Python SVD.similarity方法代码示例

# 需要导入模块: from recsys.algorithm.factorize import SVD [as 别名]
# 或者: from recsys.algorithm.factorize.SVD import similarity [as 别名]
def loadSVD():
    
        
    filename = 'doubanRate.dat'
    svd = SVD()
    svd.load_data(filename=filename, sep='::', format={'col':0, 'row':1, 'value':2, 'ids': int})
    
    svd.save_data("svd.dat", False)
    
    K=25
    svd.compute(k=K, min_values=1, pre_normalize="rows", mean_center=False, post_normalize=True, savefile='.')
    
    
    #svd.recommend(USERID, n=10, only_unknowns=True, is_row=False)
    
    sparse_matrix = svd.get_matrix()
    
    sim_matrix = svd.get_matrix_similarity()
    
    
    
    print sparse_matrix
    
    print sim_matrix
    
    #1173893,1396943(borne identity),1251131(kong)
    sim = svd.similar(1396943, 10)
    simi = svd.similarity(1396943, 1429174)
    
    filename = 'swoffering.yaml'
    titleStream = file(filename, 'r')
    titleList = yaml.load(titleStream)
    
    for row in sim:
        
        (offid, similar) = row
        
        print offid, titleList[str(offid)], similar

# 需要导入模块: from recsys.algorithm.factorize import SVD [as 别名]
# 或者: from recsys.algorithm.factorize.SVD import similarity [as 别名]

# 2. Compute Singular Value Decomposition (SVD), M=U Sigma V^t:
k = 100
svd.compute(k=k,
            min_values=10,
            pre_normalize=None,
            mean_center=True,
            post_normalize=True,
            savefile='/tmp/movielens')

# 3. Get similarity between two movies:
ITEMID1 = 1    # Toy Story (1995)
ITEMID2 = 2355 # A bug's life (1998)

print svd.similarity(ITEMID1, ITEMID2)
# 0.67706936677315799


"""

# 4. Get movies similar to Toy Story:
svd.similar(ITEMID1)


# 5. Predict the rating a user (USERID) would give to a movie (ITEMID):
MIN_RATING = 0.0
MAX_RATING = 5.0
ITEMID = 1
USERID = 1

# 需要导入模块: from recsys.algorithm.factorize import SVD [as 别名]
# 或者: from recsys.algorithm.factorize.SVD import similarity [as 别名]
    svd.set_data(train)
    svd.compute(k=K, min_values=None, pre_normalize=None, mean_center=True, post_normalize=True)

    # save
    # svd.set_data(None)  # clear data before saving
    # pickle.dump(svd, open('./model/svd.obj', 'w'))
    svd.save_model('./model/svd.obj.zip',
                   {'k': K, 'min_values': 5,
                    'pre_normalize': None, 'mean_center': True, 'post_normalize': True})


    # similarity between items x and y
    print '-------- SIMILARITIES:'
    for prodid1 in [0, 1, 3, 4]:
        for prodid2 in [0, 1, 3, 4]:
            print prodid1, prodid2, svd.similarity(prodid1, prodid2)

    # similar to item x
    # svd.similar(1)
    #
    # # predict ratings
    # evaluate(svd, test, True)
    #
    # # recommend products to a user
    # for userid in [0, 1, 2]:
    #     print 'User #', userid
    #     print svd.recommend(userid, is_row=False, only_unknowns=True)
    #
    # # which users should use a given product?
    # for prodid in [0, 1, 3, 4]:
    #     print 'Product #', prodid

# 需要导入模块: from recsys.algorithm.factorize import SVD [as 别名]
# 或者: from recsys.algorithm.factorize.SVD import similarity [as 别名]
# In[4]:

# compute svd
k = 100
svd.compute(k=k, min_values=10, pre_normalize=None, mean_center=True,
    post_normalize=True)


# In[5]:

# movie id's
ITEMID1 = 1      # toy story
ITEMID2 = 1221   # godfather II

# How similar are these films
svd.similarity(ITEMID1, ITEMID2)


# In[6]:

# What about 
ITEMID3 = 2355 # A bug's life
svd.similarity(ITEMID1, ITEMID3)


# In[7]:

# We cen get films similar to Toy Story
svd.similar(ITEMID1)

# 需要导入模块: from recsys.algorithm.factorize import SVD [as 别名]
# 或者: from recsys.algorithm.factorize.SVD import similarity [as 别名]
import numpy
import scipy
from recsys.algorithm.factorize import SVD
svd = SVD()
svd.load_data(filename='m-medium/ratings.dat',
	    sep='::',
	    format={'col':0, 'row':1, 'value':2, 'ids': int})
k = 100
svd.compute(k=k,
	    min_values=10,
	    pre_normalize=None,
	    mean_center=True,
	    post_normalize=True,
	    savefile='/tmp/movielens')
sims=[]
for i in range(0,89000):
	for j in range(i+1,90000):
		try:
			l=[[i,j],[svd.similarity(i,j),1]]
			print(l)
		except:
			pass

# 需要导入模块: from recsys.algorithm.factorize import SVD [as 别名]
# 或者: from recsys.algorithm.factorize.SVD import similarity [as 别名]
               })

item2 = Item(1)
item2.add_data({'name': 'project1',
               'popularity': 0.5,
               'tags': [0, 0, 1]
               })

# create a user
userId = 0
user = User(userId)

# link an item with a user
rating = 1
user.add_item(itemId, rating)

data = Data()
data.add_tuple((rating, itemId, userId))
data.add_tuple((10, 1, 2))


svd = SVD()
svd.set_data(data)
svd.compute(k=100, min_values=0, pre_normalize=None, mean_center=True, post_normalize=True)

svd.similarity(0, 0)

l1 = ['a', 0, 1, 1]
l2 = ['b', 0, 1, 1]
print 1- spatial.distance.cosine(l1, l2)
cosine_similarity(l1, l2)