Version 2.6.7 of OpenMx now available

The newest release of *OpenMx*, version 2.6.7, is now available through CRAN and through our own repository. The most important change for users to be aware of is that, whereas previously OpenMx defaulted to using one less than the number of cores, it now defaults to using 2 threads. This change was made at the insistence of the CRAN Team to reduce their test-server burden. However, the number of threads can be changed via mxOption, with the `Number of Threads` option. As a reminder, multithreading in *OpenMx* is only supported with Linux and with the *OpenMx* Team's build for Mac OS X.

New features and bug-fixes include:

• The CSOLNP optimizer is now better at calculating confidence intervals.
• It is now possible to augment the GREML fitfunction with an arbitrary scalar-valued function, to be evaluated and added to the fitfunction value. This can be used to regularize model-fitting with a prior loglikelihood. The GREML fitfunction can also use analytic derivatives of the augmentation function.
• When using SLSQP to optimize the maximum-likelihood fitfunction, it is now possible to use multithreading to evaluate the elements of the numerical gradient in parallel, by using argument rowwiseParallel=FALSE to mxFitFunctionML(). By default, in an analysis of raw data, mxFitFunctionML() parallelizes evaluation of the row likelihoods, and not the evaluation of the gradient elements.
• There is a new mxOption, "Parallel diagnostics", which can be turned on (set to "Yes") to make OpenMx provide diagnostic messages about the use of multiple threads.
• Function mxRestore() now behaves correctly with argument strict=TRUE.
• A subtle bug in the GREML fitfunction has been repaired. Under certain circumstances, this bug caused analytic derivatives of the covariance matrix to fail to be recalculated after changes in the values of the free parameters upon which they depend.
• Functions `dnbinom()`, `pnbinom()`, `dpois()`, and `ppois()` (from the `stats` package) are now usable in MxAlgebras.