class SklearnClassifier(BaseClassifier):
def load_dataset(self, dataset_path=None):
self.dspath = dataset_path if dataset_path is not None else "invalid path!"
try:
self.X = load_sparse_matrix(os.path.join(dataset_path, "X.npz"))
self.Y = np.load(os.path.join(dataset_path, "Y.npy"))
except IOError:
print "Seems model files (X.npz, Y.npy) are not found..."
def trainSGD(self):
sgd = SGDClassifier(
loss=self.loss,
penalty=self.reg,
alpha=self.alpha,
n_iter=self.epochs,
shuffle=True,
n_jobs=self.multicpu,
class_weight="auto",
)
# print "Classifier (sklearn SGD): training the model \t(%s)"%self.dspath
if self.kernel_approx is True:
rbf_feature = RBFSampler(gamma=1, n_components=100.0, random_state=1)
Xk = rbf_feature.fit_transform(self.X)
self.glm = OneVsRestClassifier(sgd).fit(Xk, self.Y)
else:
self.glm = OneVsRestClassifier(sgd).fit(self.X, self.Y)
print "Classifier (sklearn SGD): Done. \t(%s)" % self.dspath
def trainSVM(self):
# svm = NuSVC(nu=0.5, kernel='rbf', degree=3, gamma=0.0, coef0=0.0, shrinking=True, probability=True, tol=0.001, cache_size=500, verbose=False, max_iter=-1)
svm = SVC(
C=1.0,
kernel="rbf",
degree=3,
gamma=0.0,
coef0=0.0,
shrinking=True,
probability=True,
tol=0.001,
cache_size=200,
class_weight=None,
verbose=True,
max_iter=-1,
)
print "Classifier (sklearn NuSVC): training the model \t(%s)" % self.dspath
self.glm = OneVsRestClassifier(svm).fit(self.X, self.Y)
print "Classifier (sklearn NuSVC): Done. \t(%s)" % self.dspath
def predict_f(self, testset_path=None, X=None, Y=None):
fn_x = os.path.join(testset_path, "X.npz")
fn_y = os.path.join(testset_path, "Y.npy")
Xtest = load_sparse_matrix(fn_x)
Ytest = np.load(fn_y)
pred = self.glm.predict(Xtest)
return pred
def predict_f_prob(self, testset_path=None, X=None, Y=None):
"""
Parameters
----------
testset_path: path prefix for testset .npz and .npy file
X: {array-like, sparse matrix}, shape = [n_samples, n_features]
Data.
Y : numpy array of shape [n_samples]
Multi-class targets.
Returns
-------
pred_p: {array-like}, shape = {n_samples, n_classes}
"""
if hasattr(self.glm, "predict_prob"):
fn_x = os.path.join(testset_path, "X.npz")
fn_y = os.path.join(testset_path, "Y.npy")
Xtest = load_sparse_matrix(fn_x)
Ytest = np.load(fn_y)
pred_p = self.glm.predict_prob(Xtest)
return pred_p
else:
raise NotImplementedError
