def __init__(self):
super(Window, self).__init__()
animes = readDataframe.Animes()
self.itemBased = WindowFiltering(
CollaborativeFiltering(animes.animes, animes.rating))
self.contentBased = WindowFiltering(ContentFiltering(animes.animes))
self.addTab(self.itemBased, "Pessoas tambem assistiram")
self.addTab(self.contentBased, "Porque voce assistiu")
def getUserMovies(self, method, userId, noMovies):
if method == "Collaborative":
recommender = CollaborativeFiltering(self._ratingsFilename, self._trainSetPath)
# recommender.train()
return recommender.getImpersonatedUserMovies(userId, noMovies)
elif method == "Content":
recommender = ContentFiltering(self._ratingsFilename, self._tagsFilename, self._trainSetPath)
# recommender.train()
return recommender.getImpersonatedUserMovies(userId, noMovies)
elif method == "Hybrid":
collaborativeTrainer = CollaborativeFiltering(self._ratingsFilename, self._trainSetPath)
# collaborativeTrainer.train()
collaborativeMovies = collaborativeTrainer.getImpersonatedUserMovies(userId, noMovies*2)
contentTrainer = ContentFiltering(self._ratingsFilename, self._tagsFilename, self._trainSetPath)
# contentTrainer.train()
contentMovies = contentTrainer.getImpersonatedUserMovies(userId, noMovies*2)
return self._hybridise(noMovies, collaborativeMovies, contentMovies)
else:
pass
import pandas as pd
a = pd.read_csv("movieLens.csv")
from CollaborativeFiltering import CollaborativeFiltering
b= CollaborativeFiltering(formatizer={'user': '******', 'item': 'movieName', 'value': 'rating'})
from sklearn.cross_validation import train_test_split
from sklearn.metrics import mean_squared_error
x,y = train_test_split(a, test_size= 0.1)
x.reset_index(inplace=True)
y.reset_index(inplace=True)
b.fit(data=x)
m= y['rating'].tolist()
n= b.predict(X=y)
print mean_squared_error(m, n)
out = []
true= []
for i in range(len(m)):
print m[i], n[i]
if n[i]!=2.5:
out.append(n[i])
true.append(m[i])
print mean_squared_error(true, out)
"""
bb=transformPrefs(ratings)
baseline_recommend = BaselineRecommendations('arbitrary')
baseline_recommend.run_baseline()
#***************Item-Item Collaborative Filtering Recommendation***************
collab_filt = raw_input("Do you want to run the collaborative filtering recommendation? (yes or no) ")
while collab_filt != 'yes' and collab_filt != 'no':
collab_filt = raw_input("Do you want to run the collaborative filtering recommendation? (yes or no) ")
if collab_filt == 'yes':
sample_type = raw_input("Do you want to run the randomly sampled data, arbitrarily sampled data or both? (r, a, b) ")
while sample_type != 'r' and sample_type != 'a' and sample_type != 'b':
sample_type = raw_input("Do you want to run the randomly sampled data, arbitrarily sampled data or both? (r, a, b) ")
print "Note: this is a fairly time consuming process--please allow for approximately 2+ hours to calculate RMSE on a single dataset."
if sample_type == 'r':
print "Calculating RMSE for random dataset split."
collab_recommend = CollaborativeFiltering()
collab_recommend.collaborative_filter()
elif sample_type == 'a':
print "Calculating RMSE for arbitrary dataset split."
collab_recommend = CollaborativeFiltering(dataset='arbitrary')
collab_recommend.collaborative_filter()
elif sample_type == 'b':
print "Calculating RMSE for random dataset split."
collab_recommend = CollaborativeFiltering()
collab_recommend.collaborative_filter()
print "Calculating RMSE for arbitrary dataset split."
collab_recommend = CollaborativeFiltering(dataset='arbitrary')
collab_recommend.collaborative_filter()
#***************Latent Factor Model Recommendation***************
from CollaborativeFiltering import CollaborativeFiltering
file_path = "21B_tag_views_dataset.csv"
cf = CollaborativeFiltering()
cf.load_data(file_path)
print("data loaded")
current_tag_id = "ff0d3fb21c00bc33f71187a2beec389e9eff5332"
similar_items = cf.similar_items(current_tag_id)
print("Users that viewed \"" + cf.products[current_tag_id] + "\" also viewed:")
print(similar_items)
import pandas as pd
from CollaborativeFiltering import CollaborativeFiltering
if __name__ == '__main__':
ratings = pd.read_csv('./ml-1m/ratings.dat',
engine='python',
sep='::',
names=['user_id', 'movie_id', 'rating', 'timestamp'])
collaborative_filtering = CollaborativeFiltering(ratings)
user_id = 1
similarity_user_count = 5
recommend_user_count = 5
result = collaborative_filtering.recommend_item(user_id,
similarity_user_count,
recommend_user_count)
print(result)