0th subset disappearing problem when getting all solutions

[lsjbh45]

Hello, while testing randomly generated test cases, I came across a strange test case that seems to be a bug.
The test case found is as follow:

import numpy as np
import exact_cover as ec

tc = ((1, 1, 0,), (1, 0, 1,), (1, 0, 1,), (1, 1, 0,), (1, 1, 0,), (0, 0, 0,),
      (0, 0, 0,), (0, 0, 1,), (1, 1, 0,), (1, 0, 1,), (0, 1, 0,), (1, 0, 0,),)

S = np.array(tc, dtype=bool)
sols = ec.get_all_solutions(S)
print(sols)
# result: {(9, 10), (7, 4), (2, 10), (7, 11, 10), (7, 3), (7,), (1, 10), (7, 8)}
# expected: {(9, 10), (7, 4), (2, 10), (7, 11, 10), (7, 3), (7, 0), (1, 10), (7, 8)}

Additional information:

• If the 0th subset is chosen as the "last" subset of S* to form an exact cover (�i.e. selected at last, while selecting subsets to cover from the preceding elements of X), this problem is presumed to occur.
• It has been confirmed that this problem does not occur in the get_exact_cover method, but only for the get_all_solutions method.

import numpy as np
import exact_cover as ec

tc = ((0, 0, 0, 1, 0), (1, 1, 0, 0, 0), (0, 0, 1, 0, 1))

S = np.array(tc, dtype=bool)
sols = ec.get_all_solutions(S)
print(sols)
# result: {(1, 2)}
# expected: {(1, 2, 0)}

sol = ec.get_exact_cover(S)
print(sol)
# result: [1 2 0], as expected

Thanks for reading!

[kmohanram]

looks like i opened a new issue that is similar to (or at least overlaps) this one ...

[kmohanram]

one way to temporarily address this issue/bug is to add a dummy first row that is all zeros to the input ... it will never show up in the cover and one just has to adjust indexing from the returned results. it looks like the encoding of the results and the trunc step in _solutions_array_to_set in wrapper.py will have to be reconciled to fix this subtle issue/bug.

[lsjbh45]

@kmohanram Thanks for additional confirmation on this issue! :)
The way you temporarily resolved the problem was exactly in line with the action I took; i.e. use 1-based indexing instead of 0-based indexing.
I hope that the underlying cause of this problem will be resolved soon.

[jwg4]

looking at this