print(len(data))
# print(vals)
# print(len(data[999]))
data = check_array(data, accept_sparse=True)
# print("d1: ",data[0])
svd.fit(data)
d = svd.fit_transform(data)
# print("d1: ",data[0])
#DummyClassifier(strategy='most_frequent').fit(d,[styledict[s] for s in style])
# classifier1 = MLPClassifier(random_state=1,max_iter=1000).fit(d,[styledict[s] for s in style])
classifier = GradientBoostingClassifier(
n_estimators=100,
init=MLPClassifier(max_iter=1000)).fit(d,
[styledict[s] for s in style])
if type(classifier) == type(GaussianNB()):
classifier.partial_fit(d, [styledict[s] for s in style], classes=vals)
for i in range(3):
data, style = grab_rand_data(cur, length)
data = svd.fit_transform(data)
classifier.partial_fit(data, [styledict[s] for s in style])
elif type(classifier) == type(MLPClassifier()):
pass
test_size = 50000
# print(cur.fetchone())
data, style = grab_rand_data(cur, test_size)
print(
classifier.score(svd.fit_transform(data),
[styledict[s] for s in style]))
# print(classifier1.score(svd.fit_transform(data),[styledict[s] for s in style]))
# print(test_clf(classifier,classifier1,svd.fit_transform(data),[styledict[s] for s in style]))
