[Use NicolPlus instead] this is the implementation of the nicol's algorithm as it is provided in "Ali Pinar and Cevdet Aykanat, Fast Optimal Load Balancing Algorithms for 1D Partitioning, JPDC, 2004" in figure 5a. NicolPlus should be preferred over NicolMinus due to speed.
More...
#include <oned/nicol_minus.hpp>
Detailed Description
template<typename T, typename Pr>
class oned::NicolMinus< T, Pr >
[Use NicolPlus instead] this is the implementation of the nicol's algorithm as it is provided in "Ali Pinar and Cevdet Aykanat, Fast Optimal Load Balancing Algorithms for 1D Partitioning, JPDC, 2004" in figure 5a. NicolPlus should be preferred over NicolMinus due to speed.
List of all members.
Static Public Member Functions |
static T | nicol_minus (int procCount, const Pr &prefixSumArray, int length, int *cutIndexes, T max) |
Static Private Member Functions |
static T | nicol_minus_internal (const Pr &w, int n, int p) |
Member Function Documentation
template<typename T , typename Pr >
T oned::NicolMinus< T, Pr >::nicol_minus |
( |
int |
procCount, |
|
|
const Pr & |
prefixSumArray, |
|
|
int |
length, |
|
|
int * |
cutIndexes, |
|
|
T |
max | |
|
) |
| | [static] |
- Parameters:
-
[in] | procCount | number of processors |
[in] | prefixSumArray | Prefix sum array that always starts with value 0 in the 0th index |
[in] | length | is the length of w |
[out] | cutIndexes | cut index points |
[in] | max | maximum load of a given array. If unknown, use -1. |
- Returns:
- Load of the max. loaded processor
template<typename T , typename Pr >
The documentation for this class was generated from the following files: