polyominator 1.0.1

Polyomino task solver.

Polyomino-solver

Solving the problem of accurately paving polyominoes with a given set of shapes. The program translates the polymino problem into the language of the X-algorithm. The algorithm-x library is used to solve Algorithm X.

Training manual

In general, using the package looks like this:

1. Connect the package and numpy

import numpy as np
from polyominator import polyomino

2. Set the field as an array, where 1 is a cell for paving, 0 is not for paving (like a Boolean array). For example, a 3x3 square field will look like this:

area = np.array([
    [1,1,1],
    [1,1,1],
    [1,1,1]
])

3. Set the shapes as arrays, where 1 is the element of the shape, 0 is the free space (again a Boolean array). For example, here is a 3-segment corner, a 2-segment stick, and a single segment, respectively:

corner = np.array([
    [1,1],
    [1,0],
])
stick = np.array([
    [1,1]
])
ceil = np.array([
    [1]
])

Please note that there is no fixed size of the array for the shapes. The main thing is that the shape fits into the array.

4. Pass the field and the list of shapes to the `polyomino' class.

figures = [corner,corner,stick,ceil]
pm = polyomino(figures=figures,area=area)

5. Get a single solution (with the argument get_only_one_solve = True) or a generator of all solutions (get_only_one_solve = False) using the solve method.

By default, get_only_one_solve = True.

solution = pm.solve()
#solutions_generator = pm.solve(False)

>>> [[0 0 2]
>>>  [0 1 2]
>>>  [3 1 1]]

What does the solution look like

The solution that (or which) the program will return to you, they look like an array with the same dimension as the field, and filled with numbers. If it seems to you that these numbers form shapes, it doesn't seem to you. When transferring figures to a class, they are numbered in the order of transfer (from 0). In this case:

• 0: corner,
• 1: corner,
• 2: stick,
• 3: ceil.

And these numbers form the silhouette of the figure in the place of the field in which this figure should be inserted.

Example

The field and the shapes for paving are given:

Then the search for the correct paving is presented in the file examples\example from readme.py . The solution looks like this:

>>> [[0 0 0 5]
>>>  [1 0 5 5]
>>>  [1 1 5 4]
>>>  [1 4 4 4]
>>>  [3 2 2 2]
>>>  [3 3 3 2]]

That is, the final arrangement of the shapes is:

Notes

• Theoretically, holes and protrusions in the fields are allowed, but they have not been tested for operability. For example:

example_area = [
    [1,1,1,0],
    [1,0,1,1],
    [1,1,1,0]
]

• There is an 'is_rotate` flag in the class. In theory, it regulates whether the pieces can rotate. By default, rotates are allowed, but I have not tested how the program behaves when it is forbidden.

Thanks

Thanks to the Talos Principle and its beautiful puzzles, in particular the last yellow one, for inspiring the development. I sent rays of goodness to Elohim and Milton, may they be healthy forever and ever.