xmu_update_covar for that pesky 9999 trick that saves your analyses…

[lf-araujo]

…from doom.

@jelman

[lf-araujo]

I added a experimental umxDoC2 it should work with the umxDoC example below. Note that I spent a lot of time trying to get umxSummary to work with this, I tried to keep matrices named as close as the original. But in the end, I just don't understand properly how umxSummary works.

The advantage of this version is that in my tests, it works and fits better than the current implementation.

 ```
 data(docData)
 docData = umx_scale_wide_twin_data(c(var1, var2), docData, sep= "_T")
 mzData  = subset(docData, zygosity %in% c("MZFF", "MZMM"))
 dzData  = subset(docData, zygosity %in% c("DZFF", "DZMM"))
 
 # =======================================
 # = 2. Define manifests for var 1 and 2 =
 # =======================================
 var1 = paste0("varA", 1:3)
 var2 = paste0("varB", 1:3)
 
 # =======================================================
 # = 3. Make the non-causal (Cholesky) and causal models =
 # =======================================================
 Chol = umxDoC2(var1= var1, var2= var2, mzData= mzData, dzData= dzData, causal= FALSE)
 # nb: DoC initially has causal paths fixed @0
 DoC  = umxDoC2(var1= var1, var2= var2, mzData= mzData, dzData= dzData, causal= TRUE)
 a2b   = umxModify(DoC, "a2b", free = TRUE, name = "a2b"); summary(a2b)
 b2a   = umxModify(DoC, "b2a", free = TRUE, name = "b2a"); summary(b2a)
 Recip = umxModify(DoC, c("a2b", "b2a"), free = TRUE, name = "Recip"); summary(Recip)
 
 # Compare fits
 umxCompare(Chol, c(a2b, b2a, Recip))
 
 # ==========================================
 # = Alternative call with data in one file =
 # ==========================================
 data(docData)
 docData = umx_scale_wide_twin_data(varsToScale=c("varA1", "varA2"), data = docData, sep= "_T")
 DoC = umxDoC2(var1= paste0("varA", 1:3), var2= paste0("varB", 1:2),  covar="varA3",             mzData= c("MZFF", "MZMM"), dzData= c("DZFF", "DZMM"), data = docData, verbose = TRUE)

[tbates]

You should make the new model a class, then you can make a custom umxSummary that will just work

#model = as(model, "MxModelDoC")

[lf-araujo]

Thanks, That is what I was trying to do, the following fails with an error in the dimension of selDVs apparently?

Chol = as(Chol, "MxModelDoC") umxSummary(Chol)

Summary support for DoC models not complete yet. Feedback welcome at http://github.com/tbates/umx/issues if you are using this. Chol -2 × log(Likelihood) = 45911.55 CIs not supported for DoC models yet Error in `colnames<-`(`*tmp*`, value = selDVs[1:nVar]) : attempt to set 'colnames' on an object with less than two dimensions L Sent from Mailspring (https://getmailspring.com/), the best free email app for work

…

On Mar 5 2025, at 9:38 pm, Tim Bates ***@***.***> wrote: tbates left a comment (tbates/umx#257) (#257 (comment)) You should make the new model a class, then you can make a custom umxSummary that will just work #model = as(model, "MxModelDoC") — Reply to this email directly, view it on GitHub (#257 (comment)), or unsubscribe (https://github.com/notifications/unsubscribe-auth/ACNLSLVVTDNATQM354DBCB32S6YLJAVCNFSM6AAAAABYNGLAAGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOMBSGYZTKMZWGU). You are receiving this because you authored the thread.

[tbates]

So colnames(means) = selDVs[1:nVar] is failing.
It worked for umxDoC but fails for umxDoC2, so you've changed the structure of the means matrix I guess?
Yeah, you now have
$matrices : 'FacLoad', 'unitM', 'Mean', 'as', 'cs', 'es', 'beta', 'I', 'psi_a', 'psi_c', and 'psi_e'
$algebras : 'VC', 'V', 'FacMean', 'ManMean', 'A', 'C', and 'E'

So Mean not means? or something?
Need to edit umxSummary.MxModelDoC for the new names. That will break umxDoC support, but it's pretty broken anyhow?

[tbates]

FYI, on CRAN you can't use any non-ascii characters lambda delta etc.
Also it's unsafe to rely on "F" being FALSE.

Had to go through a bunch of finding all these

[lf-araujo]

Oh, sorry Tim. Won't happen again! L Sent from Mailspring ***@***.***/0?redirect=https%3A%2F%2Fgetmailspring.com%2F&recipient=cmVwbHkrQUNOTFNMVVdESzVNQ1BITjJHT1pCWFdGMjdXRFRFVkJOSEhLWlJSNDZVQHJlcGx5LmdpdGh1Yi5jb20%3D), the best free email app for work

…

On Mar 8 2025, at 3:17 am, Tim Bates ***@***.***> wrote: tbates left a comment (tbates/umx#257) ***@***.***/1?redirect=https%3A%2F%2Fgithub.com%2Ftbates%2Fumx%2Fpull%2F257%23issuecomment-2708133977&recipient=cmVwbHkrQUNOTFNMVVdESzVNQ1BITjJHT1pCWFdGMjdXRFRFVkJOSEhLWlJSNDZVQHJlcGx5LmdpdGh1Yi5jb20%3D) FYI, on CRAN you can't use any non-ascii characters lambda delta etc. Also it's unsafe to rely on "F" being FALSE. Had to go through a bunch of finding all these — Reply to this email directly, view it on GitHub ***@***.***/2?redirect=https%3A%2F%2Fgithub.com%2Ftbates%2Fumx%2Fpull%2F257%23issuecomment-2708133977&recipient=cmVwbHkrQUNOTFNMVVdESzVNQ1BITjJHT1pCWFdGMjdXRFRFVkJOSEhLWlJSNDZVQHJlcGx5LmdpdGh1Yi5jb20%3D), or unsubscribe ***@***.***/3?redirect=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FACNLSLUO4YCS4EIPCIIB2LL2TKRSTAVCNFSM6AAAAABYNGLAAGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOMBYGEZTGOJXG4&recipient=cmVwbHkrQUNOTFNMVVdESzVNQ1BITjJHT1pCWFdGMjdXRFRFVkJOSEhLWlJSNDZVQHJlcGx5LmdpdGh1Yi5jb20%3D). You are receiving this because you authored the thread.