Now sets classes when fitting committee

[OskarLiew]

This pull requests aims to solve the second issue mentioned in #97. When creating a committee without any training data, calling Committee.fit(X, y) would cause an error because Committee.classes_ was None.

The following code would cause an error in vote_entropy and other disagreement functions:

import numpy as np
from sklearn.svm import SVC
from modAL.uncertainty import uncertainty_sampling
from modAL.disagreement import vote_entropy_sampling
from modAL.models import ActiveLearner, Committee

# Create data
X = np.array([[0, 0],
              [0, 1],
              [1, 0],
              [1, 1]])

y = np.array([1, 1, 0 ,0])

X_teach = X[:3]
y_teach = y[:3]

X_query = X[3].reshape((-1, 2))
y_query = y[3]

# ActiveLearner
svm = SVC(probability=True, gamma='auto')
learner = ActiveLearner(svm, uncertainty_sampling)

learner.fit(X_teach, y_teach)

query_idx, query_inst = learner.query(X_query)

print('Index %d, Coordinates %s' % (query_idx, query_inst))

# Committee
estimators = [SVC(probability=True, gamma='auto') for _ in range(20)]
learner_list = [ActiveLearner(estimator, uncertainty_sampling) for estimator in estimators]
committee = Committee(learner_list, vote_entropy_sampling)

committee.fit(X_teach, y_teach)

query_idx, query_inst = committee.query(X_query)

print('Index %d, Coordinates %s' % (query_idx, query_inst))

p_vote = np.zeros(shape=(X.shape[0], len(committee.classes_)))
TypeError: object of type 'NoneType' has no len()

With the fix, it is able to run smoothly.

[cosmic-cortex]

Thanks for the fix! Sorry to be almost two weeks late, I am a little behind on modAL maintainance :)