I have a 26*26 matrix for 10 studies with many missing values. It costed a few hours to run the random effect model of tssem1. When I summarized the results by "summary(random1)" , the error "Error in chol.default(V) : 'a' must have dims > 0" appereared. I also noiced that many correlications are presented as "0.1".

Could anyone help to fix the error? May I know that those "0.1"s in the correlaiton matrix are resulted from the many "NA"s in the input matrices? Thank you for your great help.

The data and code are attached.

Regrads

Annyu

There is almost no data in the dataset. It is impossible to make it work.

Moreover, the correlation matrices are not symmetric. The upper triangles are all zeros.

Mike