Ejemplo n.º 1
0
def test_1_2_expression_parser_parsing_exceptions():
    """Testing error handling of different erroneous parser instantiations:

    See Also
    --------
    :class:`ExpressionParser`: Detailed documentation of expression parser attributes and methods.
    :class:`LambdaExpressionParser`: Documentation of a non-symbolic expression parser.
    """

    # list parsers to be tested
    parsers = [ExpressionParser, ExpressionParser]
    backends = [TensorflowBackend, NumpyBackend]

    # test expected exceptions
    ##########################

    for Parser, backend in zip(parsers, backends):

        b = backend()

        # undefined variables
        with pytest.raises(ValueError):
            Parser("a + b", {}, backend=b).parse_expr()

        # wrong dimensionality of dictionary arguments
        args = parse_dict({'a': {'vtype': 'constant', 'value': np.ones((3, 3)), 'dtype': 'float32', 'shape': (3, 3)},
                           'b': {'vtype': 'constant', 'value': np.ones((4, 4)), 'dtype': 'float32', 'shape': (4, 4)}},
                          backend=b)
        with pytest.raises(Exception):
            Parser("a + b", {'vars': args}, backend=b).parse_expr()

        # wrong data type of dictionary arguments
        args = parse_dict({'a': {'vtype': 'constant', 'value': np.ones((3, 3)), 'dtype': 'float32', 'shape': (3, 3)},
                           'b': {'vtype': 'constant', 'value': np.ones((3, 3)), 'dtype': 'float32', 'shape': (3, 3)}},
                          backend=b)
        with pytest.raises(Exception):
            Parser("bool(a) + float32(b)", {'vars': args}, backend=b).parse_expr()()

        # undefined mathematical operator
        args = parse_dict({'a': {'vtype': 'constant', 'value': np.ones((3, 3)), 'dtype': 'float32', 'shape': (3, 3)},
                           'b': {'vtype': 'constant', 'value': 5., 'dtype': 'float32', 'shape': ()}},
                          backend=b)
        with pytest.raises(ValueError):
            Parser("a $ b", {'vars': args}, backend=b).parse_expr()()

        # undefined function
        args = parse_dict({'a': {'vtype': 'constant', 'value': np.ones((3, 3)), 'dtype': 'float32', 'shape': (3, 3)},
                           'b': {'vtype': 'constant', 'value': 5., 'dtype': 'float32', 'shape': ()}},
                          backend=b)
        with pytest.raises((KeyError, IndexError)):
            Parser("a / b(5.)", {'vars': args}, backend=b).parse_expr()()
Ejemplo n.º 2
0
def test_1_7_equation_parsing():
    """Tests equation parsing functionalities.

    See Also
    --------
    :func:`parse_equation`: Detailed documentation of parse_equation arguments.
    """

    # define test equations
    equations = ["a = 5. + 2.",              # simple update of variable
                 "d/dt * a = 5. + 2.",       # simple differential equation
                 ]

    # define equation variables
    a = np.zeros(shape=(), dtype=np.float32)

    # define equation results
    results = [7., 0.7]

    # test equation solving of tensorflow-based parser
    ##################################################

    # define backend
    b = TensorflowBackend()
    args = {'node/op/a': {'vtype': 'state_var', 'value': a, 'shape': a.shape, 'dtype': a.dtype},
            'all/all/dt': {'vtype': 'constant', 'value': 0.1, 'shape': (), 'dtype': 'float32'}}
    arguments = [parse_dict(args, backend=b), parse_dict(args, backend=b)]

    # test equation parser on different test equations
    for eq, args, target in zip(equations, arguments, results):

        # tensorflow-based parsing
        result_tmp = parse_equation_system(equations=[[(eq, 'node/op')]], equation_args=args, backend=b)['node/op/a']
        result = result_tmp.numpy() if hasattr(result_tmp, 'numpy') else result_tmp.eval().numpy()
        #assert result == pytest.approx(target, rel=1e-6)

    # test equation solving of numpy-based parser
    #############################################

    # define backend
    b = NumpyBackend()
    args = {'node/op/a': {'vtype': 'state_var', 'value': a, 'shape': a.shape, 'dtype': a.dtype},
            'all/all/dt': {'vtype': 'constant', 'value': 0.1, 'shape': (), 'dtype': 'float32'}}
    arguments = [parse_dict(args, backend=b), parse_dict(args, backend=b)]

    # test equation parser on different test equations
    for eq, args, target in zip(equations, arguments, results):
        # tensorflow-based parsing
        result = parse_equation_system(equations=[[(eq, 'node/op')]], equation_args=args, backend=b)['node/op/a']
Ejemplo n.º 3
0
def test_1_6_expression_parser_funcs():
    """Testing handling of function calls by expression parsers:

    See Also
    --------
    :class:`ExpressionParser`: Detailed documentation of expression parser attributes and methods.
    :class:`LambdaExpressionParser`: Documentation of a non-symbolic expression parser.
    """

    # define variables
    A = np.array(np.random.randn(10, 10), dtype=np.float32)

    # define valid test cases
    expressions = [("abs(5.)", 5.),  # simple function call
                   ("abs(-5.)", 5.),  # function call of negative arg
                   ("abs(4. * -2. + 1)", 7.),  # function call on mathematical expression
                   ("int64(4 > 5)", 0),  # function call on boolean expression
                   ("abs(A[2, :])",
                    np.abs(A[2, :])),  # function call on indexed variable
                   ("abs(sin(1.5))",
                    np.abs(np.sin(1.5))),  # nested function calls
                   ]

    # define invalid test cases
    expressions_wrong = ["abs((4.)",       # wrong parentheses I
                         "abs[4.]",        # wrong parentheses II
                         "abs(0. True)",   # no comma separation on arguments
                         "abs(0.,1,5,3)",  # wrong argument number
                         ]

    # test function calling on tensorflow-based parser
    ##################################################

    # define backend
    b = TensorflowBackend()
    args = parse_dict({'A': {'vtype': 'constant', 'value': A, 'shape': A.shape, 'dtype': A.dtype}}, backend=b)

    # start testing: valid cases
    for expr, target in expressions:
        # tensorflow-based parser
        p = ExpressionParser(expr_str=expr, args=args, backend=b)
        p.parse_expr()
        result = p.rhs_eval.numpy() if hasattr(p.rhs_eval, 'numpy') else p.rhs_eval
        if hasattr(result, 'eval'):
            result = result.eval()
        assert result == pytest.approx(target, rel=1e-6)

    # invalid cases
    for expr in expressions_wrong:

        # tensorflow-based parser
        with pytest.raises((IndexError, ValueError, SyntaxError, TypeError, BaseException)):
            p = ExpressionParser(expr_str=expr, args=args, backend=b)
            p.parse_expr()

    # test function calling on numpy-based parser
    #############################################

    # define backend
    b = NumpyBackend()

    # define variables
    args = parse_dict({'A': {'vtype': 'constant', 'value': A, 'shape': A.shape, 'dtype': A.dtype}}, backend=b)

    # start testing: valid cases
    for expr, target in expressions:
        # tensorflow-based parser
        p = ExpressionParser(expr_str=expr, args=args, backend=b)
        p.parse_expr()
        result = p.rhs_eval.numpy() if hasattr(p.rhs_eval, 'numpy') else p.rhs_eval
        if hasattr(result, 'eval'):
            result = result.eval()
        assert result == pytest.approx(target, rel=1e-6)

    # invalid cases
    for expr in expressions_wrong[:-1]:
        # tensorflow-based parser
        with pytest.raises((IndexError, ValueError, SyntaxError, TypeError, BaseException)):
            p = ExpressionParser(expr_str=expr, args=args, backend=b)
            p.parse_expr()
Ejemplo n.º 4
0
def test_1_5_expression_parser_indexing():
    """Testing handling of indexing operations by expression parsers:

    See Also
    --------
    :class:`ExpressionParser`: Detailed documentation of expression parser attributes and methods.
    :class:`LambdaExpressionParser`: Documentation of a non-symbolic expression parser.
    """

    # test indexing ability of tensorflow-based parser
    ##################################################

    # define backend
    b = TensorflowBackend()

    # define variables
    A = np.array(np.random.randn(10, 10), dtype=np.float32)
    B = np.eye(10, dtype=np.float32) == 1
    arg_dict = {'A': {'vtype': 'constant', 'value': A, 'shape': A.shape, 'dtype': A.dtype},
                'B': {'vtype': 'constant', 'value': B, 'shape': B.shape, 'dtype': B.dtype},
                'd': {'vtype': 'constant', 'value': 4, 'Shape': (), 'dtype': 'int32'}}
    args = parse_dict(arg_dict, backend=b)

    # define valid test cases
    indexed_expressions = [("A[:]", A[:]),               # single-dim indexing I
                           ("A[0]", A[0]),               # single-dim indexing II
                           ("A[-2]", A[-2]),             # single-dim indexing III
                           ("A[0:5]", A[0:5]),           # single-dim slicing I
                           ("A[-1:0:-1]", A[-1:0:-1]),   # single-dim slicing II
                           ("A[4,5]", A[4, 5]),          # two-dim indexing I
                           ("A[5,0:-2]", A[5, 0:-2]),    # two-dim indexing II
                           ("A[A > 0.]", A[A > 0.]),     # boolean indexing
                           ("A[B]", A[B]),               # indexing with other array
                           ("A[d:8 - 1]",                # using variables as indices
                            A[4:8 - 1]),
                           ]

    # test expression parsers on expression results
    for expr, target in indexed_expressions:

        # tensorflow-based parser
        p = ExpressionParser(expr_str=expr, args=args, backend=b)
        p.parse_expr()
        result = p.rhs_eval.numpy() if hasattr(p.rhs_eval, 'numpy') else p.rhs_eval
        if hasattr(result, 'eval'):
            result = result.eval()
        assert result == pytest.approx(target, rel=1e-6)

    # define invalid test cases
    indexed_expressions_wrong = ["A[1.2]",       # indexing with float variables
                                 "A[all]",       # indexing with undefined key words
                                 "A[-11]",       # index out of bounds
                                 "A[0:5:2:1]",   # too many arguments for slicing
                                 "A[-1::0:-1]",  # wrong slicing syntax II
                                 ]

    # test expression parsers on expression results
    for expr in indexed_expressions_wrong:

        # tensorflow-based parser
        with pytest.raises((IndexError, ValueError, SyntaxError, TypeError, BaseException)):
            p = ExpressionParser(expr_str=expr, args=args, backend=b)
            p.parse_expr()

    # test indexing ability of numpy-based parser
    #############################################

    # define backend
    b = NumpyBackend()
    args = parse_dict(arg_dict, backend=b)

    # test expression parsers on expression results
    for expr, target in indexed_expressions:
        p = ExpressionParser(expr_str=expr, args=args, backend=b)
        p.parse_expr()
        result = p.rhs_eval.numpy() if hasattr(p.rhs_eval, 'numpy') else p.rhs_eval
        if hasattr(result, 'eval'):
            result = result.eval()
        assert result == pytest.approx(target, rel=1e-6)

    # test expression parsers on expression results
    for expr in indexed_expressions_wrong:
        with pytest.raises((IndexError, ValueError, SyntaxError, TypeError, BaseException)):
            p = ExpressionParser(expr_str=expr, args=args, backend=b)
            p.parse_expr()
            p.op.eval()