[pre-commit.ci] pre-commit autoupdate

.pre-commit-config.yaml

@@ -8,7 +8,7 @@ ci:
 
 repos:
   - repo: https://github.com/lorenzwalthert/precommit
-    rev: v0.4.3.9029
+    rev: v0.4.3.9030
     hooks:
       - id: parsable-R
       - id: no-browser-statement
@@ -49,19 +49,19 @@ repos:
         exclude: '(\.Rd|python/doc/source/reference/.*|test-doctest-.*)'
 
   - repo: https://github.com/tox-dev/tox-ini-fmt
-    rev: 1.7.2
+    rev: 1.9.0
     hooks:
       - id: tox-ini-fmt
 
   - repo: https://github.com/tox-dev/pyproject-fmt
-    rev: v2.25.2
+    rev: v2.29.4
     hooks:
       - id: pyproject-fmt
         additional_dependencies: ["tox>=4.12.1"]
 
   - repo: https://github.com/astral-sh/ruff-pre-commit
     # Ruff version.
-    rev: v0.15.21
+    rev: v0.16.8
     hooks:
       # Run the formatter.
       - id: ruff-format

r/NEWS.md

@@ -1,6 +1,7 @@
-# moocore (development)
+# moocore 0.3.2.900
 
  * `is_nondominated` is up to 10x faster in some inputs thanks to a customized radixsort implementation.
+ * New shapes `"cliff-concave"` and `"cliff-convex"` added to `generate_ndset()`.
 
 
 # moocore 0.3.2

r/R/generate.R

@@ -70,6 +70,31 @@
 #'    hypersphere to the positive orthant. Thus, the sampling remains uniform.
 #'
 #'   }
+#'   \item{`'cliff-concave'`}{
+#'
+#'    Equivalent to generating a 2D set using
+#'    `method='concave-sphere'`, then generating the other `d-2` columns
+#'    uniformly at random within the unit hypercube
+#'    \citep{EmmFon2011emo,GueFon2017hv4d}. This method does not make sense
+#'    for `d=2`.
+#'
+#'    In the resulting set, the first two columns are mutually nondominated,
+#'    that is, no point can dominate another regardless of the other
+#'    objectives, while the remaining `d-2` columns do not provide any
+#'    ordering. These sets are adversarial for algorithms that aim to exploit
+#'    dominance structure.
+#'
+#'    The dimensions that are generated uniformly at random should be chosen
+#'    adversarially according to the algorithm being tested. Alternatively, the
+#'    columns may be randomly shuffled.
+#'
+#'   }
+#'   \item{`'cliff-convex'`}{
+#'
+#'    Equivalent to `1 - generate_ndset(..., method='cliff-concave')`.
+#'    This method does not make sense for `d=2`.
+#'
+#'   }
 #'   \item{`'convex-simplex'`}{
 #'
 #'    Equivalent to `generate_ndset(..., method='simplex')^2`, which is convex
@@ -114,6 +139,7 @@
 #' generate_ndset(5, 3, "simplex", seed = 42, integer = TRUE)
 #' generate_ndset(4, 2, "sphere", seed = 123)
 #' generate_ndset(3, 5, "convex-sphere", seed = 123)
+#' generate_ndset(3, 5, "cliff-convex", seed = 42)
 #' generate_ndset(4, 4, "convex-simplex", seed = 123)
 #'
 #' @export
@@ -129,14 +155,23 @@ generate_ndset <- function(n, d, method, seed = NULL, integer = FALSE)
     x
   }
 
-  sample_sphere <- function() {
-    x <- abs(rnorm(n * d))
-    dim(x) <- c(n, d)
+  sample_sphere <- function(dims = d) {
+    x <- abs(rnorm(n * dims))
+    dim(x) <- c(n, dims)
     x <- x / sqrt(rowSums(x * x))
     x
   }
 
+  sample_cliff_concave <- function() {
+    x <- matrix(0, nrow = n, ncol = d)
+    x[, 1:2] <- sample_sphere(2L)
+    cols <- seq.int(3L, d)
+    x[, cols] <- matrix(runif(n * (d - 2L)), nrow = n, ncol = d - 2L)
+    x
+  }
+
   sample_convex_sphere <- function() 1. - sample_sphere()
+  sample_cliff_convex <- function() 1. - sample_cliff_concave()
   sample_convex_simplex <- function() sample_simplex()^2
   sample_inverted_simplex <- function() 1. - sample_simplex()
   sample_concave_simplex <- function() 1. - sample_convex_simplex()
@@ -145,6 +180,11 @@ generate_ndset <- function(n, d, method, seed = NULL, integer = FALSE)
     if (method %in% c("simplex", "linear", "L")) sample_simplex
     else if (method %in% c("concave-sphere", "sphere", "C")) sample_sphere
     else if (method %in% c("convex-sphere", "X")) sample_convex_sphere
+    else if (method %in% c("cliff-concave", "cliff-convex")) {
+      if (d <= 2L)
+        stop("method='", method, "' requires at least 3 dimensions")
+      if (method == "cliff-concave") sample_cliff_concave else sample_cliff_convex
+    }
     else if (method == "convex-simplex") sample_convex_simplex
     else if (method %in% c("inverted-simplex", "inverted-linear")) sample_inverted_simplex
     else if (method == "concave-simplex") sample_concave_simplex

r/tests/testthat/test-generate.R

@@ -4,10 +4,30 @@ test_that("generate_ndset", {
   one <- replicate(n,1)
   for (dim in seq(5,10)) {
     points <- generate_ndset(n, dim, "simplex")
+    expect_false(any_dominated(points))
     expect_equal(rowSums(points), one)
     points <- generate_ndset(n, dim, "concave-sphere")
+    expect_false(any_dominated(points))
     expect_equal(rowSums(points**2), one)
     points <- generate_ndset(n, dim, "convex-simplex")
+    expect_false(any_dominated(points))
     expect_equal(rowSums(sqrt(points)), one)
+    points <- generate_ndset(n, dim, "cliff-concave")
+    expect_false(any_dominated(points))
+    expect_equal(rowSums(points[, 1:2]^2), one)
+    expect_true(all(points[, 3:dim] >= 0 & points[, 3:dim] <= 1))
+    points <- generate_ndset(n, dim, "cliff-convex")
+    expect_false(any_dominated(points))
+    expect_equal(rowSums((1 - points[, 1:2])^2), one)
+    expect_true(all(points[, 3:dim] >= 0 & points[, 3:dim] <= 1))
   }
+
+  expect_error(
+    generate_ndset(n, 2L, "cliff-concave"),
+    "requires at least 3 dimensions"
+  )
+  expect_error(
+    generate_ndset(n, 2L, "cliff-convex"),
+    "requires at least 3 dimensions"
+  )
 })

r/vignettes/articles/generate.Rmd

@@ -229,6 +229,19 @@ fig2 <- plot_3d("simplex", moocore::generate_ndset(n, 3, "concave-simplex", seed
 plotly_side_by_side(fig1, fig2)
 ```
 
+# Cliff sets (3D)
+
+The `cliff` type is often used to test the performance of algorithms for
+computing the hypervolume [@EmmFon2011emo; @GueFon2017hv4d], due to its
+particular structure.
+
+```{r generate_3d_cliff}
+fig1 <- plot_3d("simplex", moocore::generate_ndset(n, 3, "cliff-concave", seed = 42), title = 'method="cliff-concave"')
+fig2 <- plot_3d("simplex", moocore::generate_ndset(n, 3, "cliff-convex", seed = 42), title = 'method="cliff-convex"')
+
+plotly_side_by_side(fig1, fig2)
+```
+
 # Uniform sampling (moocore) vs projections of uniform samples (naive)
 
 Naive methods for sampling such sets usually sample points uniformly in the