class Bot(object):
def __init__(self):
self.recommendation = Recommendation()
def respond_to(self, sender, message):
# Enregistre l'utilisateur s'il n'existe pas déjà
user = self.recommendation.register_user(sender)
# Donne le message pour que l'utilisateur l'utilise
user.give_message(message)
# Si le chatbot doit faire une recommandation ou pas
self.recommendation.make_recommendation(user)
if user.should_make_recommendation():
return self.recommendation.make_recommendation(user)
else:
intro = ""
print("bon,bad, neutre")
print(user.good_ratings)
print(user.bad_ratings)
print(user.neutral_ratings)
# Si l'utilisateur parle pour la première fois, affiche un message d'intro
if not user.has_been_asked_a_question():
intro = "Bonjour ! Je vais vous poser des questions puis vous faire une recommandation.\n"
else:
if (message == "oui"):
user.answer_yes()
elif (message == "non"):
user.answer_no()
else:
user.answer_neutral()
message = self.recommendation.ask_question(user)
return intro + message
def home():
if "recommend" in request.args:
try:
title = request.args["recommend"]
rec = ContentBased()
did_you_mean = False
df = rec.recommend(title,
DEFAULT_LIMIT,
full_search=True,
keywords_and_desc=False,
critics=False)
poster_paths = get_poster_paths(df["id"].tolist(),
df["original_title"].tolist())
if rec.changed_title != title and rec.changed_title != str():
did_you_mean = True
else:
rec.changed_title = title
rec_title_meta = get_meta(rec.changed_title, None)
rec_id = rec_title_meta[0]["id"]
return render_template('recommendations.html',
titles=df["original_title"].tolist(),
images=poster_paths,
votes=df["vote_average"].tolist(),
m_id=df["id"].tolist(),
rec_title=rec.changed_title,
rec_id=rec_id,
did_you_mean=did_you_mean)
except:
abort(404)
elif "genres" in request.args:
genre = request.args["genres"]
if genre == "All":
genre = None
offset = int(request.args["offset"])
gen_rec = Recommendation()
gen_rec.filter_genres()
df = gen_rec.top_movies(gen_rec.md,
percentile=0.85,
limit=DEFAULT_LIMIT,
offset=offset,
genre=genre)
poster_paths = get_poster_paths(df["id"].tolist(),
df["original_title"].tolist())
return render_template('recommendations.html',
titles=df["original_title"].tolist(),
images=poster_paths,
votes=df["vote_average"].tolist(),
m_id=df["id"].tolist(),
rec_title=request.args["genres"],
offset=offset,
next_offset=offset + DEFAULT_LIMIT,
prev_offset=offset - DEFAULT_LIMIT,
rec_id=None,
did_you_mean=None)
else:
return render_template('homepage.html')
def menu():
recommendation = Recommendation()
name_movie = None
while name_movie != "":
print("#################################")
print("RECOMMENDATION CONTENT BASED IMDB")
name_movie = input("Enter name movie: ")
if name_movie != "":
recommendation.recommendation_movies(name_movie)
def compute_recommendation(user):
"""
Compute the recommendation for a user. Return the recommended events
"""
recommendation = Recommendation()
recommendation.init_basic_matrix()
recommendation.init_frequency_matrix()
recommendation.compute_matrix()
return get_features_event(recommendation.compute_recommended_events(user))
def main():
if (sys.argv[-1] == '--collect'):
url_pois = 'https://pt.foursquare.com/explore?mode=url&ne=-29.358988%2C-50.837817&q=Sele%C3%A7%C3%B5es%20principais&sw=-29.41889%2C-50.887942'
url_city = 'http://www.dataviva.info/pt/location/5rs020102'
e = Extraction(url_pois, url_city)
e.poi_data_extraction()
e.city_data_extraction()
# Gera relatório do dataset
file = 'foursquare_data.csv'
df = pd.read_csv(file, parse_dates=True, encoding='UTF-8')
profile = pandas_profiling.ProfileReport(df)
profile.to_file(outputfile='dataset_report.html')
P = Process(file)
df = P.method()
df_report = pd.read_csv('preprocessed.csv',
parse_dates=True,
encoding='UTF-8')
profile = pandas_profiling.ProfileReport(df_report)
profile.to_file(outputfile='preprocessed_dataset_report.html')
R = Recommendation(df)
R.pattern_recommendation()
R.new_recommendation()
R.compare()
# R.test_rec()
ont = Ontology()
ont.write_owl()
def run(self, category=None, problems_solved=[]):
solutions_df = self.load_solutions(category)
thetas = Recommendation.get_thetas(solutions_df,
nr_features=self.NR_FEATURES,
nr_epochs=self.NR_EPOCHS,
nr_iterations=self.NR_ITERATIONS)
# Thetas already calculated for math category(5)
# thetas = pd.read_csv('data_files/thetas.csv')
# thetas = thetas.to_dict('list')
next_problem = Recommendation.get_next_problem(
problems_solved,
thetas,
nr_features=self.NR_FEATURES,
aggressivity_radius=self.AGGRESSIVITY_RADIUS)
return next_problem
def setUp(self):
clean()
c = Category(name="C", external_id="1")
w = Weight(name="t", weight=1.0)
w.save()
c.save()
for i in range(0, 3):
User(name="U"+str(i), external_id=str(i), token="t"+str(i)).save()
users = User.objects.all()
for i in range(0, 5):
Feature(name="F"+str(i)).save()
features = Feature.objects.all()
for i in range(0, 5):
Event(category=c, external_id=str(i), name='E'+str(i), website='', description='').save()
events = Event.objects.all()
Rating(event=events[0], user=users[0], rating=1).save()
Rating(event=events[1], user=users[0], rating=2).save()
Rating(event=events[2], user=users[0], rating=4).save()
Rating(event=events[3], user=users[0], rating=5).save()
Rating(event=events[4], user=users[0], rating=3).save()
Rating(event=events[0], user=users[1], rating=3).save()
Rating(event=events[1], user=users[1], rating=2).save()
Rating(event=events[2], user=users[1], rating=5).save()
Rating(event=events[0], user=users[2], rating=1).save()
Rating(event=events[1], user=users[2], rating=4).save()
Rating(event=events[4], user=users[2], rating=3).save()
EventFeature(event=events[0], feature=features[0], tf_idf=0.2, weight=w).save()
EventFeature(event=events[0], feature=features[2], tf_idf=0.8, weight=w).save()
EventFeature(event=events[1], feature=features[1], tf_idf=0.8, weight=w).save()
EventFeature(event=events[1], feature=features[2], tf_idf=1.0, weight=w).save()
EventFeature(event=events[1], feature=features[4], tf_idf=0.3, weight=w).save()
EventFeature(event=events[2], feature=features[0], tf_idf=1.0, weight=w).save()
EventFeature(event=events[2], feature=features[2], tf_idf=0.3, weight=w).save()
EventFeature(event=events[3], feature=features[1], tf_idf=0.2, weight=w).save()
EventFeature(event=events[3], feature=features[3], tf_idf=0.4, weight=w).save()
EventFeature(event=events[4], feature=features[3], tf_idf=0.8, weight=w).save()
FeatureUser(user=users[0], feature=features[1], weight=0.4).save()
FeatureUser(user=users[0], feature=features[3], weight=0.8).save()
FeatureUser(user=users[0], feature=features[4], weight=0.3).save()
FeatureUser(user=users[1], feature=features[0], weight=1.0).save()
FeatureUser(user=users[1], feature=features[2], weight=0.5).save()
FeatureUser(user=users[2], feature=features[1], weight=0.2).save()
FeatureUser(user=users[2], feature=features[4], weight=0.6).save()
from recommendation import Recommendation
self.recommendation = Recommendation()
def get_meta(title, m_id):
rec = Recommendation()
rec.filter_genres()
rec.filter_productions()
df_movies = rec.md
df_credits = pd.read_csv(PATH_CREDITS)
df_imdb_link = pd.read_csv(PATH_MOVIELENS_TO_TMDB)
attributes = [
"id", "original_title", "genres", "homepage", "overview",
"release_date", "production_companies", "runtime", "tagline",
"vote_average", "vote_count"
]
df_title = df_movies.iloc[df_movies.index[df_movies["original_title"] ==
title][0]][attributes]
df_crew = df_credits.iloc[df_credits.index[df_credits["title"] == title]
[0]][["cast", "crew"]]
cast = [cast["name"] for cast in literal_eval(df_crew["cast"])[0:5]]
crew = [
crew["name"] for crew in literal_eval(df_crew["crew"])
if crew["job"] in ["Director"]
]
try:
imdb_link = str(df_imdb_link.iloc[df_imdb_link.index[
df_imdb_link["tmdbId"] == int(m_id)][0]]["imdbId"])[:-2]
imdb_link = ("https://www.imdb.com/title/tt" + "0" *
(IMDB_ID_LEN - len(imdb_link)) + imdb_link)
except:
imdb_link = "https://www.imdb.com/search/title/?title=" + title
return df_title, cast, crew, imdb_link
class Bot(object):
def __init__(self):
self.recommendation = Recommendation()
def respond_to(self, sender, message):
# Enregistre l'utilisateur s'il n'existe pas déjà
user = self.recommendation.register_user(sender)
# Donne le message pour que l'utilisateur l'utilise
user.give_message(message)
# Si le chatbot doit faire une recommandation ou pas
if user.should_make_recommendation():
return self.recommendation.make_recommendation(user)
else:
intro = ""
# Si l'utilisateur parle pour la première fois, affiche un message d'intro
if not user.has_been_asked_a_question():
intro = "Bonjour ! Je vais vous poser des questions puis vous faire une recommandation.\n"
message = self.recommendation.ask_question(user)
return intro + message
def compare_list(l1, l2):
if len(l1) != len(l2):
return False
from recommendation import Recommendation
try:
isSameSize = True
for ll1 in l1:
for ll2 in l2:
isSameSize &= len(ll1) == len(ll2)
if not isSameSize:
return False
for i in range(0, len(l1)):
for j in range(0, len(l1[0])):
if not Recommendation.compare_float(l1[i][j], l2[i][j]):
print '(', i, ';', j, ') ', l1[i][j], ' not eq ', l2[i][j]
return False
except:
for i in range(0, len(l1)):
if not Recommendation.compare_float(l1[i], l2[i]):
print '(', i, ') ', l1[i], ' not eq ', l2[i]
return False
return True
def compare_list(l1, l2):
if len(l1) != len(l2):
return False
from recommendation import Recommendation
try:
isSameSize = True
for ll1 in l1:
for ll2 in l2:
isSameSize &= len(ll1) == len(ll2)
if not isSameSize:
return False
for i in range(0, len(l1)):
for j in range(0, len(l1[0])):
if not Recommendation.compare_float(l1[i][j], l2[i][j]):
print '(', i, ';', j, ') ', l1[i][j], ' not eq ', l2[i][j]
return False
except:
for i in range(0, len(l1)):
if not Recommendation.compare_float(l1[i], l2[i]):
print '(', i, ') ', l1[i], ' not eq ', l2[i]
return False
return True
def recommend(data):
recs = Recommendation(data, 5, 5)
return (recs.allRecommendations())
class RecommendationTest1:
def setUp(self):
clean()
c = Category(name="C", external_id="1")
w = Weight(name="t", weight=1.0)
w.save()
c.save()
for i in range(0, 3):
User(name="U"+str(i), external_id=str(i), token="t"+str(i)).save()
users = User.objects.all()
for i in range(0, 5):
Feature(name="F"+str(i)).save()
features = Feature.objects.all()
for i in range(0, 5):
Event(category=c, external_id=str(i), name='E'+str(i), website='', description='').save()
events = Event.objects.all()
Rating(event=events[0], user=users[0], rating=1).save()
Rating(event=events[1], user=users[0], rating=2).save()
Rating(event=events[2], user=users[0], rating=4).save()
Rating(event=events[3], user=users[0], rating=5).save()
Rating(event=events[4], user=users[0], rating=3).save()
Rating(event=events[0], user=users[1], rating=3).save()
Rating(event=events[1], user=users[1], rating=2).save()
Rating(event=events[2], user=users[1], rating=5).save()
Rating(event=events[0], user=users[2], rating=1).save()
Rating(event=events[1], user=users[2], rating=4).save()
Rating(event=events[4], user=users[2], rating=3).save()
EventFeature(event=events[0], feature=features[0], tf_idf=0.2, weight=w).save()
EventFeature(event=events[0], feature=features[2], tf_idf=0.8, weight=w).save()
EventFeature(event=events[1], feature=features[1], tf_idf=0.8, weight=w).save()
EventFeature(event=events[1], feature=features[2], tf_idf=1.0, weight=w).save()
EventFeature(event=events[1], feature=features[4], tf_idf=0.3, weight=w).save()
EventFeature(event=events[2], feature=features[0], tf_idf=1.0, weight=w).save()
EventFeature(event=events[2], feature=features[2], tf_idf=0.3, weight=w).save()
EventFeature(event=events[3], feature=features[1], tf_idf=0.2, weight=w).save()
EventFeature(event=events[3], feature=features[3], tf_idf=0.4, weight=w).save()
EventFeature(event=events[4], feature=features[3], tf_idf=0.8, weight=w).save()
FeatureUser(user=users[0], feature=features[1], weight=0.4).save()
FeatureUser(user=users[0], feature=features[3], weight=0.8).save()
FeatureUser(user=users[0], feature=features[4], weight=0.3).save()
FeatureUser(user=users[1], feature=features[0], weight=1.0).save()
FeatureUser(user=users[1], feature=features[2], weight=0.5).save()
FeatureUser(user=users[2], feature=features[1], weight=0.2).save()
FeatureUser(user=users[2], feature=features[4], weight=0.6).save()
from recommendation import Recommendation
self.recommendation = Recommendation()
def test_RFSP(self):
RR = [[1, 2, 4, 5, 3],[3, 2, 5, -1, -1],[1, 4, -1, -1, 3]]
FF = [[0.2, -1, 0.8, -1, -1],[-1, 0.8, 1, -1, 0.3],[1, -1, 0.3, -1, -1],[-1, 0.2, -1, 0.4, -1],[-1, -1, -1, 0.8, -1]]
SS = [[0, 0.4, 0, 0.8, 0.3],[1, 0, 0.5, 0, 0],[0, 0.2, 0, 0, 0.6]]
PP = [[1, 0.2, 0.3, 1.2, 0],[1.2, 0, 1.1, 0, 0],[0, 0.8, 1, 0.8, 0.3]]
self.recommendation.init_basic_matrix()
if not compare_list(RR, transform_matrix_list(self.recommendation.get_matrix_r())):
print 'error RR test_RFSP'
print RR, '\n', transform_matrix_list(self.recommendation.get_matrix_r())
if not compare_list(FF, transform_matrix_list(self.recommendation.get_matrix_f())):
print 'error FF test_RFSP'
print FF, '\n', transform_matrix_list(self.recommendation.get_matrix_f())
if not compare_list(SS, transform_matrix_list(self.recommendation.get_matrix_s())):
print 'error SS test_RFSP'
print SS, '\n', transform_matrix_list(self.recommendation.get_matrix_s())
if not compare_list(PP, transform_matrix_list(self.recommendation.get_matrix_p())):
print 'error PP test_RFSP'
print PP, '\n', transform_matrix_list(self.recommendation.get_matrix_p())
def test_FF_UF_IUF(self):
FF = [[1, 0.6, 0.3, 2.0, 0.3],[2.2, 0, 1.6, 0, 0],[0, 1.0, 1, 0.8, 0.9]]
UF = [2, 2, 3, 2, 2]
IUF = [0.176091259056, 0.176091259056, 0.0, 0.176091259056, 0.176091259056]
self.recommendation.init_frequency_matrix()
if not compare_list(FF, transform_matrix_list(self.recommendation.get_matrix_ff())):
print 'error FF test_RFSP'
print FF, '\n', transform_matrix_list(self.recommendation.get_matrix_ff())
if not compare_list(UF, transform_matrix_list(self.recommendation.get_matrix_uf())):
print 'error UF test_RFSP'
print UF, '\n', transform_matrix_list(self.recommendation.get_matrix_uf())
if not compare_list(IUF, transform_matrix_list(self.recommendation.get_matrix_iuf())):
print 'error IUF test_RFSP'
print IUF, '\n', transform_matrix_list(self.recommendation.get_matrix_iuf())
def test_W_UU(self):
W = [[0.176091259056, 0.105654755433, -1, 0.352182518111, 0.0528273777167],[0.387400769922, -1, -1, -1, -1],[-1, 0.176091259056, -1, 0.140873007245, 0.15848213315]]
UU = [[-1, 1.0, 0.822730338093],[1.0, -1, -1],[0.822730338093, -1, -1]]
self.recommendation.compute_matrix_w()
self.recommendation.compute_matrix_uu()
if not compare_list(W, transform_matrix_list(self.recommendation.get_matrix_w())):
print 'error W test_RFSP'
print W, '\n', transform_matrix_list(self.recommendation.get_matrix_w())
if not compare_list(UU, transform_matrix_list(self.recommendation.get_matrix_uu())):
print 'error UU test_RFSP'
print UU, '\n', transform_matrix_list(self.recommendation.get_matrix_uu())
def test_final(self):
events = Event.objects.all()
res = {events[3]: 3, events[4]: 2}
res = OrderedDict(sorted(res.items(), key=lambda t: t[1], reverse=True))
if not compare_ordered_dict(self.recommendation.compute_recommended_events(User.objects.all()[1]), res):
print 'error final result'
print res
print self.recommendation.compute_recommended_events(User.objects.all()[1])
class RecommendationTest2:
def setUp(self):
clean()
c = Category(name="C", external_id="1")
w = Weight(name="t", weight=1.0)
w.save()
c.save()
for i in range(0, 4):
User(name="U"+str(i), external_id=str(i), token="t"+str(i)).save()
users = User.objects.all()
for i in range(0, 4):
Feature(name="F"+str(i)).save()
features = Feature.objects.all()
for i in range(0, 6):
Event(category=c, external_id=str(i), name='E'+str(i), website='', description='').save()
events = Event.objects.all()
Rating(event=events[1], user=users[0], rating=4).save()
Rating(event=events[4], user=users[0], rating=5).save()
Rating(event=events[1], user=users[1], rating=3).save()
Rating(event=events[3], user=users[1], rating=4).save()
Rating(event=events[5], user=users[2], rating=4).save()
Rating(event=events[0], user=users[3], rating=5).save()
Rating(event=events[2], user=users[3], rating=3).save()
EventFeature(event=events[0], feature=features[1], tf_idf=1, weight=w).save()
EventFeature(event=events[1], feature=features[0], tf_idf=1, weight=w).save()
EventFeature(event=events[1], feature=features[1], tf_idf=1, weight=w).save()
EventFeature(event=events[2], feature=features[1], tf_idf=1, weight=w).save()
EventFeature(event=events[2], feature=features[2], tf_idf=1, weight=w).save()
EventFeature(event=events[3], feature=features[1], tf_idf=1, weight=w).save()
EventFeature(event=events[4], feature=features[0], tf_idf=1, weight=w).save()
EventFeature(event=events[4], feature=features[1], tf_idf=1, weight=w).save()
EventFeature(event=events[4], feature=features[2], tf_idf=1, weight=w).save()
EventFeature(event=events[5], feature=features[3], tf_idf=1, weight=w).save()
from recommendation import Recommendation
self.recommendation = Recommendation()
def test_RFSP(self):
RR = [[-1,4,-1,-1,5,-1],[-1,3,-1,4,-1,-1],[-1,-1,-1,-1,-1,4],[5,-1,3,-1,-1,-1]]
FF = [[-1,1,-1,-1],[1,1,-1,-1],[-1,1,1,-1],[-1,1,-1,-1],[1,1,1,-1],[-1,-1,-1,1]]
SS = [[0,0,0,0],[0,0,0,0],[0,0,0,0],[0,0,0,0]]
PP = [[2,2,1,0],[1,2,0,0],[0,0,0,1],[0,2,1,0]]
self.recommendation.init_basic_matrix()
if not compare_list(RR, transform_matrix_list(self.recommendation.get_matrix_r())):
print 'error RR test_RFSP'
print RR, '\n', transform_matrix_list(self.recommendation.get_matrix_r())
if not compare_list(FF, transform_matrix_list(self.recommendation.get_matrix_f())):
print 'error FF test_RFSP'
print FF, '\n', transform_matrix_list(self.recommendation.get_matrix_f())
if not compare_list(SS, transform_matrix_list(self.recommendation.get_matrix_s())):
print 'error SS test_RFSP'
print SS, '\n', transform_matrix_list(self.recommendation.get_matrix_s())
if not compare_list(PP, transform_matrix_list(self.recommendation.get_matrix_p())):
print 'error PP test_RFSP'
print PP, '\n', transform_matrix_list(self.recommendation.get_matrix_p())
def test_FF_UF_IUF(self):
FF = [[2,2,1,0],[1,2,0,0],[0,0,0,1],[0,2,1,0]]
UF = [2,3,2,1]
IUF = [0.30102999566398114, 0.1249387366082999, 0.30102999566398114, 0.6020599913279623]
self.recommendation.init_frequency_matrix()
if not compare_list(FF, transform_matrix_list(self.recommendation.get_matrix_ff())):
print 'error FF test_RFSP'
print FF, '\n', transform_matrix_list(self.recommendation.get_matrix_ff())
if not compare_list(UF, transform_matrix_list(self.recommendation.get_matrix_uf())):
print 'error UF test_RFSP'
print UF, '\n', transform_matrix_list(self.recommendation.get_matrix_uf())
if not compare_list(IUF, transform_matrix_list(self.recommendation.get_matrix_iuf())):
print 'error IUF test_RFSP'
print IUF, '\n', transform_matrix_list(self.recommendation.get_matrix_iuf())
def test_W_UU(self):
W = [[0.6020599913279623, 0.2498774732165998, 0.30102999566398114, -1], [0.30102999566398114, 0.2498774732165998, -1, -1], [-1, -1, -1, 0.6020599913279623], [-1, 0.2498774732165998, 0.30102999566398114, -1]]
UU = [[-1, 0.95551065333456875, -1.0, 1.0], [0.95551065333456875, -1, -1.0, 1], [-1.0, -1.0, -1, -1.0], [1, 1, -1.0, -1]]
self.recommendation.compute_matrix_w()
self.recommendation.compute_matrix_uu()
if not compare_list(W, transform_matrix_list(self.recommendation.get_matrix_w())):
print 'error W test_RFSP'
print W, '\n', transform_matrix_list(self.recommendation.get_matrix_w())
if not compare_list(UU, transform_matrix_list(self.recommendation.get_matrix_uu())):
print 'error UU test_RFSP'
print UU, '\n', transform_matrix_list(self.recommendation.get_matrix_uu())
def test_final(self):
events = Event.objects.all()
res = {events[0]: 3, events[2]: 5, events[4] : 6}
res = OrderedDict(sorted(res.items(), key=lambda t: t[1], reverse=True))
if not compare_ordered_dict(self.recommendation.compute_recommended_events(User.objects.all()[1]), res):
print 'error final result'
print res
print self.recommendation.compute_recommended_events(User.objects.all()[1])
def recommend(data): recs = Recommendation(data, 5, 5) return (recs.allRecommendations())
def setUp(self):
clean()
c = Category(name="C", external_id="1")
w = Weight(name="t", weight=1.0)
w.save()
c.save()
for i in range(0, 4):
User(name="U" + str(i), external_id=str(i),
token="t" + str(i)).save()
users = User.objects.all()
for i in range(0, 4):
Feature(name="F" + str(i)).save()
features = Feature.objects.all()
for i in range(0, 6):
Event(category=c,
external_id=str(i),
name='E' + str(i),
website='',
description='').save()
events = Event.objects.all()
Rating(event=events[1], user=users[0], rating=4).save()
Rating(event=events[4], user=users[0], rating=5).save()
Rating(event=events[1], user=users[1], rating=3).save()
Rating(event=events[3], user=users[1], rating=4).save()
Rating(event=events[5], user=users[2], rating=4).save()
Rating(event=events[0], user=users[3], rating=5).save()
Rating(event=events[2], user=users[3], rating=3).save()
EventFeature(event=events[0], feature=features[1], tf_idf=1,
weight=w).save()
EventFeature(event=events[1], feature=features[0], tf_idf=1,
weight=w).save()
EventFeature(event=events[1], feature=features[1], tf_idf=1,
weight=w).save()
EventFeature(event=events[2], feature=features[1], tf_idf=1,
weight=w).save()
EventFeature(event=events[2], feature=features[2], tf_idf=1,
weight=w).save()
EventFeature(event=events[3], feature=features[1], tf_idf=1,
weight=w).save()
EventFeature(event=events[4], feature=features[0], tf_idf=1,
weight=w).save()
EventFeature(event=events[4], feature=features[1], tf_idf=1,
weight=w).save()
EventFeature(event=events[4], feature=features[2], tf_idf=1,
weight=w).save()
EventFeature(event=events[5], feature=features[3], tf_idf=1,
weight=w).save()
from recommendation import Recommendation
self.recommendation = Recommendation()
class RecommendationTest2:
def setUp(self):
clean()
c = Category(name="C", external_id="1")
w = Weight(name="t", weight=1.0)
w.save()
c.save()
for i in range(0, 4):
User(name="U" + str(i), external_id=str(i),
token="t" + str(i)).save()
users = User.objects.all()
for i in range(0, 4):
Feature(name="F" + str(i)).save()
features = Feature.objects.all()
for i in range(0, 6):
Event(category=c,
external_id=str(i),
name='E' + str(i),
website='',
description='').save()
events = Event.objects.all()
Rating(event=events[1], user=users[0], rating=4).save()
Rating(event=events[4], user=users[0], rating=5).save()
Rating(event=events[1], user=users[1], rating=3).save()
Rating(event=events[3], user=users[1], rating=4).save()
Rating(event=events[5], user=users[2], rating=4).save()
Rating(event=events[0], user=users[3], rating=5).save()
Rating(event=events[2], user=users[3], rating=3).save()
EventFeature(event=events[0], feature=features[1], tf_idf=1,
weight=w).save()
EventFeature(event=events[1], feature=features[0], tf_idf=1,
weight=w).save()
EventFeature(event=events[1], feature=features[1], tf_idf=1,
weight=w).save()
EventFeature(event=events[2], feature=features[1], tf_idf=1,
weight=w).save()
EventFeature(event=events[2], feature=features[2], tf_idf=1,
weight=w).save()
EventFeature(event=events[3], feature=features[1], tf_idf=1,
weight=w).save()
EventFeature(event=events[4], feature=features[0], tf_idf=1,
weight=w).save()
EventFeature(event=events[4], feature=features[1], tf_idf=1,
weight=w).save()
EventFeature(event=events[4], feature=features[2], tf_idf=1,
weight=w).save()
EventFeature(event=events[5], feature=features[3], tf_idf=1,
weight=w).save()
from recommendation import Recommendation
self.recommendation = Recommendation()
def test_RFSP(self):
RR = [[-1, 4, -1, -1, 5, -1], [-1, 3, -1, 4, -1, -1],
[-1, -1, -1, -1, -1, 4], [5, -1, 3, -1, -1, -1]]
FF = [[-1, 1, -1, -1], [1, 1, -1, -1], [-1, 1, 1, -1], [-1, 1, -1, -1],
[1, 1, 1, -1], [-1, -1, -1, 1]]
SS = [[0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0]]
PP = [[2, 2, 1, 0], [1, 2, 0, 0], [0, 0, 0, 1], [0, 2, 1, 0]]
self.recommendation.init_basic_matrix()
if not compare_list(
RR, transform_matrix_list(self.recommendation.get_matrix_r())):
print 'error RR test_RFSP'
print RR, '\n', transform_matrix_list(
self.recommendation.get_matrix_r())
if not compare_list(
FF, transform_matrix_list(self.recommendation.get_matrix_f())):
print 'error FF test_RFSP'
print FF, '\n', transform_matrix_list(
self.recommendation.get_matrix_f())
if not compare_list(
SS, transform_matrix_list(self.recommendation.get_matrix_s())):
print 'error SS test_RFSP'
print SS, '\n', transform_matrix_list(
self.recommendation.get_matrix_s())
if not compare_list(
PP, transform_matrix_list(self.recommendation.get_matrix_p())):
print 'error PP test_RFSP'
print PP, '\n', transform_matrix_list(
self.recommendation.get_matrix_p())
def test_FF_UF_IUF(self):
FF = [[2, 2, 1, 0], [1, 2, 0, 0], [0, 0, 0, 1], [0, 2, 1, 0]]
UF = [2, 3, 2, 1]
IUF = [
0.30102999566398114, 0.1249387366082999, 0.30102999566398114,
0.6020599913279623
]
self.recommendation.init_frequency_matrix()
if not compare_list(
FF, transform_matrix_list(
self.recommendation.get_matrix_ff())):
print 'error FF test_RFSP'
print FF, '\n', transform_matrix_list(
self.recommendation.get_matrix_ff())
if not compare_list(
UF, transform_matrix_list(
self.recommendation.get_matrix_uf())):
print 'error UF test_RFSP'
print UF, '\n', transform_matrix_list(
self.recommendation.get_matrix_uf())
if not compare_list(
IUF, transform_matrix_list(
self.recommendation.get_matrix_iuf())):
print 'error IUF test_RFSP'
print IUF, '\n', transform_matrix_list(
self.recommendation.get_matrix_iuf())
def test_W_UU(self):
W = [[0.6020599913279623, 0.2498774732165998, 0.30102999566398114, -1],
[0.30102999566398114, 0.2498774732165998, -1, -1],
[-1, -1, -1, 0.6020599913279623],
[-1, 0.2498774732165998, 0.30102999566398114, -1]]
UU = [[-1, 0.95551065333456875, -1.0, 1.0],
[0.95551065333456875, -1, -1.0, 1], [-1.0, -1.0, -1, -1.0],
[1, 1, -1.0, -1]]
self.recommendation.compute_matrix_w()
self.recommendation.compute_matrix_uu()
if not compare_list(
W, transform_matrix_list(self.recommendation.get_matrix_w())):
print 'error W test_RFSP'
print W, '\n', transform_matrix_list(
self.recommendation.get_matrix_w())
if not compare_list(
UU, transform_matrix_list(
self.recommendation.get_matrix_uu())):
print 'error UU test_RFSP'
print UU, '\n', transform_matrix_list(
self.recommendation.get_matrix_uu())
def test_final(self):
events = Event.objects.all()
res = {events[0]: 3, events[2]: 5, events[4]: 6}
res = OrderedDict(sorted(res.items(), key=lambda t: t[1],
reverse=True))
if not compare_ordered_dict(
self.recommendation.compute_recommended_events(
User.objects.all()[1]), res):
print 'error final result'
print res
print self.recommendation.compute_recommended_events(
User.objects.all()[1])
class RecommendationTest1:
def setUp(self):
clean()
c = Category(name="C", external_id="1")
w = Weight(name="t", weight=1.0)
w.save()
c.save()
for i in range(0, 3):
User(name="U" + str(i), external_id=str(i),
token="t" + str(i)).save()
users = User.objects.all()
for i in range(0, 5):
Feature(name="F" + str(i)).save()
features = Feature.objects.all()
for i in range(0, 5):
Event(category=c,
external_id=str(i),
name='E' + str(i),
website='',
description='').save()
events = Event.objects.all()
Rating(event=events[0], user=users[0], rating=1).save()
Rating(event=events[1], user=users[0], rating=2).save()
Rating(event=events[2], user=users[0], rating=4).save()
Rating(event=events[3], user=users[0], rating=5).save()
Rating(event=events[4], user=users[0], rating=3).save()
Rating(event=events[0], user=users[1], rating=3).save()
Rating(event=events[1], user=users[1], rating=2).save()
Rating(event=events[2], user=users[1], rating=5).save()
Rating(event=events[0], user=users[2], rating=1).save()
Rating(event=events[1], user=users[2], rating=4).save()
Rating(event=events[4], user=users[2], rating=3).save()
EventFeature(event=events[0],
feature=features[0],
tf_idf=0.2,
weight=w).save()
EventFeature(event=events[0],
feature=features[2],
tf_idf=0.8,
weight=w).save()
EventFeature(event=events[1],
feature=features[1],
tf_idf=0.8,
weight=w).save()
EventFeature(event=events[1],
feature=features[2],
tf_idf=1.0,
weight=w).save()
EventFeature(event=events[1],
feature=features[4],
tf_idf=0.3,
weight=w).save()
EventFeature(event=events[2],
feature=features[0],
tf_idf=1.0,
weight=w).save()
EventFeature(event=events[2],
feature=features[2],
tf_idf=0.3,
weight=w).save()
EventFeature(event=events[3],
feature=features[1],
tf_idf=0.2,
weight=w).save()
EventFeature(event=events[3],
feature=features[3],
tf_idf=0.4,
weight=w).save()
EventFeature(event=events[4],
feature=features[3],
tf_idf=0.8,
weight=w).save()
FeatureUser(user=users[0], feature=features[1], weight=0.4).save()
FeatureUser(user=users[0], feature=features[3], weight=0.8).save()
FeatureUser(user=users[0], feature=features[4], weight=0.3).save()
FeatureUser(user=users[1], feature=features[0], weight=1.0).save()
FeatureUser(user=users[1], feature=features[2], weight=0.5).save()
FeatureUser(user=users[2], feature=features[1], weight=0.2).save()
FeatureUser(user=users[2], feature=features[4], weight=0.6).save()
from recommendation import Recommendation
self.recommendation = Recommendation()
def test_RFSP(self):
RR = [[1, 2, 4, 5, 3], [3, 2, 5, -1, -1], [1, 4, -1, -1, 3]]
FF = [[0.2, -1, 0.8, -1, -1], [-1, 0.8, 1, -1, 0.3],
[1, -1, 0.3, -1, -1], [-1, 0.2, -1, 0.4, -1],
[-1, -1, -1, 0.8, -1]]
SS = [[0, 0.4, 0, 0.8, 0.3], [1, 0, 0.5, 0, 0], [0, 0.2, 0, 0, 0.6]]
PP = [[1, 0.2, 0.3, 1.2, 0], [1.2, 0, 1.1, 0, 0],
[0, 0.8, 1, 0.8, 0.3]]
self.recommendation.init_basic_matrix()
if not compare_list(
RR, transform_matrix_list(self.recommendation.get_matrix_r())):
print 'error RR test_RFSP'
print RR, '\n', transform_matrix_list(
self.recommendation.get_matrix_r())
if not compare_list(
FF, transform_matrix_list(self.recommendation.get_matrix_f())):
print 'error FF test_RFSP'
print FF, '\n', transform_matrix_list(
self.recommendation.get_matrix_f())
if not compare_list(
SS, transform_matrix_list(self.recommendation.get_matrix_s())):
print 'error SS test_RFSP'
print SS, '\n', transform_matrix_list(
self.recommendation.get_matrix_s())
if not compare_list(
PP, transform_matrix_list(self.recommendation.get_matrix_p())):
print 'error PP test_RFSP'
print PP, '\n', transform_matrix_list(
self.recommendation.get_matrix_p())
def test_FF_UF_IUF(self):
FF = [[1, 0.6, 0.3, 2.0, 0.3], [2.2, 0, 1.6, 0, 0],
[0, 1.0, 1, 0.8, 0.9]]
UF = [2, 2, 3, 2, 2]
IUF = [
0.176091259056, 0.176091259056, 0.0, 0.176091259056, 0.176091259056
]
self.recommendation.init_frequency_matrix()
if not compare_list(
FF, transform_matrix_list(
self.recommendation.get_matrix_ff())):
print 'error FF test_RFSP'
print FF, '\n', transform_matrix_list(
self.recommendation.get_matrix_ff())
if not compare_list(
UF, transform_matrix_list(
self.recommendation.get_matrix_uf())):
print 'error UF test_RFSP'
print UF, '\n', transform_matrix_list(
self.recommendation.get_matrix_uf())
if not compare_list(
IUF, transform_matrix_list(
self.recommendation.get_matrix_iuf())):
print 'error IUF test_RFSP'
print IUF, '\n', transform_matrix_list(
self.recommendation.get_matrix_iuf())
def test_W_UU(self):
W = [[
0.176091259056, 0.105654755433, -1, 0.352182518111, 0.0528273777167
], [0.387400769922, -1, -1, -1, -1],
[-1, 0.176091259056, -1, 0.140873007245, 0.15848213315]]
UU = [[-1, 1.0, 0.822730338093], [1.0, -1, -1],
[0.822730338093, -1, -1]]
self.recommendation.compute_matrix_w()
self.recommendation.compute_matrix_uu()
if not compare_list(
W, transform_matrix_list(self.recommendation.get_matrix_w())):
print 'error W test_RFSP'
print W, '\n', transform_matrix_list(
self.recommendation.get_matrix_w())
if not compare_list(
UU, transform_matrix_list(
self.recommendation.get_matrix_uu())):
print 'error UU test_RFSP'
print UU, '\n', transform_matrix_list(
self.recommendation.get_matrix_uu())
def test_final(self):
events = Event.objects.all()
res = {events[3]: 3, events[4]: 2}
res = OrderedDict(sorted(res.items(), key=lambda t: t[1],
reverse=True))
if not compare_ordered_dict(
self.recommendation.compute_recommended_events(
User.objects.all()[1]), res):
print 'error final result'
print res
print self.recommendation.compute_recommended_events(
User.objects.all()[1])
def recommend(self,
title,
limit,
critics=False,
full_search=False,
use_pickle=True,
keywords_and_desc=False):
"""
param: title - movie title (as in TMDB dataset)
limit - no. of movies to display
critics - True - will display critically acclaimed movies
False - will not sort movies on basis of their imdb rankings
(DEFAULT - False)
full_search - True - will search using cast, crew, keywords
and genre as metadata
False - will search using overview and tagline
as metadata
(DEFAULT - False)
use_pickle - True - will use pickled results
False - will compute the results from scratch
(DEFAULT - True)
keywords_and_desc - True - will merge results of keywords
and description
False - will not merge results of keywords
and description
return: pandas DataFrame object with attributes -
original_title, id, vote_average, vote_count, popularity, release_date
"""
rec = Recommendation()
rec.filter_genres()
title_index = self.verify_title(rec.md, title)
if keywords_and_desc:
if isfile(PATH_PICKLE_KEYWORDS) and isfile(
PATH_PICKLE_DESC) and use_pickle:
df_keywords = pd.read_pickle(PATH_PICKLE_KEYWORDS)
df_desc = pd.read_pickle(PATH_PICKLE_DESC)
rec_matrix_keywords = self.countvectorize(df_keywords)
rec_matrix_desc = self.tfidf(df_desc)
rec_matrix = rec_matrix_keywords + rec_matrix_desc
df = df_keywords
elif full_search:
if isfile(PATH_PICKLE_KEYWORDS) and use_pickle:
df = pd.read_pickle(PATH_PICKLE_KEYWORDS)
else:
df = self.make_keywords(rec.md)
df.to_pickle(PATH_PICKLE_KEYWORDS)
rec_matrix = self.countvectorize(df)
else:
if isfile(PATH_PICKLE_DESC) and use_pickle:
df = pd.read_pickle(PATH_PICKLE_DESC)
else:
df = self.make_desc(rec.md)
df.to_pickle(PATH_PICKLE_DESC)
rec_matrix = self.tfidf(df)
rec_movie = rec_matrix[title_index]
ids = rec_movie.argsort()[::-1][1:SCAN_SIZE + 1]
if critics:
return rec.top_movies(df.iloc[ids],
percentile=0.50,
limit=limit,
offset=0)
else:
return df.iloc[ids[:limit]][[
"original_title",
"id",
"vote_average",
"vote_count",
"popularity",
"release_date",
]]
def __init__(self):
self.recommendation = Recommendation()
def test_recommendation(self):
recommendation = Recommendation(self.client)
recommendation.forward_to("book1", "book2")
recommendation.forward_to("book1", "book2")
recommendation.forward_to("book1", "book2")
recommendation.forward_to("book1", "book3")
recommendation.forward_to("book1", "book3")
recommendation.forward_to("book1", "book4")
recommendation.forward_to("book1", "book5")
recommendation.forward_to("book1", "book6")
recommendation.forward_to("book1", "book6")
assert recommendation.get_recommendations("book1", 1, 5) == [
"book2",
"book6",
"book3",
"book5",
"book4",
]
assert recommendation.get_recommendations("book1",
1,
5,
with_time=True) == [
("book2", 3),
("book6", 2),
("book3", 2),
("book5", 1),
("book4", 1),
]
