Tim Bates

any updates?

```R mxGetExpected(model, component = c('covariance', 'means', 'thresholds'), type = c('observed', 'latent', 'total'), defvar.row = 1, subname=model$name) ``` Moving type forward should help users find it. Hiding parameters seems seldom a...

Just a note, I was improving the example for mxGetExpected, and I see that the Rd talks about options for component called 'vector' and 'standVector' but these aren't given as...

agree - that should trump everything, in the end. Add new parameters to the end of the list.

so, change to Option 1 ```R mxGetExpected(model, component, defvar.row=1, subname=model$name, type=c('observed', 'latent', 'total')) ```

And offering up ref models for this bad-fitting supermodel gives CFI -0.83 (!) while again yielding perfect TLI and RMSEA... (and RMSEA CI and p are not computed...) ```R m1Sat...

Sorry: pasted output of the 4-parameter model (parameters for each group), not the 2-parameter model. The problem is that for the 2-parameter model (weight equated across the groups), `summary(m2)` reports...

yeah... so perhaps mxConstraint should fail, reporting the string as not a legal right hand side in a constraint algebra?

Seems like the `MxAlgebra.R` fn `algebraErrorChecking` (called in `mxConstraint`) should be catching that the whole rhs is a string? So this would be a requires to enhance the checking by...

current pretty printing is as follows. Not sure what (if anything) would be nicer? maybe close issue? ```R omxDefaultComputePlan() MxComputeSequence 'compute' $freeSet : NA independent : FALSE steps[[ 'GD' ]]...