Solomon Bothwell
A 2D bin packing library based on on Jukka Jylänki's article "A Thousand Ways to Pack the Bin - A Practical Approach to Two-Dimensional Rectangle Bin Packing."
This library is intended for offline packing. All algorithms heuristics and optimizations from Jukka's article are included.
A web demo made with Flask and ReactJS is available "here" Packing performance varies drastically with different combinations of optimizations and datasets, so its important to under the settings and test a variety of them.
In [1]: import greedypacker
In [2]: M = greedypacker.BinManager(8, 4, pack_algo='shelf', heuristic='best_width_fit', wastemap=True, rotation=True)
In [3]: ITEM = greedypacker.Item(4, 2)
In [4]: ITEM2 = greedypacker.Item(5, 2)
In [5]: ITEM3 = greedypacker.Item(2, 2)
In [6]: M.add_items(ITEM, ITEM2, ITEM3)
In [7]: M.execute()
In [8]: M.bins
Out[8]: [Sheet(width=8, height=4, shelves=[{'y': 2, 'x': 8, 'available_width': 0, 'area': 6, 'vertical_offset': 0, 'items': [Item(width=5, height=2, x=0, y=0)]}, {'y': 2, 'x': 8, 'available_width': 4, 'area': 8, 'vertical_offset': 2, 'items': [Item(width=4, height=2, x=0, y=2)]}])]
All optimizations are passed in as keyword arguments when the GreedyPacker instance is created:
Item rotation can be disabled with the keyword argument rotation=False
Items can be pre-sorted according to a number of settings for the 'sorting_heuristic' keyword argument:
- ASCA: Sort By Area Ascending
- DESCA: Sort By Area Descending (This is the default setting)
- ASCSS: Sort By Shorter Side Ascending
- DESCSS: Sort By Shorter Side Descending
- ASCLS: Sort By Longer Side Ascending
- DESCLS: Sort By Longer Side Descending
- ASCPERIM: Sort By Perimeter Ascending
- DESCPERIM: Sort By Perimeter Descending
- ASCDIFF: Sort by The ABS Difference Between Sides Ascending
- DESCDIFF: Sort By The ABS Difference Between Sides Descending
- ASCRATIO: Sort By The Ratio of The Sides Ascending
- DESCRATIO: Sort By The Ratio of The Sides Descending
- False: Pack in the order added to the binmanager
See the algorithm specific pages linked above.
Requires Python>=3.0.
python -m unittest test