def test_grammar_json(self):
self.sut.append_step_estimation(0, self.mk_evolution_step(tp=3, tn=2, fp=3, fn=2))
self.sut.append_step_estimation(0, self.mk_evolution_step(tp=5, tn=5, fp=0, fn=0))
self.sut.append_step_estimation(1, self.mk_evolution_step(tp=0, tn=1, fp=2, fn=0))
self.common_final_assert()
jsonized = self.sut.json_coder()
self.sut = GrammarEstimator()
self.sut.json_decoder(jsonized)
self.common_final_assert()
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
self.sut = GrammarEstimator()
class TestGrammarEstimator(unittest.TestCase):
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
self.sut = GrammarEstimator()
@staticmethod
def mk_evolution_step(tp, tn, fp, fn):
step = EvolutionStepEstimator()
step._true_positive = tp
step._true_negative = tn
step._false_positive = fp
step._false_negative = fn
return step
def assert_estimation(self, step, fitness, miss_rate, fallout, sensitivity, specificity,
accuracy,
min_fitness, max_fitness, min_miss_rate, max_miss_rate,
min_fallout, max_fallout, min_sensitivity, max_sensitivity,
min_specificity, max_specificity, min_accuracy, max_accuracy,
average_fitness, average_miss_rate, average_fallout, average_sensitivity,
average_specificity, average_accuracy, global_min_fitness,
global_min_miss_rate, global_min_fallout, global_min_sensitivity,
global_min_specificity, global_min_accuracy, global_max_fitness,
global_max_miss_rate, global_max_fallout, global_max_sensitivity,
global_max_specificity, global_max_accuracy, estimator=None):
fit = 'fitness'
mis = 'missrate'
fal = 'fallout'
sen = 'sensitivity'
spe = 'specificity'
acc = 'accuracy'
estimator = estimator if estimator is not None else self.sut
assert_nearly_equal_or_both_nan(estimator[fit].get(step), fitness)
assert_nearly_equal_or_both_nan(estimator[mis].get(step), miss_rate)
assert_nearly_equal_or_both_nan(estimator[fal].get(step), fallout)
assert_nearly_equal_or_both_nan(estimator[sen].get(step), sensitivity)
assert_nearly_equal_or_both_nan(estimator[spe].get(step), specificity)
assert_nearly_equal_or_both_nan(estimator[acc].get(step), accuracy)
assert_nearly_equal_or_both_nan(estimator[fit].get_min(step), min_fitness)
assert_nearly_equal_or_both_nan(estimator[fit].get_max(step), max_fitness)
assert_nearly_equal_or_both_nan(estimator[mis].get_min(step), min_miss_rate)
assert_nearly_equal_or_both_nan(estimator[mis].get_max(step), max_miss_rate)
assert_nearly_equal_or_both_nan(estimator[fal].get_min(step), min_fallout)
assert_nearly_equal_or_both_nan(estimator[fal].get_max(step), max_fallout)
assert_nearly_equal_or_both_nan(estimator[sen].get_min(step), min_sensitivity)
assert_nearly_equal_or_both_nan(estimator[sen].get_max(step), max_sensitivity)
assert_nearly_equal_or_both_nan(estimator[spe].get_min(step), min_specificity)
assert_nearly_equal_or_both_nan(estimator[spe].get_max(step), max_specificity)
assert_nearly_equal_or_both_nan(estimator[acc].get_min(step), min_accuracy)
assert_nearly_equal_or_both_nan(estimator[acc].get_max(step), max_accuracy)
assert_nearly_equal_or_both_nan(estimator[fit].get_global_average(), average_fitness)
assert_nearly_equal_or_both_nan(estimator[mis].get_global_average(), average_miss_rate)
assert_nearly_equal_or_both_nan(estimator[fal].get_global_average(), average_fallout)
assert_nearly_equal_or_both_nan(estimator[sen].get_global_average(), average_sensitivity)
assert_nearly_equal_or_both_nan(estimator[spe].get_global_average(), average_specificity)
assert_nearly_equal_or_both_nan(estimator[acc].get_global_average(), average_accuracy)
assert_nearly_equal_or_both_nan(estimator[fit].get_global_min(), global_min_fitness)
assert_nearly_equal_or_both_nan(estimator[mis].get_global_min(), global_min_miss_rate)
assert_nearly_equal_or_both_nan(estimator[fal].get_global_min(), global_min_fallout)
assert_nearly_equal_or_both_nan(estimator[sen].get_global_min(), global_min_sensitivity)
assert_nearly_equal_or_both_nan(estimator[spe].get_global_min(), global_min_specificity)
assert_nearly_equal_or_both_nan(estimator[acc].get_global_min(), global_min_accuracy)
assert_nearly_equal_or_both_nan(estimator[fit].get_global_max(), global_max_fitness)
assert_nearly_equal_or_both_nan(estimator[mis].get_global_max(), global_max_miss_rate)
assert_nearly_equal_or_both_nan(estimator[fal].get_global_max(), global_max_fallout)
assert_nearly_equal_or_both_nan(estimator[sen].get_global_max(), global_max_sensitivity)
assert_nearly_equal_or_both_nan(estimator[spe].get_global_max(), global_max_specificity)
assert_nearly_equal_or_both_nan(estimator[acc].get_global_max(), global_max_accuracy)
def test_grammar_estimation(self):
self.assert_estimation(step=0, fitness=float('nan'), miss_rate=float('nan'),
fallout=float('nan'), sensitivity=float('nan'),
specificity=float('nan'), accuracy=float('nan'),
min_fitness=float('nan'), max_fitness=float('nan'),
min_miss_rate=float('nan'), max_miss_rate=float('nan'),
min_fallout=float('nan'), max_fallout=float('nan'),
min_sensitivity=float('nan'), max_sensitivity=float('nan'),
min_specificity=float('nan'), max_specificity=float('nan'),
min_accuracy=float('nan'), max_accuracy=float('nan'),
average_fitness=float('nan'),
average_miss_rate=float('nan'),
average_fallout=float('nan'),
average_sensitivity=float('nan'),
average_specificity=float('nan'),
average_accuracy=float('nan'),
global_min_fitness=float('nan'),
global_min_miss_rate=float('nan'),
global_min_fallout=float('nan'),
global_min_sensitivity=float('nan'),
global_min_specificity=float('nan'),
global_min_accuracy=float('nan'),
global_max_fitness=float('nan'),
global_max_miss_rate=float('nan'),
global_max_fallout=float('nan'),
global_max_sensitivity=float('nan'),
global_max_specificity=float('nan'),
global_max_accuracy=float('nan'))
self.sut.append_step_estimation(0, self.mk_evolution_step(tp=3, tn=2, fp=3, fn=2))
self.assert_estimation(step=0, fitness=0.5, miss_rate=0.4, fallout=0.6, sensitivity=0.6,
specificity=0.4, accuracy=0.5,
min_fitness=0.5, max_fitness=0.5,
min_miss_rate=0.4, max_miss_rate=0.4,
min_fallout=0.6, max_fallout=0.6,
min_sensitivity=0.6, max_sensitivity=0.6,
min_specificity=0.4, max_specificity=0.4,
min_accuracy=0.5, max_accuracy=0.5,
average_fitness=0.5,
average_miss_rate=0.4,
average_fallout=0.6,
average_sensitivity=0.6,
average_specificity=0.4,
average_accuracy=0.5,
global_min_fitness=0.5,
global_min_miss_rate=0.4,
global_min_fallout=0.6,
global_min_sensitivity=0.6,
global_min_specificity=0.4,
global_min_accuracy=0.5,
global_max_fitness=0.5,
global_max_miss_rate=0.4,
global_max_fallout=0.6,
global_max_sensitivity=0.6,
global_max_specificity=0.4,
global_max_accuracy=0.5)
self.sut.append_step_estimation(0, self.mk_evolution_step(tp=5, tn=5, fp=0, fn=0))
self.assert_estimation(step=0, fitness=0.75, miss_rate=0.2, fallout=0.3, sensitivity=0.8,
specificity=0.7, accuracy=0.75,
min_fitness=0.5, max_fitness=0.75,
min_miss_rate=0.2, max_miss_rate=0.4,
min_fallout=0.3, max_fallout=0.6,
min_sensitivity=0.6, max_sensitivity=0.8,
min_specificity=0.4, max_specificity=0.7,
min_accuracy=0.5, max_accuracy=0.75,
average_fitness=0.75,
average_miss_rate=0.2,
average_fallout=0.3,
average_sensitivity=0.8,
average_specificity=0.7,
average_accuracy=0.75,
global_min_fitness=0.5,
global_min_miss_rate=0.2,
global_min_fallout=0.3,
global_min_sensitivity=0.6,
global_min_specificity=0.4,
global_min_accuracy=0.5,
global_max_fitness=0.75,
global_max_miss_rate=0.4,
global_max_fallout=0.6,
global_max_sensitivity=0.8,
global_max_specificity=0.7,
global_max_accuracy=0.75)
self.sut.append_step_estimation(1, self.mk_evolution_step(tp=0, tn=1, fp=2, fn=0))
self.common_final_assert()
def common_grammar_estimation_adding(self, left, right, first, second, third):
first.append_step_estimation(0, self.mk_evolution_step(tp=7, tn=2, fp=1, fn=0))
self.assert_estimation(step=0, fitness=0.9, miss_rate=0, fallout=0.33, sensitivity=1,
specificity=0.67, accuracy=0.9,
min_fitness=0.9, max_fitness=0.9,
min_miss_rate=0, max_miss_rate=0,
min_fallout=0.33, max_fallout=0.33,
min_sensitivity=1, max_sensitivity=1,
min_specificity=0.67, max_specificity=0.67,
min_accuracy=0.9, max_accuracy=0.9,
average_fitness=0.9,
average_miss_rate=0,
average_fallout=0.33,
average_sensitivity=1,
average_specificity=0.67,
average_accuracy=0.9,
global_min_fitness=0.9,
global_min_miss_rate=0,
global_min_fallout=0.33,
global_min_sensitivity=1,
global_min_specificity=0.67,
global_min_accuracy=0.9,
global_max_fitness=0.9,
global_max_miss_rate=0,
global_max_fallout=0.33,
global_max_sensitivity=1,
global_max_specificity=0.67,
global_max_accuracy=0.9,
estimator=left + right)
second.append_step_estimation(0, self.mk_evolution_step(tp=3, tn=2, fp=0, fn=1))
self.assert_estimation(step=0, fitness=0.87, miss_rate=0.13, fallout=0.17, sensitivity=0.88,
specificity=0.83, accuracy=0.87,
min_fitness=0.87, max_fitness=0.9,
min_miss_rate=0, max_miss_rate=0.13,
min_fallout=0.17, max_fallout=0.33,
min_sensitivity=0.88, max_sensitivity=1,
min_specificity=0.67, max_specificity=0.83,
min_accuracy=0.87, max_accuracy=0.9,
average_fitness=0.87,
average_miss_rate=0.13,
average_fallout=0.17,
average_sensitivity=0.88,
average_specificity=0.83,
average_accuracy=0.87,
global_min_fitness=0.87,
global_min_miss_rate=0,
global_min_fallout=0.17,
global_min_sensitivity=0.88,
global_min_specificity=0.67,
global_min_accuracy=0.87,
global_max_fitness=0.9,
global_max_miss_rate=0.13,
global_max_fallout=0.33,
global_max_sensitivity=1,
global_max_specificity=0.83,
global_max_accuracy=0.9,
estimator=left + right)
third.append_step_estimation(1, self.mk_evolution_step(tp=2, tn=7, fp=1, fn=5))
self.assert_estimation(step=0, fitness=0.87, miss_rate=0.13, fallout=0.17, sensitivity=0.88,
specificity=0.83, accuracy=0.87,
min_fitness=0.87, max_fitness=0.9,
min_miss_rate=0, max_miss_rate=0.13,
min_fallout=0.17, max_fallout=0.33,
min_sensitivity=0.88, max_sensitivity=1,
min_specificity=0.67, max_specificity=0.83,
min_accuracy=0.87, max_accuracy=0.9,
average_fitness=0.73,
average_miss_rate=0.42,
average_fallout=0.15,
average_sensitivity=0.58,
average_specificity=0.85,
average_accuracy=0.73,
global_min_fitness=0.6,
global_min_miss_rate=0,
global_min_fallout=0.13,
global_min_sensitivity=0.29,
global_min_specificity=0.67,
global_min_accuracy=0.6,
global_max_fitness=0.9,
global_max_miss_rate=0.71,
global_max_fallout=0.33,
global_max_sensitivity=1,
global_max_specificity=0.88,
global_max_accuracy=0.9,
estimator=left + right)
self.assert_estimation(step=1, fitness=0.6, miss_rate=0.71, fallout=0.13, sensitivity=0.29,
specificity=0.88, accuracy=0.6,
min_fitness=0.6, max_fitness=0.6,
min_miss_rate=0.71, max_miss_rate=0.71,
min_fallout=0.13, max_fallout=0.13,
min_sensitivity=0.29, max_sensitivity=0.29,
min_specificity=0.88, max_specificity=0.88,
min_accuracy=0.6, max_accuracy=0.6,
average_fitness=0.73,
average_miss_rate=0.42,
average_fallout=0.15,
average_sensitivity=0.58,
average_specificity=0.85,
average_accuracy=0.73,
global_min_fitness=0.6,
global_min_miss_rate=0,
global_min_fallout=0.13,
global_min_sensitivity=0.29,
global_min_specificity=0.67,
global_min_accuracy=0.6,
global_max_fitness=0.9,
global_max_miss_rate=0.71,
global_max_fallout=0.33,
global_max_sensitivity=1,
global_max_specificity=0.88,
global_max_accuracy=0.9,
estimator=left + right)
@staticmethod
def _select(left, right, decision):
return left if not decision else right
def test_grammar_estimation_adding(self):
test_data = [
(0, 0, 0) # 3/0
, (0, 0, 1) # false
, (0, 1, 0) # 2/1
, (0, 1, 1) # 1/2
]
for decisions in test_data:
left = GrammarEstimator()
right = GrammarEstimator()
args = [left, right] + [self._select(left, right, x) for x in decisions]
self.common_grammar_estimation_adding(*args)
def common_final_assert(self):
self.assert_estimation(step=1, fitness=0.33, miss_rate=float('nan'), fallout=0.67,
sensitivity=float('nan'), specificity=0.33, accuracy=0.33,
min_fitness=0.33, max_fitness=0.33,
min_miss_rate=float('nan'), max_miss_rate=float('nan'),
min_fallout=0.67, max_fallout=0.67,
min_sensitivity=float('nan'), max_sensitivity=float('nan'),
min_specificity=0.33, max_specificity=0.33,
min_accuracy=0.33, max_accuracy=0.33,
average_fitness=0.54,
average_miss_rate=0.2,
average_fallout=0.48,
average_sensitivity=0.8,
average_specificity=0.52,
average_accuracy=0.54,
global_min_fitness=0.33,
global_min_miss_rate=0.2,
global_min_fallout=0.3,
global_min_sensitivity=0.6,
global_min_specificity=0.33,
global_min_accuracy=0.33,
global_max_fitness=0.75,
global_max_miss_rate=0.4,
global_max_fallout=0.67,
global_max_sensitivity=0.8,
global_max_specificity=0.7,
global_max_accuracy=0.75)
self.assert_estimation(step=0, fitness=0.75, miss_rate=0.2, fallout=0.3, sensitivity=0.8,
specificity=0.7, accuracy=0.75,
min_fitness=0.5, max_fitness=0.75,
min_miss_rate=0.2, max_miss_rate=0.4,
min_fallout=0.3, max_fallout=0.6,
min_sensitivity=0.6, max_sensitivity=0.8,
min_specificity=0.4, max_specificity=0.7,
min_accuracy=0.5, max_accuracy=0.75,
average_fitness=0.54,
average_miss_rate=0.2,
average_fallout=0.48,
average_sensitivity=0.8,
average_specificity=0.52,
average_accuracy=0.54,
global_min_fitness=0.33,
global_min_miss_rate=0.2,
global_min_fallout=0.3,
global_min_sensitivity=0.6,
global_min_specificity=0.33,
global_min_accuracy=0.33,
global_max_fitness=0.75,
global_max_miss_rate=0.4,
global_max_fallout=0.67,
global_max_sensitivity=0.8,
global_max_specificity=0.7,
global_max_accuracy=0.75)
def test_grammar_json(self):
self.sut.append_step_estimation(0, self.mk_evolution_step(tp=3, tn=2, fp=3, fn=2))
self.sut.append_step_estimation(0, self.mk_evolution_step(tp=5, tn=5, fp=0, fn=0))
self.sut.append_step_estimation(1, self.mk_evolution_step(tp=0, tn=1, fp=2, fn=0))
self.common_final_assert()
jsonized = self.sut.json_coder()
self.sut = GrammarEstimator()
self.sut.json_decoder(jsonized)
self.common_final_assert()
