def fit(self, X, y=None):
self._sklearn_model = SKLModel(**self._hyperparams)
if (y is not None):
self._sklearn_model.fit(X, y)
else:
self._sklearn_model.fit(X)
return self
class LabelPropagationImpl():
def __init__(self,
kernel='rbf',
gamma=20,
n_neighbors=7,
alpha=None,
max_iter=1000,
tol=0.001,
n_jobs=None):
self._hyperparams = {
'kernel': kernel,
'gamma': gamma,
'n_neighbors': n_neighbors,
'alpha': alpha,
'max_iter': max_iter,
'tol': tol,
'n_jobs': n_jobs
}
self._wrapped_model = Op(**self._hyperparams)
def fit(self, X, y=None):
if (y is not None):
self._wrapped_model.fit(X, y)
else:
self._wrapped_model.fit(X)
return self
def predict(self, X):
return self._wrapped_model.predict(X)
def predict_proba(self, X):
return self._wrapped_model.predict_proba(X)
def __init__(self,
kernel='rbf',
gamma=20,
n_neighbors=7,
alpha=None,
max_iter=1000,
tol=0.001,
n_jobs=None):
self._hyperparams = {
'kernel': kernel,
'gamma': gamma,
'n_neighbors': n_neighbors,
'alpha': alpha,
'max_iter': max_iter,
'tol': tol,
'n_jobs': n_jobs
}
self._wrapped_model = Op(**self._hyperparams)
warnings.filterwarnings('ignore')
classifiers = [
AdaBoostClassifier(),
BaggingClassifier(),
BernoulliNB(),
CalibratedClassifierCV(),
DecisionTreeClassifier(),
ExtraTreeClassifier(),
ExtraTreesClassifier(),
GaussianNB(),
GaussianProcessClassifier(),
GradientBoostingClassifier(),
KNeighborsClassifier(),
LabelPropagation(),
LabelSpreading(),
LinearDiscriminantAnalysis(),
LogisticRegression(),
LogisticRegressionCV(),
MLPClassifier(),
NuSVC(probability=True),
QuadraticDiscriminantAnalysis(),
RandomForestClassifier(),
SGDClassifier(loss='log'),
SVC(probability=True),
XGBClassifier()
]
names = [
'AdaBoostClassifier', 'BaggingClassifier', 'BernoulliNB',
'GraphLasso':GraphLasso(), 'GraphLassoCV':GraphLassoCV(), 'HuberRegressor':HuberRegressor(), 'Imputer':Imputer(), 'IncrementalPCA':IncrementalPCA(), 'IsolationForest':IsolationForest(), 'Isomap':Isomap(), 'KMeans':KMeans(), 'KNeighborsClassifier':KNeighborsClassifier(), 'KNeighborsRegressor':KNeighborsRegressor(), 'KernelCenterer':KernelCenterer(), 'KernelDensity':KernelDensity(), 'KernelPCA':KernelPCA(), 'KernelRidge':KernelRidge(), 'LSHForest':LSHForest(), 'LabelPropagation':LabelPropagation(), 'LabelSpreading':LabelSpreading(), 'Lars':Lars(), 'LarsCV':LarsCV(), 'Lasso':Lasso(), 'LassoCV':LassoCV(), 'LassoLars':LassoLars(), 'LassoLarsCV':LassoLarsCV(), 'LassoLarsIC':LassoLarsIC(), 'LatentDirichletAllocation':LatentDirichletAllocation(), 'LedoitWolf':LedoitWolf(), 'LinearDiscriminantAnalysis':LinearDiscriminantAnalysis(), 'LinearRegression':LinearRegression(), 'LinearSVC':LinearSVC(), 'LinearSVR':LinearSVR(), 'LocallyLinearEmbedding':LocallyLinearEmbedding(),
def compare(X, Y, n, index, kernel='rbf', tol=0.01, max_iter=100, gamma=20):
"""
Compare Propagation between our method and random method
Parameters
----------
X : 2D array
The dataset
Y : 1d array
True label of every points in X
n : int
number of point to label first
index :1d array of int
array of index of the point to label
kernel : str
rbf
tol : float
tol for labelpropagation
max_iter : int
max_iter for labelpropagation
gamma : int
gamma for labelpropagation
"""
plt.figure(figsize=(17, 8))
# start with randomness
ok = False
c = 0
# make sure we have a point of each class (at least)
while not ok:
label = np.random.randint(Y.shape[0], size=n)
c += 1
if len(np.unique(Y[label])) == len(np.unique(Y)):
ok = True
Y_unlabel = -np.ones(Y.shape[0])
Y_unlabel[label] = Y[label]
label_spread = LabelPropagation(kernel=kernel,
tol=tol,
max_iter=max_iter,
n_jobs=-1,
gamma=gamma)
label_spread.fit(X, Y_unlabel)
Y_predict = label_spread.transduction_
iter_random = label_spread.n_iter_
plt.subplot(1, 2, 1)
oui = np.where(Y_predict == 0)[0]
non = np.where(Y_predict == 1)[0]
plt.scatter(X[oui, 0], X[oui, 1], color='b', marker='s', lw=0, s=10)
plt.scatter(X[non, 0], X[non, 1], color='g', marker='s', lw=0, s=10)
plot_interesting_points(X, label, 1, 0)
# now active part
label = index[:n]
if len(np.unique(Y[label])) < len(np.unique(Y)):
raise Exception("we have less points than classes")
Y_unlabel = -np.ones(Y.shape[0])
Y_unlabel[label] = Y[label]
label_spread = LabelPropagation(kernel=kernel,
tol=tol,
max_iter=max_iter,
n_jobs=-1,
gamma=gamma)
label_spread.fit(X, Y_unlabel)
Y_predict = label_spread.transduction_
iter_active = label_spread.n_iter_
# plotting
plt.subplot(1, 2, 2)
oui = np.where(Y_predict == 0)[0]
non = np.where(Y_predict == 1)[0]
plt.scatter(X[oui, 0], X[oui, 1], color='b', marker='s', lw=0, s=10)
plt.scatter(X[non, 0], X[non, 1], color='g', marker='s', lw=0, s=10)
plot_interesting_points(X, label, 0.5, 0)
return (iter_random, iter_active)