cosapp.utils.testing

Utility functions for testing purposes

Functions

DummySystemFactory(classname[, base])	Factory creating a dummy system class with custom attributes.
are_same(o1, o2)	Utility function to compare two objects for equality.
assert_all_type(collection, dtype)	Asserts that all the elements of a collecion are of a given type
assert_close_dict(actual, expected[, abs, rel])	Assert that actual and expected dictionaries are identical, within given tolerance bounds.
assert_keys(dictionary, *keys)	Utility function to test dictionary keys
get_args(*args, **kwargs)	Utility function to collect args and kwargs in a tuple
has_keys(dictionary, *keys)	Utility function to test if a dictionary has the given keys
no_exception(**expected_data)	No-op context manager yielding expected data as a dictionary.
not_raised(exception_cls)	Context manager to assert that a block does not raise exception_cls.
pickle_roundtrip(s)	Utility function to pickle and unpickle a System object.
rel_error(actual, expected)	Computes the relative error of actual compared to expected

cosapp.utils.testing.DummySystemFactory(classname: str, base: type[System] | None = None, **settings) → type[System]

Factory creating a dummy system class with custom attributes. System is “dummy” in the sense it has no compute, and no connectors.

Parameters:

• classname [str]: Output class name

• base [type[System], optional]:

  Base class, derived from System. If not provided (default), base is System.

• **settings [dict[str, args_kwargs]]:

  Class characteristics, as a dictionary. Keys are attribute names (e.g. inputs); values are (args, kwargs) forwarded to the associated method (e.g. add_input).

Possible Attributes:

• inputs

• outputs

• inwards

• outwards

• modevars_in

• modevars_out

• transients

• rates

• properties

• children

• events

• unknowns

• equations

• targets

• design_methods

Examples:

>>> from cosapp.utils.testing import DummySystemFactory, get_args
>>>
>>> Dummy = DummySystemFactory(
>>>     inwards=[
>>>         get_args('h', 0.1, unit='m'),
>>>         get_args('L', 2.0, unit='m'),
>>>     ],
>>>     outwards=[
>>>         get_args('b_ratio', 0.0),
>>>     ],
>>>     events=[
>>>         get_args('kaboom', trigger='h > L / 2')
>>>     ],
>>>     properties=[
>>>         get_args('n', 12),
>>>     ],
>>>     equations=[
>>>         "b_ratio == 1",
>>>     ],
>>>     unknowns=[
>>>         "h",
>>>     ],
>>> )
>>>
>>> s = Dummy('s')
>>> assert s.assembled_problem().shape == (1, 1)

cosapp.utils.testing.are_same(o1: Driver | System | Distribution, o2: Driver | System | Distribution) → bool

Utility function to compare two objects for equality. It serializes both objects to JSON and compares the resulting strings.

cosapp.utils.testing.assert_all_type(collection, dtype)

Asserts that all the elements of a collecion are of a given type

cosapp.utils.testing.assert_close_dict(actual: dict, expected: dict, abs=None, rel=None) → None

Assert that actual and expected dictionaries are identical, within given tolerance bounds. Works recursively with nested dictionaries.

cosapp.utils.testing.assert_keys(dictionary, *keys)

Utility function to test dictionary keys

cosapp.utils.testing.get_args(*args, **kwargs) → tuple[tuple[Any, ...], dict[str, Any]]

Utility function to collect args and kwargs in a tuple

cosapp.utils.testing.has_keys(dictionary: dict, *keys) → bool

Utility function to test if a dictionary has the given keys

cosapp.utils.testing.no_exception(**expected_data)

No-op context manager yielding expected data as a dictionary. Raises AssertionError if any exception is raised within the context.

cosapp.utils.testing.not_raised(exception_cls: type[Exception])

Context manager to assert that a block does not raise exception_cls.

cosapp.utils.testing.pickle_roundtrip(s: System) → System

Utility function to pickle and unpickle a System object. This is useful for testing serialization and deserialization of systems.

cosapp.utils.testing.rel_error(actual: Number | Iterable, expected: Number | Iterable) → float | ndarray

Computes the relative error of actual compared to expected