forked from James-Durant/experimental-design
-
Notifications
You must be signed in to change notification settings - Fork 4
/
Copy pathtest_optimise.py
228 lines (184 loc) · 8.4 KB
/
test_optimise.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
"""Unit tests for the methods in the optimise module"""
import pytest
import numpy as np
from hogben.optimise import Optimiser, optimise_parameters
from refnx.analysis import Parameter
from refnx.reflect import SLD
from hogben.models.samples import Sample
from hogben.models.bilayers import BilayerDMPC
from unittest.mock import MagicMock, patch
@pytest.fixture
def refnx_sample():
"""Defines a structure describing a simple sample."""
air = SLD(0, name='Air')
layer1_thick = Parameter(100, 'Layer 1 Thickness', (50, 120))
layer1_sld = Parameter(4, 'Layer 1 SLD', (50, 120))
layer1_thick.optimize = True
layer1_sld.optimize = True
layer1 = SLD(layer1_sld, name='Layer 1')(thick=layer1_thick, rough=2)
layer2 = SLD(8, name='Layer 2')(thick=150, rough=2)
substrate = SLD(2.047, name='Substrate')(thick=0, rough=2)
structure = air | layer1 | layer2 | substrate
sample = Sample(structure)
sample._vary_structure()
return sample
class MockSample:
"""
Mocks a HOGBEN sample, so that class methods can be mocked
and tracked
"""
def __init__(self):
"""Iniitalize the mocked sample"""
self.params = [MagicMock(name='param1'), MagicMock(name='param2')]
self.simulate_reflectivity = MagicMock()
self.sld_profile = MagicMock()
self.get_models = MagicMock()
self.is_magnetic = MagicMock()
self.polarised = False
def get_param_by_attribute(self, _attribute):
"""
Return mocked parameters when requesting parameters from attribute
"""
return self.params
class MockFisher:
"""Mocks the Fisher class, to return a simple minimum eigenvalue of 1.0 """
def __init__(self, _sample, _angle_times, inst_or_path='OFFSPEC'):
"""Initialize the mocked Fisher class and set eigenval to 1"""
self.min_eigenval = 1.0
@staticmethod
def from_sample(sample, angle_times, inst_or_path='OFFSPEC'):
"""Create a Mocked Fisher object from sample"""
return MockFisher(sample, angle_times, inst_or_path)
@patch('hogben.optimise.Optimiser._Optimiser__optimise')
def test_optimise_angle_times_length(mock_optimise, refnx_sample):
"""
Tests that the optimise_angle_times method outputs the correct amount of
angles and counting times.
"""
num_angles = 2
optimiser = Optimiser(refnx_sample)
# Mock values retrieved from previous run
mock_optimise.return_value = np.array([0.8847156, 0.88834418,
0.00139696,
0.99860304]), -0.7573710562837207
angles, splits, _ = optimiser.optimise_angle_times(num_angles,
angle_bounds=(0.2, 2.3),
verbose=False)
assert len(angles) == num_angles and len(splits) == num_angles
@patch('hogben.optimise.Optimiser._Optimiser__optimise')
def test_optimise_contrasts(mock_optimise):
"""
Tests that the optimise_contrasts method outputs the correct amount of
contrasts and counting times.
"""
optimiser = Optimiser(BilayerDMPC())
num_contrasts = 3
angle_times = [(0.7, 100, 10), (2.3, 100, 40)]
# Mock values retreived from previous run
mock_optimise.return_value = (
np.array([-0.56, 2.15, 6.36, 0.17, 0.28, 0.56]), -0.18
)
contrasts, splits, _ = optimiser.optimise_contrasts(num_contrasts,
angle_times,
workers=-1,
verbose=False)
assert len(contrasts) == num_contrasts and len(splits) == num_contrasts
@patch('hogben.optimise.Optimiser._Optimiser__optimise')
def test_optimise_underlayers(mock_optimise):
"""
Tests that the optimise_contrasts method outputs the correct amount of
contrasts and counting times.
"""
optimiser = Optimiser(BilayerDMPC())
num_underlayers = 3
angle_times = [(0.7, 100, 10), (2.3, 100, 40)]
contrasts = [-0.56, 6.36]
thick_bounds = (0, 500)
sld_bounds = (1, 9)
# Mock values retreived from previous run
mock_optimise.return_value = (
np.array([-0.56, 2.15, 6.36, 0.17, 0.28, 0.56]), -0.18
)
contrasts, splits, _ = optimiser.optimise_underlayers(num_underlayers,
angle_times,
contrasts,
thick_bounds,
sld_bounds,
verbose=False)
assert len(contrasts) == num_underlayers and len(splits) == num_underlayers
def test_angle_times_func_result(refnx_sample):
"""Checks that the angle_times_func method gives the correct result"""
angle_time_split = [0.3, 1.3, 0.8, 0.2] # [angle, angle, time, time]
num_angles = 2
contrasts = [3, 14, -2]
points = 100
total_time = 10000
optimiser = Optimiser(refnx_sample)
result = optimiser._angle_times_func(angle_time_split, num_angles,
contrasts, points, total_time)
expected_result = -1.7716709840530174
np.testing.assert_allclose(result, expected_result, rtol=1e-06)
def test_contrasts_func_result():
"""Checks that the _contrasts_func method gives the correct result"""
contrasts_time = [0.3, 9.3, 0.8, 0.2] # [SLD, SLD, time, time]
num_contrasts = 2
angle_splits = [(0.7, 100, 0.6), (2.3, 100, 0.4)]
total_time = 100000
optimiser = Optimiser(BilayerDMPC())
result = optimiser._contrasts_func(contrasts_time, num_contrasts,
angle_splits, total_time)
expected_result = -0.199884
np.testing.assert_allclose(result, expected_result, rtol=1e-06)
def test_underlayers_func():
"""Checks that the _underlayers_func method gives the correct result"""
thickness_SLD = [50, 20, -5, 10] # [thickness, thickness, SLD, SLD]
bilayer = BilayerDMPC()
optimiser = Optimiser(bilayer)
num_underlayers = 2
contrasts = [-0.56, 6.36]
angle_times = [(0.7, 100, 10000), (2.3, 100, 10000)]
result = optimiser._underlayers_func(thickness_SLD, num_underlayers,
angle_times, contrasts)
expected_result = -1.500100628
np.testing.assert_allclose(result, expected_result, rtol=1e-06)
@patch('hogben.optimise.Optimiser.optimise_parameters')
@patch('hogben.optimise.scan_parameters')
@patch('hogben.optimise.Fisher.from_sample', new=MockFisher)
def test_optimise_parameters(mock_scan_parameters, mock_optimise_parameters):
"""
Runs the general `optimise_parameters` workflow, and makes sure the output
is handled correctly, and data visualization is called.
"""
mock_optimise_parameters.return_value = ([0.5, 0.5], 2.0)
mock_scan_parameters.return_value = None
sample = MockSample()
angle_times = [1, 2, 3]
result = optimise_parameters(sample, angle_times, visualise=True)
# Check if the sample parameters were updated correctly
assert result.params[0].value == 0.5
assert result.params[1].value == 0.5
# Check whether the visualisation functions are called
sample.sld_profile.assert_called_once()
sample.simulate_reflectivity.assert_called_once()
mock_scan_parameters.assert_called_once()
def test_optimise_params_length(refnx_sample):
"""
Runs the optimiser workflow and tests that the optimise_parameters method
outputs the correct amount of counting parameters.
"""
angle_times = [(0.7, 100, 10000), (2.3, 100, 10000)]
params = refnx_sample.get_param_by_attribute('optimize')
optimiser = Optimiser(refnx_sample)
# Mock values retrieved from previous run
values, _ = optimiser.optimise_parameters(angle_times, verbose=False)
assert len(values) == len(params)
def test_parameter_func(refnx_sample):
"""Checks that the _underlayers_func method gives the correct result"""
optimiser = Optimiser(refnx_sample)
params = refnx_sample.get_param_by_attribute('optimize')
params.sort(key=lambda x: x.name)
values = [3.3, 109]
angle_times = [(0.7, 100, 10000), (2.3, 100, 10000)]
result = optimiser._parameter_func(values, params, angle_times)
expected_result = -15.438106789796723
np.testing.assert_allclose(result, expected_result, rtol=1e-06)