litvinenko / combinatorics
Classes for generating combinatorial sets and solving some combinatorial problems
Installs: 16
Dependents: 0
Suggesters: 0
Security: 0
Stars: 0
Watchers: 1
Forks: 0
Open Issues: 0
pkg:composer/litvinenko/combinatorics
Requires
- php: >=5.4.0
- litvinenko/app: 0.0.3
- litvinenko/object: *
- nicmart/tree: ~0.2
This package is not auto-updated.
Last update: 2025-10-15 15:39:16 UTC
README
This code can:
- generate permutations and some special cases of permutations
- solve combinatorial problems by means of branch and bound and beam search
PDP problem
Main interface class is Pdp.php. Uses https://bitbucket.org/absent1706/grebennik-pdp-python to check 3D constraints Uses https://github.com/absent1706/common_php_app to handle event-observer pattern Uses https://github.com/absent1706/extended_varien_object as a foundation of all objects
####Usage example
require_once 'vendor/autoload.php'; $xmlConfigFile = __DIR__.'/config.xml'; $paramsFile = __DIR__.'/data/params.json'; \Litvinenko\Common\App::init($xmlConfigFile); $launcher = new \Litvinenko\Combinatorics\Pdp\Pdp; $params = json_decode(file_get_contents($paramsFile), true); echo json_encode($launcher->getSolution($params['data'], $params['config']));
####Input params Example if input JSON:
{
"method": "gen", // hardcoded to use one solution method for now
"config": { // config section. sets how to solve pdp problem
"check_final_loading": false, // check 3D constraints for finally generated 'best' paths
"check_transitional_loading_probability": "20", // probability of checking 3D constraints in partial paths
"python_file": "/home/litvinenko/www/pdp-php/public_html/demo/pdphelper/pdphelper.py", // where script for checking 3D constraints is located
"precise": "5", // beam width in % of generated path variants
"weight_capacity": "1000", // vehicle weight_capacity
"load_area": { // vehicle load area size
"x": "100",
"y": "100",
"z": "100"
}
},
"data": {
"depot": [
"200", // depot x
"200" // depot y
],
"points": {
"1": [ // point id. Points with id < <point count> (first 2 points in this example) will be treated as pickups, other half of points - as deliveries
"117463", // point X coord
"476120", // point Y coord
"27", // point box size x (for pickups)
"38", // point box size y (for pickups)
"47", // point box size z (for pickups)
"38" // point weight (for pickups)
],
"2": [
"400.39",
"152.36",
"9.37",
"9.62",
"9.34",
"13.43"
],
"3": [
"345.42",
"414.24",
null,
null,
null,
null
],
"4": [
"10.35",
"17.68",
null,
null,
null,
null
]
}
}
}
On base of input points, point ojects are created in Litvinenko\Combinatorics\Pdp\Pdp::createPointsFromArray method. To correctly create PDP-pairs, points must have ids from 1 to 2N , where N - number of PDP pairs. Pickup with number i (i<N) will be associated to delivery with number i+N. I.g., for 2 PDP pairs we have 2 pickups (ids 1 and 2) and 2 deliveries (ids 3 and 4). Pdp pairs are: 1-3 and 2-4. So we must deliver one box from 1 to 3 and another from 2 to 4.
XML config for app:
<?xml version="1.0"?> <config> <events> <!-- <new_path_generated> <observers> <solutioninfocollector_new_path_generated> <class>\SolutionInfoCollector</class> <method>logGeneratedPath</method> <singleton>1</singleton> </solutioninfocollector_new_path_generated> </observers> </new_path_generated> <cant_load> <observers> <solutioninfocollector_cant_load> <class>\SolutionInfoCollector</class> <method>logNotLoadedPath</method> <singleton>1</singleton> </solutioninfocollector_cant_load> </observers> </cant_load> --> <point_add_before> <observers> <precise_generation_solver_point_add_before> <class>\Litvinenko\Combinatorics\Pdp\Solver\PreciseGenerationSolver</class> <method>canLoadObserver</method> <singleton>1</singleton> </precise_generation_solver_point_add_before> </observers> </point_add_before> </events> <developer_mode>1</developer_mode> </config>