|
|
The Cholesky decomposition model is a fully parametrised descriptive model without any restrictions on the structure of latent genetic and residual influences. This saturated model can be fit to the data through a decomposition of both the genetic variance and residual variance into as many latent factors A and E as there are observed phenotypes k. However, model coefficients can only be interpreted, if there is an inherent longitudinal order of measured phenotypes that is reflected in the order of phenotypes entered in the model.
|
|
|
|
|
|
Example of a quad-variate Cholesky decomposition model (k=4):
|
|
|
|
|
|
![model_quadvariate_path_diagram_Cholesky](uploads/c90895fc107efa170ca41fcbb4ca2548/model_quadvariate_path_diagram_Cholesky.png) |
|
|
\ No newline at end of file |