![]() ![]() The other TWO refers to the additional structures ( and ee). ![]() This is because the errors are independent between sites. The 2 1 2 after the model equation refers to TWO error structures, which are identity matrices of size number of observations (represented by the first value of the line). This creates results for each factor in both sites. The first one (!part 3) uses site interacting with all the terms of the model equation, including tree. The previous example uses two alternative ways to code the problem. Two sites single trait treated as multivariateĪ few explanations now. Finally, the use of !GP to enforce positive definite covariancematrices because, in this specific dataset, some of the parameters tended to be outside the parametric space (e.g. A few points to highlight: the numbers after the model equation are now 1 2 2, because there is still only ONE error structure that is the product of TWO matrices, but now there are TWO additional covariance structures to be defined (Trait.mom and Trait.fam). This overlay creates a unique factor, something that you can see in part 3, where the structure Trait.mom refers to mom and(dad). Because we are using a parental model ratherthan an individual model, we use mom and(dad) to overlay the design matrices and create a unique prediction for each parent, rather than one as male and one as male. In the case of the Controlled Pollinated trial we know both parents, so we can get an estimate of dominance using the fam term, which refers to the family code. There are several elements in common with the previous example, so I will mention the few new issues. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |