Structure with type defintions for all reduction methods.
More...
#include <LAMAInterface.hpp>
Public Types |
typedef ValueType(* | sum )(const ValueType array[], const IndexType n) |
| Sum n contiguously stored values.
|
typedef ValueType(* | maxval )(const ValueType array[], const IndexType n) |
typedef ValueType(* | absMaxVal )(const ValueType array[], const IndexType n) |
| Absolute maximum of n contiguously stored values.
|
typedef ValueType(* | absMaxDiffVal )(const ValueType array1[], const ValueType array2[], const IndexType n) |
| Building absolute maximum of element-wise difference of vector elements.
|
Detailed Description
template<typename ValueType>
struct lama::UtilsInterface::Reductions< ValueType >
Structure with type defintions for all reduction methods.
- Template Parameters:
-
ValueType | specifies the value type used in the reduction. |
Member Typedef Documentation
template<typename ValueType >
Building absolute maximum of element-wise difference of vector elements.
- Parameters:
-
array1i[in] | first array |
array2i[in] | second array |
n | size of array1 and array2 |
- Returns:
- max( abs( array1[i] - array2[i] ) ),

Function is helpful to compute maximum norm for vectors and matrices
template<typename ValueType >
Absolute maximum of n contiguously stored values.
template<typename ValueType >
template<typename ValueType >
Sum n contiguously stored values.
- Parameters:
-
x | is an array of values |
n | is the size of x |
- Returns:
- sum of all values in x
The documentation for this struct was generated from the following file: