dd60ae31ae
Extends explanation for recursive bisection ids Cleans up Bisection State Removes license boilerplate from partitioner config Sorts Spatially and picks Sources and Sinks Uses sets for sources and sinks for now; see how large they will get Runs n cuts in parallel changing the slope and uses the best Clarifies balance <-> ratio naming
92 lines
2.8 KiB
C++
92 lines
2.8 KiB
C++
#ifndef OSRM_PARTITION_RECURSIVE_BISECTION_STATE_HPP_
|
||
#define OSRM_PARTITION_RECURSIVE_BISECTION_STATE_HPP_
|
||
|
||
#include <cstddef>
|
||
#include <vector>
|
||
|
||
#include "partition/bisection_graph.hpp"
|
||
#include "util/typedefs.hpp"
|
||
|
||
namespace osrm
|
||
{
|
||
namespace partition
|
||
{
|
||
|
||
// The recursive state describes the relation between our global graph and the recursive state in a
|
||
// recursive bisection.
|
||
//
|
||
// We describe the state with the use of two arrays:
|
||
//
|
||
// The id-arrays keeps nodes in order, based on their partitioning. Initially, it contains all nodes
|
||
// in sorted order:
|
||
//
|
||
// ids: [0,1,2,3,4,5,6,7,8,9]
|
||
//
|
||
// When partitioned (for this example we use even : odd), the arrays is re-ordered to group all
|
||
// elements within a single cell of the partition:
|
||
//
|
||
// ids: [0,2,4,6,8,1,3,5,7,9]
|
||
//
|
||
// This can be performed recursively by interpreting the arrays [0,2,4,6,8] and [1,3,5,7,9] as new
|
||
// input.
|
||
//
|
||
// The partitioning array describes all results of the partitioning in form of `left` or `right`.
|
||
// It is a sequence of bits for every id that describes if it is moved to the `front` or `back`
|
||
// during the split of the ID array. In the example, we would get the result
|
||
//
|
||
// bisection-ids: [0,1,0,1,0,1,0,1,0,1]
|
||
//
|
||
// Further partitioning [0,2,4,6,8] into [0,4,8], [2,6] and [1,3,5,7,9] into [1,3,9] and [5,7]
|
||
// the result would be:
|
||
//
|
||
// bisection-ids: [00,10,01,10,00,11,01,11,00,10]
|
||
|
||
/* Written out in a recursive tree form:
|
||
|
||
ids: [0,1,2,3,4,5,6,7,8,9]
|
||
mask: [0,1,0,1,0,1,0,1,0,1]
|
||
/ \
|
||
ids: [0,2,4,6,8] [1,3,5,7,9]
|
||
mask: [0,1,0,1,0] [0,0,1,1,0]
|
||
/ \ / \
|
||
ids: [0,4,8] [2,6] [1,3,9] [5,7]
|
||
|
||
The bisection ids then trace the path (left: 0, right: 1) through the tree:
|
||
|
||
ids: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9 ]
|
||
path: [00, 10, 01, 10, 00, 11, 01, 11, 00, 10]
|
||
|
||
*/
|
||
class RecursiveBisectionState
|
||
{
|
||
public:
|
||
// The ID in the partition array
|
||
using BisectionID = std::uint32_t;
|
||
using IDIterator = std::vector<NodeID>::const_iterator;
|
||
|
||
RecursiveBisectionState(const BisectionGraph &bisection_graph);
|
||
~RecursiveBisectionState();
|
||
|
||
BisectionID GetBisectionID(const NodeID nid) const;
|
||
|
||
// Bisects the node id array's sub-range based on the partition mask.
|
||
// Returns: partition point of the bisection: iterator to the second group's first element.
|
||
IDIterator ApplyBisection(const IDIterator begin,
|
||
const IDIterator end,
|
||
const std::vector<bool> &partition);
|
||
|
||
const IDIterator Begin() const;
|
||
const IDIterator End() const;
|
||
|
||
private:
|
||
const BisectionGraph &bisection_graph;
|
||
|
||
std::vector<NodeID> id_array;
|
||
std::vector<BisectionID> bisection_ids;
|
||
};
|
||
|
||
} // namespace partition
|
||
} // namespace osrm
|
||
|
||
#endif // OSRM_PARTITION_RECURSIVE_BISECTION_STATE_HPP_
|