OpenMx General Help

Hi OpenMxers,
I want to determine automatically which order to put models into mxCompare so that the model with more degrees of freedom comes first.

Both models have been run, and I can access a models number of estimated parameters from summary(model).

summary() however, can be costly in terms of execution time: Is there a way to access the EP (or df) property from the model directly without calling summary?

Besten Wünsche,
tim

FYI, the situation I am trying to automate is avoiding this:

|Model        | EP|Delta-2LL |Delta df|p  |      AIC|Compare with Model |

Dear all,

I have been running a model within a loop over several variables, which does not converge well for all variables. Therefore, I applied mxTryHard. There are two things I noticed:

1) for models that run well without calling mxTryHard the best fit does not seem to change using mxTryHard, but the status code changes from 0 to -1

> fit1 <- mxRun(model,intervals =F)
> fit1$output$Minus2LogLikelihood
[1] 4595.342
> fit1$output$status
$code
[1] 0
$status
[1] 0

fit2<-mxTryHard(model,intervals=F)
> fit2$output$Minus2LogLikelihood
[1] 4595.342
fit2$output$status
$code

Hi, I'm using Onyx, which I think is based on OpenMx, for a confirmatory factor analysis.

I'm using a Macbook Air, and I can't work out how to modify the path arrows so they look neat. Can anyone help?

There is nothing in the documentation. The models in the examples all look nice, whereas everytime I add a path or covariance path, the program sticks them in anywhere and I can't move the end-points around to different parts of the observed or latent variable.

Please help!

Jay.

I created a model (using a function) saved as "x". I would like to change a label in the A matrix. Here are the labels for the A matrix:

> x$matrices$A$labels
v_1 v_2 v_3 mu
v_1 NA NA NA "b_1"
v_2 NA NA NA "b_2"
v_3 NA NA NA "b_3"
mu NA NA NA NA

I would like to change "b_3" to "b_2":

> x$matrices$A$labels[3,4] <- "b_2"
Error: Right-hand side of assignment operator has illegal value FullMatrix, SymmMatrix, FullMatrix, and FullMatrix

Why does this not work?

Hi,

This is actually quite a quick question and the title says most of it already.
Anyway, is there anyway to constrain mxAlgebra objects across multiple submodels?

I would like to constrain estimated heritability of two separate twin models to be the same while allowing each submodel to have separate parameter estimations.

Unfortunately I was unable to find a way to do so. I have tried mxConstraint(modelA.algebra1==modelA.algebra2) but that seemed not to work.

I would be thankful for any reply!

I am a bit confused what the latest stable OpenMx Version is. The last news announcement is for 2.0.1, the install script installs 2.2.4 on my linux supercomputer and fails to install 2.2.2 on my mac laptop (see http://openmx.psyc.virginia.edu/thread/3997).

Are the default confidences intervals for mxCI 95% or 90%? Or is there something funky in the documentation of mxCI? My understanding of a 95% confidence interval is that it is equivalent to a two-tailed test with alpha = .05. That is, it delimits the 2.5% most extreme values at the low end and the 2.5% most extreme values at the high end, leaving the 95% most likely values between the confidence intervals.

Attached are two outputs, one from OpenMx 2.0.1.4157 using OS X on a Mac and the other from OpenMx 2.2.2 on Red Hat Linux. In both situations the models converge when I use mxExpectationNormal and mxFitFunctionML. But when I manually set the fit function to -log(likelihood) using mxFitFunctionAlgebra, mxRun crashes in both operating systems (albeit with different errors).
Am perplexed.
Greg

Why does this give an error when the code is taken directly from the documentation to mxExpectationGREML?

> require(OpenMx)
> mxVersion()
OpenMx version: 2.0.1.4157
R version: R version 3.1.2 (2014-10-31)
Platform: x86_64-apple-darwin10.8.0
Default optimiser: NPSOL
> # the following are the first three commands from the Examples from
> # the documentation to mxExpectationGREML
> dat <- cbind(rnorm(100),rep(1,100))
> colnames(dat) <- c("y","x")
>
> ge <- mxExpectationGREML(V="V",yvars="y",Xvars=list("X"),addOnes=FALSE)