Update v6_pedigree_model_fitting.Rmd

Author: smasongarrison

vignettes/v6_pedigree_model_fitting.Rmd

@@ -224,7 +224,7 @@ Before fitting a model, we need observed data. In practice, this would be measur
 
 We define "true" variance components and use the relatedness matrices to construct the population covariance matrix, then sample from it.
 
-```{r simulate-phenotype, eval = run_models}
+```{r simulate-phenotype, eval = has_mvtnorm}
 # True variance components (proportions of total variance)
 true_var <- list(
   ad2 = 0.50,  # additive genetic
@@ -250,13 +250,23 @@ V_true <- true_var$ad2 * add_matrix +
 set.seed(123)
 y <- mvtnorm::rmvnorm(1, sigma = V_true)
 
+
 # Create named variable labels (required by OpenMx)
 ytemp <- paste("S", rownames(add_matrix))
+```
+
+```{r show-phenotype}
+
+if (is.null(y)) {
+  y <- rep(NA, nrow(add_matrix))
+
+}
 
-cat("Simulated phenotype for", ncol(y), "individuals\n")
-cat("Mean:", round(mean(y), 3), " SD:", round(sd(y), 3), "\n")
 ```
 
+
+We simulated phenotypic data for`r ncol(y)` individuals, with a mean of `r round(mean(y), 3)` and a standard deviation of `r round(sd(y), 3)`. The variance in this simulated phenotype arises from the specified genetic and environmental components according to the covariance structure we defined.
+
 In practice, you would have data from multiple independently ascertained families. Here we simulate data from a single pedigree for simplicity, but the model-fitting functions support multiple pedigrees (shown in a later section).
 
 
@@ -295,6 +305,8 @@ vc_model <- buildPedigreeModelCovariance(
   Vam = FALSE,  # do not estimate A x Mt interaction
   Ver = TRUE    # estimate unique environment
 )
+vc_model
+
 summary(vc_model)
 ```
 
@@ -336,6 +348,7 @@ full_model <- buildPedigreeMx(
   vars = start_vars,
   group_models = list(family_group)
 )
+full_model$submodels 
 ```