def train_trim_nmf(data, R):
kfold = KFold(n_splits=10)
rmse_list = [[], [], []]
for k in range(2, 52, 2):
print("using k = %d" % k)
p_rmse = []
u_rmse = []
hv_rmse = []
nmf = NMF(n_factors=k)
for trainset, testset in kfold.split(data):
nmf.fit(trainset)
(p_testset, u_testset, hv_testset) = trim(testset, R)
p_pred = nmf.test(p_testset)
u_pred = nmf.test(u_testset)
hv_pred = nmf.test(hv_testset)
p_rmse.append(accuracy.rmse(p_pred))
u_rmse.append(accuracy.rmse(u_pred))
hv_rmse.append(accuracy.rmse(hv_pred))
rmse_list[0].append(np.mean(p_rmse))
rmse_list[1].append(np.mean(u_rmse))
rmse_list[2].append(np.mean(hv_rmse))
print("NMF with trim is finished!!")
return rmse_list
def nmf_ratings_predicate(observed_ratings_df,
truth_ratings_df,
fold='0',
phase='eval'):
"""
nmf_ratings Predicates
"""
print("NMF predicates")
nmf_model = NMF()
reader = Reader(rating_scale=(0.2, 1))
train_dataset = Dataset.load_from_df(df=observed_ratings_df.reset_index(
).loc[:, ['userId', 'movieId', 'rating']],
reader=reader)
nmf_model.fit(train_dataset.build_full_trainset())
# make predictions
predictions = pd.DataFrame(index=truth_ratings_df.index,
columns=['rating'])
for row in truth_ratings_df.loc[:, ['rating']].iterrows():
uid = row[0][0]
iid = row[0][1]
predictions.loc[(uid, iid), 'rating'] = nmf_model.predict(uid, iid).est
write(predictions, 'nmf_rating_obs', fold, phase)
def nmf_compute_high_var_trim_rmse(k):
nmf = NMF(n_factors=k, random_state=42)
rmse = []
for trainset, testset in KFold(n_splits=10, random_state=42).split(R):
nmf.fit(trainset)
testset_trimmed = high_variance_trimming(testset, frequency, variance)
pred = nmf.test(testset_trimmed)
rmse.append(accuracy.rmse(pred, verbose=False))
print('k: %s | RMSE: %f' % (k, np.mean(rmse)))
return np.mean(rmse)
def nmf_compute_prec_rec(t):
precision, recall = [], []
for trainset, testset in KFold(n_splits=10, random_state=42).split(R):
nmf = NMF(n_factors=nmf_best_k, random_state=42)
nmf.fit(trainset)
trimmed_testset = trim_unpopular_user(testset, t, threshold)
pred = nmf.test(trimmed_testset)
precision_dict, recall_dict = calculate_precision_recall(
pred, t, threshold)
precision.append(np.mean([prec for prec in precision_dict.values()]))
recall.append(np.mean([rec for rec in recall_dict.values()]))
return np.mean(precision), np.mean(recall)
def Q23(col=0):
print('Chosen column is ' + str(col))
data = np.loadtxt('ml-latest-small/ratings.csv',
delimiter=',',
skiprows=1,
usecols=(0, 1, 2))
row_userId = data[:, :1].astype(int)
row_movieId = data[:, 1:2].astype(int)
row_rating = data[:, 2:3]
sortedId = np.sort(row_movieId.transpose()[0])
m = {}
idx = 0
last = None
for i in sortedId.tolist():
if i != last:
m[i] = idx
idx += 1
last = i
data = load_data()
model = NMF(n_factors=20)
trainset, testset = train_test_split(data, test_size=0.0001)
model.fit(trainset)
U = model.pu
V = model.qi
import csv
dict_ID_to_genre = {}
with open('ml-latest-small/movies.csv', 'rt') as csvfile:
reader = csv.reader(csvfile, delimiter=',', quotechar='|')
cnt = 0
for row in reader:
if cnt != 0:
dict_ID_to_genre[row[0]] = row[1:]
cnt += 1
dict_col_to_ID = {}
for key in m:
dict_col_to_ID[m[key]] = key
V_col = V[:, col]
V_col_sort_top10 = np.sort(V_col)[::-1][:10]
V_col_list = V_col.tolist()
for val in V_col_sort_top10:
ind = V_col_list.index(val)
m_id = dict_col_to_ID[ind]
genre = dict_ID_to_genre[str(m_id)]
print(genre[-1])
def NMF_trim_filter(ratings, dims, func, mv_dict):
reader = Reader(rating_scale=(0.0, 5.0))
data = Dataset.load_from_df(ratings[['userId', 'movieId', 'rating']],
reader)
RMSE = np.empty([len(dims)])
MAE = np.empty([len(dims)])
min_RMSE = False
min_MAE = False
fac_num_RMSE = 0
fac_num_MAE = 0
kf = KFold(n_splits=10, random_state=42)
for k in range(len(dims)):
nmf = NMF(n_factors=dims[k], random_state=42)
test_rmse = np.array([])
test_mae = np.array([])
for trainset, testset in kf.split(data):
nmf.fit(trainset)
full_data = trainset.build_testset() + testset
func(mv_dict, testset)
pred = nmf.test(testset)
test_rmse = np.append(test_rmse, accuracy.rmse(pred,
verbose=False))
test_mae = np.append(test_mae, accuracy.mae(pred, verbose=False))
RMSE[k] = np.mean(test_rmse)
if ((not min_RMSE) or RMSE[k] < min_RMSE):
min_RMSE = RMSE[k]
fac_num_RMSE = dims[k]
MAE[k] = np.mean(test_mae)
if ((not min_MAE) or MAE[k] < min_MAE):
min_MAE = MAE[k]
fac_num_MAE = dims[k]
print('For k = %i :' % dims[k])
print('RMSE: ', RMSE[k])
print('MAE: ', MAE[k])
plt.plot(dims, RMSE)
plt.plot(dims, MAE)
plt.legend(['RMSE', 'MAE'])
plt.show()
print('Finishing Plotting...')
print('For RMSE:')
print('\t---Optimal number of latent factors is ', fac_num_RMSE)
print('\t---Minumun Average RMSE is ', min_RMSE)
print('\nFor MAE:')
print('\t---Optimal number of latent factors is ', fac_num_MAE)
print('\t---Minumun Average MAE is ', min_MAE)
def Q19to21(qNum):
data = load_data()
kf = KFold(n_splits=10)
trimFun = {12: popularTrim, 13: unpopularTrim, 14: highVarTrim}
RMSE = []
for k in range(2, 20, 2):
nmf = NMF()
subRMSE = []
for trainSet, testSet in kf.split(data):
subsubRMSE = 0
nmf.fit(trainSet)
testSet = trimFun[qNum](testSet)
nTest = len(testSet)
print("test set size after trimming: %d", nTest)
predictions = nmf.test(testSet)
for p in predictions:
subsubRMSE += pow(p.est - p.r_ui, 2)
# average of all train-test splits of k-NN
RMSE.append(np.mean(subRMSE))
return RMSE
def use_nmf():
start = time.time()
performance = []
data = Dataset.load_builtin('ml-100k')
trainset = data.build_full_trainset()
print('Using NMF')
algo_NMF = NMF()
algo_NMF.fit(trainset)
testset = trainset.build_anti_testset()
predictions_NMF = algo_NMF.test(testset)
accuracy_rmse = accuracy.rmse(predictions_NMF)
accuracy_mae = accuracy.mae(predictions_NMF)
performance.append(accuracy_rmse)
performance.append(accuracy_mae)
end = time.time()
performance.append(end - start)
return performance
def NMF_bin_pre(ratings, ts, nmf_fac, thrd):
reader = Reader(rating_scale=(0.0, 5.0))
data = Dataset.load_from_df(ratings[['userId', 'movieId', 'rating']],
reader)
trainset, testset = train_test_split(data, test_size=ts)
algo = NMF(n_factors=nmf_fac, random_state=42)
algo.fit(trainset)
pre = algo.test(testset)
true_rating = np.empty(len(pre))
pred_rating = np.empty(len(pre))
for i in range(len(pre)):
true_rating[i] = pre[i][2]
pred_rating[i] = pre[i][3]
bi_rating = np.empty(len(pre))
one_idx = true_rating >= thrd
zero_idx = true_rating < thrd
bi_rating[one_idx] = 1.0
bi_rating[zero_idx] = 0.0
return bi_rating, pred_rating
bi_rating[zero_idx] = 0.0
return bi_rating, pred_rating
threshold = np.array([2.5, 3, 3.5, 4])
for td in threshold:
tar, pre = NMF_bin_pre(ratings, 0.1, 18, td)
plot_roc(pre, tar)
# Q23
reader = Reader(rating_scale=(0.0, 5.0))
data = Dataset.load_from_df(ratings[['userId', 'movieId', 'rating']], reader)
data = data.build_full_trainset()
nmf = NMF(n_factors=20, random_state=42)
nmf.fit(data)
for i in range(10):
col = nmf.qi[:, i]
top_movie = col.argsort()[::-1][:10]
print('For the %i th column, the top 10 movie genres are:' % (i + 1))
for j in range(10):
raw_iid = nmf.trainset.to_raw_iid(top_movie[j])
gen = movies.loc[movies['movieId'] == raw_iid]['genres'].values
print('\t--%i :' % (j + 1), gen)
# MF With Bias Filter
# Q24 & 25
def MF_bias_filter(ratings, dims):
reader = Reader(rating_scale=(0.0, 5.0))
plt.show()
fig3.savefig(path + 'fig/Part_8_knn_preVSrec.png')
#define model for training
k_min_rmse = 18
nnmf = NMF(n_factors=k_min_rmse, random_state=1)
#train, test and rank
top_t_list = range(1, 26)
pre_list_nnmf = []
rec_list_nnmf = []
for top_t in top_t_list:
pre = 0
rec = 0
for trainset, testset in kf.split(data_raw):
nnmf.fit(trainset)
prediction = nnmf.test(testset)
G = create_dict(testset)
G_s = create_dict(prediction, if_pred=1)
R, R_s = threshold_rank_filter(G, G_s, thre=3, top_t=top_t)
#precision and recall for each fold
pre_fold = 0
rec_fold = 0
for key in R.keys():
pre_temp, rec_temp = precision_recall(R[key], R_s[key])
pre_fold += pre_temp
rec_fold += rec_temp
pre += pre_fold / len(R)
rec += rec_fold / len(R)
pre_list_nnmf.append(pre / num_fold)
y_score = [prediction.est for prediction in predictions]
fpr, tpr, thresholds = roc_curve(y_true=y_true, y_score=y_score)
roc_auc = auc(fpr, tpr)
plotROC(fpr, tpr, roc_auc, threshold)
"""
Question 23: Movie-Latent Factor Interaction
"""
reader = Reader(rating_scale=(0.5, 5))
data = Dataset.load_from_df(df[['userId', 'movieId', 'rating']], reader)
data = data.build_full_trainset()
movieDat = pd.read_csv('ml-latest-small/movies.csv')
nmf = NMF(n_factors=20, biased=False)
nmf.fit(data)
movies = df['movieId'].unique(
) # identify unique movie IDs from the ratings CSV (9724, already sorted)
V = nmf.qi
# get top 10 movie genres for the first 20 columns of the V matrix
for i in range(20):
Vcol = V[:, i]
# convert column of V into a list for processing
VcolOrig = []
VcolSort = []
for j in range(len(Vcol)):
VcolOrig.append(Vcol[j]) # original array for looking up movie index
VcolSort.append(Vcol[j]) # sorted array for getting top movies
class DataLoader(object):
def __init__(self,category,save_name):
self.category=category
self.max_user=10000 #maximum number of user
self.price_dict={}
self.price_dict_temp={}
self.cate_dict={}
self.cate_dict_temp={}
self.top_value=15 # top x features in SVD
self.model=NMF()
self.topk=500 #maximum items in each category, finding the top k popular
self.max_price={}
self.save_path= os.path.join("..", "feature", save_name)
if not os.path.isfile(self.save_path):
self.load_data() #load raw data
#self.create_user_item_matrix()
self.create_ratings()
self.gen_new_price_dict()
self.save_data(self.save_path) #save the feature
else:
self.load(self.save_path) #load the feature
def load_ratings(self, filename):
with open(os.path.join("..", "data", filename), "rb") as f:
ratings = pd.read_csv(f,names=("user","item","rating","timestamp"))
return ratings
def load_prices(self,filename):
price_dict = {}
num_no_price=0
for review in parse(os.path.join("..", "data", filename)):
try:
price=review['price']
asin=review['asin']
v=list(review['salesRank'].values())[0]
if v<self.topk:
price_dict[asin]=price
except:
num_no_price+=1
continue
print("filename:",filename)
print("length of price dict:", len(price_dict))
print("# of items without price", num_no_price)
return price_dict
def load_data(self):
print("Loading data:")
for i in self.category:
ratings_name= "ratings_"+i+".csv"
price_name="meta_"+i+".json.gz"
ratings_temp=self.load_ratings(ratings_name)
print(len(ratings_temp))
price_temp=self.load_prices(price_name)
ratings_temp=ratings_temp[ratings_temp['item'].isin(price_temp.keys())]
print(len(ratings_temp))
self.price_dict_temp.update(price_temp)
self.max_price[i]=max(list(price_temp.values()))
cate_temp={}
for j in price_temp.keys():
cate_temp[j]=i
self.cate_dict_temp.update(cate_temp)
price_temp.clear()
try:
self.ratings=pd.merge(self.ratings,ratings_temp, how='outer')
except:
self.ratings=ratings_temp
print(self.max_price)
#old method
def create_user_item_matrix(self, user_key="user",item_key="item"):
n = len(set(self.ratings[user_key]))
d = len(set(self.ratings[item_key]))
self.user_mapper = dict(zip(np.unique(self.ratings[user_key]), list(range(n))))
self.item_mapper = dict(zip(np.unique(self.ratings[item_key]), list(range(d))))
self.user_inverse_mapper = dict(zip(list(range(n)), np.unique(self.ratings[user_key])))
self.item_inverse_mapper = dict(zip(list(range(d)), np.unique(self.ratings[item_key])))
self.user_ind = [self.user_mapper[i] for i in self.ratings[user_key]]
self.item_ind = [self.item_mapper[i] for i in self.ratings[item_key]]
self.ratings_matrix = sparse_matrix((self.ratings["rating"]-3, (self.user_ind, self.item_ind)), shape=(n,d))
print("user-item matrix generated.")
def create_ratings(self):
#C=MBRecsys(self.ratings_matrix,top_value)
S=set(self.ratings['user'])
S=sample(S,self.max_user)
n = len(S)
d = len(set(self.ratings['item']))
self.ratings=self.ratings[self.ratings['user'].isin(S)]
reader=Reader(rating_scale=(1,5))
data = Dataset.load_from_df(self.ratings[['user', 'item', 'rating']], reader)
train_set=data.build_full_trainset()
self.model.fit(train_set)
self.inv_cate_dict={} #{'categoryA':[],'categoryB':[]}
for i in self.category:
self.inv_cate_dict[i]=[]
for j in train_set.all_items():
item_raw=train_set.to_raw_iid(j)
self.inv_cate_dict[self.cate_dict_temp[item_raw]].append(j)
self.price_dict[j]=self.price_dict_temp[item_raw]
self.cate_dict[j]=self.cate_dict_temp[item_raw]
self.cate_dict_temp.clear()
self.price_dict_temp.clear()
print("inv_cate_dict constructed.")
d=0
for i in self.category:
d+=len(self.inv_cate_dict[i])
print(i,':',len(self.inv_cate_dict[i]))
self.ratings_predict=np.zeros([n,d])
for i in train_set.all_users():
user_raw=train_set.to_raw_uid(i)
for j in train_set.all_items():
item_raw=train_set.to_raw_iid(j)
self.ratings_predict[i][j]=self.model.predict(user_raw, item_raw)[3]
print("predicted ratings generated.")
self.ranking=np.zeros([n,d])
temp={}
for i in range(n):
for c in self.category:
temp[c]=sorted(self.ratings_predict[i][self.inv_cate_dict[c]],reverse=True)
for j in range(d):
c=self.cate_dict[j]
self.ranking[i][j]= temp[c].index(self.ratings_predict[i][j])+1
print("user_item rankings generated.")
def save_data(self,save_path):
self.dict_all={'prices':self.price_dict,#'raw_ratings':self.ratings_matrix,
'new_ratings':self.ratings_predict,'cate':self.cate_dict,
'rankings': self.ranking,'max_price':self.max_price,
'new_price':self.new_price_dict}
#'user_mapper':self.user_mapper, 'item_mapper':self.item_mapper,
#'user_inverse_mapper':self.user_inverse_mapper, 'item_inverse_mapper':self.item_inverse_mapper}
with open(save_path,'wb') as f:
pickle.dump(self.dict_all, f)
print("data saved in ", save_path)
def load(self,save_path):
with open(save_path,'rb') as f:
self.dict_all=pickle.load(f)
#self.ratings_matrix =self.dict_all['raw_ratings']
self.ratings_predict=self.dict_all['new_ratings']
self.price_dict=self.dict_all['prices']
self.cate_dict=self.dict_all['cate']
self.ranking=self.dict_all['rankings']
self.max_price=self.dict_all['max_price']
self.new_price_dict=self.dict_all['new_price']
#self.user_mapper=self.dict_all['user_mapper']
#self.item_mapper=self.dict_all['item_mapper']
#self.user_inverse_mapper=self.dict_all['user_inverse_mapper']
#self.item_inverse_mapper=self.dict_all['item_inverse_mapper']
self.dict_all.clear()
del self.dict_all
print("Saved data loaded.")
def gen_new_price_dict(self):
self.new_price_dict={}
for i in self.category:
self.new_price_dict[i]={}
for i in range(len(self.cate_dict)):
self.new_price_dict[self.cate_dict[i]][i]=self.price_dict[i]
print("new price dictionary generated.")
plt.plot(fpr,
tpr,
color='darkorange',
lw=lw,
label='ROC curve (area = %0.4f)' % roc_auc)
plt.plot([0, 1], [0, 1], color='navy', lw=lw, linestyle='--')
plt.xlim([0.0, 1.0])
plt.ylim([0.0, 1.05])
plt.xlabel('False Positive Rate')
plt.ylabel('True Positive Rate')
plt.title('ROC curve for threshold = %0.2f' % thresh)
plt.legend(loc="lower right")
plt.show()
#23
trainset, testset = train_test_split(data, test_size=.1, random_state=100)
nmf = NMF(n_factors=20, verbose=False, random_state=100)
nmf.fit(trainset).test(testset)
V = nmf.qi
k = [item for item in range(0, 20)]
df = pd.read_csv(movies_file_path,
names=['movieid', 'title', 'genres'],
header=0)
for i in k:
print(i)
mov = V[:, i]
mov1 = [(n, j) for n, j in enumerate(mov)]
mov1.sort(key=lambda x: x[1], reverse=True)
for a in mov1[:10]:
print(df['genres'][a[0]])
'RMSE after High Variance Movie Trimming')
# In[42]:
print("Minimum average RMSE after high variance movie trimming: %.4f" %
np.min(nmf_rmse_high_var_trim))
# <font size=4>**Question 22:** Plot the ROC curves for the NNMF-based collaborative filter designed in Question 17 for threshold values [2.5,3,3.5,4]. For the ROC plotting use the optimal number of latent factors found in Question 18. For each of the plots, also report the area under the curve (AUC) value.</font>
# In[43]:
nmf_best_k = ks[np.argmin(nmf_rmse)]
trainset, testset = train_test_split(R, test_size=0.1, random_state=42)
nmf_best = NMF(n_factors=nmf_best_k, random_state=42)
nmf_best.fit(trainset)
nmf_best_pred = nmf_best.test(testset)
plot_roc_curves(testset, nmf_best_pred, 'NNMF')
# <font size=4>**Question 23:** Perform Non-negative matrix factorization on the ratings matrix R to obtain the factor matrices U and V , where U represents the user-latent factors interaction and V represents the movie-latent factors interaction (use k = 20). For each column of V , sort the movies in descending order and report the genres of the top 10 movies. Do the top 10 movies belong to a particular or a small collection of genre? Is there a connection between the latent factors and the movie genres?</font>
# In[44]:
nmf_k20 = NMF(n_factors=20, random_state=42)
nmf_k20.fit(R.build_full_trainset())
# In[45]:
item_factors = nmf_k20.qi
def main():
# Load data
reader = Reader(sep=',', rating_scale=(0.0, 5.0), skip_lines=1)
allMoives = Dataset.load_from_file('ratings.csv', reader=reader)
popMoives = Dataset.load_from_file('popular.csv', reader=reader)
unpopMoives = Dataset.load_from_file('unpopular.csv', reader=reader)
varMoives = Dataset.load_from_file('variance.csv', reader=reader)
binary = []
binary.append(Dataset.load_from_file('bin2.5.csv', reader=reader))
binary.append(Dataset.load_from_file('bin3.csv', reader=reader))
binary.append(Dataset.load_from_file('bin3.5.csv', reader=reader))
binary.append(Dataset.load_from_file('bin4.csv', reader=reader))
with open('movies.csv', 'r', encoding='utf8') as f:
reader = csv.reader(f, delimiter=',', quotechar='"')
next(reader, None)
movies = {int(movie[0]): movie[2] for movie in reader}
# NMFs
ks = range(2, 52, 2)
mae, rmse = [0] * len(ks), [0] * len(ks)
def nmf(dataName, data, biased=True):
print('Start building NMF with ' + dataName + '!')
for i, k in enumerate(ks):
nmf = NMF(n_factors=k, biased=biased)
scores = cross_validate(nmf, data, cv=10)
mae[i] = scores['test_mae'].mean()
rmse[i] = scores['test_rmse'].mean()
print('k = ' + str(k) + ' finished!')
plt.figure()
plt.subplot(211)
plt.plot(ks, mae)
plt.xlabel('k')
plt.ylabel('mean absolute error')
plt.title('Mean absolute error vs. k of ' + dataName)
plt.subplot(212)
plt.plot(ks, rmse)
plt.xlabel('k')
plt.ylabel('root mean squared error')
plt.title('Root mean squared error vs. k of ' + dataName)
print('mae:')
print(mae)
print('rmse:')
print(rmse)
print('Finish building NMF with ' + dataName + '!')
# Q17
nmf('all movies', allMoives)
# Q18
optimalK = 4
print('The optimal number of latent factors is ' + str(optimalK))
# Q19
nmf('popular movies', popMoives)
# Q20
nmf('unpopular movies', unpopMoives)
# Q21
nmf('high variance movies', varMoives)
# Draw ROC Curve
thresholds = [2.5, 3, 3.5, 4]
def drawRoc(model, i, k):
print('Start drawing ROC curve of NMF with optimal k = ' + str(k) +
', threshold = ' + str(thresholds[i]) + '!')
train, test = train_test_split(binary[i],
train_size=0.9,
test_size=0.1)
model.fit(train)
labels = model.test(test)
y_true = [label.r_ui for label in labels]
y_pred = [label.est for label in labels]
fpr, tpr, _ = roc_curve(y_true, y_pred)
roc_auc = auc(fpr, tpr)
plt.figure()
plt.plot(fpr,
tpr,
color='darkorange',
lw=2,
label='ROC curve (area = %0.2f)' % roc_auc)
plt.plot([0, 1], [0, 1], color='navy', lw=2, linestyle='--')
plt.xlim([-0.05, 1.05])
plt.ylim([-0.05, 1.05])
plt.xlabel('False Positive Rate')
plt.ylabel('True Positive Rate')
plt.title('ROC curve of NMF with optimal k = ' + str(k) +
', threshold = ' + str(thresholds[i]))
plt.legend(loc="lower right")
print('Finish drawing ROC curve of NMF with optimal k = ' + str(k) +
', threshold = ' + str(thresholds[i]) + '!')
# Q22
nmf = NMF(n_factors=optimalK)
for i in range(len(thresholds)):
drawRoc(nmf, i, optimalK)
# Q23
print("Start finding top K!")
k, col = 20, 5
nmf = NMF(n_factors=k)
trainAllMovies = allMoives.build_full_trainset()
nmf.fit(trainAllMovies)
ids = [[] for _ in range(col)]
for i in range(col):
factors = nmf.qi[:, i]
s = sorted([[i, factor] for i, factor in enumerate(factors)],
key=lambda x: x[1],
reverse=True)
for k in range(10):
ids[i].append(s[k][0])
genres = [[] for _ in range(col)]
for i in range(col):
for j in range(10):
genres[i].append(movies[int(trainAllMovies.to_raw_iid(ids[i][j]))])
for i in range(col):
print('Col ' + str(i + 1) + ':')
for genre in genres[i]:
print(genre, end=', ')
print('')
print("Finish finding top K!")
# Q24
nmf('all movies', allMoives, True)
# Q25
optimalKBiased = 2
print('The optimal number of latent factors is ' + optimalKBiased)
# Q26
nmf('popular movies', popMoives, True)
# Q27
nmf('unpopular movies', unpopMoives, True)
# Q28
nmf('high variance movies', varMoives, True)
# Q29
optimalKBiased = 2
nmfBiased = NMF(n_factors=optimalKBiased, biased=True)
for i in range(len(thresholds)):
drawRoc(nmfBiased, i, optimalKBiased)
plt.show()
plt.savefig('plot/q22_nmf_roc_' + str(threshold) + '.png')
plt.clf()
if __name__ == "__main__":
threshold = [2.5, 3, 3.5, 4]
file_path = os.path.expanduser("ml-latest-small/ratings_new.csv")
reader = Reader(sep=',')
data = Dataset.load_from_file(file_path, reader=reader)
sim_options = {'name': 'pearson', 'user_based': True}
trainset, testset = train_test_split(data, test_size=0.1)
for th in threshold:
algo = NMF(n_factors=16)
algo.fit(trainset)
predictions = algo.test(testset)
y_true = []
y_estimate = []
for row in predictions:
if row[2] >= th:
y_true.append(1)
else:
y_true.append(0)
y_estimate.append(row[3])
plot_roc(y_true, y_estimate, th)
def problems_19_20_21_rmse_pop_unpop_hv():
x_axis = range(2, 52, 2)
ratings = {}
for r in data.raw_ratings:
if r[1] not in ratings:
ratings[r[1]] = []
ratings[r[1]].append(r[2])
###############################################################################################
popular_movies = [x for x in ratings if len(ratings[x]) > 2]
unpopular_movies = [x for x in ratings if len(ratings[x]) <= 2]
###################################################################################
kf = KFold(n_splits=10)
rmse_popular_store = []
for i in x_axis:
algo = NMF(i, verbose=False)
accu = []
for trainset, testset in kf.split(data):
algo.fit(trainset)
test_trim = [x for x in testset if x[1] in popular_movies]
predictions = algo.test(test_trim)
accu.append(accuracy.rmse(predictions, verbose=True))
s = np.mean(accu)
rmse_popular_store.append(s)
plotgraphs(x_axis, rmse_popular_store, 'K', 'Mean RMSE scores',
'Plot of popular movies', 'q19_rmse_popular_movies.png')
plotgraphs(x_axis, rmse_popular_store, 'K', 'Mean RMSE scores',
'Plot of popular movies')
##########################################################################################
kf = KFold(n_splits=10)
rmse_unpopular_store = []
for i in x_axis:
algo = NMF(i, verbose=False)
accu = []
for trainset, testset in kf.split(data):
algo.fit(trainset)
test_trim = [x for x in testset if x[1] in unpopular_movies]
predictions = algo.test(test_trim)
accu.append(accuracy.rmse(predictions, verbose=True))
s = np.mean(accu)
rmse_unpopular_store.append(s)
plotgraphs(x_axis, rmse_unpopular_store, 'K', 'Mean RMSE scores',
'Plot of unpopular movies', 'q20_rmse_unpopular_movies.png')
plotgraphs(x_axis, rmse_unpopular_store, 'K', 'Mean RMSE scores',
'Plot of unpopular movies')
############ rates "key" id, values are ratings #######################################
movie_var = {}
for k in ratings:
# print(k)
movie_var[k] = np.var(ratings[k])
####################################################################################
highvar_movies = [
x for x in ratings if len(ratings[x]) >= 5 and movie_var[x] >= 2
]
##################################################################################
kf = KFold(n_splits=10)
rmse_highvar_store = []
for i in x_axis:
algo = NMF(i, verbose=False)
accu = []
for trainset, testset in kf.split(data):
algo.fit(trainset)
test_trim = [x for x in testset if x[1] in highvar_movies]
predictions = algo.test(test_trim)
accu.append(accuracy.rmse(predictions, verbose=True))
s = np.mean(accu)
rmse_highvar_store.append(s)
pd.DataFrame(rmse_highvar_store).to_csv("rmse_highvar_store_21.csv")
plotgraphs(x_axis, rmse_highvar_store, 'K', 'Mean RMSE scores',
'Plot of high variance movies', 'q21_rmse_high_var_movies.png')
