reduction

group Built-in reduction operators

All built-in operators are defined for signed and unsigned integer types. Floating point types (Half, float, and double) are supported by all but bitwise operators. Arithmetic operators also cover complex types Complex<Half> and Complex<float>.

For details about reduction operators, See Library::register_reduction_operator.

template<typename T>
class SumReduction : public Legion::SumReduction<T>

#include <legate/redop/base.h>

Reduction with addition.

See Legion::SumReduction.

template<typename T>
class ProdReduction : public Legion::ProdReduction<T>

#include <legate/redop/base.h>

Reduction with multiplication.

See Legion::ProdReduction.

template<typename T>
class MaxReduction : public Legion::MaxReduction<T>

#include <legate/redop/base.h>

Reduction with the binary max operator.

See Legion::MaxReduction.

template<typename T>
class MinReduction : public Legion::MinReduction<T>

#include <legate/redop/base.h>

Reduction with the binary min operator.

See Legion::MinReduction.

template<typename T>
class OrReduction : public Legion::OrReduction<T>

#include <legate/redop/base.h>

Reduction with bitwise or.

See Legion::OrReduction.

template<typename T>
class AndReduction : public Legion::AndReduction<T>

#include <legate/redop/base.h>

Reduction with bitwise and.

See Legion::AndReduction.

template<typename T>
class XORReduction : public Legion::XorReduction<T>

#include <legate/redop/base.h>

Reduction with bitwise xor.

See Legion::XorReduction.