std::experimental::parallel::transform_reduce

From cppreference.com
Defined in header <experimental/numeric>
template<class InputIt, class UnaryOp, class T, class BinaryOp>

T transform_reduce(InputIt first, InputIt last,

                   UnaryOp unary_op, T init, BinaryOp binary_op);
(1) (parallelism TS)
template<class ExecutionPolicy,

         class InputIt, class UnaryOp, class T, class BinaryOp>
T transform_reduce(ExecutionPolicy&& policy,
                   InputIt first, InputIt last,

                   UnaryOp unary_op, T init, BinaryOp binary_op);
(2) (parallelism TS)

Applies unary_op to each element in the range [first; last) and reduces the results (possibly permuted and aggregated in unspecified manner) along with the initial value init over binary_op.

The behavior is non-deterministic if binary_op is not associative or not commutative.

The behavior is undefined if unary_op or binary_op modifies any element or invalidates any iterator in [first; last).

Parameters

first, last - the range of elements to apply the algorithm to
init - the initial value of the generalized sum
policy - the execution policy
unary_op - unary FunctionObject that will be applied to each element of the input range. The return type must be acceptable as input to binary_op
binary_op - binary FunctionObject that will be applied in unspecified order to the results of unary_op, the results of other binary_op and init.
Type requirements
-
InputIt must meet the requirements of InputIterator.

Return value

Generalized sum of init and unary_op(*first), unary_op(*(first+1)), ... unary_op(*(last-1)) over binary_op,

where generalized sum GSUM(op, a
1
, ..., a
N
)
is defined as follows:

  • if N=1, a
    1
  • if N > 1, op(GSUM(op, b
    1
    , ..., b
    K
    ), GSUM(op, b
    M
    , ..., b
    N
    ))
    where
  • b
    1
    , ..., b
    N
    may be any permutation of a1, ..., aN and
  • 1 < K+1 = M ≤ N

in other words, the results of unary_op may be grouped and arranged in arbitrary order.

Complexity

O(last - first) applications each of unary_op and binary_op.

Exceptions

  • If execution of a function invoked as part of the algorithm throws an exception,
  • if policy is parallel_vector_execution_policy, std::terminate is called
  • if policy is sequential_execution_policy or parallel_execution_policy, the algorithm exits with an exception_list containing all uncaught exceptions. If there was only one uncaught exception, the algorithm may rethrow it without wrapping in exception_list. It is unspecified how much work the algorithm will perform before returning after the first exception was encountered.
  • if policy is some other type, the behavior is implementation-defined
  • If the algorithm fails to allocate memory (either for itself or to construct an exception_list when handling a user exception), std::bad_alloc is thrown.

Notes

unary_op is not applied to init

If the range is empty, init is returned, unmodified

  • If policy is an instance of sequential_execution_policy, all operations are performed in the calling thread.
  • If policy is an instance of parallel_execution_policy, operations may be performed in unspecified number of threads, indeterminately sequenced with each other
  • If policy is an instance of parallel_vector_execution_policy, execution may be both parallelized and vectorized: function body boundaries are not respected and user code may be overlapped and combined in arbitrary manner (in particular, this implies that a user-provided Callable must not acquire a mutex to access a shared resource)

Example

transform_reduce can be used to parallelize std::inner_product:

#include <vector>
#include <iterator>
#include <functional>
#include <iostream>
#include <experimental/numeric>
#include <experimental/execution_policy>
#include <boost/iterator/zip_iterator.hpp>
#include <boost/tuple.hpp>
 
int main()
{
    std::vector<double> xvalues(10007, 1.0), yvalues(10007, 1.0);
 
    double result = std::experimental::parallel::transform_reduce(
        std::experimental::parallel::par,
        boost::iterators::make_zip_iterator(
            boost::make_tuple(std::begin(xvalues), std::begin(yvalues))),
        boost::iterators::make_zip_iterator(
            boost::make_tuple(std::end(xvalues), std::end(yvalues))),
        [](auto r) { return boost::get<0>(r) * boost::get<1>(r); }
        0.0,
        std::plus<>()
    );
    std::cout << result << '\n';
}

Output:

10007

See also

sums up a range of elements
(function template)
applies a function to a range of elements
(function template)
(parallelism TS)
similar to std::accumulate, except out of order
(function template)