243
4 A 4 6 0.70277427 125.43848
5 A 5 6 0.79625063 53.94078
6 S 1 1 0.04081422 510.59245

Clean up the three images in SimpleReg, MultReg and MultiReg:
-keep the notation of (particularly intercept) terms consistant with text
-make parameter notation text bigger

Give Hermine some bug reports.

The script models/failing/MultipleRegression_MatrixRawReverse.R is identical to models/passing/MultipleRegression_MatrixRaw.R except the 1st and 3rd columns of observed data have been switched (along with their column names). The dataColumns slot is assigned c(2,1,0) in MultipleRegression_MatrixRawReverse.R, but the correct answer is not being computed.

After an mxAlgebra expression is computed, the dimnames of the mxAlgebra object should be applied to the result slot. This can be done in the front-end, after the results from the optimizer are returned.

A DGEMM error is being thrown by some uses of the quadratic multiply operator. This needs to either stop being an error or start being an interpretable error.

Script throwing the error is attached to http://openmx.psyc.virginia.edu/thread/103

When the user provides vectors as arguments to mxMatrix() and type is 'Full', only one of nrow and ncol is needed. This fix applies only to 'Full' matrices. The following matrices are all square: "Diag", "Iden", "Lower", "Stand", "Sdiag", "Symm". And "Unit" and "Zero" need both row and column information.

See http://openmx.psyc.virginia.edu/thread/83 comments #4 and #5.

The following model has been updated in the /demo directory, but not in the documentation. The AE model should be created by modifying the sub-models. On the documentation page, currently the super-model is being modified to create the AE model (this doesn't work).

http://openmx.psyc.virginia.edu/docs/OpenMx/latest/Examples_Path.html#ace-model-a-twin-analysis