Wed, 02/09/2011 - 11:07

anyone explain this?

> model2 <- mxModel(model1, mxAlgebra(.5*CTf, name="CTDZf"))

> model2 <- mxRun(model2)

Running SameMZDZf

Error in runHelper(model, frontendStart, intervals, silent, suppressWarnings, :

BLAS/LAPACK routine 'DPOTRF' gave error code -4

> model2 <- mxModel(model1, mxAlgebra(.5 %x% CTf, name="CTDZf"))

> model2 <- mxRun(model2)

Running SameMZDZf

Warning message:

In model 'SameMZDZf' NPSOL returned a non-zero status code 1. blah blah ...

yet

> mat <- matrix(rnorm(16), 4)

> .5 * mat == .5 %x% mat

[,1] [,2] [,3] [,4]

[1,] TRUE TRUE TRUE TRUE

[2,] TRUE TRUE TRUE TRUE

[3,] TRUE TRUE TRUE TRUE

[4,] TRUE TRUE TRUE TRUE

suspect something simple. if not, i can send the MxModel.

greg

Which version of OpenMx are you running? I get an error message under v. 1.0.4:

There are three multiplication operators in OpenMx (operators and functions). "*" performs elementwise-multiplication. "%*%" performs matrix-multiplication. And "%x%" is the kronecker product. The reason for these three operators is that OpenMx is a matrix processing language. In R, the "*" operator will perform a element-scalar multiplication if one argument is scalar, or elementwise-multiplication if both arguments are matrices. In OpenMx, the scalar value is transformed into a 1 x 1 matrix.

In the OpenMx 1.1 release, the expression

`mxEval(bar, model, compute=TRUE)`

will return an error. It currently does not return an error, but this is a bug. We had to rewrite a significant portion of the evaluation routines in order to clean this up.merci, mike

am using version 1.0.3-1505.

suggest that you change the documentation for mxAlgebra from

expression An R expression of matrix operators and matrix functions.

to

expression An expression of OpenMx-supported R matrix operators and matrix functions

"*" is a valid matrix operator in R and in other matrix algebra languages because a 1 x 1 matrix is, by definition, a scalar and scalar*matrix is a legitimate operation.

> expression An expression of OpenMx-supported R matrix operators and matrix functions

done