Fix issue #154. (#265)

R/cv_varsel.R

@@ -108,7 +108,7 @@ cv_varsel.refmodel <- function(
   nclusters = NULL,
   ndraws_pred = 400,
   nclusters_pred = NULL,
-  cv_search = !inherits(object, "datafit"),
+  refit_prj = !inherits(object, "datafit"),
   nterms_max = NULL,
   penalty = NULL,
   verbose = TRUE,
@@ -126,13 +126,13 @@ cv_varsel.refmodel <- function(
   refmodel <- object
   ## resolve the arguments similar to varsel
   args <- parse_args_varsel(
-    refmodel = refmodel, method = method, cv_search = cv_search,
+    refmodel = refmodel, method = method, refit_prj = refit_prj,
     nterms_max = nterms_max, nclusters = nclusters,
     ndraws = ndraws, nclusters_pred = nclusters_pred, ndraws_pred = ndraws_pred,
     search_terms = search_terms
   )
   method <- args$method
-  cv_search <- args$cv_search
+  refit_prj <- args$refit_prj
   nterms_max <- args$nterms_max
   nclusters <- args$nclusters
   ndraws <- args$ndraws
@@ -154,7 +154,7 @@ cv_varsel.refmodel <- function(
     sel_cv <- loo_varsel(
       refmodel = refmodel, method = method, nterms_max = nterms_max,
       ndraws = ndraws, nclusters = nclusters, ndraws_pred = ndraws_pred,
-      nclusters_pred = nclusters_pred, cv_search = cv_search, penalty = penalty,
+      nclusters_pred = nclusters_pred, refit_prj = refit_prj, penalty = penalty,
       verbose = verbose, opt = opt, nloo = nloo,
       validate_search = validate_search, seed = seed,
       search_terms = search_terms
@@ -163,7 +163,7 @@ cv_varsel.refmodel <- function(
     sel_cv <- kfold_varsel(
       refmodel = refmodel, method = method, nterms_max = nterms_max,
       ndraws = ndraws, nclusters = nclusters, ndraws_pred = ndraws_pred,
-      nclusters_pred = nclusters_pred, cv_search = cv_search, penalty = penalty,
+      nclusters_pred = nclusters_pred, refit_prj = refit_prj, penalty = penalty,
       verbose = verbose, opt = opt, K = K, seed = seed,
       search_terms = search_terms
     )
@@ -179,7 +179,7 @@ cv_varsel.refmodel <- function(
     sel <- varsel(refmodel,
                   method = method, ndraws = ndraws, nclusters = nclusters,
                   ndraws_pred = ndraws_pred, nclusters_pred = nclusters_pred,
-                  cv_search = cv_search, nterms_max = nterms_max - 1,
+                  refit_prj = refit_prj, nterms_max = nterms_max - 1,
                   penalty = penalty, verbose = verbose,
                   lambda_min_ratio = lambda_min_ratio, nlambda = nlambda,
                   regul = regul, search_terms = search_terms, seed = seed)
@@ -287,7 +287,7 @@ parse_args_cv_varsel <- function(refmodel, cv_method, K) {
 }
 
 loo_varsel <- function(refmodel, method, nterms_max, ndraws,
-                       nclusters, ndraws_pred, nclusters_pred, cv_search,
+                       nclusters, ndraws_pred, nclusters_pred, refit_prj,
                        penalty, verbose, opt, nloo = NULL,
                        validate_search = TRUE, seed = NULL,
                        search_terms = NULL) {
@@ -379,7 +379,7 @@ loo_varsel <- function(refmodel, method, nterms_max, ndraws,
       search_path = search_path,
       nterms = c(0, seq_along(search_path$solution_terms)),
       p_ref = p_pred, refmodel = refmodel, regul = opt$regul,
-      cv_search = cv_search
+      refit_prj = refit_prj
     )
 
     if (verbose) {
@@ -459,7 +459,7 @@ loo_varsel <- function(refmodel, method, nterms_max, ndraws,
         search_path = search_path,
         nterms = c(0, seq_along(search_path$solution_terms)),
         p_ref = p_pred, refmodel = refmodel, regul = opt$regul,
-        cv_search = cv_search
+        refit_prj = refit_prj
       )
       summaries_sub <- .get_sub_summaries(
         submodels = submodels, test_points = c(i), refmodel = refmodel
@@ -512,7 +512,7 @@ loo_varsel <- function(refmodel, method, nterms_max, ndraws,
 
 kfold_varsel <- function(refmodel, method, nterms_max, ndraws,
                          nclusters, ndraws_pred, nclusters_pred,
-                         cv_search, penalty, verbose, opt, K, seed = NULL,
+                         refit_prj, penalty, verbose, opt, K, seed = NULL,
                          search_terms = NULL) {
   # Fetch the K reference model fits (or fit them now if not already done) and
   # create objects of class `refmodel` from them (and also store the `omitted`
@@ -571,7 +571,7 @@ kfold_varsel <- function(refmodel, method, nterms_max, ndraws,
 
   # Re-project along the solution path (or fetch the projections from the search
   # results) for each fold:
-  if (verbose && cv_search) {
+  if (verbose && refit_prj) {
     print("Computing projections..")
     pb <- utils::txtProgressBar(min = 0, max = K, style = 3, initial = 0)
   }
@@ -581,16 +581,16 @@ kfold_varsel <- function(refmodel, method, nterms_max, ndraws,
       search_path = search_path,
       nterms = c(0, seq_along(search_path$solution_terms)),
       p_ref = fold$p_pred, refmodel = fold$refmodel, regul = opt$regul,
-      cv_search = cv_search
+      refit_prj = refit_prj
     )
-    if (verbose && cv_search) {
+    if (verbose && refit_prj) {
       utils::setTxtProgressBar(pb, fold_index)
     }
     return(submodels)
   }
   submodels_cv <- mapply(get_submodels_cv, search_path_cv, seq_along(list_cv),
                          SIMPLIFY = FALSE)
-  if (verbose && cv_search) {
+  if (verbose && refit_prj) {
     close(pb)
   }
 

R/project.R

@@ -19,16 +19,16 @@
 #'   of predictor terms for the submodel onto which the projection will be
 #'   performed. Argument `nterms` is ignored in that case. For an `object` which
 #'   is not of class `vsel`, `solution_terms` must not be `NULL`.
-#' @param cv_search A single logical value indicating whether to fit the
+#' @param refit_prj A single logical value indicating whether to fit the
 #'   submodels (again) (`TRUE`) or to retrieve the fitted submodels from
 #'   `object` (`FALSE`). For an `object` which is not of class `vsel`,
-#'   `cv_search` must be `TRUE`.
-#' @param ndraws Only relevant if `cv_search` is `TRUE`. Number of posterior
+#'   `refit_prj` must be `TRUE`.
+#' @param ndraws Only relevant if `refit_prj` is `TRUE`. Number of posterior
 #'   draws to be projected. **Caution:** For `ndraws <= 20`, the value of
 #'   `ndraws` is passed to `nclusters` (so that clustering is used). Ignored if
 #'   `nclusters` is not `NULL` or if the reference model is of class `datafit`
 #'   (in which case one cluster is used). See also section "Details" below.
-#' @param nclusters Only relevant if `cv_search` is `TRUE`. Number of clusters
+#' @param nclusters Only relevant if `refit_prj` is `TRUE`. Number of clusters
 #'   of posterior draws to be projected. Ignored if the reference model is of
 #'   class `datafit` (in which case one cluster is used). For the meaning of
 #'   `NULL`, see argument `ndraws`. See also section "Details" below.
@@ -103,7 +103,7 @@
 #'
 #' @export
 project <- function(object, nterms = NULL, solution_terms = NULL,
-                    cv_search = TRUE, ndraws = 400, nclusters = NULL,
+                    refit_prj = TRUE, ndraws = 400, nclusters = NULL,
                     seed = NULL, regul = 1e-4, ...) {
   if (inherits(object, "datafit")) {
     stop("project() does not support an `object` of class \"datafit\".")
@@ -112,32 +112,32 @@ project <- function(object, nterms = NULL, solution_terms = NULL,
     stop("Please provide an `object` of class \"vsel\" or use argument ",
          "`solution_terms`.")
   }
-  if (!inherits(object, "vsel") && !cv_search) {
+  if (!inherits(object, "vsel") && !refit_prj) {
     stop("Please provide an `object` of class \"vsel\" or use ",
-         "`cv_search = TRUE`.")
+         "`refit_prj = TRUE`.")
   }
 
   refmodel <- get_refmodel(object, ...)
 
-  if (cv_search && inherits(refmodel, "datafit")) {
-    warning("Automatically setting `cv_search` to `FALSE` since the reference ",
+  if (refit_prj && inherits(refmodel, "datafit")) {
+    warning("Automatically setting `refit_prj` to `FALSE` since the reference ",
             "model is of class \"datafit\".")
-    cv_search <- FALSE
+    refit_prj <- FALSE
   }
 
-  if (!cv_search &&
+  if (!refit_prj &&
       !is.null(solution_terms) &&
       any(
         solution_terms(object)[seq_along(solution_terms)] != solution_terms
       )) {
     warning("The given `solution_terms` are not part of the solution path ",
-            "(from `solution_terms(object)`), so `cv_search` is automatically ",
+            "(from `solution_terms(object)`), so `refit_prj` is automatically ",
             "set to `TRUE`.")
-    cv_search <- TRUE
+    refit_prj <- TRUE
   }
 
-  if (!cv_search) {
-    warning("Currently, `cv_search = FALSE` requires some caution, see GitHub ",
+  if (!refit_prj) {
+    warning("Currently, `refit_prj = FALSE` requires some caution, see GitHub ",
             "issues #168 and #211.")
   }
 
@@ -224,7 +224,7 @@ project <- function(object, nterms = NULL, solution_terms = NULL,
       submodls = object$search_path$submodls
     ),
     nterms = nterms, p_ref = p_ref, refmodel = refmodel, regul = regul,
-    cv_search = cv_search
+    refit_prj = refit_prj
   )
 
   # Output:

R/projfun.R

@@ -35,8 +35,8 @@ project_submodel <- function(solution_terms, p_ref, refmodel, regul = 1e-4) {
 # sizes `nterms`. Returns a list of submodels (each processed by
 # .init_submodel()).
 .get_submodels <- function(search_path, nterms, p_ref, refmodel, regul,
-                           cv_search = FALSE) {
-  if (!cv_search) {
+                           refit_prj = FALSE) {
+  if (!refit_prj) {
     # In this case, simply fetch the already computed projections, so don't
     # project again.
     fetch_submodel <- function(nterms) {

R/refmodel.R

@@ -90,7 +90,7 @@
 #'     returned by [project()] in its output element `submodl` (which in turn is
 #'     the same as the return value of `div_minimizer`, except if [project()]
 #'     was used with an `object` of class `vsel` based on an L1 search as well
-#'     as with `cv_search = FALSE`).
+#'     as with `refit_prj = FALSE`).
 #'     + `newdata` accepts data for new observations (at least in the form of a
 #'     `data.frame`).
 #' * `div_minimizer` does not need to have a specific prototype, but it needs to

R/varsel.R

@@ -17,7 +17,7 @@
 #'   L1 search and `"forward"` for forward search. If `NULL`, then `"forward"`
 #'   is used if the reference model has multilevel or additive terms and `"L1"`
 #'   otherwise. See also section "Details" below.
-#' @param cv_search A single logical value indicating whether to fit the
+#' @param refit_prj A single logical value indicating whether to fit the
 #'   submodels along the solution path again (`TRUE`) or to retrieve their fits
 #'   from the search part (`FALSE`) before using those (re-)fits in the
 #'   evaluation part.
@@ -30,12 +30,12 @@
 #'   part. Ignored in case of L1 search (because L1 search always uses a single
 #'   cluster). For the meaning of `NULL`, see argument `ndraws`. See also
 #'   section "Details" below.
-#' @param ndraws_pred Only relevant if `cv_search` is `TRUE`. Number of
+#' @param ndraws_pred Only relevant if `refit_prj` is `TRUE`. Number of
 #'   posterior draws used in the evaluation part. **Caution:** For `ndraws_pred
 #'   <= 20`, the value of `ndraws_pred` is passed to `nclusters_pred` (so that
 #'   clustering is used). Ignored if `nclusters_pred` is not `NULL`. See also
 #'   section "Details" below.
-#' @param nclusters_pred Only relevant if `cv_search` is `TRUE`. Number of
+#' @param nclusters_pred Only relevant if `refit_prj` is `TRUE`. Number of
 #'   clusters of posterior draws used in the evaluation part. For the meaning of
 #'   `NULL`, see argument `ndraws_pred`. See also section "Details" below.
 #' @param nterms_max Maximum number of predictor terms until which the search is
@@ -140,7 +140,7 @@ varsel.default <- function(object, ...) {
 varsel.refmodel <- function(object, d_test = NULL, method = NULL,
                             ndraws = 20, nclusters = NULL, ndraws_pred = 400,
                             nclusters_pred = NULL,
-                            cv_search = !inherits(object, "datafit"),
+                            refit_prj = !inherits(object, "datafit"),
                             nterms_max = NULL, verbose = TRUE,
                             lambda_min_ratio = 1e-5, nlambda = 150,
                             thresh = 1e-6, regul = 1e-4, penalty = NULL,
@@ -149,13 +149,13 @@ varsel.refmodel <- function(object, d_test = NULL, method = NULL,
 
   ## fetch the default arguments or replace them by the user defined values
   args <- parse_args_varsel(
-    refmodel = refmodel, method = method, cv_search = cv_search,
+    refmodel = refmodel, method = method, refit_prj = refit_prj,
     nterms_max = nterms_max, nclusters = nclusters, ndraws = ndraws,
     nclusters_pred = nclusters_pred, ndraws_pred = ndraws_pred,
     search_terms = search_terms
   )
   method <- args$method
-  cv_search <- args$cv_search
+  refit_prj <- args$refit_prj
   nterms_max <- args$nterms_max
   nclusters <- args$nclusters
   ndraws <- args$ndraws
@@ -191,7 +191,7 @@ varsel.refmodel <- function(object, d_test = NULL, method = NULL,
   submodels <- .get_submodels(search_path = search_path,
                               nterms = c(0, seq_along(solution_terms)),
                               p_ref = p_pred, refmodel = refmodel,
-                              regul = regul, cv_search = cv_search)
+                              regul = regul, refit_prj = refit_prj)
   sub <- .get_sub_summaries(
     submodels = submodels, test_points = seq_along(refmodel$y),
     refmodel = refmodel
@@ -276,7 +276,7 @@ select <- function(method, p_sel, refmodel, nterms_max, penalty, verbose, opt,
   }
 }
 
-parse_args_varsel <- function(refmodel, method, cv_search, nterms_max,
+parse_args_varsel <- function(refmodel, method, refit_prj, nterms_max,
                               nclusters, ndraws, nclusters_pred, ndraws_pred,
                               search_terms) {
   ##
@@ -311,11 +311,11 @@ parse_args_varsel <- function(refmodel, method, cv_search, nterms_max,
     stop("Unknown search method")
   }
 
-  stopifnot(!is.null(cv_search))
-  if (cv_search && inherits(refmodel, "datafit")) {
-    warning("For an `object` of class \"datafit\", `cv_search` is ",
+  stopifnot(!is.null(refit_prj))
+  if (refit_prj && inherits(refmodel, "datafit")) {
+    warning("For an `object` of class \"datafit\", `refit_prj` is ",
             "automatically set to `FALSE`.")
-    cv_search <- FALSE
+    refit_prj <- FALSE
   }
 
   stopifnot(!is.null(ndraws))
@@ -353,7 +353,7 @@ parse_args_varsel <- function(refmodel, method, cv_search, nterms_max,
   nterms_max <- min(max_nv_possible, nterms_max + 1)
 
   return(nlist(
-    method, cv_search, nterms_max, nclusters, ndraws, nclusters_pred,
+    method, refit_prj, nterms_max, nclusters, ndraws, nclusters_pred,
     ndraws_pred, search_terms
   ))
 }