From b8e7aab8f12d9fda37054c2e649bbc975733b4c2 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Florian=20G=C3=A4rber?= <florian.gaerber@uni-hamburg.de>
Date: Thu, 20 Jul 2023 13:07:09 +0200
Subject: [PATCH] Fix failing `\donttest` examples in `devtools::check()`
---
R/addSurrogates.R | 6 +++---
R/build_clusters.R | 20 +++++++++++++++++---
R/reduce.surrogates.R | 22 +++++++++++++++++++---
R/var.relations.R | 10 +++++++++-
R/var.relations.mfi.R | 9 ++++++++-
R/variable_selection_md.R | 2 +-
R/variable_selection_mir.R | 2 +-
R/variable_selection_smd.R | 2 +-
man/build.clusters.Rd | 20 +++++++++++++++++---
man/reduce.surrogates.Rd | 22 +++++++++++++++++++---
man/var.relations.Rd | 10 +++++++++-
man/var.relations.mfi.Rd | 9 ++++++++-
man/var.select.md.Rd | 2 +-
man/var.select.mir.Rd | 2 +-
man/var.select.smd.Rd | 2 +-
15 files changed, 115 insertions(+), 25 deletions(-)
diff --git a/R/addSurrogates.R b/R/addSurrogates.R
index 40351af..ed3fafd 100644
--- a/R/addSurrogates.R
+++ b/R/addSurrogates.R
@@ -2,8 +2,6 @@
#'
#' This function adds surrogate variables and adjusted agreement values to a forest that was created by getTreeranger.
#'
-#' @useDynLib RFSurrogates, .registration = TRUE, .fixes = "C_"
-#'
#' @param RF random forest object created by ranger (with keep.inbag=TRUE).
#' @param trees list of trees created by getTreeranger.
#' @param s Predefined number of surrogate splits (it may happen that the actual number of surrogate splits differes in individual nodes). Default is 1 \% of no. of variables.
@@ -74,6 +72,8 @@ getSurrogate = function(surr.par, k = 1, maxsurr) {
#'
#' This is an internal function
#'
+#' @useDynLib RFSurrogates, .registration = TRUE
+#'
#' @keywords internal
SurrTree = function(j,wt,Xdata,controls,column.names,tree,maxsurr,ncat) {
node = tree[j,]
@@ -92,7 +92,7 @@ SurrTree = function(j,wt,Xdata,controls,column.names,tree,maxsurr,ncat) {
}
- surrogate.parameters = .Call(C_getSurrogates,
+ surrogate.parameters = .Call("getSurrogates",
ncat = as.integer(ncat),
wt = as.numeric(wt),
X = as.matrix(Xdata),
diff --git a/R/build_clusters.R b/R/build_clusters.R
index af66275..dcf428b 100644
--- a/R/build_clusters.R
+++ b/R/build_clusters.R
@@ -19,13 +19,27 @@
#' allvariables = colnames(x)# extract variables names
#' nvar = length(allvariables) # count number of variables
#' set.seed(42)
-#' RF = ranger::ranger(data = SMD_example_data, dependent.variable.name = "y", num.trees = 10, keep.inbag = TRUE,mtry = floor(nvar^(3/4)), min.node.size = 1)
+#' RF = ranger::ranger(
+#' data = SMD_example_data,
+#' dependent.variable.name = "y",
+#' num.trees = 10,
+#' keep.inbag = TRUE,
+#' mtry = floor(nvar^(3/4)),
+#' min.node.size = 1,
+#' num.threads = 1)
#' trees = getTreeranger(RF = RF, num.trees = 10)
#' trees.lay = addLayer(trees)
-#' trees.surr = addSurrogates(RF = RF, trees = trees.lay, s = 10, Xdata = x, num.threads = NULL)
+#' trees.surr = addSurrogates(RF = RF, trees = trees.lay, s = 10, Xdata = x, num.threads = 1)
#'
#' # investigate variable relations
-#' rel=var.relations(forest = list(trees = trees.surr, allvariables = allvariables), variables = allvariables, candidates = allvariables, t = 10)
+#' rel=var.relations(
+#' x = data.frame(),
+#' create.forest = FALSE,
+#' forest = list(trees = trees.surr, allvariables = allvariables),
+#' variables = allvariables,
+#' candidates = allvariables,
+#' t = 10,
+#' num.threads = 1)
#' groups = build.clusters(rel)
#' }
#'
diff --git a/R/reduce.surrogates.R b/R/reduce.surrogates.R
index a5745eb..18abf0f 100644
--- a/R/reduce.surrogates.R
+++ b/R/reduce.surrogates.R
@@ -16,15 +16,31 @@
#' ###### use result of SMD variable importance and reduce surrogate variables to 10
#' # select variables with smd variable importance (usually more trees are needed)
#' set.seed(42)
-#' res = var.select.smd(x = SMD_example_data[,2:ncol(SMD_example_data)], y = SMD_example_data[,1], s = 100, num.trees = 10)
+#' res = var.select.smd(
+#' x = as.data.frame(SMD_example_data[,2:ncol(SMD_example_data)]),
+#' y = SMD_example_data[,1],
+#' s = 100,
+#' num.trees = 10,
+#' num.threads = 1)
#' forest.new = reduce.surrogates(forest = res$forest, s = 10)
#'
#' # execute SMD on tree with reduced number of surrogates
-#' res.new = var.select.smd(create.forest = FALSE, forest = forest.new)
+#' res.new = var.select.smd(
+#' x = data.frame(),
+#' create.forest = FALSE,
+#' forest = forest.new,
+#' num.threads = 1)
#' res.new$var
#'
#' #' # investigate variable relations
-#' rel = var.relations(forest = forest.new, variables=c("X1","X7"), candidates = res$forest[["allvariables"]][1:100], t = 5)
+#' rel = var.relations(
+#' x = data.frame(),
+#' create.forest = FALSE,
+#' forest = forest.new,
+#' variables=c("X1","X7"),
+#' candidates = res$forest[["allvariables"]][1:100],
+#' t = 5,
+#' num.threads = 1)
#' rel$var
#'}
#' @export
diff --git a/R/var.relations.R b/R/var.relations.R
index 75fee16..9edc76a 100644
--- a/R/var.relations.R
+++ b/R/var.relations.R
@@ -28,7 +28,15 @@
#' \donttest{
#' # calculate variable relations
#' set.seed(42)
-#' res = var.relations(x = x, y = y, s = 10, num.trees = 100, variables = c("X1","X7"), candidates = colnames(x)[1:100], t = 5)
+#' res = var.relations(
+#' x = x,
+#' y = y,
+#' s = 10,
+#' num.trees = 10,
+#' variables = c("X1","X7"),
+#' candidates = colnames(x)[1:100],
+#' t = 5,
+#' num.threads = 1)
#' res$var
#' }
#'
diff --git a/R/var.relations.mfi.R b/R/var.relations.mfi.R
index bab78e8..563fb23 100644
--- a/R/var.relations.mfi.R
+++ b/R/var.relations.mfi.R
@@ -31,7 +31,14 @@
#' \donttest{
#' # calculate variable relations
#' set.seed(42)
-#' res = var.relations.mfi(x = x, y = y, s = 10, num.trees = 100, variables = c("X1","X7"), candidates = colnames(x)[1:100])
+#' res = var.relations.mfi(
+#' x = x,
+#' y = y,
+#' s = 10,
+#' num.trees = 10,
+#' variables = c("X1","X7"),
+#' candidates = colnames(x)[1:100],
+#' num.threads = 1)
#' res$var.rel[[1]]
#' }
#'
diff --git a/R/variable_selection_md.R b/R/variable_selection_md.R
index 42ec1fd..eb2aa7b 100644
--- a/R/variable_selection_md.R
+++ b/R/variable_selection_md.R
@@ -44,7 +44,7 @@
#' \donttest{
#' # select variables (usually more trees are needed)
#' set.seed(42)
-#' res = var.select.md(x = SMD_example_data[,2:ncol(SMD_example_data)], y = SMD_example_data[,1], num.trees = 10)
+#' res = var.select.md(x = SMD_example_data[,2:ncol(SMD_example_data)], y = SMD_example_data[,1], num.trees = 10, num.threads = 1)
#' res$var
#' }
#'
diff --git a/R/variable_selection_mir.R b/R/variable_selection_mir.R
index 93d65ae..e053bc8 100644
--- a/R/variable_selection_mir.R
+++ b/R/variable_selection_mir.R
@@ -50,7 +50,7 @@
#' \donttest{
#' # select variables (usually more trees are needed)
#' set.seed(42)
-#' res = var.select.mir(x = SMD_example_data[,2:ncol(SMD_example_data)], y = SMD_example_data[,1],s = 10, num.trees = 10)
+#' res = var.select.mir(x = SMD_example_data[,2:ncol(SMD_example_data)], y = SMD_example_data[,1],s = 10, num.trees = 10, num.threads = 1)
#' res$var
#' }
#'@references
diff --git a/R/variable_selection_smd.R b/R/variable_selection_smd.R
index e8c3ad4..179f9d9 100644
--- a/R/variable_selection_smd.R
+++ b/R/variable_selection_smd.R
@@ -49,7 +49,7 @@
#' \donttest{
#' # select variables (usually more trees are needed)
#' set.seed(42)
-#' res = var.select.smd(x = SMD_example_data[,2:ncol(SMD_example_data)], y = SMD_example_data[,1],s = 10, num.trees = 10)
+#' res = var.select.smd(x = SMD_example_data[,2:ncol(SMD_example_data)], y = SMD_example_data[,1],s = 10, num.trees = 10, num.threads = 1)
#' res$var
#' }
#'@references
diff --git a/man/build.clusters.Rd b/man/build.clusters.Rd
index 965c0c1..206ca09 100644
--- a/man/build.clusters.Rd
+++ b/man/build.clusters.Rd
@@ -29,13 +29,27 @@ data("SMD_example_data")
allvariables = colnames(x)# extract variables names
nvar = length(allvariables) # count number of variables
set.seed(42)
- RF = ranger::ranger(data = SMD_example_data, dependent.variable.name = "y", num.trees = 10, keep.inbag = TRUE,mtry = floor(nvar^(3/4)), min.node.size = 1)
+ RF = ranger::ranger(
+ data = SMD_example_data,
+ dependent.variable.name = "y",
+ num.trees = 10,
+ keep.inbag = TRUE,
+ mtry = floor(nvar^(3/4)),
+ min.node.size = 1,
+ num.threads = 1)
trees = getTreeranger(RF = RF, num.trees = 10)
trees.lay = addLayer(trees)
- trees.surr = addSurrogates(RF = RF, trees = trees.lay, s = 10, Xdata = x, num.threads = NULL)
+ trees.surr = addSurrogates(RF = RF, trees = trees.lay, s = 10, Xdata = x, num.threads = 1)
# investigate variable relations
- rel=var.relations(forest = list(trees = trees.surr, allvariables = allvariables), variables = allvariables, candidates = allvariables, t = 10)
+ rel=var.relations(
+ x = data.frame(),
+ create.forest = FALSE,
+ forest = list(trees = trees.surr, allvariables = allvariables),
+ variables = allvariables,
+ candidates = allvariables,
+ t = 10,
+ num.threads = 1)
groups = build.clusters(rel)
}
diff --git a/man/reduce.surrogates.Rd b/man/reduce.surrogates.Rd
index b21ab58..ff5f4b7 100644
--- a/man/reduce.surrogates.Rd
+++ b/man/reduce.surrogates.Rd
@@ -25,15 +25,31 @@ data("SMD_example_data")
###### use result of SMD variable importance and reduce surrogate variables to 10
# select variables with smd variable importance (usually more trees are needed)
set.seed(42)
-res = var.select.smd(x = SMD_example_data[,2:ncol(SMD_example_data)], y = SMD_example_data[,1], s = 100, num.trees = 10)
+res = var.select.smd(
+ x = as.data.frame(SMD_example_data[,2:ncol(SMD_example_data)]),
+ y = SMD_example_data[,1],
+ s = 100,
+ num.trees = 10,
+ num.threads = 1)
forest.new = reduce.surrogates(forest = res$forest, s = 10)
# execute SMD on tree with reduced number of surrogates
-res.new = var.select.smd(create.forest = FALSE, forest = forest.new)
+res.new = var.select.smd(
+ x = data.frame(),
+ create.forest = FALSE,
+ forest = forest.new,
+ num.threads = 1)
res.new$var
#' # investigate variable relations
-rel = var.relations(forest = forest.new, variables=c("X1","X7"), candidates = res$forest[["allvariables"]][1:100], t = 5)
+rel = var.relations(
+ x = data.frame(),
+ create.forest = FALSE,
+ forest = forest.new,
+ variables=c("X1","X7"),
+ candidates = res$forest[["allvariables"]][1:100],
+ t = 5,
+ num.threads = 1)
rel$var
}
}
diff --git a/man/var.relations.Rd b/man/var.relations.Rd
index f056eae..c34efdf 100644
--- a/man/var.relations.Rd
+++ b/man/var.relations.Rd
@@ -88,7 +88,15 @@ y = SMD_example_data[,1]
\donttest{
# calculate variable relations
set.seed(42)
-res = var.relations(x = x, y = y, s = 10, num.trees = 100, variables = c("X1","X7"), candidates = colnames(x)[1:100], t = 5)
+res = var.relations(
+ x = x,
+ y = y,
+ s = 10,
+ num.trees = 10,
+ variables = c("X1","X7"),
+ candidates = colnames(x)[1:100],
+ t = 5,
+ num.threads = 1)
res$var
}
diff --git a/man/var.relations.mfi.Rd b/man/var.relations.mfi.Rd
index 94cf036..82cc354 100644
--- a/man/var.relations.mfi.Rd
+++ b/man/var.relations.mfi.Rd
@@ -93,7 +93,14 @@ y = SMD_example_data[,1]
\donttest{
# calculate variable relations
set.seed(42)
-res = var.relations.mfi(x = x, y = y, s = 10, num.trees = 100, variables = c("X1","X7"), candidates = colnames(x)[1:100])
+res = var.relations.mfi(
+ x = x,
+ y = y,
+ s = 10,
+ num.trees = 10,
+ variables = c("X1","X7"),
+ candidates = colnames(x)[1:100],
+ num.threads = 1)
res$var.rel[[1]]
}
diff --git a/man/var.select.md.Rd b/man/var.select.md.Rd
index f0dcc70..c0f9318 100644
--- a/man/var.select.md.Rd
+++ b/man/var.select.md.Rd
@@ -80,7 +80,7 @@ data("SMD_example_data")
\donttest{
# select variables (usually more trees are needed)
set.seed(42)
-res = var.select.md(x = SMD_example_data[,2:ncol(SMD_example_data)], y = SMD_example_data[,1], num.trees = 10)
+res = var.select.md(x = SMD_example_data[,2:ncol(SMD_example_data)], y = SMD_example_data[,1], num.trees = 10, num.threads = 1)
res$var
}
diff --git a/man/var.select.mir.Rd b/man/var.select.mir.Rd
index 8932956..cde216d 100644
--- a/man/var.select.mir.Rd
+++ b/man/var.select.mir.Rd
@@ -103,7 +103,7 @@ data("SMD_example_data")
\donttest{
# select variables (usually more trees are needed)
set.seed(42)
-res = var.select.mir(x = SMD_example_data[,2:ncol(SMD_example_data)], y = SMD_example_data[,1],s = 10, num.trees = 10)
+res = var.select.mir(x = SMD_example_data[,2:ncol(SMD_example_data)], y = SMD_example_data[,1],s = 10, num.trees = 10, num.threads = 1)
res$var
}
}
diff --git a/man/var.select.smd.Rd b/man/var.select.smd.Rd
index 488d2a9..613b0b9 100644
--- a/man/var.select.smd.Rd
+++ b/man/var.select.smd.Rd
@@ -87,7 +87,7 @@ data("SMD_example_data")
\donttest{
# select variables (usually more trees are needed)
set.seed(42)
-res = var.select.smd(x = SMD_example_data[,2:ncol(SMD_example_data)], y = SMD_example_data[,1],s = 10, num.trees = 10)
+res = var.select.smd(x = SMD_example_data[,2:ncol(SMD_example_data)], y = SMD_example_data[,1],s = 10, num.trees = 10, num.threads = 1)
res$var
}
}
--
GitLab