What happens if you run `modelE` with `mxRun()` instead? Also, what are the free parameter values of `modelE` before it is run (you can just apply `summary()` or `coef()` to `modelE`)?
You could also try doing incomplete "test runs" of `modelE`, for example:
fitE <- mxRun( modelE, onlyFrontEnd=T )
fitE <- mxRun( modelE, useOptimizer=F )
Thanks for the answer. I applied coef () and summary () to model E. The findings are as follows and I tried using Mxrun as you mentioned, but I got an error like this. What do I have to do in this situation?
Thank you so much for your time and help!
> fitE <- mxRun( modelE, onlyFrontEnd=T )
Error: mxRun does not accept ... arguments. The first parameter in ... was named 'onlyFrontEnd' with value 'TRUE'
> fitE <- mxRun( modelE, useOptimizer=F )
Running twoEvj with 6 parameters
Error: The job for model 'twoEvj' exited abnormally with the error message: fit is not finite (Ordinal covariance is not positive definite in data 'DZ.data' row 64 (loc1))
The model does not satisfy the first-order optimality conditions to the required accuracy, and no improved point for the merit function could be found during the final linesearch (Mx status RED)
Your ordinal model may converge if you reduce mvnRelEps
try: model <- mxOption(model, 'mvnRelEps', mxOption(model, 'mvnRelEps')/5)
free parameters:
name matrix row col Estimate lbound ubound
1 meanno_2_FA meanG 1 1 0.51030080789
2 t1thno_2_FA thinG 1 1 0.98087891180 -3
3 t1thcannabis_dummy thinG 1 2 0.95914007926 -3
4 VA11 VA 1 1 0.00061060707
5 VA21 VA 1 2 0.00785928083
6 VA22 VA 2 2 1.76357137423
7 VE11 VE 1 1 0.00245181741
8 VE21 VE 1 2 -0.00282031911
9 VE22 VE 2 2 -0.76357137423
confidence intervals:
lbound estimate ubound note
twoAEvj.US[1,7] -0.079641611 0.19938682 0.44356784
twoAEvj.US[1,9] 0.000000000 0.00000000 0.00000000 !!!
twoAEvj.US[1,11] 0.556432214 0.80061318 1.07964766
twoAEvj.US[2,7] NA 1.55970243 NA !!!
twoAEvj.US[2,9] 0.000000000 0.00000000 0.00000000 !!!
twoAEvj.US[2,11] NA -0.55970243 NA !!!
twoAEvj.US[2,8] 1.378359239 1.76357155 1.93985753
twoAEvj.US[2,10] 0.000000000 0.00000000 0.00000000 !!!
twoAEvj.US[2,12] -0.939845346 -0.76357155 -0.37836482
To investigate missing CIs, run summary() again, with verbose=T, to see CI details.
Model Statistics:
| Parameters | Degrees of Freedom | Fit (-2lnL units)
Model: 9 407 -404.64357
Saturated: NA NA NA
Independence: NA NA NA
Number of observations/statistics: 105/416
> fitE <- mxRun( modelE, onlyFrontEnd=T )
Error: mxRun does not accept ... arguments. The first parameter in ... was named 'onlyFrontEnd' with value 'TRUE'
My mistake. The argument is `onlyFrontend`, not `onlyFrontEnd`.
Based on what I see in your post, try setting 'VE21' to zero before running the E-only model.
Thanks, it worked. Another question I have is this: I am trying to add another continuous variable to the same model. In this case, I will have two continuous and an ordinal variable. I could not find a script with multiple variables containing both ordinal and continuous variables.
Can I modify it to this model? If possible, should the oc<- (0,0,1) and SvMe <- (0, 10,15) variables be 3 digits like this?
I'm just trying to learn, sorry for so many questions.
more info?
You could also try doing incomplete "test runs" of `modelE`, for example:
fitE <- mxRun( modelE, onlyFrontEnd=T )
fitE <- mxRun( modelE, useOptimizer=F )
Log in or register to post comments
In reply to more info? by AdminRobK
Thanks for the answer. I
Thank you so much for your time and help!
> fitE <- mxRun( modelE, onlyFrontEnd=T )
Error: mxRun does not accept ... arguments. The first parameter in ... was named 'onlyFrontEnd' with value 'TRUE'
> fitE <- mxRun( modelE, useOptimizer=F )
Running twoEvj with 6 parameters
Error: The job for model 'twoEvj' exited abnormally with the error message: fit is not finite (Ordinal covariance is not positive definite in data 'DZ.data' row 64 (loc1))
> coef(modelE)
meanno_2_FA t1thno_2_FA t1thcannabis_dummy VA11 VA21 VA22
0.51030080789 0.98087891180 0.95914007926 0.00061060707 0.00785928083 1.76357137423
VE11 VE21 VE22
0.00245181741 -0.00282031911 -0.76357137423
Summary of twoEvj
The model does not satisfy the first-order optimality conditions to the required accuracy, and no improved point for the merit function could be found during the final linesearch (Mx status RED)
Your ordinal model may converge if you reduce mvnRelEps
try: model <- mxOption(model, 'mvnRelEps', mxOption(model, 'mvnRelEps')/5)
free parameters:
name matrix row col Estimate lbound ubound
1 meanno_2_FA meanG 1 1 0.51030080789
2 t1thno_2_FA thinG 1 1 0.98087891180 -3
3 t1thcannabis_dummy thinG 1 2 0.95914007926 -3
4 VA11 VA 1 1 0.00061060707
5 VA21 VA 1 2 0.00785928083
6 VA22 VA 2 2 1.76357137423
7 VE11 VE 1 1 0.00245181741
8 VE21 VE 1 2 -0.00282031911
9 VE22 VE 2 2 -0.76357137423
confidence intervals:
lbound estimate ubound note
twoAEvj.US[1,7] -0.079641611 0.19938682 0.44356784
twoAEvj.US[1,9] 0.000000000 0.00000000 0.00000000 !!!
twoAEvj.US[1,11] 0.556432214 0.80061318 1.07964766
twoAEvj.US[2,7] NA 1.55970243 NA !!!
twoAEvj.US[2,9] 0.000000000 0.00000000 0.00000000 !!!
twoAEvj.US[2,11] NA -0.55970243 NA !!!
twoAEvj.US[2,8] 1.378359239 1.76357155 1.93985753
twoAEvj.US[2,10] 0.000000000 0.00000000 0.00000000 !!!
twoAEvj.US[2,12] -0.939845346 -0.76357155 -0.37836482
To investigate missing CIs, run summary() again, with verbose=T, to see CI details.
Model Statistics:
| Parameters | Degrees of Freedom | Fit (-2lnL units)
Model: 9 407 -404.64357
Saturated: NA NA NA
Independence: NA NA NA
Number of observations/statistics: 105/416
Constraint 'Var1' contributes 1 observed statistic.
Information Criteria:
| df Penalty | Parameters Penalty | Sample-Size Adjusted
AIC: -1218.6436 -386.64357 -384.74884
BIC: -2298.8054 -362.75793 -391.19060
CFI: NA
TLI: 1 (also known as NNFI)
RMSEA: 0 [95% CI (NA, NA)]
Prob(RMSEA <= 0.05): NA
Log in or register to post comments
In reply to Thanks for the answer. I by handeezgia
some suggestions
My mistake. The argument is `onlyFrontend`, not `onlyFrontEnd`.
Based on what I see in your post, try setting 'VE21' to zero before running the E-only model.
Log in or register to post comments
In reply to some suggestions by AdminRobK
another question
Can I modify it to this model? If possible, should the oc<- (0,0,1) and SvMe <- (0, 10,15) variables be 3 digits like this?
I'm just trying to learn, sorry for so many questions.
Thank you in advance!
Log in or register to post comments