Python numpy.product() Examples

def adjoint_acton(self, state):
        """ Act the adjoint of this gate map on an input state """
        #NOTE: Same as acton except uses 'adjoint_acton(...)' below
        output_state = DMStateRep(_np.zeros(state.base.shape, 'd'))
        offset = self.offset  # if relToBlock else self.offset (relToBlock == False here)

        for b in _itertools.product(*self.basisInds_noop_blankaction):  # zeros in all action-index locations
            vec_index_noop = _np.dot(self.multipliers, tuple(b))
            inds = []
            for op_b in _itertools.product(*self.basisInds_action):
                vec_index = vec_index_noop
                for i, bInd in zip(self.actionInds, op_b):
                    #b[i] = bInd #don't need to do this; just update vec_index:
                    vec_index += self.multipliers[i] * bInd
                inds.append(offset + vec_index)
            embedded_instate = DMStateRep(state.base[inds])
            embedded_outstate = self.embedded.adjoint_acton(embedded_instate)
            output_state.base[inds] += embedded_outstate.base

        #act on other blocks trivially:
        self._acton_other_blocks_trivially(output_state, state)
        return output_state 

def dot(self, other):
        """
        Computes the Hilbert-Schmidt dot product (normed to 1) between this
        Pauli operator and `other`.

        Parameters
        ----------
        other : NQPauliOp
            The other operator to take a dot product with.

        Returns
        -------
        integer
            Either 0, 1, or -1.
        """
        assert(len(self) == len(other)), "Length mismatch!"
        if other.rep == self.rep:
            return self.sign * other.sign
        else:
            return 0 

def test_basic_use(self, simple_linear_model):
        """ Test the basic use of `ConstantParameter` using the Python API """
        model = simple_linear_model
        # Add two scenarios
        scA = Scenario(model, 'Scenario A', size=2)
        scB = Scenario(model, 'Scenario B', size=5)

        p = ConstantParameter(model, np.pi, name='pi', comment='Mmmmm Pi!')

        assert not p.is_variable
        assert p.double_size == 1
        assert p.integer_size == 0

        model.setup()
        ts = model.timestepper.current
        # Now ensure the appropriate value is returned for all scenarios
        for i, (a, b) in enumerate(itertools.product(range(scA.size), range(scB.size))):
            si = ScenarioIndex(i, np.array([a, b], dtype=np.int32))
            np.testing.assert_allclose(p.value(ts, si), np.pi) 

def test_parameter_constant_scenario(simple_linear_model):
    """
    Test ConstantScenarioParameter

    """
    model = simple_linear_model
    # Add two scenarios
    scA = Scenario(model, 'Scenario A', size=2)
    scB = Scenario(model, 'Scenario B', size=5)

    p = ConstantScenarioParameter(model, scB, np.arange(scB.size, dtype=np.float64))
    model.setup()
    ts = model.timestepper.current
    # Now ensure the appropriate value is returned for the Scenario B indices.
    for i, (a, b) in enumerate(itertools.product(range(scA.size), range(scB.size))):
        si = ScenarioIndex(i, np.array([a, b], dtype=np.int32))
        np.testing.assert_allclose(p.value(ts, si), float(b)) 

def test_parameter_constant_scenario(simple_linear_model):
    """
    Test ConstantScenarioIndexParameter

    """
    model = simple_linear_model
    # Add two scenarios
    scA = Scenario(model, 'Scenario A', size=2)
    scB = Scenario(model, 'Scenario B', size=5)

    p = ConstantScenarioIndexParameter(model, scB, np.arange(scB.size, dtype=np.int32))
    model.setup()
    ts = model.timestepper.current
    # Now ensure the appropriate value is returned for the Scenario B indices.
    for i, (a, b) in enumerate(itertools.product(range(scA.size), range(scB.size))):
        si = ScenarioIndex(i, np.array([a, b], dtype=np.int32))
        np.testing.assert_allclose(p.index(ts, si), b) 

def get_low_pmi_bigrams(threshold_coef, word_freq_df):
	# type: (float, pd.DataFrame) -> object
	is_bigram = np.array([' ' in word for word in word_freq_df.index])
	unigram_freq = word_freq_df[~is_bigram].sum(axis=1)
	bigram_freq = word_freq_df[is_bigram].sum(axis=1)
	bigram_prob = bigram_freq / bigram_freq.sum()
	unigram_prob = unigram_freq / unigram_freq.sum()

	def get_pmi(bigram):
		try:
			return np.log(
				bigram_prob[bigram] / np.product([unigram_prob[word] for word in bigram.split(' ')])
			) / np.log(2)
		except:
			return 0

	low_pmi_bigrams = bigram_prob[bigram_prob.index.map(get_pmi) < threshold_coef * 2]
	return low_pmi_bigrams 

def test_load(self, simple_linear_model):
        """ Test load from JSON dict"""
        model = simple_linear_model
        data = {
            "type": "aggregated",
            "agg_func": "product",
            "parameters": [
                0.8,
                {
                    "type": "monthlyprofile",
                    "values": list(range(12))
                }
            ]
        }

        p = load_parameter(model, data)
        # Correct instance is loaded
        assert isinstance(p, AggregatedParameter)

        @assert_rec(model, p)
        def expected(timestep, scenario_index):
            return (timestep.month - 1) * 0.8

        model.run() 

def test_addsumprod(self):
        # Tests add, sum, product.
        (x, y, a10, m1, m2, xm, ym, z, zm, xf) = self.d
        assert_equal(np.add.reduce(x), add.reduce(x))
        assert_equal(np.add.accumulate(x), add.accumulate(x))
        assert_equal(4, sum(array(4), axis=0))
        assert_equal(4, sum(array(4), axis=0))
        assert_equal(np.sum(x, axis=0), sum(x, axis=0))
        assert_equal(np.sum(filled(xm, 0), axis=0), sum(xm, axis=0))
        assert_equal(np.sum(x, 0), sum(x, 0))
        assert_equal(np.product(x, axis=0), product(x, axis=0))
        assert_equal(np.product(x, 0), product(x, 0))
        assert_equal(np.product(filled(xm, 1), axis=0), product(xm, axis=0))
        s = (3, 4)
        x.shape = y.shape = xm.shape = ym.shape = s
        if len(s) > 1:
            assert_equal(np.concatenate((x, y), 1), concatenate((xm, ym), 1))
            assert_equal(np.add.reduce(x, 1), add.reduce(x, 1))
            assert_equal(np.sum(x, 1), sum(x, 1))
            assert_equal(np.product(x, 1), product(x, 1)) 

def word_list_to_embedding_product(words, embeddings, embedding_dimension=50):
    '''
    :param words: an n x (2*window_size + 1) matrix from data_to_mat
    :param embeddings: an embedding dictionary where keys are strings and values
    are embeddings; the output from embeddings_to_dict
    :param embedding_dimension: the dimension of the values in embeddings; in this
    assignment, embedding_dimension=50
    :return: an n x embedding_dimension matrix where the embeddings of an example were
    the hadamard product of the embeddings occupies a row in this matrix
    '''
    m, n = words.shape
    words = words.reshape((-1))

    out = np.array([embeddings[w] for w in words], dtype=np.float32).reshape(m, n, embedding_dimension)
    return np.product(out, 1)

# Problem 1 Part 2 

def _load_one_param_type(self, conv_params, param_category, suffix):
        """Deserializes the weights from a file stream in the DarkNet order.

        Keyword arguments:
        conv_params -- a ConvParams object
        param_category -- the category of parameters to be created ('bn' or 'conv')
        suffix -- a string determining the sub-type of above param_category (e.g.,
        'weights' or 'bias')
        """
        param_name = conv_params.generate_param_name(param_category, suffix)
        channels_out, channels_in, filter_h, filter_w = conv_params.conv_weight_dims
        if param_category == 'bn':
            param_shape = [channels_out]
        elif param_category == 'conv':
            if suffix == 'weights':
                param_shape = [channels_out, channels_in, filter_h, filter_w]
            elif suffix == 'bias':
                param_shape = [channels_out]
        param_size = np.product(np.array(param_shape))
        param_data = np.ndarray(
            shape=param_shape,
            dtype='float32',
            buffer=self.weights_file.read(param_size * 4))
        param_data = param_data.flatten().astype(float)
        return param_name, param_data, param_shape 

def _load_one_param_type(self, conv_params, param_category, suffix):
        """Deserializes the weights from a file stream in the DarkNet order.

        Keyword arguments:
        conv_params -- a ConvParams object
        param_category -- the category of parameters to be created ('bn' or 'conv')
        suffix -- a string determining the sub-type of above param_category (e.g.,
        'weights' or 'bias')
        """
        param_name = conv_params.generate_param_name(param_category, suffix)
        channels_out, channels_in, filter_h, filter_w = conv_params.conv_weight_dims
        if param_category == 'bn':
            param_shape = [channels_out]
        elif param_category == 'conv':
            if suffix == 'weights':
                param_shape = [channels_out, channels_in, filter_h, filter_w]
            elif suffix == 'bias':
                param_shape = [channels_out]
        param_size = np.product(np.array(param_shape))
        param_data = np.ndarray(
            shape=param_shape,
            dtype='float32',
            buffer=self.weights_file.read(param_size * 4))
        param_data = param_data.flatten().astype(float)
        return param_name, param_data, param_shape 

def test_testAddSumProd(self):
        # Test add, sum, product.
        (x, y, a10, m1, m2, xm, ym, z, zm, xf, s) = self.d
        assert_(eq(np.add.reduce(x), add.reduce(x)))
        assert_(eq(np.add.accumulate(x), add.accumulate(x)))
        assert_(eq(4, sum(array(4), axis=0)))
        assert_(eq(4, sum(array(4), axis=0)))
        assert_(eq(np.sum(x, axis=0), sum(x, axis=0)))
        assert_(eq(np.sum(filled(xm, 0), axis=0), sum(xm, axis=0)))
        assert_(eq(np.sum(x, 0), sum(x, 0)))
        assert_(eq(np.product(x, axis=0), product(x, axis=0)))
        assert_(eq(np.product(x, 0), product(x, 0)))
        assert_(eq(np.product(filled(xm, 1), axis=0),
                           product(xm, axis=0)))
        if len(s) > 1:
            assert_(eq(np.concatenate((x, y), 1),
                               concatenate((xm, ym), 1)))
            assert_(eq(np.add.reduce(x, 1), add.reduce(x, 1)))
            assert_(eq(np.sum(x, 1), sum(x, 1)))
            assert_(eq(np.product(x, 1), product(x, 1))) 

def evaluate(self, variable_values):
        """
        Evaluate this polynomial for a given set of variable values.

        Parameters
        ----------
        variable_values : array-like
            An object that can be indexed so that `variable_values[i]` gives the
            numerical value for i-th variable (x_i).

        Returns
        -------
        float or complex
            Depending on the types of the coefficients and `variable_values`.
        """
        #FUTURE: make this function smarter (Russian peasant)
        ret = 0
        for ivar, coeff in self.items():
            ret += coeff * _np.product([variable_values[i] for i in ivar])
        assert(_np.isclose(ret, self.fastpoly.evaluate(variable_values)))
        self.check_fastpoly()
        return ret 

def repeat_sum(u,shape,rep_axes):
    """
    Computes sum of a repeated matrix
    
    In effect, this routine computes 
    code:`np.sum(repeat(u,shape,rep_axes))`.  However, it performs
    this without having to perform the full repetition.
    
    """
    # Must convert to np.array to perform slicing
    shape_vec = np.array(shape,dtype=int)
    rep_vec = np.array(rep_axes,dtype=int)
    
    # repeat and sum
    urep = repeat_axes(u,shape,rep_axes,rep=False)
    usum = np.sum(urep)*np.product(shape_vec[rep_vec])
    return usum 

def tensor_product_block_dims(self, iTPB):  # unused
        """
        Get the dimension corresponding to each label in the
        `iTBP`-th tensor-product block.  The dimension of the
        entire block is the product of these.

        Parameters
        ----------
        iTPD : int
           The index of the tensor product block whose state-space
           dimensions you wish to retrieve.

        Returns
        -------
        tuple
        """
        return tuple((self.labeldims[lbl] for lbl in self.labels[iTPB])) 

def cost_adjust(self,r,z,rvar,zvar,shape,var_axes):
        """
        Computes the cost adjustment term for the
        Bethe Free Energy:
            
            J = beta*[log(2*pi*rvar) + ((z-r)**2 + xvar)/rvar]
            
        where beta = 1 for complex problems and 0 for real problems
        """    
        J0 = np.mean(np.log(2*np.pi*rvar))*np.product(shape)
        rvar_rep = common.repeat_axes(rvar,shape,\
                                      var_axes,rep=False)
        J1 = np.sum(np.abs(r-z)**2/rvar_rep)
        J2 = np.mean(zvar/rvar)*np.product(shape)
        J = J0 + J1 + J2
        if not self.is_complex:
            J = J / 2
        return J 

def affine_simplified(h_num_hidden):

    def compile_fn(di, dh):
        shape = di['in'].get_shape().as_list()
        n = np.product(shape[1:])

        def forward_fn(di):
            In = di['in']
            if len(shape) > 2:
                In = tf.reshape(In, [-1, n])
            return {'out': tf.layers.dense(In, dh['num_hidden'])}

        return forward_fn

    return siso_tensorflow_module('AffineSimplified', compile_fn,
                                  {'num_hidden': h_num_hidden}) 

def affine(h_num_hidden, h_W_init_fn, h_b_init_fn):

    def compile_fn(di, dh):
        m = dh['num_hidden']
        shape = di['in'].get_shape().as_list()
        n = np.product(shape[1:])
        W = tf.Variable(dh['W_init_fn']([n, m]))
        b = tf.Variable(dh['b_init_fn']([m]))

        def forward_fn(di):
            In = di['in']
            if len(shape) > 2:
                In = tf.reshape(In, [-1, n])
            return {'out': tf.add(tf.matmul(In, W), b)}

        return forward_fn

    return siso_tensorflow_module(
        'Affine', compile_fn, {
            'num_hidden': h_num_hidden,
            'W_init_fn': h_W_init_fn,
            'b_init_fn': h_b_init_fn
        }) 

def adjoint_acton(self, state):
        """ Act the adjoint of this gate map on an input state """
        #NOTE: Same as acton except uses 'adjoint_acton(...)' below
        output_state = SVStateRep(_np.zeros(state.base.shape, complex))
        offset = self.offset  # if relToBlock else self.offset (relToBlock == False here)

        for b in _itertools.product(*self.basisInds_noop_blankaction):  # zeros in all action-index locations
            vec_index_noop = _np.dot(self.multipliers, tuple(b))
            inds = []
            for op_b in _itertools.product(*self.basisInds_action):
                vec_index = vec_index_noop
                for i, bInd in zip(self.actionInds, op_b):
                    #b[i] = bInd #don't need to do this; just update vec_index:
                    vec_index += self.multipliers[i] * bInd
                inds.append(offset + vec_index)
            embedded_instate = SVStateRep(state.base[inds])
            embedded_outstate = self.embedded.adjoint_acton(embedded_instate)
            output_state.base[inds] += embedded_outstate.base

        #act on other blocks trivially:
        self._acton_other_blocks_trivially(output_state, state)
        return output_state 

def acton(self, state):
        output_state = SVStateRep(_np.zeros(state.base.shape, complex))
        offset = self.offset  # if relToBlock else self.offset (relToBlock == False here)

        for b in _itertools.product(*self.basisInds_noop_blankaction):  # zeros in all action-index locations
            vec_index_noop = _np.dot(self.multipliers, tuple(b))
            inds = []
            for op_b in _itertools.product(*self.basisInds_action):
                vec_index = vec_index_noop
                for i, bInd in zip(self.actionInds, op_b):
                    #b[i] = bInd #don't need to do this; just update vec_index:
                    vec_index += self.multipliers[i] * bInd
                inds.append(offset + vec_index)
            embedded_instate = SVStateRep(state.base[inds])
            embedded_outstate = self.embedded.acton(embedded_instate)
            output_state.base[inds] += embedded_outstate.base

        #act on other blocks trivially:
        self._acton_other_blocks_trivially(output_state, state)
        return output_state 

def __init__(self, embedded_op, numBasisEls, actionInds,
                 blocksizes, embedded_dim, nComponentsInActiveBlock,
                 iActiveBlock, nBlocks, dim):

        self.embedded = embedded_op
        self.numBasisEls = numBasisEls
        self.actionInds = actionInds
        self.blocksizes = blocksizes

        numBasisEls_noop_blankaction = numBasisEls.copy()
        for i in actionInds: numBasisEls_noop_blankaction[i] = 1
        self.basisInds_noop_blankaction = [list(range(n)) for n in numBasisEls_noop_blankaction]

        # multipliers to go from per-label indices to tensor-product-block index
        # e.g. if map(len,basisInds) == [1,4,4] then multipliers == [ 16 4 1 ]
        self.multipliers = _np.array(_np.flipud(_np.cumprod([1] + list(
            reversed(list(numBasisEls[1:]))))), _np.int64)
        self.basisInds_action = [list(range(numBasisEls[i])) for i in actionInds]

        self.embeddedDim = embedded_dim
        self.nComponents = nComponentsInActiveBlock
        self.iActiveBlock = iActiveBlock
        self.nBlocks = nBlocks
        self.offset = sum(blocksizes[0:iActiveBlock])
        super(SVOpRep_Embedded, self).__init__(dim) 

def __init__(self, povm_to_marginalize, all_sslbls, sslbls_after_marginalizing):
        """ TODO: docstring """
        self.povm_to_marginalize = povm_to_marginalize

        if isinstance(all_sslbls, _ld.StateSpaceLabels):
            assert(len(all_sslbls.labels) == 1), "all_sslbls should only have a single tensor product block!"
            all_sslbls = all_sslbls.labels[0]

        #now all_sslbls is a tuple of labels, like sslbls_after_marginalizing
        self.sslbls_to_marginalize = all_sslbls
        self.sslbls_after_marginalizing = sslbls_after_marginalizing
        indices_to_keep = set([list(all_sslbls).index(l) for l in sslbls_after_marginalizing])
        indices_to_remove = set(range(len(all_sslbls))) - indices_to_keep
        self.indices_to_marginalize = sorted(indices_to_remove, reverse=True)

        elements_to_sum = {}
        for k in self.povm_to_marginalize.keys():
            mk = self.marginalize_effect_label(k)
            if mk in elements_to_sum:
                elements_to_sum[mk].append(k)
            else:
                elements_to_sum[mk] = [k]
        self._elements_to_sum = {k: tuple(v) for k, v in elements_to_sum.items()}  # convert to tuples
        super(MarginalizedPOVM, self).__init__(self.povm_to_marginalize.dim, self.povm_to_marginalize._evotype) 

def setUp(self):
        # Make an X that looks somewhat like a small tf-idf matrix.
        # XXX newer versions of SciPy >0.16 have scipy.sparse.rand for this.
        shape = 60, 55
        n_samples, n_features = shape
        rng = check_random_state(42)
        X = rng.randint(-100, 20, np.product(shape)).reshape(shape)
        X = sp.csr_matrix(np.maximum(X, 0), dtype=np.float64)
        X.data[:] = 1 + np.log(X.data)
        self.X = X
        self.Xdense = X.A
        self.n_samples = n_samples
        self.n_features = n_features

        self.session = new_session().as_default()
        self._old_executor = self.session._sess._executor
        self.executor = self.session._sess._executor = \
            ExecutorForTest('numpy', storage=self.session._sess._context) 

def __init__(self, embedded_op, numBasisEls, actionInds,
                 blocksizes, embedded_dim, nComponentsInActiveBlock,
                 iActiveBlock, nBlocks, dim):

        self.embedded = embedded_op
        self.numBasisEls = numBasisEls
        self.actionInds = actionInds
        self.blocksizes = blocksizes

        numBasisEls_noop_blankaction = numBasisEls.copy()
        for i in actionInds: numBasisEls_noop_blankaction[i] = 1
        self.basisInds_noop_blankaction = [list(range(n)) for n in numBasisEls_noop_blankaction]

        # multipliers to go from per-label indices to tensor-product-block index
        # e.g. if map(len,basisInds) == [1,4,4] then multipliers == [ 16 4 1 ]
        self.multipliers = _np.array(_np.flipud(_np.cumprod([1] + list(
            reversed(list(numBasisEls[1:]))))), _np.int64)
        self.basisInds_action = [list(range(numBasisEls[i])) for i in actionInds]

        self.embeddedDim = embedded_dim
        self.nComponents = nComponentsInActiveBlock
        self.iActiveBlock = iActiveBlock
        self.nBlocks = nBlocks
        self.offset = sum(blocksizes[0:iActiveBlock])
        super(DMOpRep_Embedded, self).__init__(dim) 

def evaluate(self, variable_values):
        """
        Evaluate this polynomial for a given set of variable values.

        Parameters
        ----------
        variable_values : array-like
            An object that can be indexed so that `variable_values[i]` gives the
            numerical value for i-th variable (x_i).

        Returns
        -------
        float or complex
            Depending on the types of the coefficients and `variable_values`.
        """
        #FUTURE: make this function smarter (Russian peasant)
        ret = 0
        for ivar, coeff in self.coeffs.items():
            ret += coeff * _np.product([variable_values[i] for i in ivar])
        return ret 

def _lazy_build_elements(self):
        #LAZY building of elements (in case we never need them)
        compMxs = _np.zeros((self.size,) + self.elshape, 'complex')

        #Take kronecker product of *natural* reps of component-basis elements
        # then reshape to vectors at the end.  This requires that the vector-
        # dimension of the component spaces equals the "natural space" dimension.
        comp_els = [c.elements for c in self.component_bases]
        for i, factors in enumerate(_itertools.product(*comp_els)):
            M = _np.identity(1, 'complex')
            for f in factors:
                M = _np.kron(M, f)
            compMxs[i] = M
        self._elements = compMxs 

def flatten(data):
    if data.ndim == 2:
        return data
    if not data.ndim >= 3:
        print 'data shape', data.shape
        assert data.ndim >= 3
    s = data.shape
    out = data.reshape((np.product(s[0:2]), np.product(s[2:])))

    return out


# Load data for a given pipeline. This wraps load_data_mp to also provide FeatureConcatPipeline support.
# See load_data_mp for description of check_only and meta_only parameters. 

def apply(self, data, meta=None):
        if data.ndim == 2:
            return data.ravel()
        elif data.ndim == 3:
            s = data.shape
            return data.reshape((s[0], np.product(s[1:])))
        else:
            raise NotImplementedError() 

def _lazy_build_labels(self):
        self._labels = []
        comp_lbls = [c.labels for c in self.component_bases]
        for i, factor_lbls in enumerate(_itertools.product(*comp_lbls)):
            self._labels.append(''.join(factor_lbls)) 

def central_diff_weights(Np, ndiv=1):
    """
    Return weights for an Np-point central derivative.

    Assumes equally-spaced function points.

    If weights are in the vector w, then
    derivative is w[0] * f(x-ho*dx) + ... + w[-1] * f(x+h0*dx)

    Parameters
    ----------
    Np : int
        Number of points for the central derivative.
    ndiv : int, optional
        Number of divisions.  Default is 1.

    Notes
    -----
    Can be inaccurate for large number of points.

    """
    if Np < ndiv + 1:
        raise ValueError("Number of points must be at least the derivative order + 1.")
    if Np % 2 == 0:
        raise ValueError("The number of points must be odd.")
    from scipy import linalg
    ho = Np >> 1
    x = arange(-ho,ho+1.0)
    x = x[:,newaxis]
    X = x**0.0
    for k in range(1,Np):
        X = hstack([X,x**k])
    w = product(arange(1,ndiv+1),axis=0)*linalg.inv(X)[ndiv]
    return w