new parameters to docs, test vignette output on package site

R/as_data_frame.R

@@ -15,6 +15,8 @@
 #' * `delta`\cr Time-varying regression coefficients.
 #' * `nu`\cr Group-level random effects.
 #' * `lambda`\cr Factor loadings.
+#' * `kappa`\cr Contribution of the latent factor loadings in the total
+#' variation.
 #' * `psi`\cr Latent factors.
 #' * `tau`\cr Standard deviations of the spline coefficients of `delta`.
 #' * `tau_alpha`\cr Standard deviations of the spline coefficients of
@@ -42,6 +44,8 @@
 #' in case of `nonzero_lambda = FALSE`, mean of these are used to flip the
 #' sign of `psi` to avoid multimodality due to sign-switching, but
 #' `omega_psi` variables are not modified.
+#' * `zeta`\cr Total variation of latent factors, i.e., the sum
+#' of `sigma_lambda` and `tau_psi`.
 #'
 #' @export
 #' @family output

R/as_data_table.R

@@ -639,21 +639,25 @@ as_data_table_lambda <- function(x, draws, n_draws, response, ...) {
 as_data_table_sigma_lambda <- function(draws, response, ...) {
  as_data_table_default("sigma_lambda", draws, response)
 }
+
 #' @describeIn as_data_table_default Data Table for a "tau_psi" Parameter
 #' @noRd
 as_data_table_tau_psi <- function(draws, response, ...) {
  as_data_table_default("tau_psi", draws, response)
 }
+
 #' @describeIn as_data_table_default Data Table for a "kappa" Parameter
 #' @noRd
 as_data_table_kappa <- function(draws, response, ...) {
  as_data_table_default("kappa", draws, response)
 }
+
 #' @describeIn as_data_table_default Data Table for a "zeta" Parameter
 #' @noRd
 as_data_table_zeta <- function(draws, response, ...) {
  as_data_table_default("zeta", draws, response)
 }
+
 #' @describeIn as_data_table_default Data Table for a "psi" Parameter
 #' @noRd
 as_data_table_psi <- function(x, draws, n_draws, response,

R/utilities.R

@@ -301,7 +301,7 @@ topological_order <- function(A) {
 #'
 #' @param message See [cli::cli_abort()].
 #' @param ... See [cli::cli_abort()].
-#' @param call See See [cli::cli_abort()].
+#' @param call See [cli::cli_abort()].
 #' @noRd
 stop_ <- function(message, ..., call = rlang::caller_env()) {
  cli::cli_abort(message, ..., .envir = parent.frame(), call = call)

vignettes/dynamite.Rmd

@@ -192,7 +192,7 @@ seatbelt_formula <-
 
 Next, we fit the model
 
-```{r seatbeltmodelfit, echo=TRUE, eval=TRUE, message=FALSE}
+```{r seatbeltmodelfit, echo=TRUE, eval=TRUE, message=FALSE, error=FALSE, warning=FALSE}
 fit <- dynamite(
  dformula = seatbelt_formula,
  data = seatbelt, time = "year", group = "state",
@@ -282,7 +282,7 @@ Here, the `aux()` function creates a deterministic transformation of \(p_t\) def
 
 ![(#fig:multichanneldag) A directed acyclic graph for the multichannel model with arrows corresponding to the assumed direct causal effects. A cross-section at times \(t\), \(t+1\), and \(t+2\) is shown. The vertices and edges corresponding to the deterministic tranformation \(\log(p_t)\) are omitted for clarity.](dag.png){width=80%}
 
-```{r multichannelfit, echo=TRUE, eval=TRUE, message=FALSE}
+```{r multichannelfit, echo=TRUE, eval=TRUE, message=FALSE, error=FALSE, warning=FALSE}
 multichannel_fit <- dynamite(
  dformula = multi_formula,
  data = multichannel_example, time = "time", group = "id",

vignettes/dynamite_custom.Rmd

@@ -56,7 +56,7 @@ to_vector(nu_raw) ~ student_t(df, 0, 1);
 ```
 and add `real<lower=2> df;` to the parameters block. Now we can call the `dynamite` with our new model code, using the `custom_stan_model` argument, which accepts either a `character` string containing the model code or a path to a `.stan` file.
 
-```{r}
+```{r customfit, message=FALSE, error=FALSE, warning=FALSE}
 fit <- dynamite(
  dformula = f, data = gaussian_example, time = "time", group = "id", 
  custom_stan_model = "custom_code.stan", chains = 1, refresh = 0
@@ -107,7 +107,7 @@ d <- get_data(fit)
 
 Like `get_code()`, this method is available for both `dynamiteformula` and `dynamitefit` objects. With the custom code and the corresponding input data, we can call the variational algorithm of `rstan`:
 
-```{r}
+```{r vbfit, message=FALSE, error=FALSE, warning=FALSE}
 model <- rstan::stan_model("custom_code.stan")
 fit_vb <- rstan::vb(model, data = d, iter = 1e5, refresh = 0)
 ```
@@ -123,7 +123,8 @@ Now the methods for `dynamitefit` objects such as `summary()` and `predict()` wo
 
 ```{r}
 summary(fit_vb_dynamite, types = "beta")
-
+```
+```{r}
 pp2 <- predict(
  fit_vb_dynamite, newdata = newdata, new_levels = "original", n_draws = 50
 )

vignettes/dynamite_simulation.Rmd

@@ -175,7 +175,7 @@ init <- list(
 
 We fit the model with these fixed values.
 
-```{r gaussianfit}
+```{r gaussianfit, message=FALSE, error=FALSE, warning=FALSE}
 gaussian_simulation_fit <- dynamite(
  dformula = f,
  data = d,