I've modified an example given by the computational geometry cgal
library (link) that demonstrates incremental searching on a 2D plane (Section 49.3.2). The example uses a functor to set spatial bounds for searching only positive points on the plane.
I would like to modify the functor so that k
can be passed in, as demonstrated by struct X_not_positive_k
. The complete example program below below shows the modified code and the original code.
#include <CGAL/Simple_cartesian.h>
#include <CGAL/Orthogonal_incremental_neighbor_search.h>
#include <CGAL/Search_traits_2.h>
typedef CGAL::Simple_cartesian<double> K;
typedef K::Point_2 Point_d;
typedef CGAL::Search_traits_2<K> TreeTraits;
typedef CGAL::Orthogonal_incremental_neighbor_search<TreeTraits> NN_incremental_search;
typedef NN_incremental_search::iterator NN_iterator;
typedef NN_incremental_search::Tree Tree;
int main() {
Tree tree;
tree.insert(Point_d(0,0));
tree.insert(Point_d(1,1));
tree.insert(Point_d(0,1));
tree.insert(Point_d(10,110));
tree.insert(Point_d(45,0));
tree.insert(Point_d(0,2340));
tree.insert(Point_d(0,30));
Point_d query(0,0);
// A functor that returns true, iff the x-coordinate of a dD point is not positive
// [ORIGINAL CODE]
struct X_not_positive {
bool operator()(const NN_iterator& it) { return ((*it).first)[0]<0; }
};
// [MODIFIED CODE]
// This does not work when used below.
struct X_not_positive_k {
public:
void assign_k(int k) {this->k = k; }
bool operator()(const NN_iterator& it) { return ((*it).first)[0] < k; }
private:
int k;
};
X_not_positive_k Xk;
Xk.assign_k(1);
// An iterator that only enumerates dD points with positive x-coordinate
// [ORIGINAL CODE]
// typedef CGAL::Filter_iterator<NN_iterator, X_not_positive> NN_positive_x_iterator;
// [MODIFIED CODE]
typedef CGAL::Filter_iterator<NN_iterator, X_not_positive_k> NN_positive_x_iterator;
NN_incremental_search NN(tree, query);
// [ORIGINAL CODE]
// NN_positive_x_iterator it(NN.end(), X_not_positive(), NN.begin()), end(NN.end(), X_not_positive());
// [MODIFIED CODE]
NN_positive_x_iterator it(NN.end(), Xk(), NN.begin()), end(NN.end(), Xk());
// error occurs here
std::cout << "The first 5 nearest neighbours with positive x-coord are: " << std::endl;
for (int j=0; (j < 5)&&(it!=end); ++j,++it)
std::cout << (*it).first << " at squared distance = " << (*it).second << std::endl;
return 0;
}
However, compiling this program (Windows 7 with Visual Studio 2010) causes the following compiler error. The location of the error is marked in the code above.
2>..\main.cpp(56): error C2064: term does not evaluate to a function taking 0 arguments
2> class does not define an 'operator()' or a user defined conversion operator to a pointer-to-function or reference-to-function that takes appropriate number of arguments
2>..\main.cpp(56): error C2064: term does not evaluate to a function taking 0 arguments
2> class does not define an 'operator()' or a user defined conversion operator to a pointer-to-function or reference-to-function that takes appropriate number of arguments
2>
What can be done to get rid of the error? Is there another way to be able to set the k
variable?
Just replace your Xk()
with Xk
in this line 56
alternatively replace assign(k)
with constructor argument, then replace Xk()
with X_not_positive_k(1)
actually real trouble is functor's name! replace name X_not_positive_k
with X_less_than
- so call to X_less_than(1)
looks nice in line 56
like this:
struct X_less_than {
public:
X_less_than(int i)
: k(i)
{
}
bool operator()(const NN_iterator& it) { return ((*it).first)[0] < k; }
private:
int k;
};