RMKirkpatrick

In my opinion, no.

Note that lavaan's behavior when `cov.std=TRUE` is defined in terms of LISREL specification. It may not have a general-case equivalent under RAM specification.

@mcneale `mxStandardizeRAMpaths()` is incompatible with LISREL expectation.

Although not high-priority, implementing `mxStandardizeLISRELpaths()` would be useful. For such a function, imitating the functionality of `cov.std` would be trivial. However, it should default to `FALSE`, i.e. by default it...

This is a GenomicSEM issue.

@tbates Hmm, you're right. Good catch.

BTW, please don't refer to it as "GSEM". The name of the package in question is "GenomicSEM". "GSEM" is the name of [Beate St Pourcain's package](https://www.mpi.nl/people/st-pourcain-beate/tool).

As I said in an email today, I think four of the five fit statistics that GenomicSEM likes to report can be obtained from OpenMx if the "independence" and null...

> At first blush, setting an arbitrary sample size seems to me idiotic. IDK what is going on there. Seems to lack any statistical justification. That's done merely to give...

Let me copy-paste the pertinent part of an email I sent to the list on Friday, 5/3/19: > The instructions for using GenomicSEM with OpenMx, https://github.com/MichelNivard/GenomicSEM/wiki/6.-GenomicSEM-and-OpenMx , say to set...