Python numpy.vectorize() Examples

def test_static_nonlinearity(self):
        # Linear dynamical system
        linsys = self.siso_linsys
        ioslin = ios.LinearIOSystem(linsys)

        # Nonlinear saturation
        sat = lambda u: u if abs(u) < 1 else np.sign(u)
        sat_output = lambda t, x, u, params: sat(u)
        nlsat =  ios.NonlinearIOSystem(None, sat_output, inputs=1, outputs=1)

        # Set up parameters for simulation
        T, U, X0 = self.T, 2 * self.U, self.X0
        Usat = np.vectorize(sat)(U)

        # Make sure saturation works properly by comparing linear system with
        # saturated input to nonlinear system with saturation composition
        lti_t, lti_y, lti_x = ct.forced_response(linsys, T, Usat, X0)
        ios_t, ios_y, ios_x = ios.input_output_response(
            ioslin * nlsat, T, U, X0, return_x=True)
        np.testing.assert_array_almost_equal(lti_t, ios_t)
        np.testing.assert_array_almost_equal(lti_y, ios_y, decimal=2) 

def test_keywords2_ticket_2100(self):
        # Test kwarg support: enhancement ticket 2100

        def foo(a, b=1):
            return a + b

        f = vectorize(foo)
        args = np.array([1, 2, 3])
        r1 = f(a=args)
        r2 = np.array([2, 3, 4])
        assert_array_equal(r1, r2)
        r1 = f(b=1, a=args)
        assert_array_equal(r1, r2)
        r1 = f(args, b=2)
        r2 = np.array([3, 4, 5])
        assert_array_equal(r1, r2) 

def dict_lookup(dict_, keys, default=None):
    """Looks up :attr:`keys` in the dict, returns the corresponding values.

    The :attr:`default` is used for keys not present in the dict.

    Args:
        dict_ (dict): A dictionary for lookup.
        keys: A numpy array or a (possibly nested) list of keys.
        default (optional): Value to be returned when a key is not in
            :attr:`dict_`. Error is raised if :attr:`default` is not given and
            key is not in the dict.

    Returns:
        A numpy array of values with the same structure as :attr:`keys`.

    Raises:
        TypeError: If key is not in :attr:`dict_` and :attr:`default` is `None`.
    """
    return np.vectorize(lambda x: dict_.get(x, default))(keys) 

def testRelativePositionalEmbeddingLayer(self):
    with self.session(use_gpu=False):
      radius = 3
      p = layers.RelativePositionalEmbeddingLayer.Params().Set(
          name='rel_position_emb', radius=radius, dim=4)
      layer = p.Instantiate()
      indices = np.array([-5, -2, 0, 1, 4], dtype=np.int32)
      pos_emb = layer.FPropDefaultTheta(tf.convert_to_tensor(indices))

      self.evaluate(tf.global_variables_initializer())
      actual_pos_emb, full_emb = self.evaluate([pos_emb, layer.vars.w])

      clipped_indices = np.vectorize(lambda x: max(-radius, min(radius, x)))(
          indices) + radius
      expected_output = np.take_along_axis(full_emb,
                                           np.expand_dims(clipped_indices, -1),
                                           0)
      print('expected_position_embs:', expected_output)
      print('actual_position_embs:', actual_pos_emb)
      self.assertAllClose(actual_pos_emb, expected_output) 

def plot3d(self, scale=0.32):
		r"""Plot 3d scatter plot of benchmark function.

		Args:
			scale (float): Scale factor for points.
		"""
		fig = plt.figure()
		ax = Axes3D(fig)
		func = self.function()
		Xr, Yr = arange(self.Lower, self.Upper, scale), arange(self.Lower, self.Upper, scale)
		X, Y = meshgrid(Xr, Yr)
		Z = vectorize(self.__2dfun)(X, Y, func)
		ax.plot_surface(X, Y, Z, rstride=8, cstride=8, alpha=0.3)
		ax.contourf(X, Y, Z, zdir='z', offset=-10, cmap=cm.coolwarm)
		ax.set_xlabel('X')
		ax.set_ylabel('Y')
		ax.set_zlabel('Z')
		plt.show()

# vim: tabstop=3 noexpandtab shiftwidth=3 softtabstop=3 

def xindex(array, row_num, col_num=None, area_num=1):
    is_reference = isinstance(array, Ranges)
    if is_reference:
        arrays = [Ranges((rng,), array.values).value for rng in array.ranges]
    else:
        arrays = [array]

    row_num, col_num, area_num = parse_ranges(row_num, col_num, area_num)[0]

    res = np.vectorize(_index, excluded={0}, otypes=[object])(
        arrays, row_num, col_num, area_num, is_reference,
        isinstance(row_num, np.ndarray)
    )
    if not res.shape:
        res = res.reshape(1, 1)
    return res.view(Array) 

def _munp(self, n, beta, m):
        """
        Returns the n-th non-central moment of the crystalball function.
        """
        N = 1.0 / (m/beta / (m-1) * np.exp(-beta**2 / 2.0) + _norm_pdf_C * _norm_cdf(beta))

        def n_th_moment(n, beta, m):
            """
            Returns n-th moment. Defined only if n+1 < m
            Function cannot broadcast due to the loop over n
            """
            A = (m/beta)**m * np.exp(-beta**2 / 2.0)
            B = m/beta - beta
            rhs = 2**((n-1)/2.0) * sc.gamma((n+1)/2) * (1.0 + (-1)**n * sc.gammainc((n+1)/2, beta**2 / 2))
            lhs = np.zeros(rhs.shape)
            for k in range(n + 1):
                lhs += sc.binom(n, k) * B**(n-k) * (-1)**k / (m - k - 1) * (m/beta)**(-m + k + 1)
            return A * lhs + rhs

        return N * _lazywhere(np.atleast_1d(n + 1 < m),
                              (n, beta, m),
                              np.vectorize(n_th_moment, otypes=[np.float]),
                              np.inf) 

def _rvs(self, low, high):
        """An array of *size* random integers >= ``low`` and < ``high``."""
        if self._size is not None:
            # Numpy's RandomState.randint() doesn't broadcast its arguments.
            # Use `broadcast_to()` to extend the shapes of low and high
            # up to self._size.  Then we can use the numpy.vectorize'd
            # randint without needing to pass it a `size` argument.
            low = broadcast_to(low, self._size)
            high = broadcast_to(high, self._size)
        randint = np.vectorize(self._random_state.randint, otypes=[np.int_])
        return randint(low, high) 

def get_param_df(self):
        ddd = {}
        for i, trial in enumerate(self.trials):
            dd = {}
            dd["marginal_ll"] = trial["result"]["loss"]
            for item in trial["result"]["space"].values():
                for key, value in item.items():
                    dd[key] = value
            ddd[i] = dd
        df_space = pd.DataFrame(ddd)
        df_space = df_space.T
        n_params_dataset = np.vectorize(
            partial(n_params, self.trainer.gene_dataset.nb_genes)
        )
        df_space["n_params"] = n_params_dataset(
            df_space["n_layers"], df_space["n_hidden"], df_space["n_latent"]
        )
        df_space = df_space[
            [
                "marginal_ll",
                "n_layers",
                "n_hidden",
                "n_latent",
                "reconstruction_loss",
                "dropout_rate",
                "lr",
                "n_epochs",
                "n_params",
            ]
        ]
        df_space = df_space.sort_values(by="marginal_ll")
        df_space["run index"] = df_space.index
        df_space.index = np.arange(1, df_space.shape[0] + 1)
        return df_space 

def knn_purity(latent, label, n_neighbors=30):
    nbrs = NearestNeighbors(n_neighbors=n_neighbors + 1).fit(latent)
    indices = nbrs.kneighbors(latent, return_distance=False)[:, 1:]
    neighbors_labels = np.vectorize(lambda i: label[i])(indices)

    # pre cell purity scores
    scores = ((neighbors_labels - label.reshape(-1, 1)) == 0).mean(axis=1)
    res = [
        np.mean(scores[label == i]) for i in np.unique(label)
    ]  # per cell-type purity

    return np.mean(res) 

def test_ufunc(self):
        import math
        f = vectorize(math.cos)
        args = np.array([0, 0.5 * np.pi, np.pi, 1.5 * np.pi, 2 * np.pi])
        r1 = f(args)
        r2 = np.cos(args)
        assert_array_almost_equal(r1, r2) 

def test_keywords(self):

        def foo(a, b=1):
            return a + b

        f = vectorize(foo)
        args = np.array([1, 2, 3])
        r1 = f(args)
        r2 = np.array([2, 3, 4])
        assert_array_equal(r1, r2)
        r1 = f(args, 2)
        r2 = np.array([3, 4, 5])
        assert_array_equal(r1, r2) 

def test_scalar(self):
        def addsubtract(a, b):
            if a > b:
                return a - b
            else:
                return a + b

        f = vectorize(addsubtract)
        r = f([0, 3, 6, 9], 5)
        assert_array_equal(r, [5, 8, 1, 4]) 

def test_large(self):
        x = np.linspace(-3, 2, 10000)
        f = vectorize(lambda x: x)
        y = f(x)
        assert_array_equal(y, x) 

def test_keywords_no_func_code(self):
        # This needs to test a function that has keywords but
        # no func_code attribute, since otherwise vectorize will
        # inspect the func_code.
        import random
        try:
            vectorize(random.randrange)  # Should succeed
        except:
            raise AssertionError() 

def test_keywords3_ticket_2100(self):
        # Test excluded with mixed positional and kwargs: ticket 2100
        def mypolyval(x, p):
            _p = list(p)
            res = _p.pop(0)
            while _p:
                res = res * x + _p.pop(0)
            return res

        vpolyval = np.vectorize(mypolyval, excluded=['p', 1])
        ans = [3, 6]
        assert_array_equal(ans, vpolyval(x=[0, 1], p=[1, 2, 3]))
        assert_array_equal(ans, vpolyval([0, 1], p=[1, 2, 3]))
        assert_array_equal(ans, vpolyval([0, 1], [1, 2, 3])) 

def _create_arrays(broadcast_shape, dim_sizes, list_of_core_dims, dtypes):
    """Helper for creating output arrays in vectorize."""
    shapes = _calculate_shapes(broadcast_shape, dim_sizes, list_of_core_dims)
    arrays = tuple(np.empty(shape, dtype=dtype)
                   for shape, dtype in zip(shapes, dtypes))
    return arrays 

def xsumproduct(*args):
    # Check all arrays are the same length
    # Excel returns #VAlUE! error if they don't match
    raise_errors(args)
    assert len(set(arg.size for arg in args)) == 1
    inputs = []
    for a in args:
        a = a.ravel()
        x = np.zeros_like(a, float)
        b = np.vectorize(is_number)(a)
        x[b] = a[b]
        inputs.append(x)

    return np.sum(np.prod(inputs, axis=0)) 

def assert_property_unique(src, tiles, size, occx, occy, boundary_effect_locus):
    tls = np.vectorize(operator.attrgetter('tl'), signature='()->(2)')(tiles.flatten())
    assert np.unique(tls, axis=0).shape[0] == tiles.size 

def assert_property_tile_size(src, tiles, size, occx, occy, boundary_effect_locus):
    w = np.vectorize(operator.attrgetter('w'))(tiles.flatten())
    assert np.unique(w).size == 1
    h = np.vectorize(operator.attrgetter('h'))(tiles.flatten())
    assert np.unique(h).size == 1 

def assert_property_unique(fps, tiles, tx, ty, ox, oy, param2):
    tls = np.vectorize(operator.attrgetter('tl'), signature='()->(2)')(tiles.flatten())
    assert np.unique(tls, axis=0).shape[0] == tiles.size 

def assert_property_overlap(fps, tiles, tx, ty, ox, oy, param2):
    if param2 == 'br':
        stride = tiles[0, 0].rsize - (ox, oy)
        refptx, refpty = 'lx', 'ty'
    elif param2 == 'tl':
        stride = tiles[-1, -1].rsize - (ox, oy)
        refptx, refpty = 'rx', 'by'
    elif param2 == 'tr':
        stride = tiles[-1, 0].rsize - (ox, oy)
        refptx, refpty = 'lx', 'by'
    elif param2 == 'bl':
        stride = tiles[0, -1].rsize - (ox, oy)
        refptx, refpty = 'rx', 'ty'
    else:
        assert False
    refptx = operator.attrgetter(refptx)
    refpty = operator.attrgetter(refpty)
    if tiles.shape[0] > 1:
        diffs_vert = np.vectorize(
            lambda a, b: abs(refpty(a) - refpty(b)))(
                tiles[0:-1], tiles[1:])
        assert (diffs_vert == stride[1]).all()
    if tiles.shape[1] > 1:
        diffs_horiz = np.vectorize(
            lambda a, b: abs(refptx(a) - refptx(b)))(
                tiles[:, 0:-1], tiles[:, 1:])
        assert (diffs_horiz == stride[0]).all() 

def assert_property_tile_size(fps, tiles, tx, ty, ox, oy, param2):
    w = np.vectorize(operator.attrgetter('w'))(tiles.flatten())
    assert np.unique(w).size == 1
    h = np.vectorize(operator.attrgetter('h'))(tiles.flatten())
    assert np.unique(h).size == 1 

def _assign_unique_in_plate(self, plate):
        preds = np.array(list(map(itemgetter(1), plate)))
        preds = np.vectorize(lambda x: x if x != -np.inf else -1e10)(preds)
        _, indices = linear_sum_assignment(-preds)
        return [(k, v.item()) for (k, _), v in zip(plate, indices)] 

def _compare_utc_to_local(tz_didx):
    def f(x):
        return conversion.tz_convert_single(x, UTC, tz_didx.tz)

    result = conversion.tz_convert(tz_didx.asi8, UTC, tz_didx.tz)
    expected = np.vectorize(f)(tz_didx.asi8)

    tm.assert_numpy_array_equal(result, expected) 

def _create_arrays(broadcast_shape, dim_sizes, list_of_core_dims, dtypes):
    """Helper for creating output arrays in vectorize."""
    shapes = _calculate_shapes(broadcast_shape, dim_sizes, list_of_core_dims)
    arrays = tuple(np.empty(shape, dtype=dtype)
                   for shape, dtype in zip(shapes, dtypes))
    return arrays 

def _parse_input_dimensions(args, input_core_dims):
    """
    Parse broadcast and core dimensions for vectorize with a signature.

    Arguments
    ---------
    args : Tuple[ndarray, ...]
        Tuple of input arguments to examine.
    input_core_dims : List[Tuple[str, ...]]
        List of core dimensions corresponding to each input.

    Returns
    -------
    broadcast_shape : Tuple[int, ...]
        Common shape to broadcast all non-core dimensions to.
    dim_sizes : Dict[str, int]
        Common sizes for named core dimensions.
    """
    broadcast_args = []
    dim_sizes = {}
    for arg, core_dims in zip(args, input_core_dims):
        _update_dim_sizes(dim_sizes, arg, core_dims)
        ndim = arg.ndim - len(core_dims)
        dummy_array = np.lib.stride_tricks.as_strided(0, arg.shape[:ndim])
        broadcast_args.append(dummy_array)
    broadcast_shape = np.lib.stride_tricks._broadcast_shape(*broadcast_args)
    return broadcast_shape, dim_sizes 

def get_forces(self):
        # Return the force vector for the problem
        topx_to_id = np.vectorize(
            lambda x: xy_to_id(x, 0, self.nelx, self.nely))
        topx = 2 * topx_to_id(np.arange((self.nelx + 1) // 2)) + 1
        f = np.zeros((2 * (self.nelx + 1) * (self.nely + 1), 1))
        f[topx, 0] = -100
        return f 

def test_isin(self):
        # the tests for in1d cover most of isin's behavior
        # if in1d is removed, would need to change those tests to test
        # isin instead.
        def _isin_slow(a, b):
            b = np.asarray(b).flatten().tolist()
            return a in b
        isin_slow = np.vectorize(_isin_slow, otypes=[bool], excluded={1})
        def assert_isin_equal(a, b):
            x = isin(a, b)
            y = isin_slow(a, b)
            assert_array_equal(x, y)

        #multidimensional arrays in both arguments
        a = np.arange(24).reshape([2, 3, 4])
        b = np.array([[10, 20, 30], [0, 1, 3], [11, 22, 33]])
        assert_isin_equal(a, b)

        #array-likes as both arguments
        c = [(9, 8), (7, 6)]
        d = (9, 7)
        assert_isin_equal(c, d)

        #zero-d array:
        f = np.array(3)
        assert_isin_equal(f, b)
        assert_isin_equal(a, f)
        assert_isin_equal(f, f)

        #scalar:
        assert_isin_equal(5, b)
        assert_isin_equal(a, 6)
        assert_isin_equal(5, 6)

        #empty array-like:
        x = []
        assert_isin_equal(x, b)
        assert_isin_equal(a, x)
        assert_isin_equal(x, x) 

def test_size_zero_output(self):
        # see issue 5868
        f = np.vectorize(lambda x: x)
        x = np.zeros([0, 5], dtype=int)
        with assert_raises_regex(ValueError, 'otypes'):
            f(x)

        f.otypes = 'i'
        assert_array_equal(f(x), x)

        f = np.vectorize(lambda x: x, signature='()->()')
        with assert_raises_regex(ValueError, 'otypes'):
            f(x)

        f = np.vectorize(lambda x: x, signature='()->()', otypes='i')
        assert_array_equal(f(x), x)

        f = np.vectorize(lambda x: x, signature='(n)->(n)', otypes='i')
        assert_array_equal(f(x), x)

        f = np.vectorize(lambda x: x, signature='(n)->(n)')
        assert_array_equal(f(x.T), x.T)

        f = np.vectorize(lambda x: [x], signature='()->(n)', otypes='i')
        with assert_raises_regex(ValueError, 'new output dimensions'):
            f(x)