From 7fa9f37e8647afe71111eaddd9d4fb0056fbfec1 Mon Sep 17 00:00:00 2001
From: topepo <mxkuhn@gmail.com>
Date: Sun, 18 Nov 2018 09:09:15 -0500
Subject: [PATCH 1/5] Changes requested by CRAN

Examples for `step_isomap` fail due to missing dependency:

```r
> im_estimates <- prep(im_trans, training = biomass_tr)
Error in chckpkg("RSpectra") :
  require 'RSpectra' package, install it using install.packages('RSpectra')
Calls: prep ... embed -> .local -> do.call -> <Anonymous> -> chckpkg
Execution halted
```
---
 DESCRIPTION        |  3 ++-
 R/ica.R            | 23 +++++++++++++----------
 R/isomap.R         | 26 ++++++++++++++------------
 R/kpca.R           | 26 ++++++++++++++------------
 R/nnmf.R           | 20 ++++++++++----------
 man/step_ica.Rd    | 15 +++++++++------
 man/step_isomap.Rd | 16 +++++++++-------
 man/step_kpca.Rd   | 16 +++++++++-------
 man/step_nnmf.Rd   | 20 ++++++++++----------
 9 files changed, 90 insertions(+), 75 deletions(-)

diff --git a/DESCRIPTION b/DESCRIPTION
index f3c2b44bf..9ea0a4495 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,6 +1,6 @@
 Package: recipes
 Title: Preprocessing Tools to Create Design Matrices
-Version: 0.1.3.9002
+Version: 0.1.4
 Authors@R: c(
     person("Max", "Kuhn", , "max@rstudio.com", c("aut", "cre")),
     person("Hadley", "Wickham", , "hadley@rstudio.com", "aut"),
@@ -48,6 +48,7 @@ Suggests:
     rmarkdown,
     rpart,
     rsample,
+    RSpectra,
     testthat
 License: GPL-2
 VignetteBuilder: knitr
diff --git a/R/ica.R b/R/ica.R
index 7d0ea89fa..2aeb41d15 100644
--- a/R/ica.R
+++ b/R/ica.R
@@ -25,9 +25,9 @@
 #' @param res The [fastICA::fastICA()] object is stored
 #'  here once this preprocessing step has be trained by
 #'  [prep.recipe()].
-#' @param num The number of components to retain (this will be 
-#'  deprecated in factor of `num_comp` in version 0.1.5). `num_comp` 
-#'  will override this option. 
+#' @param num The number of components to retain (this will be
+#'  deprecated in factor of `num_comp` in version 0.1.5). `num_comp`
+#'  will override this option.
 #' @param prefix A character string that will be the prefix to the
 #'  resulting new variables. See notes below.
 #' @return An updated version of `recipe` with the new step
@@ -82,14 +82,17 @@
 #' ica_trans <- step_center(rec,  V1, V2)
 #' ica_trans <- step_scale(ica_trans, V1, V2)
 #' ica_trans <- step_ica(ica_trans, V1, V2, num_comp = 2)
-#' # ica_estimates <- prep(ica_trans, training = tr)
-#' # ica_data <- bake(ica_estimates, te)
 #'
-#' # plot(te$V1, te$V2)
-#' # plot(ica_data$IC1, ica_data$IC2)
+#' if (require(dimRed) & require(fastICA)) {
+#'   ica_estimates <- prep(ica_trans, training = tr)
+#'   ica_data <- bake(ica_estimates, te)
 #'
-#' # tidy(ica_trans, number = 3)
-#' # tidy(ica_estimates, number = 3)
+#'   plot(te$V1, te$V2)
+#'   plot(ica_data$IC1, ica_data$IC2)
+#'
+#'   tidy(ica_trans, number = 3)
+#'   tidy(ica_estimates, number = 3)
+#' }
 #' @seealso [step_pca()] [step_kpca()]
 #'   [step_isomap()] [recipe()] [prep.recipe()]
 #'   [bake.recipe()]
@@ -108,7 +111,7 @@ step_ica <-
 
 
     recipes_pkg_check(c("dimRed", "fastICA"))
-    if (!is.null(num)) 
+    if (!is.null(num))
       message("The argument `num` is deprecated in factor of `num_comp`. ",
               "`num` will be removed in next version.", call. = FALSE)
     add_step(
diff --git a/R/isomap.R b/R/isomap.R
index f426d2d9a..486e5d9ed 100644
--- a/R/isomap.R
+++ b/R/isomap.R
@@ -20,9 +20,9 @@
 #'  used.
 #' @param neighbors The number of neighbors.
 #' @param options A list of options to [dimRed::Isomap()].
-#' @param num The number of isomap dimensions (this will be deprecated 
-#'  in factor of  `num_terms` in version 0.1.5). `num_terms` will 
-#'  override this option. 
+#' @param num The number of isomap dimensions (this will be deprecated
+#'  in factor of  `num_terms` in version 0.1.5). `num_terms` will
+#'  override this option.
 #' @param res The [dimRed::Isomap()] object is stored
 #'  here once this preprocessing step has be trained by
 #'  [prep.recipe()].
@@ -85,16 +85,18 @@
 #'               neighbors = 100,
 #'               num_terms = 2)
 #'
-#' # im_estimates <- prep(im_trans, training = biomass_tr)
+#' if (require(dimRed) & require(RSpectra)) {
+#'   im_estimates <- prep(im_trans, training = biomass_tr)
 #'
-#' # im_te <- bake(im_estimates, biomass_te)
+#'   im_te <- bake(im_estimates, biomass_te)
 #'
-#' # rng <- extendrange(c(im_te$Isomap1, im_te$Isomap2))
-#' # plot(im_te$Isomap1, im_te$Isomap2,
-#' #      xlim = rng, ylim = rng)
+#'   rng <- extendrange(c(im_te$Isomap1, im_te$Isomap2))
+#'   plot(im_te$Isomap1, im_te$Isomap2,
+#'        xlim = rng, ylim = rng)
 #'
-#' # tidy(im_trans, number = 4)
-#' # tidy(im_estimates, number = 4)
+#'   tidy(im_trans, number = 4)
+#'   tidy(im_estimates, number = 4)
+#' }
 #' }
 #' @seealso [step_pca()] [step_kpca()]
 #'   [step_ica()] [recipe()] [prep.recipe()]
@@ -115,7 +117,7 @@ step_isomap <-
            id = rand_id("isomap")) {
 
     recipes_pkg_check(c("dimRed", "RSpectra", "igraph", "RANN"))
-    if (!is.null(num)) 
+    if (!is.null(num))
       message("The argument `num` is deprecated in factor of `num_terms`. ",
               "`num` will be removed in next version.", call. = FALSE)
     add_step(
@@ -137,7 +139,7 @@ step_isomap <-
   }
 
 step_isomap_new <-
-  function(terms, role, trained, num_terms, neighbors, options, res, num, 
+  function(terms, role, trained, num_terms, neighbors, options, res, num,
            prefix, skip, id) {
     step(
       subclass = "isomap",
diff --git a/R/kpca.R b/R/kpca.R
index 949256c72..8db42c850 100644
--- a/R/kpca.R
+++ b/R/kpca.R
@@ -26,9 +26,9 @@
 #' @param res An S4 [kernlab::kpca()] object is stored
 #'  here once this preprocessing step has be trained by
 #'  [prep.recipe()].
-#' @param num The number of components to retain (this will be 
-#'  deprecated in factor of `num_comp` in version 0.1.5). `num_comp` 
-#'  will override this option. 
+#' @param num The number of components to retain (this will be
+#'  deprecated in factor of `num_comp` in version 0.1.5). `num_comp`
+#'  will override this option.
 #' @param prefix A character string that will be the prefix to the
 #'  resulting new variables. See notes below.
 #' @return An updated version of `recipe` with the new step
@@ -94,16 +94,18 @@
 #'   step_scale(all_predictors()) %>%
 #'   step_kpca(all_predictors())
 #'
-#' # kpca_estimates <- prep(kpca_trans, training = biomass_tr)
+#' if (require(dimRed) & require(kernlab)) {
+#'   kpca_estimates <- prep(kpca_trans, training = biomass_tr)
 #'
-#' # kpca_te <- bake(kpca_estimates, biomass_te)
+#'   kpca_te <- bake(kpca_estimates, biomass_te)
 #'
-#' # rng <- extendrange(c(kpca_te$kPC1, kpca_te$kPC2))
-#' # plot(kpca_te$kPC1, kpca_te$kPC2,
-#' #      xlim = rng, ylim = rng)
+#'   rng <- extendrange(c(kpca_te$kPC1, kpca_te$kPC2))
+#'   plot(kpca_te$kPC1, kpca_te$kPC2,
+#'        xlim = rng, ylim = rng)
 #'
-#' # tidy(kpca_trans, number = 4)
-#' # tidy(kpca_estimates, number = 4)
+#'   tidy(kpca_trans, number = 4)
+#'   tidy(kpca_estimates, number = 4)
+#' }
 #' @seealso [step_pca()] [step_ica()]
 #'   [step_isomap()] [recipe()] [prep.recipe()]
 #'   [bake.recipe()]
@@ -123,10 +125,10 @@ step_kpca <-
            id = rand_id("kpca")) {
 
     recipes_pkg_check(c("dimRed", "kernlab"))
-    if (!is.null(num)) 
+    if (!is.null(num))
       message("The argument `num` is deprecated in factor of `num_comp`. ",
               "`num` will be removed in next version.", call. = FALSE)
-    
+
     add_step(
       recipe,
       step_kpca_new(
diff --git a/R/nnmf.R b/R/nnmf.R
index 345de1dc5..cd96d07a9 100644
--- a/R/nnmf.R
+++ b/R/nnmf.R
@@ -52,20 +52,20 @@
 #'
 #' @examples
 #' \donttest{
-#' library(dimRed)
-#' library(NMF)
 #' data(biomass)
 #'
-#' rec <- recipe(HHV ~ ., data = biomass) %>%
-#'   update_role(sample, new_role = "id var") %>%
-#'   update_role(dataset, new_role = "split variable") %>%
-#'   step_nnmf(all_predictors(), num_comp = 2, seed = 473, num_run = 2) %>%
-#'   prep(training = biomass, retain = TRUE)
+#' if (require(dimRed) & require(NMF)) {
+#'   rec <- recipe(HHV ~ ., data = biomass) %>%
+#'     update_role(sample, new_role = "id var") %>%
+#'     update_role(dataset, new_role = "split variable") %>%
+#'     step_nnmf(all_predictors(), num_comp = 2, seed = 473, num_run = 2) %>%
+#'     prep(training = biomass, retain = TRUE)
 #'
-#' # juice(rec)
+#'   juice(rec)
 #'
-#' # library(ggplot2)
-#' # ggplot(juice(rec), aes(x = NNMF2, y = NNMF1, col = HHV)) + geom_point()
+#'   library(ggplot2)
+#'   ggplot(juice(rec), aes(x = NNMF2, y = NNMF1, col = HHV)) + geom_point()
+#' }
 #' }
 #' @seealso [step_pca()], [step_ica()], [step_kpca()],
 #'   [step_isomap()], [recipe()], [prep.recipe()],
diff --git a/man/step_ica.Rd b/man/step_ica.Rd
index 6b4192645..d9362e0a9 100644
--- a/man/step_ica.Rd
+++ b/man/step_ica.Rd
@@ -113,14 +113,17 @@ rec <- recipe( ~ ., data = tr)
 ica_trans <- step_center(rec,  V1, V2)
 ica_trans <- step_scale(ica_trans, V1, V2)
 ica_trans <- step_ica(ica_trans, V1, V2, num_comp = 2)
-# ica_estimates <- prep(ica_trans, training = tr)
-# ica_data <- bake(ica_estimates, te)
 
-# plot(te$V1, te$V2)
-# plot(ica_data$IC1, ica_data$IC2)
+if (require(dimRed) & require(fastICA)) {
+  ica_estimates <- prep(ica_trans, training = tr)
+  ica_data <- bake(ica_estimates, te)
 
-# tidy(ica_trans, number = 3)
-# tidy(ica_estimates, number = 3)
+  plot(te$V1, te$V2)
+  plot(ica_data$IC1, ica_data$IC2)
+
+  tidy(ica_trans, number = 3)
+  tidy(ica_estimates, number = 3)
+}
 }
 \references{
 Hyvarinen, A., and Oja, E. (2000). Independent
diff --git a/man/step_isomap.Rd b/man/step_isomap.Rd
index 469d6dde4..93a0224fb 100644
--- a/man/step_isomap.Rd
+++ b/man/step_isomap.Rd
@@ -114,16 +114,18 @@ im_trans <- rec \%>\%
               neighbors = 100,
               num_terms = 2)
 
-# im_estimates <- prep(im_trans, training = biomass_tr)
+if (require(dimRed) & require(RSpectra)) {
+  im_estimates <- prep(im_trans, training = biomass_tr)
 
-# im_te <- bake(im_estimates, biomass_te)
+  im_te <- bake(im_estimates, biomass_te)
 
-# rng <- extendrange(c(im_te$Isomap1, im_te$Isomap2))
-# plot(im_te$Isomap1, im_te$Isomap2,
-#      xlim = rng, ylim = rng)
+  rng <- extendrange(c(im_te$Isomap1, im_te$Isomap2))
+  plot(im_te$Isomap1, im_te$Isomap2,
+       xlim = rng, ylim = rng)
 
-# tidy(im_trans, number = 4)
-# tidy(im_estimates, number = 4)
+  tidy(im_trans, number = 4)
+  tidy(im_estimates, number = 4)
+}
 }
 }
 \references{
diff --git a/man/step_kpca.Rd b/man/step_kpca.Rd
index 653fb5e2e..2a244a4b1 100644
--- a/man/step_kpca.Rd
+++ b/man/step_kpca.Rd
@@ -122,16 +122,18 @@ kpca_trans <- rec \%>\%
   step_scale(all_predictors()) \%>\%
   step_kpca(all_predictors())
 
-# kpca_estimates <- prep(kpca_trans, training = biomass_tr)
+if (require(dimRed) & require(kernlab)) {
+  kpca_estimates <- prep(kpca_trans, training = biomass_tr)
 
-# kpca_te <- bake(kpca_estimates, biomass_te)
+  kpca_te <- bake(kpca_estimates, biomass_te)
 
-# rng <- extendrange(c(kpca_te$kPC1, kpca_te$kPC2))
-# plot(kpca_te$kPC1, kpca_te$kPC2,
-#      xlim = rng, ylim = rng)
+  rng <- extendrange(c(kpca_te$kPC1, kpca_te$kPC2))
+  plot(kpca_te$kPC1, kpca_te$kPC2,
+       xlim = rng, ylim = rng)
 
-# tidy(kpca_trans, number = 4)
-# tidy(kpca_estimates, number = 4)
+  tidy(kpca_trans, number = 4)
+  tidy(kpca_estimates, number = 4)
+}
 }
 \references{
 Scholkopf, B., Smola, A., and Muller, K. (1997).
diff --git a/man/step_nnmf.Rd b/man/step_nnmf.Rd
index 182c21b72..27522c835 100644
--- a/man/step_nnmf.Rd
+++ b/man/step_nnmf.Rd
@@ -89,20 +89,20 @@ If \code{num = 101}, the names would be \code{NNMF001} -
 }
 \examples{
 \donttest{
-library(dimRed)
-library(NMF)
 data(biomass)
 
-rec <- recipe(HHV ~ ., data = biomass) \%>\%
-  update_role(sample, new_role = "id var") \%>\%
-  update_role(dataset, new_role = "split variable") \%>\%
-  step_nnmf(all_predictors(), num_comp = 2, seed = 473, num_run = 2) \%>\%
-  prep(training = biomass, retain = TRUE)
+if (require(dimRed) & require(NMF)) {
+  rec <- recipe(HHV ~ ., data = biomass) \%>\%
+    update_role(sample, new_role = "id var") \%>\%
+    update_role(dataset, new_role = "split variable") \%>\%
+    step_nnmf(all_predictors(), num_comp = 2, seed = 473, num_run = 2) \%>\%
+    prep(training = biomass, retain = TRUE)
 
-# juice(rec)
+  juice(rec)
 
-# library(ggplot2)
-# ggplot(juice(rec), aes(x = NNMF2, y = NNMF1, col = HHV)) + geom_point()
+  library(ggplot2)
+  ggplot(juice(rec), aes(x = NNMF2, y = NNMF1, col = HHV)) + geom_point()
+}
 }
 }
 \seealso{

From 39b0c4e1e45a0a6fd30202ccdd15c9ac04643a4e Mon Sep 17 00:00:00 2001
From: topepo <mxkuhn@gmail.com>
Date: Sun, 18 Nov 2018 10:29:10 -0500
Subject: [PATCH 2/5] global variable false positive

---
 R/integer.R | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/R/integer.R b/R/integer.R
index 8699aada1..ec6a9ebba 100644
--- a/R/integer.R
+++ b/R/integer.R
@@ -21,7 +21,7 @@
 #' @param strict A logical for whether the values should be returned as
 #'  integers (as opposed to double).
 #' @param zero_based A logical for whether the integers should start at zero and
-#'  new values be appended as the largest integer. 
+#'  new values be appended as the largest integer.
 #' @return An updated version of `recipe` with the new step added
 #'  to the sequence of existing steps (if any). For the `tidy`
 #'  method, a tibble with columns `terms` (the selectors or
@@ -38,7 +38,7 @@
 #'  argument above). Missing values propagate.
 #'
 #' Factor inputs are ordered by their levels. All others are
-#'  ordered by `sort`. 
+#'  ordered by `sort`.
 #'
 #' Despite the name, the new values are returned as numeric unless
 #'  `strict = TRUE`, which will coerce the results to integers.
@@ -144,12 +144,12 @@ prep.step_integer <- function(x, training, info = NULL, ...) {
 map_key_to_int <- function(dat, key, strict = FALSE, zero = FALSE) {
   if (is.factor(dat))
     dat <- as.character(dat)
-  
+
   res <- full_join(tibble(value = dat, .row = seq_along(dat)), key, by = "value")
   res <- dplyr::filter(res, !is.na(.row))
   res <- arrange(res, .row)
   if (zero) {
-    res$integer[is.na(res$integer) & !is.na(res$value)] <- 
+    res$integer[is.na(res$integer) & !is.na(res$value)] <-
       max(key$integer, na.rm = TRUE) + 1
   } else {
     res$integer[is.na(res$integer) & !is.na(res$value)] <- 0
@@ -200,3 +200,6 @@ tidy.step_integer <- function(x, ...) {
   res$id <- x$id
   res
 }
+
+#' @importFrom stats runif
+utils::globalVariables(c(".row"))

From d7f5df9622f75ca81d30a35851ab9f074f50111a Mon Sep 17 00:00:00 2001
From: topepo <mxkuhn@gmail.com>
Date: Sun, 18 Nov 2018 10:29:26 -0500
Subject: [PATCH 3/5] isomap values are unique up to sign

---
 tests/testthat/test_isomap.R | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/tests/testthat/test_isomap.R b/tests/testthat/test_isomap.R
index 9417d0888..0517328b6 100644
--- a/tests/testthat/test_isomap.R
+++ b/tests/testthat/test_isomap.R
@@ -29,7 +29,7 @@ test_that('correct Isomap values', {
   skip_if_not_installed("igraph")
   skip_if_not_installed("RANN")
   skip_if_not_installed("dimRed")
-  
+
   im_rec <- rec %>%
     step_isomap(x1, x2, x3, neighbors = 3, num_terms = 3, id = "")
 
@@ -37,7 +37,8 @@ test_that('correct Isomap values', {
 
   im_pred <- bake(im_trained, new_data = dat2)
 
-  all.equal(as.matrix(im_pred), as.matrix(exp_res))
+  # unique up to sign
+  all.equal(abs(as.matrix(im_pred)), abs(as.matrix(exp_res)))
 
   im_tibble <-
     tibble(terms = c("x1", "x2", "x3"), id = "")
@@ -61,7 +62,7 @@ test_that('printing', {
   skip_if_not_installed("igraph")
   skip_if_not_installed("RANN")
   skip_if_not_installed("dimRed")
-  
+
   im_rec <- rec %>%
     step_isomap(x1, x2, x3, neighbors = 3, num_terms = 3)
   expect_output(print(im_rec))

From 9f77b639f5310360a654b26310babb452054bed1 Mon Sep 17 00:00:00 2001
From: topepo <mxkuhn@gmail.com>
Date: Sun, 18 Nov 2018 13:23:39 -0500
Subject: [PATCH 4/5] skip if early R version

---
 tests/testthat/test_isomap.R | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/tests/testthat/test_isomap.R b/tests/testthat/test_isomap.R
index 0517328b6..414719e96 100644
--- a/tests/testthat/test_isomap.R
+++ b/tests/testthat/test_isomap.R
@@ -29,6 +29,7 @@ test_that('correct Isomap values', {
   skip_if_not_installed("igraph")
   skip_if_not_installed("RANN")
   skip_if_not_installed("dimRed")
+  skip_if(getRversion() <= "3.4.4")
 
   im_rec <- rec %>%
     step_isomap(x1, x2, x3, neighbors = 3, num_terms = 3, id = "")
@@ -62,6 +63,7 @@ test_that('printing', {
   skip_if_not_installed("igraph")
   skip_if_not_installed("RANN")
   skip_if_not_installed("dimRed")
+  skip_if(getRversion() <= "3.4.4")
 
   im_rec <- rec %>%
     step_isomap(x1, x2, x3, neighbors = 3, num_terms = 3)

From 335ac68997d38c6737518c6c007a4e6fdfdf131d Mon Sep 17 00:00:00 2001
From: topepo <mxkuhn@gmail.com>
Date: Sun, 18 Nov 2018 13:48:53 -0500
Subject: [PATCH 5/5] minor format change

---
 NEWS.md | 1 +
 1 file changed, 1 insertion(+)

diff --git a/NEWS.md b/NEWS.md
index d81060fba..79760f1c8 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -1,6 +1,7 @@
 # `recipes` 0.1.3.9002
 
 ## Breaking Changes
+
  * Several argument names were changed to be consistent with other `tidymodels` packages (e.g. `dials`) and the general tidyverse naming conventions. 
    * `K` in `step_knnimpute` was changed to `neighbors`. `step_isomap` had the number of neighbors promoted to a main argument called `neighbors ` 
    *  `step_pca`, `step_pls`, `step_kpca`, `step_ica` now use  `num_comp`  instead of `num`. , `step_isomap` uses `num_terms` instead of `num`.