You can do a QR decomposition on anything and the columns of Q will be orthogonal. But this will no longer be independent normally distributed. All the location and scale information about the original distribution will be in R.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Latent orthogonal factors | 1 | 630 | January 25, 2022 | |
| Multivariate normal distribution | 4 | 2617 | March 29, 2019 | |
| LKJCholeskyCov random method | 2 | 479 | August 28, 2018 | |
| Weird posterior for correlation structure using LKJ distribution | 5 | 1404 | August 6, 2018 | |
| Cholesky decomposition and correlation among random effects | 20 | 5642 | March 11, 2019 |