def test_plot__individual_cdf(self):
x, outcomes = utilities.load_flu_data()
y = outcomes['deceased population region 1'][:, -1] > 1000000
fig, ax = plt.subplots()
unc = 'fatality ratio region 1'
regional_sa.plot_individual_cdf(ax,
unc,
x[unc],
y,
discrete=False,
legend=True,
xticklabels_on=True,
yticklabels_on=True)
fig, ax = plt.subplots()
unc = 'model'
regional_sa.plot_individual_cdf(ax,
unc,
x[unc],
y,
discrete=True,
legend=True,
xticklabels_on=True,
yticklabels_on=True)
def test_get_univariate_feature_scores(self):
x, outcomes = utilities.load_flu_data()
def classify(data):
#get the output for deceased population
result = data['deceased population region 1']
#make an empty array of length equal to number of cases
classes = np.zeros(result.shape[0])
#if deceased population is higher then 1.000.000 people, classify as 1
classes[result[:, -1] > 1000000] = 1
return classes
y = classify(outcomes)
# f classify
scores = fs.get_univariate_feature_scores(x,y,
score_func=F_CLASSIFICATION)
self.assertEqual(len(scores), len(x.columns)-2)
# chi2
scores = fs.get_univariate_feature_scores(x,y, score_func=CHI2)
self.assertEqual(len(scores), len(x.columns)-2)
# f regression
y= outcomes['deceased population region 1'][:,-1]
scores = fs.get_univariate_feature_scores(x,y, score_func=F_REGRESSION)
self.assertEqual(len(scores), len(x.columns)-2)
def test_prepare_outcomes(self):
results = utilities.load_flu_data()
# string type correct
ooi = 'nr deaths'
results[1][ooi] = results[1]['deceased population region 1'][:,-1]
y, categorical = fs._prepare_outcomes(results[1], ooi)
self.assertFalse(categorical)
self.assertTrue(len(y.shape)==1)
# string type not correct --> KeyError
with self.assertRaises(KeyError):
fs._prepare_outcomes(results[1], "non existing key")
# classify function correct
def classify(data):
result = data['deceased population region 1']
classes = np.zeros(result.shape[0])
classes[result[:, -1] > 1000000] = 1
return classes
y, categorical = fs._prepare_outcomes(results[1], classify)
self.assertTrue(categorical)
self.assertTrue(len(y.shape)==1)
# neither string nor classify function --> TypeError
with self.assertRaises(TypeError):
fs._prepare_outcomes(results[1], 1)
def test_setup_prim(self):
self.results = utilities.load_flu_data()
self.classify = flu_classify
experiments, outcomes = self.results
# test initialization, including t_coi calculation in case of searching
# for results equal to or higher than the threshold
outcomes['death toll'] = outcomes['deceased population region 1'][:,
-1]
results = experiments, outcomes
threshold = 10000
prim_obj = prim.setup_prim(results,
classify='death toll',
threshold_type=prim.ABOVE,
threshold=threshold)
value = np.ones((experiments.shape[0], ))
value = value[outcomes['death toll'] >= threshold].shape[0]
self.assertTrue(prim_obj.t_coi == value)
# test initialization, including t_coi calculation in case of searching
# for results equal to or lower than the threshold
threshold = 1000
prim_obj = prim.setup_prim(results,
classify='death toll',
threshold_type=prim.BELOW,
threshold=threshold)
value = np.ones((experiments.shape[0], ))
value = value[outcomes['death toll'] <= threshold].shape[0]
self.assertTrue(prim_obj.t_coi == value)
prim.setup_prim(self.results, self.classify, threshold=prim.ABOVE)
def test_setup_cart(self):
results = utilities.load_flu_data()
alg = cart.setup_cart(results, flu_classify,
mass_min=0.05)
self.assertTrue(alg.mode==BINARY)
x, outcomes = results
y = {k:v[:, -1] for k,v in outcomes.items()}
temp_results = (x,y)
alg = cart.setup_cart(temp_results,
'deceased population region 1',
mass_min=0.05)
self.assertTrue(alg.mode==REGRESSION)
n_cols = 5
unc = x.columns.values[0:n_cols]
alg = cart.setup_cart(results,
flu_classify,
mass_min=0.05,
incl_unc=unc)
self.assertTrue(alg.mode==BINARY)
self.assertTrue(alg.x.shape[1]==n_cols)
with self.assertRaises(TypeError):
alg = cart.setup_cart(results, 10,
mass_min=0.05)
def test_get_lasso_feature_scores(self):
x, outcomes = utilities.load_flu_data()
def classify(data):
#get the output for deceased population
result = data['deceased population region 1']
#make an empty array of length equal to number of cases
classes = np.zeros(result.shape[0])
#if deceased population is higher then 1.000.000 people, classify as 1
classes[result[:, -1] > 1000000] = 1
return classes
y = classify(outcomes)
# classification based
scores = fs.get_lasso_feature_scores(x, y, mode=CLASSIFICATION,
random_state=42)
self.assertEqual(len(scores), len(x.dtype.fields))
self.assertRaises(ValueError, fs.get_lasso_feature_scores, x,y,
mode='illegal argument')
#regression based
y = outcomes['deceased population region 1'][:,-1]
scores = fs.get_lasso_feature_scores(x, y, mode=REGRESSION,
random_state=42)
self.assertEqual(len(scores), len(x.dtype.fields))
def test_get_rf_feature_scores(self):
x, outcomes = utilities.load_flu_data()
def classify(data):
#get the output for deceased population
result = data['deceased population region 1']
#make an empty array of length equal to number of cases
classes = np.zeros(result.shape[0])
#if deceased population is higher then 1.000.000 people, classify as 1
classes[result[:, -1] > 1000000] = 1
return classes
y = classify(outcomes)
scores, forest = fs.get_rf_feature_scores(x,y,
mode=RuleInductionType.CLASSIFICATION,
random_state=10)
self.assertEqual(len(scores), len(x.columns)-2)
self.assertTrue(isinstance(forest, RandomForestClassifier))
self.assertRaises(ValueError, fs.get_rf_feature_scores, x,y,
mode='illegal argument')
y = outcomes['deceased population region 1'][:,-1]
scores, forest = fs.get_rf_feature_scores(x,y,
mode=RuleInductionType.REGRESSION,
random_state=10)
self.assertEqual(len(scores), len(x.columns)-2)
self.assertTrue(isinstance(forest, RandomForestRegressor))
def test_prim_init_select(self):
self.results = utilities.load_flu_data()
self.classify = flu_classify
experiments, outcomes = self.results
unc = experiments.columns.values.tolist()
# test initialization, including t_coi calculation in case of searching
# for results equal to or higher than the threshold
outcomes['death toll'] = outcomes['deceased population region 1'][:, -1]
results = experiments, outcomes
threshold = 10000
prim_obj = prim.setup_prim(results, classify='death toll',
threshold_type=prim.ABOVE, threshold=threshold,
incl_unc=unc)
value = np.ones((experiments.shape[0],))
value = value[outcomes['death toll'] >= threshold].shape[0]
self.assertTrue(prim_obj.t_coi==value)
# test initialization, including t_coi calculation in case of searching
# for results equal to or lower than the threshold
threshold = 1000
prim_obj = prim.setup_prim(results, classify='death toll',
threshold_type=prim.BELOW,
threshold=threshold)
value = np.ones((experiments.shape[0],))
value = value[outcomes['death toll'] <= threshold].shape[0]
self.assertTrue(prim_obj.t_coi==value)
prim.setup_prim(self.results, self.classify, threshold=prim.ABOVE)
def test_prim_exceptions(self):
results = utilities.load_flu_data()
x, outcomes = results
y = outcomes['deceased population region 1']
self.assertRaises(prim.PrimException, prim.Prim,
x, y, threshold=0.8, mode='regression')
def test_get_rf_feature_scores(self):
x, outcomes = utilities.load_flu_data()
def classify(data):
#get the output for deceased population
result = data['deceased population region 1']
#make an empty array of length equal to number of cases
classes = np.zeros(result.shape[0])
#if deceased population is higher then 1.000.000 people, classify as 1
classes[result[:, -1] > 1000000] = 1
return classes
y = classify(outcomes)
scores, forest = fs.get_rf_feature_scores(
x, y, mode=RuleInductionType.CLASSIFICATION, random_state=10)
self.assertEqual(len(scores), len(x.columns) - 3)
self.assertTrue(isinstance(forest, RandomForestClassifier))
self.assertRaises(ValueError,
fs.get_rf_feature_scores,
x,
y,
mode='illegal argument')
y = outcomes['deceased population region 1'][:, -1]
scores, forest = fs.get_rf_feature_scores(
x, y, mode=RuleInductionType.REGRESSION, random_state=10)
self.assertEqual(len(scores), len(x.columns) - 3)
self.assertTrue(isinstance(forest, RandomForestRegressor))
def test_prepare_outcomes(self):
results = utilities.load_flu_data()
# string type correct
ooi = 'nr deaths'
results[1][ooi] = results[1]['deceased population region 1'][:, -1]
y, categorical = fs._prepare_outcomes(results[1], ooi)
self.assertFalse(categorical)
self.assertTrue(len(y.shape) == 1)
# string type not correct --> KeyError
with self.assertRaises(KeyError):
fs._prepare_outcomes(results[1], "non existing key")
# classify function correct
def classify(data):
result = data['deceased population region 1']
classes = np.zeros(result.shape[0])
classes[result[:, -1] > 1000000] = 1
return classes
y, categorical = fs._prepare_outcomes(results[1], classify)
self.assertTrue(categorical)
self.assertTrue(len(y.shape) == 1)
# neither string nor classify function --> TypeError
with self.assertRaises(TypeError):
fs._prepare_outcomes(results[1], 1)
def test_get_univariate_feature_scores(self):
x, outcomes = utilities.load_flu_data()
def classify(data):
#get the output for deceased population
result = data['deceased population region 1']
#make an empty array of length equal to number of cases
classes = np.zeros(result.shape[0])
#if deceased population is higher then 1.000.000 people, classify as 1
classes[result[:, -1] > 1000000] = 1
return classes
y = classify(outcomes)
# f classify
scores = fs.get_univariate_feature_scores(x,
y,
score_func=F_CLASSIFICATION)
self.assertEqual(len(scores), len(x.columns) - 3)
# chi2
scores = fs.get_univariate_feature_scores(x, y, score_func=CHI2)
self.assertEqual(len(scores), len(x.columns) - 3)
# f regression
y = outcomes['deceased population region 1'][:, -1]
scores = fs.get_univariate_feature_scores(x,
y,
score_func=F_REGRESSION)
self.assertEqual(len(scores), len(x.columns) - 3)
def test_setup_cart(self):
results = utilities.load_flu_data()
alg = cart.setup_cart(results, flu_classify,
mass_min=0.05)
self.assertTrue(alg.mode==RuleInductionType.BINARY)
x, outcomes = results
y = {k:v[:, -1] for k,v in outcomes.items()}
temp_results = (x,y)
alg = cart.setup_cart(temp_results,
'deceased population region 1',
mass_min=0.05)
self.assertTrue(alg.mode==RuleInductionType.REGRESSION)
n_cols = 5
unc = x.columns.values[0:n_cols]
alg = cart.setup_cart(results,
flu_classify,
mass_min=0.05,
incl_unc=unc)
self.assertTrue(alg.mode==RuleInductionType.BINARY)
self.assertTrue(alg.x.shape[1]==n_cols)
with self.assertRaises(TypeError):
alg = cart.setup_cart(results, 10,
mass_min=0.05)
def test_prim_exceptions(self):
results = utilities.load_flu_data()
x, outcomes = results
y = outcomes['deceased population region 1']
self.assertRaises(prim.PrimException, prim.Prim,
x, y, threshold=0.8,
mode=RuleInductionType.REGRESSION)
def test_plot_cdfs(self):
x, outcomes = utilities.load_flu_data()
y = outcomes['deceased population region 1'][:, -1] > 1000000
regional_sa.plot_cdfs(x, y)
regional_sa.plot_cdfs(x, y, ccdf=True)
x = x.drop('scenario', axis=1)
regional_sa.plot_cdfs(x, y, ccdf=True)
def test_plot_cdfs(self):
x, outcomes = utilities.load_flu_data()
y = outcomes['deceased population region 1'][:,-1] > 1000000
regional_sa.plot_cdfs(x, y)
regional_sa.plot_cdfs(x, y, ccdf=True)
x = x.drop('scenario', axis=1)
regional_sa.plot_cdfs(x, y, ccdf=True)
def test_get_feature_scores_all(self):
x, outcomes = utilities.load_flu_data()
# we have timeseries so we need scalars
y = {'deceased population':outcomes['deceased population region 1'][:, -1],
'max. infected fraction':np.max(outcomes['infected fraction R1'], axis=1)}
scores = fs.get_feature_scores_all(x,y)
self.assertEqual(len(scores), len(x.columns)-2)
self.assertTrue(scores.ndim==2)
def test_get_feature_scores_all(self):
x, outcomes = utilities.load_flu_data()
# we have timeseries so we need scalars
y = {'deceased population':outcomes['deceased population region 1'][:, -1],
'max. infected fraction':np.max(outcomes['infected fraction R1'], axis=1)}
scores = fs.get_feature_scores_all(x,y)
self.assertEqual(len(scores), len(x.columns))
self.assertTrue(scores.ndim==2)
def test_create_pivot_plot(self):
x, outcomes = utilities.load_flu_data()
y = outcomes['deceased population region 1'][:, -1] > 1000000
dimensional_stacking.create_pivot_plot(x, y, 2)
dimensional_stacking.create_pivot_plot(x, y, 2, labels=False,
bin_labels=True )
dimensional_stacking.create_pivot_plot(x, y, 1, labels=False)
plt.draw()
plt.close('all')
def test_show_tree(self):
results = utilities.load_flu_data()
alg = cart.setup_cart(results, flu_classify,
mass_min=0.05)
alg.build_tree()
fig = alg.show_tree(mplfig=True)
bytestream = alg.show_tree(mplfig=False)
self.assertTrue(isinstance(fig, mpl.figure.Figure))
self.assertTrue(isinstance(bytestream, bytes))
def test_setup_prim_exceptions(self):
results = utilities.load_flu_data()
self.assertRaises(prim.PrimException,
prim.setup_prim,
results,
'deceased population region 1',
threshold=0.8)
def faulty_classify(outcomes):
return outcomes['deceased population region 1'][:, 0:10]
self.assertRaises(prim.PrimException, prim.setup_prim, results,
faulty_classify, threshold=0.8)
def test_build_tree(self):
results = utilities.load_flu_data()
alg = cart.setup_cart(results, flu_classify, mass_min=0.05)
alg.build_tree()
self.assertTrue(isinstance(alg.clf, cart.tree.DecisionTreeClassifier))
x, outcomes = results
y = {k: v[:, -1] for k, v in outcomes.items()}
temp_results = (x, y)
alg = cart.setup_cart(temp_results,
'deceased population region 1',
mass_min=0.05)
alg.build_tree()
self.assertTrue(isinstance(alg.clf, cart.tree.DecisionTreeRegressor))
def test_setup_prim_exceptions(self):
results = utilities.load_flu_data()
self.assertRaises(prim.PrimException,
prim.setup_prim,
results,
'deceased population region 1',
threshold=0.8)
def faulty_classify(outcomes):
return outcomes['deceased population region 1'][:, 0:10]
self.assertRaises(prim.PrimException,
prim.setup_prim,
results,
faulty_classify,
threshold=0.8)
def test_stats_to_dataframe(self):
x, outcomes = utilities.load_flu_data()
y = flu_classify(outcomes)
alg = cart.CART(x, y, mode=RuleInductionType.BINARY)
alg.build_tree()
stats = alg.stats_to_dataframe()
y = outcomes['deceased population region 1'][:, -1]
alg = cart.CART(x, y, mode=RuleInductionType.REGRESSION)
alg.build_tree()
stats = alg.stats_to_dataframe()
y = np.random.randint(1, 5, y.shape[0])
alg = cart.CART(x, y, mode=RuleInductionType.CLASSIFICATION)
alg.build_tree()
stats = alg.stats_to_dataframe()
print(stats)
def test_find_box(self):
results = utilities.load_flu_data()
classify = flu_classify
prim_obj = prim.setup_prim(results, classify, threshold=0.8)
box_1 = prim_obj.find_box()
prim_obj._update_yi_remaining()
after_find = box_1.yi.shape[0] + prim_obj.yi_remaining.shape[0]
self.assertEqual(after_find, prim_obj.y.shape[0])
box_2 = prim_obj.find_box()
prim_obj._update_yi_remaining()
after_find = box_1.yi.shape[0] +\
box_2.yi.shape[0] +\
prim_obj.yi_remaining.shape[0]
self.assertEqual(after_find, prim_obj.y.shape[0])
def test_boxes(self):
x = np.array([(0, 1, 2), (2, 5, 6), (3, 2, 1)],
dtype=[('a', np.float), ('b', np.float), ('c', np.float)])
y = {'y': np.array([0, 1, 2])}
results = (x, y)
prim_obj = prim.setup_prim(results, 'y', threshold=0.8)
boxes = prim_obj.boxes
self.assertEqual(len(boxes), 1, 'box length not correct')
# real data test case
prim_obj = prim.setup_prim(utilities.load_flu_data(),
flu_classify,
threshold=0.8)
prim_obj.find_box()
boxes = prim_obj.boxes
self.assertEqual(len(boxes), 1, 'box length not correct')
def test_build_tree(self):
results = utilities.load_flu_data()
alg = cart.setup_cart(results, flu_classify,
mass_min=0.05)
alg.build_tree()
self.assertTrue(isinstance(alg.clf,
cart.tree.DecisionTreeClassifier))
x, outcomes = results
y = {k:v[:, -1] for k,v in outcomes.items()}
temp_results = (x,y)
alg = cart.setup_cart(temp_results,
'deceased population region 1',
mass_min=0.05)
alg.build_tree()
self.assertTrue(isinstance(alg.clf,
cart.tree.DecisionTreeRegressor))
def test_find_box(self):
results = utilities.load_flu_data()
classify = flu_classify
prim_obj = prim.setup_prim(results, classify,
threshold=0.8)
box_1 = prim_obj.find_box()
prim_obj._update_yi_remaining(prim_obj)
after_find = box_1.yi.shape[0] + prim_obj.yi_remaining.shape[0]
self.assertEqual(after_find, prim_obj.y.shape[0])
box_2 = prim_obj.find_box()
prim_obj._update_yi_remaining(prim_obj)
after_find = box_1.yi.shape[0] +\
box_2.yi.shape[0] +\
prim_obj.yi_remaining.shape[0]
self.assertEqual(after_find, prim_obj.y.shape[0])
def test_boxes(self):
x = pd.DataFrame([(0,1,2),
(2,5,6),
(3,2,1)],
columns=['a', 'b', 'c'])
y = {'y':np.array([0,1,2])}
results = (x,y)
prim_obj = prim.setup_prim(results, 'y', threshold=0.8)
boxes = prim_obj.boxes
self.assertEqual(len(boxes), 1, 'box length not correct')
# real data test case
prim_obj = prim.setup_prim(utilities.load_flu_data(), flu_classify,
threshold=0.8)
prim_obj.find_box()
boxes = prim_obj.boxes
self.assertEqual(len(boxes), 1, 'box length not correct')
def test_plot__individual_cdf(self):
x, outcomes = utilities.load_flu_data()
y = outcomes['deceased population region 1'][:,-1] > 1000000
fig, ax = plt.subplots()
unc = 'fatality ratio region 1'
regional_sa.plot_individual_cdf(ax, unc, x[unc], y,
discrete=False, legend=True,
xticklabels_on=True,
yticklabels_on=True)
fig, ax = plt.subplots()
unc = 'model'
regional_sa.plot_individual_cdf(ax, unc, x[unc], y,
discrete=True, legend=True,
xticklabels_on=True,
yticklabels_on=True)
def test_get_ex_feature_scores(self):
x, outcomes = utilities.load_flu_data()
y = outcomes['deceased population region 1'][:, -1] > 1000000
scores, forest = fs.get_ex_feature_scores(x,y, mode=CLASSIFICATION,
random_state=10)
self.assertEqual(len(scores), len(x.columns))
self.assertTrue(isinstance(forest, ExtraTreesClassifier))
self.assertRaises(ValueError, fs.get_ex_feature_scores, x,y,
mode='illegal argument')
y = outcomes['deceased population region 1'][:,-1]
scores, forest = fs.get_ex_feature_scores(x,y, mode=REGRESSION,
random_state=10)
self.assertEqual(len(scores), len(x.columns))
self.assertTrue(isinstance(forest, ExtraTreesRegressor))
def test_get_ex_feature_scores(self):
x, outcomes = utilities.load_flu_data()
y = outcomes['deceased population region 1'][:, -1] > 1000000
scores, forest = fs.get_ex_feature_scores(x,y,
mode=RuleInductionType.CLASSIFICATION,
random_state=10)
self.assertEqual(len(scores), len(x.columns)-2)
self.assertTrue(isinstance(forest, ExtraTreesClassifier))
self.assertRaises(ValueError, fs.get_ex_feature_scores, x,y,
mode='illegal argument')
y = outcomes['deceased population region 1'][:,-1]
scores, forest = fs.get_ex_feature_scores(x,y,
mode=RuleInductionType.REGRESSION,
random_state=10)
self.assertEqual(len(scores), len(x.columns)-2)
self.assertTrue(isinstance(forest, ExtraTreesRegressor))
def test_get_lasso_feature_scores(self):
x, outcomes = utilities.load_flu_data()
def classify(data):
#get the output for deceased population
result = data['deceased population region 1']
#make an empty array of length equal to number of cases
classes = np.zeros(result.shape[0])
#if deceased population is higher then 1.000.000 people, classify as 1
classes[result[:, -1] > 1000000] = 1
return classes
y = classify(outcomes)
# classification based
scores = fs.get_lasso_feature_scores(x,
y,
mode=CLASSIFICATION,
random_state=42)
self.assertEqual(len(scores), len(x.dtype.fields))
self.assertRaises(ValueError,
fs.get_lasso_feature_scores,
x,
y,
mode='illegal argument')
#regression based
y = outcomes['deceased population region 1'][:, -1]
scores = fs.get_lasso_feature_scores(x,
y,
mode=REGRESSION,
random_state=42)
self.assertEqual(len(scores), len(x.dtype.fields))