def test_different_bad_design_matrix():
iv_names = ('a', 'b', 'c')
iv_values = [[1, 2], [1, 2, 3], [1, 2]]
matrix = np.array([[-1., -1., 1.], [1., -1., -1.], [-1., 1., -1.], [1., 1., 1.]]) # pyDOE.fracfact('a b ab')
d = Design(ivs=zip(iv_names, iv_values), design_matrix=matrix)
with pytest.raises(ValueError):
d.first_pass()
def test_factorial():
iv_names = ("a", "b")
iv_values = [[1, 2, 3], [1, 2]]
conditions = [
{"a": 1, "b": 1},
{"a": 1, "b": 2},
{"a": 2, "b": 1},
{"a": 2, "b": 2},
{"a": 3, "b": 1},
{"a": 3, "b": 2},
]
assert list(Design.full_cross(iv_names, iv_values)) == conditions
d = Design(zip(iv_names, iv_values))
d.first_pass()
for _ in range(3):
yield check_sequences, d.get_order(), conditions
iv_names = ("a", "b", "c")
iv_values = [[1, 2], [1, 2], [1, 2]]
conditions = [
{"a": 1, "b": 1, "c": 1},
{"a": 1, "b": 1, "c": 2},
{"a": 1, "b": 2, "c": 1},
{"a": 1, "b": 2, "c": 2},
{"a": 2, "b": 1, "c": 1},
{"a": 2, "b": 1, "c": 2},
{"a": 2, "b": 2, "c": 1},
{"a": 2, "b": 2, "c": 2},
]
assert list(Design.full_cross(iv_names, iv_values)) == conditions
d = Design(zip(iv_names, iv_values))
d.first_pass()
for _ in range(3):
yield check_sequences, d.get_order(), conditions
def test_factorial():
iv_names = ('a', 'b')
iv_values = [[1, 2, 3], [1, 2]]
conditions = [
{'a': 1, 'b': 1},
{'a': 1, 'b': 2},
{'a': 2, 'b': 1},
{'a': 2, 'b': 2},
{'a': 3, 'b': 1},
{'a': 3, 'b': 2},
]
assert list(Design.full_cross(iv_names, iv_values)) == conditions
d = Design(zip(iv_names, iv_values))
d.first_pass()
for _ in range(3):
yield check_sequences, d.get_order(), conditions
iv_names = ('a', 'b', 'c')
iv_values = [[1, 2], [1, 2], [1, 2]]
conditions = [
{'a': 1, 'b': 1, 'c': 1},
{'a': 1, 'b': 1, 'c': 2},
{'a': 1, 'b': 2, 'c': 1},
{'a': 1, 'b': 2, 'c': 2},
{'a': 2, 'b': 1, 'c': 1},
{'a': 2, 'b': 1, 'c': 2},
{'a': 2, 'b': 2, 'c': 1},
{'a': 2, 'b': 2, 'c': 2},
]
assert list(Design.full_cross(iv_names, iv_values)) == conditions
d = Design(zip(iv_names, iv_values))
d.first_pass()
for _ in range(3):
yield check_sequences, d.get_order(), conditions
def test_design_matrix_with_continuous_iv():
matrix = np.array(
[
[-1, -1],
[1, -1],
[-1, 1],
[1, 1],
[0, 0],
[0, 0],
[0, 0],
[0, 0],
[-1.41421356, 0],
[1.41421356, 0],
[0, -1.41421356],
[0, 1.41421356],
[0, 0],
[0, 0],
[0, 0],
[0, 0],
]
) # pyDOE.ccdesign(2)
iv_names = ["a", "b"]
iv_values = [None, None]
d = Design(zip(iv_names, iv_values), design_matrix=matrix)
d.first_pass()
for _ in range(3):
yield check_design_matrix, d.get_order(), iv_names, iv_values, matrix
def test_design_from_spec():
spec = {
'name': 'test1',
'order': 'Shuffle',
'n': 2,
'ivs': {'a': [True, False], 'b': [1, 2, 3]},
}
assert Design.from_dict(spec) == ('test1', Design({'a': [True, False], 'b': [1, 2, 3]}, ordering=Shuffle(2)))
spec = {
'name': 'test2',
'some_extra_field': ['blah'],
'number': 3,
}
assert Design.from_dict(spec) == ('test2', Design(extra_data={'some_extra_field': ['blah']}, ordering=Shuffle(3)))
spec = {
'name': 'test3',
'ordering': {
'class': 'Sorted',
'order': 'ascending',
},
'ivs': {'a': [1, 2, 3]},
}
assert Design.from_dict(spec) == ('test3', Design(ivs={'a': [1, 2, 3]}, ordering=Sorted(order='ascending')))
spec = {
'name': 'test4',
'order': ['CompleteCounterbalance', 3],
}
assert Design.from_dict(spec) == ('test4', Design(ordering=CompleteCounterbalance(3)))
spec.pop('name')
assert Design.from_dict(spec) == Design(ordering=CompleteCounterbalance(3))
def test_different_bad_design_matrix():
iv_names = ('a', 'b', 'c')
iv_values = [[1, 2], [1, 2, 3], [1, 2]]
matrix = np.array([[-1., -1., 1.], [1., -1., -1.], [-1., 1., -1.],
[1., 1., 1.]]) # pyDOE.fracfact('a b ab')
d = Design(ivs=zip(iv_names, iv_values), design_matrix=matrix)
with pytest.raises(ValueError):
d.first_pass()
def make_manual_exp():
tree = DesignTree.new([('participant', [Design(ordering=Shuffle(2))]),
('block', [Design(ivs={'b': [0, 1, 2]}, ordering=CompleteCounterbalance())]),
('trial', [Design({'a': [False, True]}, ordering=Shuffle(4))]),
])
exp = Experiment.new(tree)
exp.add_callback('trial', trial)
return exp
def test_design_matrix_with_continuous_iv():
matrix = np.array([[-1, -1], [1, -1], [-1, 1], [1, 1], [0, 0], [0, 0],
[0, 0], [0, 0], [-1.41421356, 0], [1.41421356, 0],
[0, -1.41421356], [0, 1.41421356], [0, 0], [0, 0],
[0, 0], [0, 0]]) # pyDOE.ccdesign(2)
iv_names = ['a', 'b']
iv_values = [None, None]
d = Design(zip(iv_names, iv_values), design_matrix=matrix)
d.first_pass()
for _ in range(3):
yield check_design_matrix, d.get_order(), iv_names, iv_values, matrix
def test_design_from_spec():
spec = {
'name': 'test1',
'order': 'Shuffle',
'n': 2,
'ivs': {
'a': [True, False],
'b': [1, 2, 3]
},
}
assert Design.from_dict(spec) == ('test1',
Design(
{
'a': [True, False],
'b': [1, 2, 3]
},
ordering=Shuffle(2)))
spec = {
'name': 'test2',
'some_extra_field': ['blah'],
'number': 3,
}
assert Design.from_dict(spec) == ('test2',
Design(extra_data={
'some_extra_field': ['blah']
},
ordering=Shuffle(3)))
spec = {
'name': 'test3',
'ordering': {
'class': 'Sorted',
'order': 'ascending',
},
'ivs': {
'a': [1, 2, 3]
},
}
assert Design.from_dict(spec) == ('test3',
Design(
ivs={'a': [1, 2, 3]},
ordering=Sorted(order='ascending')))
spec = {
'name': 'test4',
'order': ['CompleteCounterbalance', 3],
}
assert Design.from_dict(spec) == ('test4',
Design(
ordering=CompleteCounterbalance(3)))
spec.pop('name')
assert Design.from_dict(spec) == Design(ordering=CompleteCounterbalance(3))
def test_design_from_spec():
spec = {"name": "test1", "order": "Shuffle", "n": 2, "ivs": {"a": [True, False], "b": [1, 2, 3]}}
assert Design.from_dict(spec) == ("test1", Design({"a": [True, False], "b": [1, 2, 3]}, ordering=Shuffle(2)))
spec = {"name": "test2", "some_extra_field": ["blah"], "number": 3}
assert Design.from_dict(spec) == ("test2", Design(extra_data={"some_extra_field": ["blah"]}, ordering=Shuffle(3)))
spec = {"name": "test3", "ordering": {"class": "Sorted", "order": "ascending"}, "ivs": {"a": [1, 2, 3]}}
assert Design.from_dict(spec) == ("test3", Design(ivs={"a": [1, 2, 3]}, ordering=Sorted(order="ascending")))
spec = {"name": "test4", "order": ["CompleteCounterbalance", 3]}
assert Design.from_dict(spec) == ("test4", Design(ordering=CompleteCounterbalance(3)))
spec.pop("name")
assert Design.from_dict(spec) == Design(ordering=CompleteCounterbalance(3))
def test_design_matrix():
iv_names = ('a', 'b', 'c')
iv_values = [[1, 2], [1, 2], [1, 2]]
conditions = [
{'a': 1, 'b': 1, 'c': 2},
{'a': 1, 'b': 2, 'c': 1},
{'a': 2, 'b': 1, 'c': 1},
{'a': 2, 'b': 2, 'c': 2},
]
matrix = np.array([[-1., -1., 1.], [1., -1., -1.], [-1., 1., -1.], [1., 1., 1.]]) # pyDOE.fracfact('a b ab')
d = Design(ivs=zip(iv_names, iv_values), design_matrix=matrix)
d.first_pass()
for _ in range(3):
yield check_sequences, d.get_order(), conditions
yield check_design_matrix, d.get_order(), iv_names, iv_values, matrix
def test_design_matrix():
iv_names = ("a", "b", "c")
iv_values = [[1, 2], [1, 2], [1, 2]]
conditions = [
{"a": 1, "b": 1, "c": 2},
{"a": 1, "b": 2, "c": 1},
{"a": 2, "b": 1, "c": 1},
{"a": 2, "b": 2, "c": 2},
]
matrix = np.array(
[[-1.0, -1.0, 1.0], [1.0, -1.0, -1.0], [-1.0, 1.0, -1.0], [1.0, 1.0, 1.0]]
) # pyDOE.fracfact('a b ab')
d = Design(ivs=zip(iv_names, iv_values), design_matrix=matrix)
d.first_pass()
for _ in range(3):
yield check_sequences, d.get_order(), conditions
yield check_design_matrix, d.get_order(), iv_names, iv_values, matrix
def test_bad_heterogeneity():
main_structure = [
('participant',
Design(ivs={
'a': [1, 2],
'b': [1, 2]
}, ordering=Shuffle(3))),
('session',
Design(ivs={'design': ['practice', 'test']},
design_matrix=[[0], [1], [1]])),
]
other_structures = {
'practice': [
('block', Design()),
('trial', Design(ivs={'difficulty': [1, 2]},
ordering=Shuffle(20))),
],
'test': [
('block', Design(ordering=CompleteCounterbalance(2))),
('trial',
Design(ivs={'difficulty': [1, 3, 5, 7]}, ordering=Shuffle(5))),
],
}
with pytest.raises(ValueError):
DesignTree.new(main_structure, **other_structures)
def make_heterogeneous_tree():
main_structure = [
('participant',
Design(ivs={
'a': [1, 2],
'b': [1, 2]
}, ordering=Shuffle(3))),
('session',
Design(ivs={'design': ['practice', 'test']},
design_matrix=[[0], [1], [1]])),
]
other_structures = {
'practice': [
('block', Design()),
('trial', Design(ivs={'difficulty': [1, 2]},
ordering=Shuffle(20))),
],
'test': [
('block', Design(ordering=Ordering(2))),
('trial',
Design(ivs={'difficulty': [1, 3, 5, 7]}, ordering=Shuffle(5))),
],
}
return DesignTree.new(main_structure, **other_structures)
def test_design_matrix():
iv_names = ('a', 'b', 'c')
iv_values = [[1, 2], [1, 2], [1, 2]]
conditions = [
{
'a': 1,
'b': 1,
'c': 2
},
{
'a': 1,
'b': 2,
'c': 1
},
{
'a': 2,
'b': 1,
'c': 1
},
{
'a': 2,
'b': 2,
'c': 2
},
]
matrix = np.array([[-1., -1., 1.], [1., -1., -1.], [-1., 1., -1.],
[1., 1., 1.]]) # pyDOE.fracfact('a b ab')
d = Design(ivs=zip(iv_names, iv_values), design_matrix=matrix)
d.first_pass()
for _ in range(3):
yield check_sequences, d.get_order(), conditions
yield check_design_matrix, d.get_order(), iv_names, iv_values, matrix
def test_design_tree():
trial_matrix = np.array(
[[-1, -1], [1, -1], [-1, 1], [1, 1], [0, 0], [0, 0], [0, 0], [0, 0],
[-1.41421356, 0.1],
[1.41421356, 0.1], [0, -1.41421356], [0,
1.41421356],
[0, 0], [0, 0], [0,
0], [0, 0]]
) # pyDOE.ccdesign(2) with two 0.1 elements to prevent symmetry in the IVs.
trial_iv_names = ['a', 'b']
trial_iv_values = [None, None]
block_ivs = {'block': [1, 2]}
participant_iv_names = ('A', 'B')
participant_iv_values = [[1, 2], [1, 2, 3]]
trial_design = Design(ivs=zip(trial_iv_names, trial_iv_values),
ordering=Shuffle(3),
design_matrix=trial_matrix)
block_design = Design(block_ivs, ordering=CompleteCounterbalance())
practice_block_design = Design()
participant_design = Design(dict(
zip(participant_iv_names, participant_iv_values)),
ordering=Ordering(10))
levels_and_designs = OrderedDict([('participant', participant_design),
('block',
[practice_block_design, block_design]),
('trial', trial_design)])
tree = DesignTree.new(levels_and_designs)
tree.add_base_level()
levels, designs = zip(*tree.levels_and_designs)
yield check_identity, designs[1][0], participant_design
yield check_identity, designs[2][0], practice_block_design
yield check_identity, designs[2][1], block_design
yield check_identity, designs[3][0], trial_design
yield check_equality, len(tree), 4
yield check_equality, tree[3], ('trial', [trial_design])
yield check_equality, next(tree).levels_and_designs, next(
tree).levels_and_designs, tree.levels_and_designs[1:]
tree_with_participant_base = next(tree)
tree_with_block_base = next(tree_with_participant_base)
yield (check_equality, tree_with_block_base.levels_and_designs,
next(next(tree)).levels_and_designs,
[('block', [practice_block_design, block_design]),
('trial', [trial_design])])
tree_with_trial_base = next(tree_with_block_base)
yield (check_equality, tree_with_trial_base.levels_and_designs, [
('trial', [trial_design])
], next(tree_with_block_base).levels_and_designs)
with pytest.raises(StopIteration):
next(tree_with_trial_base)
for design, iv_names, iv_values, n, data, matrix in zip(
[designs[1][0], designs[2][0], designs[2][1], designs[3][0]],
[['A', 'B', CompleteCounterbalance.iv_name], [], ['block'], ['a', 'b']
], [[[1, 2], [1, 2, 3], [0, 1]], [], [[1, 2]], [None, None]],
[10 * 2 * 2 * 3, 1, 2, 3 * len(trial_matrix)],
[{}, {}, {
CompleteCounterbalance.iv_name: 0
}, {}], [None, None, None, trial_matrix]):
yield check_design, design, iv_names, iv_values, n, data, matrix
yield check_design_matrix, designs[3][0].get_order(), ['a', 'b'
], [None, None
], trial_matrix
def make_tree(levels, data):
designs = [[Design([('a', range(len(levels))), ('b', [False, True])], extra_data=data, ordering=Ordering())]
for _ in levels]
return DesignTree.new(list(zip(levels, designs)))
def test_factorial():
iv_names = ('a', 'b')
iv_values = [[1, 2, 3], [1, 2]]
conditions = [
{
'a': 1,
'b': 1
},
{
'a': 1,
'b': 2
},
{
'a': 2,
'b': 1
},
{
'a': 2,
'b': 2
},
{
'a': 3,
'b': 1
},
{
'a': 3,
'b': 2
},
]
assert list(Design.full_cross(iv_names, iv_values)) == conditions
d = Design(zip(iv_names, iv_values))
d.first_pass()
for _ in range(3):
yield check_sequences, d.get_order(), conditions
iv_names = ('a', 'b', 'c')
iv_values = [[1, 2], [1, 2], [1, 2]]
conditions = [
{
'a': 1,
'b': 1,
'c': 1
},
{
'a': 1,
'b': 1,
'c': 2
},
{
'a': 1,
'b': 2,
'c': 1
},
{
'a': 1,
'b': 2,
'c': 2
},
{
'a': 2,
'b': 1,
'c': 1
},
{
'a': 2,
'b': 1,
'c': 2
},
{
'a': 2,
'b': 2,
'c': 1
},
{
'a': 2,
'b': 2,
'c': 2
},
]
assert list(Design.full_cross(iv_names, iv_values)) == conditions
d = Design(zip(iv_names, iv_values))
d.first_pass()
for _ in range(3):
yield check_sequences, d.get_order(), conditions
def test_bad_design_matrix():
with pytest.raises(TypeError):
design = Design(ivs=[('a', [1]), ('b', [1])],
design_matrix=np.ones((3, 3)))
design.first_pass()
def test_continuous_ivs_without_design_matrix():
with pytest.raises(TypeError):
Design(ivs=[('a', None)])
def test_bad_design_matrix():
with pytest.raises(TypeError):
design = Design(ivs=[('a', [1]), ('b', [1])], design_matrix=np.ones((3, 3)))
design.first_pass()
def test_bizarre_equality():
design = Design()
assert (design == 1) is False
tree = DesignTree.new([('a', design)])
assert (tree == 1) is False
