fixes CRAN note (Found the following Rd file(s) with Rd \link{} targe…

…ts missing package anchors)

R/CreateSpacetimeFolds.R

@@ -22,7 +22,7 @@
 #' should be similar in each fold (class="LUC").
 #' @note Standard k-fold cross-validation can lead to considerable misinterpretation in spatial-temporal modelling tasks. This function can be used to prepare a Leave-Location-Out, Leave-Time-Out or Leave-Location-and-Time-Out cross-validation as target-oriented validation strategies for spatial-temporal prediction tasks. See Meyer et al. (2018) for further information.
 #' @author Hanna Meyer
-#' @seealso \code{\link{trainControl}},\code{\link{ffs}}, \code{\link{nndm}}
+#' @seealso \code{\link[caret]{trainControl}},\code{\link{ffs}}, \code{\link{nndm}}
 #' @references
 #' Meyer, H., Reudenbach, C., Hengl, T., Katurji, M., Nauß, T. (2018): Improving performance of spatio-temporal machine learning models using forward feature selection and target-oriented validation. Environmental Modelling & Software 101: 1-9.
 #' @examples

R/bss.R

@@ -33,8 +33,8 @@
 #' Note that bss is very slow since all combinations of variables are tested.
 #' A more time efficient alternative is the forward feature selection (\code{\link{ffs}}).
 #' @author Hanna Meyer
-#' @seealso \code{\link{train}},\code{\link{ffs}},
-#' \code{\link{trainControl}},\code{\link{CreateSpacetimeFolds}},
+#' @seealso \code{\link[caret]{train}},\code{\link{ffs}},
+#' \code{\link[caret]{trainControl}},\code{\link{CreateSpacetimeFolds}},
 #' \code{\link{nndm}}
 #' @examples
 #' \dontrun{

R/errorProfiles.R

@@ -5,7 +5,7 @@
 #' @param locations Optional. sf object for the training data used in model. Only used if variable=="geodist". Note that they must be in the same order as model$trainingData.
 #' @param variable Character. Which dissimilarity or distance measure to use for the error metric. Current options are "DI" or "LPD"
 #' @param multiCV Logical. Re-run model fitting and validation with different CV strategies. See details.
-#' @param window.size Numeric. Size of the moving window. See \code{\link{rollapply}}.
+#' @param window.size Numeric. Size of the moving window. See \code{\link[zoo]{rollapply}}.
 #' @param calib Character. Function to model the DI/LPD~performance relationship. Currently lm and scam are supported
 #' @param length.out Numeric. Only used if multiCV=TRUE. Number of cross-validation folds. See details.
 #' @param method Character. Method used for distance calculation. Currently euclidean distance (L2) and Mahalanobis distance (MD) are implemented but only L2 is tested. Note that MD takes considerably longer. See ?aoa for further explanation

R/ffs.R

@@ -1,18 +1,18 @@
 #' Forward feature selection
 #' @description A simple forward feature selection algorithm
-#' @param predictors see \code{\link{train}}
-#' @param response see \code{\link{train}}
-#' @param method see \code{\link{train}}
-#' @param metric see \code{\link{train}}
-#' @param maximize see \code{\link{train}}
+#' @param predictors see \code{\link[caret]{train}}
+#' @param response see \code{\link[caret]{train}}
+#' @param method see \code{\link[caret]{train}}
+#' @param metric see \code{\link[caret]{train}}
+#' @param maximize see \code{\link[caret]{train}}
 #' @param globalval Logical. Should models be evaluated based on 'global' performance? See \code{\link{global_validation}}
 #' @param withinSE Logical Models are only selected if they are better than the
 #' currently best models Standard error
 #' @param minVar Numeric. Number of variables to combine for the first selection.
 #' See Details.
-#' @param trControl see \code{\link{train}}
-#' @param tuneLength see \code{\link{train}}
-#' @param tuneGrid see \code{\link{train}}
+#' @param trControl see \code{\link[caret]{train}}
+#' @param tuneLength see \code{\link[caret]{train}}
+#' @param tuneGrid see \code{\link[caret]{train}}
 #' @param seed A random number used for model training
 #' @param cores Numeric. If > 2, mclapply will be used. see \code{\link{mclapply}}
 #' @param verbose Logical. Should information about the progress be printed?
@@ -53,8 +53,8 @@
 #' See Meyer et al. (2018) and Meyer et al. (2019) for further details.
 #'
 #' @author Hanna Meyer
-#' @seealso \code{\link{train}},\code{\link{bss}},
-#' \code{\link{trainControl}},\code{\link{CreateSpacetimeFolds}},\code{\link{nndm}}
+#' @seealso \code{\link[caret]{train}},\code{\link{bss}},
+#' \code{\link[caret]{trainControl}},\code{\link{CreateSpacetimeFolds}},\code{\link{nndm}}
 #' @references
 #' \itemize{
 #' \item Gasch, C.K., Hengl, T., Gräler, B., Meyer, H., Magney, T., Brown, D.J. (2015): Spatio-temporal interpolation of soil water, temperature, and electrical conductivity in 3D+T: the Cook Agronomy Farm data set. Spatial Statistics 14: 70-90.

R/global_validation.R

@@ -1,7 +1,7 @@
 #' Evaluate 'global' cross-validation
 #' @description Calculate validation metric using all held back predictions at once
-#' @param model an object of class \code{\link{train}}
-#' @return regression (\code{\link{postResample}}) or classification  (\code{\link{confusionMatrix}}) statistics
+#' @param model an object of class \code{\link[caret]{train}}
+#' @return regression (\code{\link[caret]{postResample}}) or classification  (\code{\link[caret]{confusionMatrix}}) statistics
 #' @details Relevant when folds are not representative for the entire area of interest.
 #' In this case, metrics like R2 are not meaningful since it doesn't reflect the general ability of
 #' the model to explain the entire gradient of the response.

R/nndm.R

@@ -23,7 +23,7 @@
 #' (distances for G function construction during LOO CV), Gjstar (distances
 #' for modified G function during NNDM LOO CV), phi (landscape autocorrelation range).
 #' indx_train and indx_test can directly be used as "index" and "indexOut" in
-#' caret's \code{\link{trainControl}} function or used to initiate a custom validation strategy in mlr3.
+#' caret's \code{\link[caret]{trainControl}} function or used to initiate a custom validation strategy in mlr3.
 #'
 #' @details NNDM proposes a LOO CV scheme such that the nearest neighbour distance distribution function between the test and training data during the CV process is matched to the nearest neighbour
 #' distance distribution function between the prediction and training points. Details of the method can be found in Milà et al. (2022).