sample covariance matrix is (near) zero or negative.

[Omar Salim]

Hello

When the model is executed with Lavaan the following error appears in the covariance matrix and no results are shown

How can this problem be solved

> fit.bis <- sem(mode22.SPE, sample.cov = S, sample.mean =

+                    sample.mean, sample.th = sample.th, sample.nobs = sample.nobs,

+                WLS.V = WLS.V, NACOV = NACOV,,se="standard",test="standard", estimator ="ML")

Error in lavData(data = data, group = group, cluster = cluster, ov.names = OV.NAMES,  : 

  lavaan ERROR: at least one variance in the sample covariance matrix is (near) zero or negative.

[balal izanloo]

On Mon, 1 Feb 2021, 00:39 Omar Salim, <omarsal...@gmail.com> wrote:

Hello

When the model is executed with Lavaan the following error appears in the covariance matrix and no results are shown

How can this problem be solved

> fit.bis <- sem(mode22.SPE, sample.cov = S, sample.mean =

+                    sample.mean, sample.th = sample.th, sample.nobs = sample.nobs,

+                WLS.V = WLS.V, NACOV = NACOV,,se="standard",test="standard", estimator ="ML")

Look to the bule comma in the above command line

Error in lavData(data = data, group = group, cluster = cluster, ov.names = OV.NAMES,  : 

  lavaan ERROR: at least one variance in the sample covariance matrix is (near) zero or negative.

That error show that one of your variable in data has fixed value. Check yor data carfully.

 You have a covarince matrix as data input may be you have introduced it un correctly. 

--
You received this message because you are subscribed to the Google Groups "lavaan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lavaan+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/lavaan/2629cbbc-82b7-4283-acc5-6daebbfee841n%40googlegroups.com.

[balal izanloo]

On Mon, Feb 1, 2021 at 12:39 AM Omar Salim <omarsal...@gmail.com> wrote:

Hello

When the model is executed with Lavaan the following error appears in the covariance matrix and no results are shown

How can this problem be solved

> fit.bis <- sem(mode22.SPE, sample.cov = S, sample.mean =

+                    sample.mean, sample.th = sample.th, sample.nobs = sample.nobs,

+                WLS.V = WLS.V, NACOV = NACOV,,se="standard",test="standard", estimator ="ML")

 

look to the double comma in the above command line 

 

Error in lavData(data = data, group = group, cluster = cluster, ov.names = OV.NAMES,  : 

  lavaan ERROR: at least one variance in the sample covariance matrix is (near) zero or negative.

That error shows that at least one of your variables in data has fixed value. Check your data carefully. i think You have a covariance matrix as data input and so maybe you have introduced it incorrectly.

[Omar Salim]

Yes, my input is a covariance matrix computed from the array data. The values ​​are correct for input How do I solve the problem of this correlation close to zero Is there a way

To view this discussion on the web visit https://groups.google.com/d/msgid/lavaan/CAKpMYbKP2tdOhuOAyp%2B%3DgZT0hm3szQ8B46KRKy97QfnuPeCA-w%40mail.gmail.com.

[balal izanloo]

One of your variable has zero variance. the error is related to this problem. Check your data

To view this discussion on the web visit https://groups.google.com/d/msgid/lavaan/CAGePCmkupfpU74DN1d90uKottL9GnEctCyxqk0sb0zEZ_LnV_g%40mail.gmail.com.

[Omar Salim]

I used a robust correlation array for the input process, omit variables, but the same problem appears
The correlation matrix is
covfch

‫في الاثنين، 1 فبراير 2021 في 8:17 م تمت كتابة ما يلي بواسطة ‪balal izanloo‬‏ <‪b.ez...@gmail.com‬‏>:‬

To view this discussion on the web visit https://groups.google.com/d/msgid/lavaan/CAKpMYbJdXr6Dkgut-4bcyZEEWunbSm2eEi44_7tEf2%3Do2wTBFw%40mail.gmail.com.

[balal izanloo]

Salim,the problem is related to your data not to an estimation method that you have used for correlation matrix. to solve the problem, use raw data instead of correlation matrix as input, if the problem continues to exist paste all of your commands to the lavaan group along with data.

 

To view this discussion on the web visit https://groups.google.com/d/msgid/lavaan/CAGePCmnU6H7XH3%3DRG0eC_L%3DTp9NMdj81-pePXa6pw8mK3Lpp4w%40mail.gmail.com.

[Tyler Moore]

I'm getting this same error:

lavaan ERROR: at least one variance in the sample covariance matrix is (near) zero or negative.

and my command is:

mod <- cfa(model, sample.cov=xcor, sample.nobs=11017, std.lv = FALSE, estimator="ULS")

I checked the correlation matrix and the diagonal is indeed 1.0 for all variables, so why would it think one of them is near-zero?  The only thing I can think of is that lavaan is glitching out because the corr matrix is 10000-by-10000.  Any ideas?

Thanks!!

[Maxim]

I had the same issue. It seems tit was caused by my covariance matrix being a data.frame object, so after I converted it using as.matrix(), the error was gone. Perhaps a little bug in the lavaan.