File size: 5,418 Bytes
d1ceb73
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
from sympy import ZZ, Matrix
from sympy.polys.matrices import DM, DomainMatrix
from sympy.polys.matrices.ddm import DDM
from sympy.polys.matrices.sdm import SDM

import pytest

zeros = lambda shape, K: DomainMatrix.zeros(shape, K).to_dense()
eye = lambda n, K: DomainMatrix.eye(n, K).to_dense()


#
# DomainMatrix.nullspace can have a divided answer or can return an undivided
# uncanonical answer. The uncanonical answer is not unique but we can make it
# unique by making it primitive (remove gcd). The tests here all show the
# primitive form. We test two things:
#
#   A.nullspace().primitive()[1] == answer.
#   A.nullspace(divide_last=True) == _divide_last(answer).
#
# The nullspace as returned by DomainMatrix and related classes is the
# transpose of the nullspace as returned by Matrix. Matrix returns a list of
# of column vectors whereas DomainMatrix returns a matrix whose rows are the
# nullspace vectors.
#


NULLSPACE_EXAMPLES = [

    (
        'zz_1',
         DM([[ 1, 2, 3]], ZZ),
         DM([[-2, 1, 0],
             [-3, 0, 1]], ZZ),
    ),

    (
        'zz_2',
         zeros((0, 0), ZZ),
         zeros((0, 0), ZZ),
    ),

    (
        'zz_3',
        zeros((2, 0), ZZ),
        zeros((0, 0), ZZ),
    ),

    (
        'zz_4',
        zeros((0, 2), ZZ),
        eye(2, ZZ),
    ),

    (
        'zz_5',
        zeros((2, 2), ZZ),
        eye(2, ZZ),
    ),

    (
        'zz_6',
        DM([[1, 2],
            [3, 4]], ZZ),
        zeros((0, 2), ZZ),
    ),

    (
        'zz_7',
        DM([[1, 1],
            [1, 1]], ZZ),
        DM([[-1, 1]], ZZ),
    ),

    (
        'zz_8',
        DM([[1],
            [1]], ZZ),
        zeros((0, 1), ZZ),
    ),

    (
        'zz_9',
        DM([[1, 1]], ZZ),
        DM([[-1, 1]], ZZ),
    ),

    (
        'zz_10',
        DM([[0, 0, 0, 0, 0, 1, 0, 0, 0, 0],
            [1, 0, 0, 0, 0, 0, 1, 0, 0, 0],
            [0, 1, 0, 0, 0, 0, 0, 1, 0, 0],
            [0, 0, 0, 1, 0, 0, 0, 0, 1, 0],
            [0, 0, 0, 0, 1, 0, 0, 0, 0, 1]], ZZ),
        DM([[ 0,  0, 1,  0,  0, 0, 0, 0, 0, 0],
            [-1,  0, 0,  0,  0, 0, 1, 0, 0, 0],
            [ 0, -1, 0,  0,  0, 0, 0, 1, 0, 0],
            [ 0,  0, 0, -1,  0, 0, 0, 0, 1, 0],
            [ 0,  0, 0,  0, -1, 0, 0, 0, 0, 1]], ZZ),
    ),

]


def _to_DM(A, ans):
    """Convert the answer to DomainMatrix."""
    if isinstance(A, DomainMatrix):
        return A.to_dense()
    elif isinstance(A, DDM):
        return DomainMatrix(list(A), A.shape, A.domain).to_dense()
    elif isinstance(A, SDM):
        return DomainMatrix(dict(A), A.shape, A.domain).to_dense()
    else:
        assert False # pragma: no cover


def _divide_last(null):
    """Normalize the nullspace by the rightmost non-zero entry."""
    null = null.to_field()

    if null.is_zero_matrix:
        return null

    rows = []
    for i in range(null.shape[0]):
        for j in reversed(range(null.shape[1])):
            if null[i, j]:
                rows.append(null[i, :] / null[i, j])
                break
        else:
            assert False # pragma: no cover

    return DomainMatrix.vstack(*rows)


def _check_primitive(null, null_ans):
    """Check that the primitive of the answer matches."""
    null = _to_DM(null, null_ans)
    cont, null_prim = null.primitive()
    assert null_prim == null_ans


def _check_divided(null, null_ans):
    """Check the divided answer."""
    null = _to_DM(null, null_ans)
    null_ans_norm = _divide_last(null_ans)
    assert null == null_ans_norm


@pytest.mark.parametrize('name, A, A_null', NULLSPACE_EXAMPLES)
def test_Matrix_nullspace(name, A, A_null):
    A = A.to_Matrix()

    A_null_cols = A.nullspace()

    # We have to patch up the case where the nullspace is empty
    if A_null_cols:
        A_null_found = Matrix.hstack(*A_null_cols)
    else:
        A_null_found = Matrix.zeros(A.cols, 0)

    A_null_found = A_null_found.to_DM().to_field().to_dense()

    # The Matrix result is the transpose of DomainMatrix result.
    A_null_found = A_null_found.transpose()

    _check_divided(A_null_found, A_null)


@pytest.mark.parametrize('name, A, A_null', NULLSPACE_EXAMPLES)
def test_dm_dense_nullspace(name, A, A_null):
    A = A.to_field().to_dense()
    A_null_found = A.nullspace(divide_last=True)
    _check_divided(A_null_found, A_null)


@pytest.mark.parametrize('name, A, A_null', NULLSPACE_EXAMPLES)
def test_dm_sparse_nullspace(name, A, A_null):
    A = A.to_field().to_sparse()
    A_null_found = A.nullspace(divide_last=True)
    _check_divided(A_null_found, A_null)


@pytest.mark.parametrize('name, A, A_null', NULLSPACE_EXAMPLES)
def test_ddm_nullspace(name, A, A_null):
    A = A.to_field().to_ddm()
    A_null_found, _ = A.nullspace()
    _check_divided(A_null_found, A_null)


@pytest.mark.parametrize('name, A, A_null', NULLSPACE_EXAMPLES)
def test_sdm_nullspace(name, A, A_null):
    A = A.to_field().to_sdm()
    A_null_found, _ = A.nullspace()
    _check_divided(A_null_found, A_null)


@pytest.mark.parametrize('name, A, A_null', NULLSPACE_EXAMPLES)
def test_dm_dense_nullspace_fracfree(name, A, A_null):
    A = A.to_dense()
    A_null_found = A.nullspace()
    _check_primitive(A_null_found, A_null)


@pytest.mark.parametrize('name, A, A_null', NULLSPACE_EXAMPLES)
def test_dm_sparse_nullspace_fracfree(name, A, A_null):
    A = A.to_sparse()
    A_null_found = A.nullspace()
    _check_primitive(A_null_found, A_null)