sparsify point set cythonization

unitary tested


git-svn-id: svn+ssh://scm.gforge.inria.fr/svnroot/gudhi/branches/ST_cythonize@1816 636b058d-ea47-450e-bf9e-a15bfbe3eedb

Former-commit-id: 170f14ffcf009ebe1169df90a4e088441eaad13f

3 files changed, 72 insertions, 11 deletions

diff --git a/src/cython/cython/subsampling.pyx b/src/cython/cython/subsampling.pyx
index c71f5810..05c0232f 100644
--- a/src/cython/cython/subsampling.pyx
+++ b/src/cython/cython/subsampling.pyx
@@ -37,6 +37,8 @@ cdef extern from "Subsampling_interface.h" namespace "Gudhi::subsampling":
     vector[vector[double]] subsampling_n_farthest_points_from_file(string off_file, unsigned nb_points, unsigned starting_point)
     vector[vector[double]] subsampling_n_random_points(vector[vector[double]] points, unsigned nb_points)
     vector[vector[double]] subsampling_n_random_points_from_file(string off_file, unsigned nb_points)
+    vector[vector[double]] subsampling_sparsify_points(vector[vector[double]] points, double min_squared_dist)
+    vector[vector[double]] subsampling_sparsify_points_from_file(string off_file, double min_squared_dist)
 
 def choose_n_farthest_points(points=[], off_file='', nb_points=0, starting_point = ''):
     """Subsample by a greedy strategy of iteratively adding the farthest point
@@ -57,7 +59,7 @@ def choose_n_farthest_points(points=[], off_file='', nb_points=0, starting_point
     point`,which is the index of the poit to start with. If not set, this \
     index is choosen randomly.
     :type starting_point: unsigned.
-    :returns:  The subsamplepoint set.
+    :returns:  The subsample point set.
     :rtype: vector[vector[double]]
     """
     if off_file is not '':
@@ -87,7 +89,7 @@ def pick_n_random_points(points=[], off_file='', nb_points=0):
 
     :param nb_points: Number of points of the subsample.
     :type nb_points: unsigned.
-    :returns:  The subsamplepoint set.
+    :returns:  The subsample point set.
     :rtype: vector[vector[double]]
     """
     if off_file is not '':
@@ -97,3 +99,27 @@ def pick_n_random_points(points=[], off_file='', nb_points=0):
             print("file " + off_file + " not found.")
     else:
         return subsampling_n_random_points(points, nb_points)
+
+def sparsify_point_set(points=[], off_file='', min_squared_dist=0.0):
+    """Subsample a point set by picking random vertices.
+
+    :param points: The input point set.
+    :type points: vector[vector[double]].
+
+    Or
+
+    :param off_file: An OFF file style name.
+    :type off_file: string
+
+    :param min_squared_dist: Number of points of the subsample.
+    :type min_squared_dist: unsigned.
+    :returns:  The subsample point set.
+    :rtype: vector[vector[double]]
+    """
+    if off_file is not '':
+        if os.path.isfile(off_file):
+            return subsampling_sparsify_points_from_file(off_file, min_squared_dist)
+        else:
+            print("file " + off_file + " not found.")
+    else:
+        return subsampling_sparsify_points(points, min_squared_dist)
diff --git a/src/cython/include/Subsampling_interface.h b/src/cython/include/Subsampling_interface.h
index 8ef4fea1..9340cd86 100644
--- a/src/cython/include/Subsampling_interface.h
+++ b/src/cython/include/Subsampling_interface.h
@@ -24,6 +24,8 @@
 #define	SUBSAMPLING_INTERFACE_H
 
 #include <gudhi/choose_n_farthest_points.h>
+#include <gudhi/pick_n_random_points.h>
+#include <gudhi/sparsify_point_set.h>
 #include <gudhi/Points_off_io.h>
 #include <CGAL/Epick_d.h>
 
@@ -40,9 +42,6 @@ using Subsampling_ft = Subsampling_dynamic_kernel::FT;
 
 // ------ choose_n_farthest_points ------
 std::vector<std::vector<double>> subsampling_n_farthest_points(std::vector<std::vector<double>>& points, unsigned nb_points) {
-  std::vector<Subsampling_point_d> input, output;
-  for (auto point : points)
-      input.push_back(Subsampling_point_d(point.size(), point.begin(), point.end()));
   std::vector<std::vector<double>> landmarks;
   Subsampling_dynamic_kernel k;
   choose_n_farthest_points(k, points, nb_points, std::back_inserter(landmarks));
@@ -51,9 +50,6 @@ std::vector<std::vector<double>> subsampling_n_farthest_points(std::vector<std::
 }
 
 std::vector<std::vector<double>> subsampling_n_farthest_points(std::vector<std::vector<double>>& points, unsigned nb_points, unsigned starting_point) {
-  std::vector<Subsampling_point_d> input, output;
-  for (auto point : points)
-      input.push_back(Subsampling_point_d(point.size(), point.begin(), point.end()));
   std::vector<std::vector<double>> landmarks;
   Subsampling_dynamic_kernel k;
   choose_n_farthest_points(k, points, nb_points, starting_point, std::back_inserter(landmarks));
@@ -75,9 +71,6 @@ std::vector<std::vector<double>> subsampling_n_farthest_points_from_file(std::st
 
 // ------ pick_n_random_points ------
 std::vector<std::vector<double>> subsampling_n_random_points(std::vector<std::vector<double>>& points, unsigned nb_points) {
-  std::vector<Subsampling_point_d> input, output;
-  for (auto point : points)
-      input.push_back(Subsampling_point_d(point.size(), point.begin(), point.end()));
   std::vector<std::vector<double>> landmarks;
   pick_n_random_points(points, nb_points, std::back_inserter(landmarks));
 
@@ -90,6 +83,26 @@ std::vector<std::vector<double>> subsampling_n_random_points_from_file(std::stri
   return subsampling_n_random_points(points, nb_points);
 }
 
+// ------ sparsify_point_set ------
+std::vector<std::vector<double>> subsampling_sparsify_points(std::vector<std::vector<double>>& points, double min_squared_dist) {
+  std::vector<Subsampling_point_d> input, output;
+  for (auto point : points)
+      input.push_back(Subsampling_point_d(point.size(), point.begin(), point.end()));
+  Subsampling_dynamic_kernel k;
+  sparsify_point_set(k, input, min_squared_dist, std::back_inserter(output));
+
+  std::vector<std::vector<double>> landmarks;
+  for (auto point : output)
+      landmarks.push_back(std::vector<double>(point.cartesian_begin(), point.cartesian_end()));
+  return landmarks;
+}
+
+std::vector<std::vector<double>> subsampling_sparsify_points_from_file(std::string& off_file, double min_squared_dist) {
+  Gudhi::Points_off_reader<std::vector<double>> off_reader(off_file);
+  std::vector<std::vector<double>> points = off_reader.get_point_cloud();
+  return subsampling_sparsify_points(points, min_squared_dist);
+}
+
 
 }  // namespace subsampling
 
diff --git a/src/cython/test/test_subsampling.py b/src/cython/test/test_subsampling.py
index 2dc12a89..2caf4ddb 100755
--- a/src/cython/test/test_subsampling.py
+++ b/src/cython/test/test_subsampling.py
@@ -109,3 +109,25 @@ def test_simple_pick_n_random_points():
                     found = True
             # Check each sub set point is existing in the point set
             assert found == True
+
+    # From off file test
+    for i in range (0, 7):
+        assert len(gudhi.pick_n_random_points(off_file = 'subsample.off', nb_points = i)) == i
+
+def test_simple_sparsify_points():
+    point_set = [[0,1], [0,0], [1,0], [1,1]]
+    # Test the limits
+    # assert gudhi.sparsify_point_set(points = [], min_squared_dist = 0.0) == []
+    # assert gudhi.sparsify_point_set(points = [], min_squared_dist = 10.0) == []
+    assert gudhi.sparsify_point_set(points = point_set, min_squared_dist = 0.0) == point_set
+    assert gudhi.sparsify_point_set(points = point_set, min_squared_dist = 1.0) == point_set
+    assert gudhi.sparsify_point_set(points = point_set, min_squared_dist = 2.0) == [[0,1], [1,0]]
+    assert gudhi.sparsify_point_set(points = point_set, min_squared_dist = 2.01) == [[0,1]]
+
+    assert len(gudhi.sparsify_point_set(off_file = 'subsample.off', min_squared_dist = 0.0)) == 7
+    assert len(gudhi.sparsify_point_set(off_file = 'subsample.off', min_squared_dist = 30.0)) == 5
+    assert len(gudhi.sparsify_point_set(off_file = 'subsample.off', min_squared_dist = 40.0)) == 4
+    assert len(gudhi.sparsify_point_set(off_file = 'subsample.off', min_squared_dist = 90.0)) == 3
+    assert len(gudhi.sparsify_point_set(off_file = 'subsample.off', min_squared_dist = 100.0)) == 2
+    assert len(gudhi.sparsify_point_set(off_file = 'subsample.off', min_squared_dist = 325.0)) == 2
+    assert len(gudhi.sparsify_point_set(off_file = 'subsample.off', min_squared_dist = 325.01)) == 1