version 0.3.0

cran · Jul 5, 2023 · 4b6332d · 4b6332d
1 parent 45b9391
commit 4b6332d
Show file tree

Hide file tree

Showing 10 changed files with 83 additions and 48 deletions.
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -1,20 +1,20 @@
 Package: RDP
 Title: The Ramer-Douglas-Peucker Algorithm
-Version: 0.2.3
+Version: 0.3.0
 Authors@R: person("Robert", "Dahl Jacobsen", role = c("aut", "cre"), email = "cran@dahl-jacobsen.dk")
 Description: Pretty fast implementation of the Ramer-Douglas-Peucker algorithm for reducing the number of points on a 2D curve.
     Urs Ramer (1972), "An iterative procedure for the polygonal approximation of plane curves" <doi:10.1016/S0146-664X(72)80017-0>.
     David H. Douglas and Thomas K. Peucker (1973), "Algorithms for the Reduction of the Number of Points Required to Represent a Digitized Line or its Caricature" <doi:10.3138/FM57-6770-U75U-7727>.
 License: GPL-3
 URL: https://github.com/robertdj/RDP
 Encoding: UTF-8
-RoxygenNote: 7.1.2
+RoxygenNote: 7.2.1
 LinkingTo: Rcpp
 Imports: Rcpp
 Suggests: testthat, withr, zeallot
 NeedsCompilation: yes
-Packaged: 2022-03-17 19:33:52 UTC; robert
+Packaged: 2023-07-05 21:57:55 UTC; robert
 Author: Robert Dahl Jacobsen [aut, cre]
 Maintainer: Robert Dahl Jacobsen <cran@dahl-jacobsen.dk>
 Repository: CRAN
-Date/Publication: 2022-03-17 20:00:02 UTC
+Date/Publication: 2023-07-05 22:13:04 UTC
diff --git a/MD5 b/MD5
@@ -1,16 +1,16 @@
-65121eeca10f7955916a9cb86033d11b *DESCRIPTION
+307eea5b522c518f3280336a284b13c0 *DESCRIPTION
 a6892181c742d67c91932b7a9ae9a1b1 *NAMESPACE
 94346c4beb59f83a249106424ee7124d *R/RDP-package.R
-0c1d68865f8d1fd242be2afd2091e969 *R/RcppExports.R
-3da93c72d3d6fdfc2deb8dd51ff28bd9 *build/partial.rdb
+15d2f288101747e10c53c61e3decd507 *R/RcppExports.R
+0c6ad5b3446721162c89589e8077ab1c *build/partial.rdb
 537e8955e966c4fcaa06a6403b734a66 *man/RDP-package.Rd
-e64af47a563757357d5238467d7bb2f3 *man/RamerDouglasPeucker.Rd
+5bd1528059d024806f80b6d56c6ef48f *man/RamerDouglasPeucker.Rd
 f83ff26c302dc4420b1b693e49719072 *man/figures/README-example-1.png
 daf246d2559a7df53698b1714f020b7a *src/Makevars
 daf246d2559a7df53698b1714f020b7a *src/Makevars.win
-e337a01300ab4bf35b015446439090f2 *src/RamerDouglasPeucker.cpp
-561f9103c77931ac481416cdec955565 *src/RamerDouglasPeucker.h
-717ec3dc1097c9e3ba69583eb232b940 *src/RcppExports.cpp
-796697c13da38f09a8b87ed424015e50 *src/wrapper.cpp
+a6a086927824b68da0ac55607f7c2c32 *src/RamerDouglasPeucker.cpp
+d8ae92e21747cce3bdc3af0ccb344dbc *src/RamerDouglasPeucker.h
+de56ffc5ab7fd9bf89c22c53f2cf6030 *src/RcppExports.cpp
+e6aca099b9b0b57001bed47704555043 *src/wrapper.cpp
 be8c94f200a61296bfa381fb42c8f65e *tests/testthat.R
-1abbbeb5341fb797ac8bca4b2f7f0368 *tests/testthat/test-RamerDouglasPeucker.R
+e6e1b02af05aa76515b04acf8849507e *tests/testthat/test-RamerDouglasPeucker.R
diff --git a/R/RcppExports.R b/R/RcppExports.R
@@ -1,25 +1,28 @@
 # Generated by using Rcpp::compileAttributes() -> do not edit by hand
 # Generator token: 10BE3573-1514-4C36-9D1C-5A225CD40393
 
-#' Ramer-Douglas-Peucker
+#' Simplify a curve using the Ramer-Douglas-Peucker algorithm.
 #'
-#' The [Ramer-Douglas-Peucker algorithm](https://en.wikipedia.org/wiki/Ramer-Douglas-Peucker_algorithm) for reducing the number of points on a curve.
+#' Implements the [Ramer-Douglas-Peucker algorithm](https://en.wikipedia.org/wiki/Ramer-Douglas-Peucker_algorithm) for reducing the number of points on a curve.
 #'
 #' @details If there are no more than two points it does not make sense to simplify.
 #' In this case the input is returned without further checks of `x` and `y`.
+#' In particular, the input is not checked for `NA` values.
 #'
-#' @param x The `x` values of the curve as a vector.
-#' @param y The `y` values of the curve as a vector.
-#' @param epsilon The threshold for filtering outliers from the simplified curve.
+#' @param x `[numeric]` The `x` values of the curve as a vector without `NA` values.
+#' @param y `[numeric]` The `y` values of the curve as a vector without `NA` values.
+#' @param epsilon `[positive numeric(1)]` The threshold for filtering outliers from the simplified curve.
+#' @param keep_index `[logical]` If `TRUE`, returns a column called `index` with the index locations of points that are kept.
 #'
 #' @return A `data.frame` with `x` and `y` values of the simplified curve.
 #'
 #' @examples
 #' RDP::RamerDouglasPeucker(x = c(0, 1, 3, 5), y = c(2, 1, 0, 1), epsilon = 0.5)
+#' RDP::RamerDouglasPeucker(x = c(0, 1, 3, 5), y = c(2, 1, 0, 1), epsilon = 0.5, keep_index = TRUE)
 #'
 #' @export
 #'
-RamerDouglasPeucker <- function(x, y, epsilon) {
-    .Call(`_RDP_RamerDouglasPeucker`, x, y, epsilon)
+RamerDouglasPeucker <- function(x, y, epsilon, keep_index = FALSE) {
+    .Call(`_RDP_RamerDouglasPeucker`, x, y, epsilon, keep_index)
 }
 
diff --git a/build/partial.rdb b/build/partial.rdb
diff --git a/man/RamerDouglasPeucker.Rd b/man/RamerDouglasPeucker.Rd
diff --git a/src/RamerDouglasPeucker.cpp b/src/RamerDouglasPeucker.cpp
@@ -23,9 +23,9 @@ double abs2(Point2D p)
 }
 
 
-// From reference points[startIndex] == points[endIndex], find the point furthest away.
+// Find the point furthest away from reference (points[startIndex] == points[endIndex])
 std::pair<double, std::size_t> findMostDistantPoint(const std::vector<Point2D> &points,
-                                                          std::size_t startIndex, std::size_t endIndex)
+                                                    std::size_t startIndex, std::size_t endIndex)
 {
     assert(startIndex < endIndex && "Start index must be smaller than end index");
     assert(endIndex < points.size() && "End index is larger than the number of points");
@@ -55,7 +55,8 @@ std::pair<double, std::size_t> findMostDistantPoint(const std::vector<Point2D> &
 // Rearranging this formula to avoid recomputing constants:
 // https://en.wikipedia.org/wiki/Distance_from_a_point_to_a_line#Line_defined_by_two_points
 std::pair<double, std::size_t> findMostDistantPointFromLine(const std::vector<Point2D> &points,
-                                                         std::size_t startIndex, std::size_t endIndex)
+                                                            std::size_t startIndex,
+                                                            std::size_t endIndex)
 {
     assert(startIndex < endIndex && "Start index must be smaller than end index");
     assert(endIndex < points.size() && "End index is larger than the number of points");
@@ -93,8 +94,9 @@ std::pair<double, std::size_t> findMostDistantPointFromLine(const std::vector<Po
 }
 
 
-void RamerDouglasPeucker(const std::vector<Point2D> &points, std::size_t startIndex, std::size_t endIndex,
-                         double epsilonSquared, std::vector<std::size_t> &indicesToKeep)
+void RamerDouglasPeucker(const std::vector<Point2D> &points, std::size_t startIndex,
+                         std::size_t endIndex, double epsilonSquared,
+                         std::vector<std::size_t> &indicesToKeep)
 {
     assert(startIndex < endIndex && "Start index must be smaller than end index");
     assert(endIndex < points.size() && "End index is larger than the number of points");
@@ -106,7 +108,8 @@ void RamerDouglasPeucker(const std::vector<Point2D> &points, std::size_t startIn
     assert(indicesToKeep.size() >= 1 && "indicesToKeep should be non-empty");
     assert(indicesToKeep[0] == 0 && "indicesToKeep should be initialized with a 0");
 
-    auto [maxDistanceSquared, maxDistanceIndex] = findMostDistantPointFromLine(points, startIndex, endIndex);
+    auto [maxDistanceSquared, maxDistanceIndex] =
+        findMostDistantPointFromLine(points, startIndex, endIndex);
 
     if (maxDistanceSquared > epsilonSquared)
     {

diff --git a/src/RamerDouglasPeucker.h b/src/RamerDouglasPeucker.h
@@ -8,8 +8,9 @@ struct Point2D
     double y;
 };
 
-void RamerDouglasPeucker(const std::vector<Point2D> &pointList, std::size_t startIndex, std::size_t endIndex,
-                         double epsilonSquared, std::vector<std::size_t> &indicesToKeep);
+void RamerDouglasPeucker(const std::vector<Point2D> &pointList, std::size_t startIndex,
+                         std::size_t endIndex, double epsilonSquared,
+                         std::vector<std::size_t> &indicesToKeep);
 }
 
 #endif
diff --git a/src/RcppExports.cpp b/src/RcppExports.cpp
@@ -11,21 +11,22 @@ Rcpp::Rostream<false>& Rcpp::Rcerr = Rcpp::Rcpp_cerr_get();
 #endif
 
 // RamerDouglasPeucker
-Rcpp::DataFrame RamerDouglasPeucker(Rcpp::NumericVector x, Rcpp::NumericVector y, double epsilon);
-RcppExport SEXP _RDP_RamerDouglasPeucker(SEXP xSEXP, SEXP ySEXP, SEXP epsilonSEXP) {
+Rcpp::DataFrame RamerDouglasPeucker(Rcpp::NumericVector x, Rcpp::NumericVector y, double epsilon, bool keep_index);
+RcppExport SEXP _RDP_RamerDouglasPeucker(SEXP xSEXP, SEXP ySEXP, SEXP epsilonSEXP, SEXP keep_indexSEXP) {
 BEGIN_RCPP
     Rcpp::RObject rcpp_result_gen;
     Rcpp::RNGScope rcpp_rngScope_gen;
     Rcpp::traits::input_parameter< Rcpp::NumericVector >::type x(xSEXP);
     Rcpp::traits::input_parameter< Rcpp::NumericVector >::type y(ySEXP);
     Rcpp::traits::input_parameter< double >::type epsilon(epsilonSEXP);
-    rcpp_result_gen = Rcpp::wrap(RamerDouglasPeucker(x, y, epsilon));
+    Rcpp::traits::input_parameter< bool >::type keep_index(keep_indexSEXP);
+    rcpp_result_gen = Rcpp::wrap(RamerDouglasPeucker(x, y, epsilon, keep_index));
     return rcpp_result_gen;
 END_RCPP
 }
 
 static const R_CallMethodDef CallEntries[] = {
-    {"_RDP_RamerDouglasPeucker", (DL_FUNC) &_RDP_RamerDouglasPeucker, 3},
+    {"_RDP_RamerDouglasPeucker", (DL_FUNC) &_RDP_RamerDouglasPeucker, 4},
     {NULL, NULL, 0}
 };
 

diff --git a/src/wrapper.cpp b/src/wrapper.cpp
@@ -1,37 +1,41 @@
 #include <Rcpp.h>
 #include "RamerDouglasPeucker.h"
 
-//' Ramer-Douglas-Peucker
+//' Simplify a curve using the Ramer-Douglas-Peucker algorithm.
 //'
-//' The [Ramer-Douglas-Peucker algorithm](https://en.wikipedia.org/wiki/Ramer-Douglas-Peucker_algorithm) for reducing the number of points on a curve.
+//' Implements the [Ramer-Douglas-Peucker algorithm](https://en.wikipedia.org/wiki/Ramer-Douglas-Peucker_algorithm) for reducing the number of points on a curve.
 //'
 //' @details If there are no more than two points it does not make sense to simplify.
 //' In this case the input is returned without further checks of `x` and `y`.
+//' In particular, the input is not checked for `NA` values.
 //'
-//' @param x The `x` values of the curve as a vector.
-//' @param y The `y` values of the curve as a vector.
-//' @param epsilon The threshold for filtering outliers from the simplified curve.
+//' @param x `[numeric]` The `x` values of the curve as a vector without `NA` values.
+//' @param y `[numeric]` The `y` values of the curve as a vector without `NA` values.
+//' @param epsilon `[positive numeric(1)]` The threshold for filtering outliers from the simplified curve.
+//' @param keep_index `[logical]` If `TRUE`, returns a column called `index` with the index locations of points that are kept.
 //'
 //' @return A `data.frame` with `x` and `y` values of the simplified curve.
 //'
 //' @examples
 //' RDP::RamerDouglasPeucker(x = c(0, 1, 3, 5), y = c(2, 1, 0, 1), epsilon = 0.5)
+//' RDP::RamerDouglasPeucker(x = c(0, 1, 3, 5), y = c(2, 1, 0, 1), epsilon = 0.5, keep_index = TRUE)
 //'
 //' @export
 //'
 // [[Rcpp::export]]
-Rcpp::DataFrame RamerDouglasPeucker(Rcpp::NumericVector x, Rcpp::NumericVector y, double epsilon)
+Rcpp::DataFrame RamerDouglasPeucker(Rcpp::NumericVector x, Rcpp::NumericVector y, double epsilon,
+                                    bool keep_index = false)
 {
-    if (epsilon < 0 || Rcpp::NumericVector::is_na(epsilon))
-        throw std::domain_error("epsilon must be a non-negative number");
-
     R_xlen_t nPoints = x.length();
     if (nPoints != y.length())
         throw std::invalid_argument("x and y vectors must be of equal length");
 
     if (nPoints <= 2)
         return Rcpp::DataFrame::create(Rcpp::Named("x") = x, Rcpp::Named("y") = y);
 
+    if (epsilon < 0 || Rcpp::NumericVector::is_na(epsilon))
+        throw std::domain_error("epsilon must be a non-negative number");
+
     std::vector<rdp::Point2D> points;
     points.reserve(nPoints);
     for (R_xlen_t i = 0; i != nPoints; ++i)
@@ -59,7 +63,15 @@ Rcpp::DataFrame RamerDouglasPeucker(Rcpp::NumericVector x, Rcpp::NumericVector y
         std::size_t index = indicesToKeep[i];
         xOut[i] = x[index];
         yOut[i] = y[index];
+        // Add 1 to index before returning to R:
+        indicesToKeep[i] += 1;
+    }
+
+    if (keep_index) {
+        return Rcpp::DataFrame::create(Rcpp::Named("x") = xOut, Rcpp::Named("y") = yOut,
+                                       Rcpp::Named("index") = indicesToKeep);
+    } else {
+        return Rcpp::DataFrame::create(Rcpp::Named("x") = xOut, Rcpp::Named("y") = yOut);
     }
 
-    return Rcpp::DataFrame::create(Rcpp::Named("x") = xOut, Rcpp::Named("y") = yOut);
 }
diff --git a/tests/testthat/test-RamerDouglasPeucker.R b/tests/testthat/test-RamerDouglasPeucker.R
@@ -76,6 +76,17 @@ test_that("Bigger example", {
     )
 })
 
+test_that("Keeps indices", {
+    x <- c(3.5, 7.3, 23.2, 37.2, 54.6, 62.2, 71.5, 101.3)
+    y <- c(21.25, 12.0, 3.1, 12.07, 18.15, 16.45, 9.7, 21.1)
+
+    expect_equal(
+        RamerDouglasPeucker(x, y, 5, keep_index = TRUE),
+        data.frame(x = c(3.5, 23.2, 54.6, 71.5, 101.3), y = c(21.25, 3.1, 18.15, 9.7, 21.1),
+                   index = c(1, 3, 5, 7, 8))
+    )
+})
+
 
 test_that("Input returned when there are too few points", {
     expect_equal(RamerDouglasPeucker(c(1, 2), c(1, 2), 5), data.frame(x = c(1, 2), y = c(1, 2)))