thrust::stable_partition

Defined in thrust/partition.h

template<typename DerivedPolicy, typename ForwardIterator, typename InputIterator, typename Predicate> ForwardIterator thrust::stable_partition(const thrust::detail::execution_policy_base<DerivedPolicy> &exec, ForwardIterator first, ForwardIterator last, InputIterator stencil, Predicate pred)

stable_partition is much like partition: it reorders the elements in the range [first, last) based on the function object pred applied to a stencil range [stencil, stencil + (last - first)), such that all of the elements whose corresponding stencil element satisfies pred precede all of the elements whose corresponding stencil element fails to satisfy it. The postcondition is that, for some iterator middle in the range [first, last), pred(*stencil_i) is true for every iterator stencil_i in the range [stencil,stencil + (middle - first)) and false for every iterator stencil_i in the range [stencil + (middle - first), stencil + (last - first)). The return value of stable_partition is middle.

stable_partition differs from partition in that stable_partition is guaranteed to preserve relative order. That is, if x and y are elements in [first, last), such that pred(x) == pred(y), and if x precedes y, then it will still be true after stable_partition that x precedes y.

The algorithm’s execution is parallelized as determined by exec.

The following code snippet demonstrates how to use stable_partition to reorder a sequence so that even numbers precede odd numbers using the thrust::host execution policy for parallelization:

#include <thrust/partition.h>
#include <thrust/execution_policy.h>
...
struct is_even
{
  __host__ __device__
  bool operator()(const int &x)
  {
    return (x % 2) == 0;
  }
};
...
int A[] = {0, 1, 0, 1, 0, 1, 0, 1, 0,  1};
int S[] = {1, 2, 3, 4, 5, 6, 7, 8, 9, 10};
const int N = sizeof(A)/sizeof(int);
thrust::stable_partition(thrust::host, A, A + N, S, is_even());
// A is now {1, 1, 1, 1, 1, 0, 0, 0, 0, 0}
// S is unmodified