def test_kfold(self):
cv = cross_validation.RandomStratifiedKFold(valid_size=1 / 50)
self.assertRaises(ValueError, cv.get_n_splits, X, y)
for split in [1, 2, 5]:
cv = cross_validation.RandomStratifiedKFold(n_splits=1 + split,
n_repeats=split,
verbose=split)
assert (cv.get_n_splits(X, y) == split * split + split)
assert (cv.verbose == split)
for idx, [tr, vl] in enumerate(cv.split(X, y)):
assert (int(tr.size / vl.size) == split)
assert (np.unique(y[vl], return_counts=True)[0].size == 5)
assert (idx + 1 == split * split + split)
def test_compare_loo_kf(self):
cv_loo = cross_validation.LeaveOneOut(random_state=12, verbose=2)
cv_kf_as_loo = cross_validation.RandomStratifiedKFold(n_splits=False,
valid_size=1,
random_state=12,
verbose=2)
for trvl_loo, trvl_kf in zip(cv_loo.split(X, y),
cv_kf_as_loo.split(X, y)):
assert (np.all(trvl_loo[0] == trvl_kf[0]))
assert (np.all(trvl_loo[1] == trvl_kf[1]))
assert (len(trvl_kf[1]) == n_class)
assert (np.unique(y[trvl_kf[1]]).size == n_class)
#to print extensions
cv_loo.get_supported_extensions()
raster, vector = datasets.load_historical_data(low_res=True)
field = 'Class'
group = 'uniquefid'
X, y, g = extract_ROI(raster, vector, field, group)
##############################################################################
# Initialize Random-Forest
# ---------------------------
classifier = RandomForestClassifier(random_state=12, n_jobs=1)
##############################################################################
# Create list of different CV
# ---------------------------
CVs = [
cross_validation.RandomStratifiedKFold(n_splits=2),
cross_validation.LeavePSubGroupOut(valid_size=0.5),
cross_validation.LeaveOneSubGroupOut(),
StratifiedKFold(n_splits=2, shuffle=True) #from sklearn
]
kappas = []
for cv in CVs:
SL = SuperLearner(classifier=classifier,
param_grid=dict(n_estimators=[50, 100]),
n_jobs=1)
SL.fit(X, y, group=g, cv=cv)
print('Kappa for ' + str(type(cv).__name__))
cvKappa = []