docs

R/check_outliers.R

@@ -205,6 +205,17 @@
 #'  observations located at `qnorm(1-0.025) * SD)` of the log-transformed
 #'  LOF distance. Requires the **dbscan** package.
 #'
+#' @section Methods for simulated residuals:
+#'
+#' The approach for detecting outliers based on simulated residuals differs
+#' from the traditional methods and may not be detecting outliers as expected.
+#' Literally, this approach compares observed to simulated values. However, we
+#' do not know the deviation of the observed data to the model expectation, and
+#' thus, the term "outlier" should be taken with a grain of salt. Basically, the
+#' comparison tests whether on observed data point is outside the simulated range.
+#' It is strongly recommended to read the related documentations in the **DHARMa**
+#' package, e.g. `?DHARMa::testOutliers`.
+#'
 #' @section Threshold specification:
 #'
 #' Default thresholds are currently specified as follows:

R/check_predictions.R

@@ -37,6 +37,8 @@
 #'
 #' @return A data frame of simulated responses and the original response vector.
 #'
+#' @seealso [`simulate_residuals()`] and [`check_residuals()`].
+#'
 #' @details An example how posterior predictive checks can also be used for model
 #'   comparison is Figure 6 from _Gabry et al. 2019, Figure 6_.
 #'

R/check_residuals.R

@@ -13,10 +13,13 @@
 #'
 #' @details Uniformity of residuals is checked using a Kolmogorov-Smirnov test.
 #' There is a `plot()` method to visualize the distribution of the residuals.
+#' The test fpr uniformity basically tests to which extent the observed values
+#' deviate from the model expectations (i.e. simulated values). In this sense,
+#' the `check_residuals()` function has similar goals like [`check_predictions()`].
 #'
 #' @inheritSection simulate_residuals Tests based on simulated residuals
 #'
-#' @seealso [`simulate_residuals()`]
+#' @seealso [`simulate_residuals()`] and [`check_predictions()`].
 #'
 #' @return The p-value of the test statistics.
 #'

R/simulate_residuals.R

@@ -13,7 +13,7 @@
 #' [`check_residuals()`]. The returned object is of class `DHARMa` and
 #' `performance_simres`.
 #'
-#' @seealso [`check_residuals()`]
+#' @seealso [`check_residuals()`] and [`check_predictions()`].
 #'
 #' @details This function is a small wrapper around [`DHARMa::simulateResiduals()`].
 #' It basically only sets `plot = FALSE` and adds an additional class attribute