thrust::exclusive_scan_by_key

Defined in thrust/scan.h

template<typename DerivedPolicy, typename InputIterator1, typename InputIterator2, typename OutputIterator, typename T, typename BinaryPredicate> OutputIterator thrust::exclusive_scan_by_key(const thrust::detail::execution_policy_base<DerivedPolicy> &exec, InputIterator1 first1, InputIterator1 last1, InputIterator2 first2, OutputIterator result, T init, BinaryPredicate binary_pred)

exclusive_scan_by_key computes an exclusive key-value or ‘segmented’ prefix sum operation. The term ‘exclusive’ means that each result does not include the corresponding input operand in the partial sum. The term ‘segmented’ means that the partial sums are broken into distinct segments. In other words, within each segment a separate exclusive scan operation is computed. Refer to the code sample below for example usage.

This version of exclusive_scan_by_key uses the value init to initialize the exclusive scan operation.

This version of exclusive_scan_by_key uses the binary predicate binary_pred to compare adjacent keys. Specifically, consecutive iterators i and i+1 in the range [first1, last1) belong to the same segment if binary_pred(*i, *(i+1)) is true, and belong to different segments otherwise.

Results are not deterministic for pseudo-associative operators (e.g., addition of floating-point types). Results for pseudo-associative operators may vary from run to run.

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

The following code snippet demonstrates how to use exclusive_scan_by_key using the thrust::host execution policy for parallelization:

#include <thrust/scan.h>
#include <thrust/functional.h>
#include <thrust/execution_policy.h>
...

int keys[10] = {0, 0, 0, 1, 1, 2, 3, 3, 3, 3};
int vals[10] = {1, 1, 1, 1, 1, 1, 1, 1, 1, 1};

int init = 5;

thrust::equal_to<int> binary_pred;

thrust::exclusive_scan_by_key(thrust::host, key, key + 10, vals, vals, init, binary_pred); // in-place scan

// vals is now {5, 6, 7, 5, 6, 5, 5, 6, 7, 8};