error with loading data into umxEFA

[mariahkahwaji]

Hi! I am having an issue with getting this error everytime I run umxEFA:
Error in h(simpleError(msg, call)) :
error in evaluating the argument 'x' in selecting a method for function 'print': argument is not a matrix

I have tried the data as both a dataframe and a matrix. I also tried running without any specifications (i.e., just putting in the data frame, number of factors, and rotation). A portion of the output runs but it stops as soon as it gets to the factor correlation matrix output. I would like to know if I need to format the dataframe a certain way before using it or if there is some other troubleshooting I can do?

[tbates]

hi @mariahkahwaji,
In the first instance can you upgrade to the newest version of umx/R just to be on a level playing field. Then post here

1. The result of traceback() run immediately after this failed run of umxEFA along with
2. The result from umxVersion()

[mcneale]

I am wondering if the raw data were provided as data, or a covariance/correlation matrix? It seems to me that the function desires raw data, not a correlation matrix, and that if a correlation matrix was entered, the sample size would seem to be the same as the number of variables, which could generate errors. Speculation here because OP didn't include their script for inspection/debugging.

[mariahkahwaji]

Thank you for the feedback. This is what was produced. myVars is a dataframe of raw data.

traceback()
7: h(simpleError(msg, call))
6: .handleSimpleError(function (cond)
.Internal(C_tryCatchHelper(addr, 1L, cond)), "argument is not a matrix",
base::quote(t.default(rm)))
5: t.default(rm)
4: t(rm)
3: solve(t(rm) %% rm)
2: print(solve(t(rm) %% rm))
1: umxEFA(x = myVars, factors = 2, scores = "Regression", minManifests = NA,
rotation = "geominQ", return = "model", report = "html",
summary = TRUE, name = "efa", digits = 2)
umxVersion()
umx version: 4.16.0
OpenMx version: 2.21.10 [GIT v2.21.10]
R version: R version 4.3.2 (2023-10-31)
Platform: x86_64-pc-linux-gnu
Default optimizer: SLSQP
NPSOL-enabled?: No
OpenMP-enabled?: Yes
You can update OpenMx with:
install.OpenMx(c("NPSOL", "travis", "CRAN", "open travis build page")

[tbates]

hi,
If you email your df "myVars" to me (tim.bates@ed.ac.uk), I'll reproduce and diagnose.
Looks like the factor correlation matrix of the rotation is invalid.

Also, I've not heard or used geominQ rotation. What happens with something like oblimin instead?