Python numpy.may_share_memory() Examples

def check_may_share_memory_exact(a, b):
    got = np.may_share_memory(a, b, max_work=MAY_SHARE_EXACT)

    assert_equal(np.may_share_memory(a, b),
                 np.may_share_memory(a, b, max_work=MAY_SHARE_BOUNDS))

    a.fill(0)
    b.fill(0)
    a.fill(1)
    exact = b.any()

    err_msg = ""
    if got != exact:
        err_msg = "    " + "\n    ".join([
            "base_a - base_b = %r" % (a.__array_interface__['data'][0] - b.__array_interface__['data'][0],),
            "shape_a = %r" % (a.shape,),
            "shape_b = %r" % (b.shape,),
            "strides_a = %r" % (a.strides,),
            "strides_b = %r" % (b.strides,),
            "size_a = %r" % (a.size,),
            "size_b = %r" % (b.size,)
        ])

    assert_equal(got, exact, err_msg=err_msg) 

def test_empty(self):
        # Tests empty/like
        datatype = [('a', int), ('b', float), ('c', '|S8')]
        a = masked_array([(1, 1.1, '1.1'), (2, 2.2, '2.2'), (3, 3.3, '3.3')],
                         dtype=datatype)
        assert_equal(len(a.fill_value.item()), len(datatype))

        b = empty_like(a)
        assert_equal(b.shape, a.shape)
        assert_equal(b.fill_value, a.fill_value)

        b = empty(len(a), dtype=datatype)
        assert_equal(b.shape, a.shape)
        assert_equal(b.fill_value, a.fill_value)

        # check empty_like mask handling
        a = masked_array([1, 2, 3], mask=[False, True, False])
        b = empty_like(a)
        assert_(not np.may_share_memory(a.mask, b.mask))
        b = a.view(masked_array)
        assert_(np.may_share_memory(a.mask, b.mask)) 

def test_creation_maskcreation(self):
        # Tests how masks are initialized at the creation of Maskedarrays.
        data = arange(24, dtype=float)
        data[[3, 6, 15]] = masked
        dma_1 = MaskedArray(data)
        assert_equal(dma_1.mask, data.mask)
        dma_2 = MaskedArray(dma_1)
        assert_equal(dma_2.mask, dma_1.mask)
        dma_3 = MaskedArray(dma_1, mask=[1, 0, 0, 0] * 6)
        fail_if_equal(dma_3.mask, dma_1.mask)

        x = array([1, 2, 3], mask=True)
        assert_equal(x._mask, [True, True, True])
        x = array([1, 2, 3], mask=False)
        assert_equal(x._mask, [False, False, False])
        y = array([1, 2, 3], mask=x._mask, copy=False)
        assert_(np.may_share_memory(x.mask, y.mask))
        y = array([1, 2, 3], mask=x._mask, copy=True)
        assert_(not np.may_share_memory(x.mask, y.mask)) 

def test_empty(self):
        # Tests empty/like
        datatype = [('a', int), ('b', float), ('c', '|S8')]
        a = masked_array([(1, 1.1, '1.1'), (2, 2.2, '2.2'), (3, 3.3, '3.3')],
                         dtype=datatype)
        assert_equal(len(a.fill_value.item()), len(datatype))

        b = empty_like(a)
        assert_equal(b.shape, a.shape)
        assert_equal(b.fill_value, a.fill_value)

        b = empty(len(a), dtype=datatype)
        assert_equal(b.shape, a.shape)
        assert_equal(b.fill_value, a.fill_value)

        # check empty_like mask handling
        a = masked_array([1, 2, 3], mask=[False, True, False])
        b = empty_like(a)
        assert_(not np.may_share_memory(a.mask, b.mask))
        b = a.view(masked_array)
        assert_(np.may_share_memory(a.mask, b.mask)) 

def check_may_share_memory_exact(a, b):
    got = np.may_share_memory(a, b, max_work=MAY_SHARE_EXACT)

    assert_equal(np.may_share_memory(a, b),
                 np.may_share_memory(a, b, max_work=MAY_SHARE_BOUNDS))

    a.fill(0)
    b.fill(0)
    a.fill(1)
    exact = b.any()

    err_msg = ""
    if got != exact:
        err_msg = "    " + "\n    ".join([
            "base_a - base_b = %r" % (a.__array_interface__['data'][0] - b.__array_interface__['data'][0],),
            "shape_a = %r" % (a.shape,),
            "shape_b = %r" % (b.shape,),
            "strides_a = %r" % (a.strides,),
            "strides_b = %r" % (b.strides,),
            "size_a = %r" % (a.size,),
            "size_b = %r" % (b.size,)
        ])

    assert_equal(got, exact, err_msg=err_msg) 

def check_may_share_memory_exact(a, b):
    got = np.may_share_memory(a, b, max_work=MAY_SHARE_EXACT)

    assert_equal(np.may_share_memory(a, b),
                 np.may_share_memory(a, b, max_work=MAY_SHARE_BOUNDS))

    a.fill(0)
    b.fill(0)
    a.fill(1)
    exact = b.any()

    err_msg = ""
    if got != exact:
        err_msg = "    " + "\n    ".join([
            "base_a - base_b = %r" % (a.__array_interface__['data'][0] - b.__array_interface__['data'][0],),
            "shape_a = %r" % (a.shape,),
            "shape_b = %r" % (b.shape,),
            "strides_a = %r" % (a.strides,),
            "strides_b = %r" % (b.strides,),
            "size_a = %r" % (a.size,),
            "size_b = %r" % (b.size,)
        ])

    assert_equal(got, exact, err_msg=err_msg) 

def may_share_memory(a, b):
        # This is Fred suggestion for a quick and dirty way of checking
        # aliasing .. this can potentially be further refined (ticket #374)
        if _is_sparse(a) and _is_sparse(b):
            return (SparseType.may_share_memory(a, b.data) or
                    SparseType.may_share_memory(a, b.indices) or
                    SparseType.may_share_memory(a, b.indptr))
        if _is_sparse(b) and isinstance(a, numpy.ndarray):
            a, b = b, a
        if _is_sparse(a) and isinstance(b, numpy.ndarray):
            if (numpy.may_share_memory(a.data, b) or
                    numpy.may_share_memory(a.indices, b) or
                    numpy.may_share_memory(a.indptr, b)):
                # currently we can't share memory with a.shape as it is a tuple
                return True
        return False 

def test_creation_maskcreation(self):
        # Tests how masks are initialized at the creation of Maskedarrays.
        data = arange(24, dtype=float)
        data[[3, 6, 15]] = masked
        dma_1 = MaskedArray(data)
        assert_equal(dma_1.mask, data.mask)
        dma_2 = MaskedArray(dma_1)
        assert_equal(dma_2.mask, dma_1.mask)
        dma_3 = MaskedArray(dma_1, mask=[1, 0, 0, 0] * 6)
        fail_if_equal(dma_3.mask, dma_1.mask)

        x = array([1, 2, 3], mask=True)
        assert_equal(x._mask, [True, True, True])
        x = array([1, 2, 3], mask=False)
        assert_equal(x._mask, [False, False, False])
        y = array([1, 2, 3], mask=x._mask, copy=False)
        assert_(np.may_share_memory(x.mask, y.mask))
        y = array([1, 2, 3], mask=x._mask, copy=True)
        assert_(not np.may_share_memory(x.mask, y.mask)) 

def test_empty(self):
        # Tests empty/like
        datatype = [('a', int), ('b', float), ('c', '|S8')]
        a = masked_array([(1, 1.1, '1.1'), (2, 2.2, '2.2'), (3, 3.3, '3.3')],
                         dtype=datatype)
        assert_equal(len(a.fill_value.item()), len(datatype))

        b = empty_like(a)
        assert_equal(b.shape, a.shape)
        assert_equal(b.fill_value, a.fill_value)

        b = empty(len(a), dtype=datatype)
        assert_equal(b.shape, a.shape)
        assert_equal(b.fill_value, a.fill_value)

        # check empty_like mask handling
        a = masked_array([1, 2, 3], mask=[False, True, False])
        b = empty_like(a)
        assert_(not np.may_share_memory(a.mask, b.mask))
        b = a.view(masked_array)
        assert_(np.may_share_memory(a.mask, b.mask)) 

def test_borrow_input(self):
        """
        Tests that the contract for io.In is respected. When borrow=False, it should be
        impossible for outputs to be aliased to the input variables provided by the user,
        either through a view-map or a destroy map. New tests should be added in the future
        when borrow=True is implemented.
        """
        a = T.dmatrix()
        aval = numpy.random.rand(3, 3)

        # when borrow=False, test that a destroy map cannot alias output to input
        f = theano.function([In(a, borrow=False)], Out(a + 1, borrow=True))
        assert numpy.all(f(aval) == aval + 1)
        assert not numpy.may_share_memory(aval, f(aval))

        # when borrow=False, test that a viewmap cannot alias output to input
        f = theano.function([In(a, borrow=False)], Out(a[0, :], borrow=True))
        assert numpy.all(f(aval) == aval[0, :])
        assert not numpy.may_share_memory(aval, f(aval)) 

def test_empty(self):
        # Tests empty/like
        datatype = [('a', int), ('b', float), ('c', '|S8')]
        a = masked_array([(1, 1.1, '1.1'), (2, 2.2, '2.2'), (3, 3.3, '3.3')],
                         dtype=datatype)
        assert_equal(len(a.fill_value.item()), len(datatype))

        b = empty_like(a)
        assert_equal(b.shape, a.shape)
        assert_equal(b.fill_value, a.fill_value)

        b = empty(len(a), dtype=datatype)
        assert_equal(b.shape, a.shape)
        assert_equal(b.fill_value, a.fill_value)

        # check empty_like mask handling
        a = masked_array([1, 2, 3], mask=[False, True, False])
        b = empty_like(a)
        assert_(not np.may_share_memory(a.mask, b.mask))
        b = a.view(masked_array)
        assert_(np.may_share_memory(a.mask, b.mask)) 

def test_no_aliasing_2(self):
        # B and A take one another's values
        # no copying is necessary since each one is updated.
        orig_a = numpy.zeros((2, 2)) + .5
        orig_b = numpy.zeros((2, 2)) - .5
        A = self.shared(orig_a)
        B = self.shared(orig_b)

        data_of_a = data_of(A)
        data_of_b = data_of(B)

        f = pfunc([], [], updates=[(A, B), (B, A)])
        f()
        # correctness
        assert numpy.all(data_of(A) == -.5)
        assert numpy.all(data_of(B) == +.5)

        # shared vars may not be aliased
        assert not numpy.may_share_memory(data_of(A), data_of(B))

        # theano should have been smart enough to not make copies
        assert numpy.may_share_memory(data_of(A), data_of_b)
        assert numpy.may_share_memory(data_of(B), data_of_a) 

def check_may_share_memory_exact(a, b):
    got = np.may_share_memory(a, b, max_work=MAY_SHARE_EXACT)

    assert_equal(np.may_share_memory(a, b),
                 np.may_share_memory(a, b, max_work=MAY_SHARE_BOUNDS))

    a.fill(0)
    b.fill(0)
    a.fill(1)
    exact = b.any()

    err_msg = ""
    if got != exact:
        err_msg = "    " + "\n    ".join([
            "base_a - base_b = %r" % (a.__array_interface__['data'][0] - b.__array_interface__['data'][0],),
            "shape_a = %r" % (a.shape,),
            "shape_b = %r" % (b.shape,),
            "strides_a = %r" % (a.strides,),
            "strides_b = %r" % (b.strides,),
            "size_a = %r" % (a.size,),
            "size_b = %r" % (b.size,)
        ])

    assert_equal(got, exact, err_msg=err_msg) 

def test_creation_maskcreation(self):
        # Tests how masks are initialized at the creation of Maskedarrays.
        data = arange(24, dtype=float)
        data[[3, 6, 15]] = masked
        dma_1 = MaskedArray(data)
        assert_equal(dma_1.mask, data.mask)
        dma_2 = MaskedArray(dma_1)
        assert_equal(dma_2.mask, dma_1.mask)
        dma_3 = MaskedArray(dma_1, mask=[1, 0, 0, 0] * 6)
        fail_if_equal(dma_3.mask, dma_1.mask)

        x = array([1, 2, 3], mask=True)
        assert_equal(x._mask, [True, True, True])
        x = array([1, 2, 3], mask=False)
        assert_equal(x._mask, [False, False, False])
        y = array([1, 2, 3], mask=x._mask, copy=False)
        assert_(np.may_share_memory(x.mask, y.mask))
        y = array([1, 2, 3], mask=x._mask, copy=True)
        assert_(not np.may_share_memory(x.mask, y.mask)) 

def test_empty(self):
        # Tests empty/like
        datatype = [('a', int), ('b', float), ('c', '|S8')]
        a = masked_array([(1, 1.1, '1.1'), (2, 2.2, '2.2'), (3, 3.3, '3.3')],
                         dtype=datatype)
        assert_equal(len(a.fill_value.item()), len(datatype))

        b = empty_like(a)
        assert_equal(b.shape, a.shape)
        assert_equal(b.fill_value, a.fill_value)

        b = empty(len(a), dtype=datatype)
        assert_equal(b.shape, a.shape)
        assert_equal(b.fill_value, a.fill_value)

        # check empty_like mask handling
        a = masked_array([1, 2, 3], mask=[False, True, False])
        b = empty_like(a)
        assert_(not np.may_share_memory(a.mask, b.mask))
        b = a.view(masked_array)
        assert_(np.may_share_memory(a.mask, b.mask)) 

def test_creation_maskcreation(self):
        # Tests how masks are initialized at the creation of Maskedarrays.
        data = arange(24, dtype=float)
        data[[3, 6, 15]] = masked
        dma_1 = MaskedArray(data)
        assert_equal(dma_1.mask, data.mask)
        dma_2 = MaskedArray(dma_1)
        assert_equal(dma_2.mask, dma_1.mask)
        dma_3 = MaskedArray(dma_1, mask=[1, 0, 0, 0] * 6)
        fail_if_equal(dma_3.mask, dma_1.mask)

        x = array([1, 2, 3], mask=True)
        assert_equal(x._mask, [True, True, True])
        x = array([1, 2, 3], mask=False)
        assert_equal(x._mask, [False, False, False])
        y = array([1, 2, 3], mask=x._mask, copy=False)
        assert_(np.may_share_memory(x.mask, y.mask))
        y = array([1, 2, 3], mask=x._mask, copy=True)
        assert_(not np.may_share_memory(x.mask, y.mask)) 

def test_creation_maskcreation(self):
        # Tests how masks are initialized at the creation of Maskedarrays.
        data = arange(24, dtype=float)
        data[[3, 6, 15]] = masked
        dma_1 = MaskedArray(data)
        assert_equal(dma_1.mask, data.mask)
        dma_2 = MaskedArray(dma_1)
        assert_equal(dma_2.mask, dma_1.mask)
        dma_3 = MaskedArray(dma_1, mask=[1, 0, 0, 0] * 6)
        fail_if_equal(dma_3.mask, dma_1.mask)

        x = array([1, 2, 3], mask=True)
        assert_equal(x._mask, [True, True, True])
        x = array([1, 2, 3], mask=False)
        assert_equal(x._mask, [False, False, False])
        y = array([1, 2, 3], mask=x._mask, copy=False)
        assert_(np.may_share_memory(x.mask, y.mask))
        y = array([1, 2, 3], mask=x._mask, copy=True)
        assert_(not np.may_share_memory(x.mask, y.mask)) 

def test_empty(self):
        # Tests empty/like
        datatype = [('a', int), ('b', float), ('c', '|S8')]
        a = masked_array([(1, 1.1, '1.1'), (2, 2.2, '2.2'), (3, 3.3, '3.3')],
                         dtype=datatype)
        assert_equal(len(a.fill_value.item()), len(datatype))

        b = empty_like(a)
        assert_equal(b.shape, a.shape)
        assert_equal(b.fill_value, a.fill_value)

        b = empty(len(a), dtype=datatype)
        assert_equal(b.shape, a.shape)
        assert_equal(b.fill_value, a.fill_value)

        # check empty_like mask handling
        a = masked_array([1, 2, 3], mask=[False, True, False])
        b = empty_like(a)
        assert_(not np.may_share_memory(a.mask, b.mask))
        b = a.view(masked_array)
        assert_(np.may_share_memory(a.mask, b.mask)) 

def test_check_array_memmap(copy):
    X = np.ones((4, 4))
    with TempMemmap(X, mmap_mode='r') as X_memmap:
        X_checked = check_array(X_memmap, copy=copy)
        assert np.may_share_memory(X_memmap, X_checked) == (not copy)
        assert X_checked.flags['WRITEABLE'] == copy 

def test_fast_power(self):
        x = np.array([1, 2, 3], np.int16)
        res = x**2.0
        assert_((x**2.00001).dtype is res.dtype)
        assert_array_equal(res, [1, 4, 9])
        # check the inplace operation on the casted copy doesn't mess with x
        assert_(not np.may_share_memory(res, x))
        assert_array_equal(x, [1, 2, 3])

        # Check that the fast path ignores 1-element not 0-d arrays
        res = x ** np.array([[[2]]])
        assert_equal(res.shape, (1, 1, 3)) 

def __array_finalize__(self, obj):
        if hasattr(obj, '_mmap') and np.may_share_memory(self, obj):
            self._mmap = obj._mmap
            self.filename = obj.filename
            self.offset = obj.offset
            self.mode = obj.mode
        else:
            self._mmap = None
            self.filename = None
            self.offset = None
            self.mode = None 

def _compare_index_result(self, arr, index, mimic_get, no_copy):
        """Compare mimicked result to indexing result.
        """
        arr = arr.copy()
        indexed_arr = arr[index]
        assert_array_equal(indexed_arr, mimic_get)
        # Check if we got a view, unless its a 0-sized or 0-d array.
        # (then its not a view, and that does not matter)
        if indexed_arr.size != 0 and indexed_arr.ndim != 0:
            assert_(np.may_share_memory(indexed_arr, arr) == no_copy)
            # Check reference count of the original array
            if no_copy:
                # refcount increases by one:
                assert_equal(sys.getrefcount(arr), 3)
            else:
                assert_equal(sys.getrefcount(arr), 2)

        # Test non-broadcast setitem:
        b = arr.copy()
        b[index] = mimic_get + 1000
        if b.size == 0:
            return # nothing to compare here...
        if no_copy and indexed_arr.ndim != 0:
            # change indexed_arr in-place to manipulate original:
            indexed_arr += 1000
            assert_array_equal(arr, b)
            return
        # Use the fact that the array is originally an arange:
        arr.flat[indexed_arr.ravel()] += 1000
        assert_array_equal(arr, b) 

def test_k_means_init_centers():
    # This test is used to check KMeans won't mutate the user provided input
    # array silently even if input data and init centers have the same type
    X_small = np.array([[1.1, 1.1], [-7.5, -7.5], [-1.1, -1.1], [7.5, 7.5]])
    init_centers = np.array([[0.0, 0.0], [5.0, 5.0], [-5.0, -5.0]])
    for dtype in [np.int32, np.int64, np.float32, np.float64]:
        X_test = dtype(X_small)
        init_centers_test = dtype(init_centers)
        assert_array_equal(init_centers, init_centers_test)
        km = KMeans(init=init_centers_test, n_clusters=3, n_init=1)
        km.fit(X_test)
        assert_equal(False, np.may_share_memory(km.cluster_centers_,
                                                init_centers)) 

def test_reindex(test_data):
    identity = test_data.series.reindex(test_data.series.index)

    # __array_interface__ is not defined for older numpies
    # and on some pythons
    try:
        assert np.may_share_memory(test_data.series.index, identity.index)
    except AttributeError:
        pass

    assert identity.index.is_(test_data.series.index)
    assert identity.index.identical(test_data.series.index)

    subIndex = test_data.series.index[10:20]
    subSeries = test_data.series.reindex(subIndex)

    for idx, val in compat.iteritems(subSeries):
        assert val == test_data.series[idx]

    subIndex2 = test_data.ts.index[10:20]
    subTS = test_data.ts.reindex(subIndex2)

    for idx, val in compat.iteritems(subTS):
        assert val == test_data.ts[idx]
    stuffSeries = test_data.ts.reindex(subIndex)

    assert np.isnan(stuffSeries).all()

    # This is extremely important for the Cython code to not screw up
    nonContigIndex = test_data.ts.index[::2]
    subNonContig = test_data.ts.reindex(nonContigIndex)
    for idx, val in compat.iteritems(subNonContig):
        assert val == test_data.ts[idx]

    # return a copy the same index here
    result = test_data.ts.reindex()
    assert not (result is test_data.ts) 

def __array_finalize__(self, obj):
        if hasattr(obj, '_mmap') and np.may_share_memory(self, obj):
            self._mmap = obj._mmap
            self.filename = obj.filename
            self.offset = obj.offset
            self.mode = obj.mode
        else:
            self._mmap = None
            self.filename = None
            self.offset = None
            self.mode = None 

def test_basic(self):
        d = np.ones((50, 60))
        d2 = np.ones((30, 60, 6))
        assert_(np.may_share_memory(d, d))
        assert_(np.may_share_memory(d, d[::-1]))
        assert_(np.may_share_memory(d, d[::2]))
        assert_(np.may_share_memory(d, d[1:, ::-1]))

        assert_(not np.may_share_memory(d[::-1], d2))
        assert_(not np.may_share_memory(d[::2], d2))
        assert_(not np.may_share_memory(d[1:, ::-1], d2))
        assert_(np.may_share_memory(d2[1:, ::-1], d2))


# Utility 

def test_may_share_memory_bad_max_work():
    x = np.zeros([1])
    assert_raises(OverflowError, np.may_share_memory, x, x, max_work=10**100)
    assert_raises(OverflowError, np.shares_memory, x, x, max_work=10**100) 

def check_may_share_memory_easy_fuzz(get_max_work, same_steps, min_count):
    # Check that overlap problems with common strides are solved with
    # little work.
    x = np.zeros([17,34,71,97], dtype=np.int16)

    feasible = 0
    infeasible = 0

    pair_iter = iter_random_view_pairs(x, same_steps)

    while min(feasible, infeasible) < min_count:
        a, b = next(pair_iter)

        bounds_overlap = np.may_share_memory(a, b)
        may_share_answer = np.may_share_memory(a, b)
        easy_answer = np.may_share_memory(a, b, max_work=get_max_work(a, b))
        exact_answer = np.may_share_memory(a, b, max_work=MAY_SHARE_EXACT)

        if easy_answer != exact_answer:
            # assert_equal is slow...
            assert_equal(easy_answer, exact_answer)

        if may_share_answer != bounds_overlap:
            assert_equal(may_share_answer, bounds_overlap)

        if bounds_overlap:
            if exact_answer:
                feasible += 1
            else:
                infeasible += 1 

def _compare_index_result(self, arr, index, mimic_get, no_copy):
        """Compare mimicked result to indexing result.
        """
        arr = arr.copy()
        indexed_arr = arr[index]
        assert_array_equal(indexed_arr, mimic_get)
        # Check if we got a view, unless its a 0-sized or 0-d array.
        # (then its not a view, and that does not matter)
        if indexed_arr.size != 0 and indexed_arr.ndim != 0:
            assert_(np.may_share_memory(indexed_arr, arr) == no_copy)
            # Check reference count of the original array
            if HAS_REFCOUNT:
                if no_copy:
                    # refcount increases by one:
                    assert_equal(sys.getrefcount(arr), 3)
                else:
                    assert_equal(sys.getrefcount(arr), 2)

        # Test non-broadcast setitem:
        b = arr.copy()
        b[index] = mimic_get + 1000
        if b.size == 0:
            return  # nothing to compare here...
        if no_copy and indexed_arr.ndim != 0:
            # change indexed_arr in-place to manipulate original:
            indexed_arr += 1000
            assert_array_equal(arr, b)
            return
        # Use the fact that the array is originally an arange:
        arr.flat[indexed_arr.ravel()] += 1000
        assert_array_equal(arr, b) 

def test_no_aliasing_1(self):
        # B is a shared variable, A is updated with B's contents
        # since B is being updated as well, we don't need to copy anything
        # to avoid aliasing shared variables.
        A = self.shared(numpy.zeros((2, 2)) + .5)
        B = self.shared(numpy.zeros((2, 2)) - .5)
        C = tensor.dmatrix()
        f = pfunc([C], [], updates=[(A, B), (B, C)])
        z = numpy.zeros((2, 2))
        f(z)
        assert not numpy.may_share_memory(data_of(A), data_of(B))
        # Theano tries to maintain its own memory space.
        assert not numpy.may_share_memory(z, data_of(B))
        assert numpy.all(data_of(B) == z)