class NMFImpl():
def __init__(self, n_components=None, init=None, solver='cd', beta_loss='frobenius', tol=0.0001, max_iter=200, random_state=None, alpha=0.0, l1_ratio=0.0, verbose=0, shuffle=False):
self._hyperparams = {
'n_components': n_components,
'init': init,
'solver': solver,
'beta_loss': beta_loss,
'tol': tol,
'max_iter': max_iter,
'random_state': random_state,
'alpha': alpha,
'l1_ratio': l1_ratio,
'verbose': verbose,
'shuffle': shuffle}
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
def transform(self, X):
return self._sklearn_model.transform(X)
class NMFRecommender(BaseEstimator, RegressorMixin):
def __init__(self, n_components):
self.nmf = NMF(n_components=2, init='random', random_state=0)
self.user_ids_dict = {}
self.book_isbns_dict = {}
def fit(self, X, y=None):
self.sparse_matrix = X['sparse_matrix']
self.user_ids_dict = X['user_ids_dict']
self.book_isbns_dict = X['book_isbns_dict']
self.nmf.fit(X['sparse_matrix'])
def predict(self, X, y=None):
ratings = X['ratings']
user_representations = self.nmf.transform(self.sparse_matrix)
book_representations = self.nmf.components_
estimations = []
for i in tqdm(range(len(ratings))):
estimation = np.dot(
user_representations[self.user_ids_dict[
ratings.iloc[i]['User-ID']]], book_representations)[
self.book_isbns_dict[ratings.iloc[i]['ISBN']]]
estimations.append(estimation)
return estimations
def fit_predict(self, X, y=None):
self.fit(X, y)
return self.predict(X, y)