def loadDataset(filename, trainingSet=[] , testSet=[]): test=[] with open(filename, 'r') as csvfile: lines = csv.reader(csvfile) dataset = list(lines) movies = di.getAllMovies() tagIds = di.getAllTags() allTagLen = len(tagIds) dataset_copy = [['' for i in range(allTagLen+1)] for j in range(len(movies))] idfMovArr = idf.idfMovieTag() for i in range(len(dataset)): idfVect = idf.tfIdfMovieTag(dataset[i][0], idfMovArr) for j in range(len(idfVect)): dataset_copy[i][j] = idfVect[j] dataset_copy[i][allTagLen]=dataset[i][1] trainingSet.append(dataset_copy[i]) train = [0 for i in range(len(dataset))] for i in range(len(dataset)): train[i] = int(dataset[i][0]) k=0 labels = ['0', '1'] testset_copy = [['' for i in range(allTagLen+1)] for j in range(len(movies)-len(train))] for i in range(len(movies)): if(int(movies[i][0]) in train): pass else: test.append(movies[i][0]) idfVect1 = idf.tfIdfMovieTag(movies[i][0], idfMovArr) for j in range(len(idfVect1)): testset_copy[k][j] = idfVect1[j] #testset_copy[k][allTagLen]=db.getMovieGenre(movies[i][0])[0] testset_copy[k][allTagLen]=random.choice(labels) testSet.append(testset_copy[k]) k=k+1 return test,trainingSet,testSet
def formSvdMat(numSemantics):
mat = np.zeros((movieLen,allTagLen))
if(len(mat)<numSemantics or len(mat[0])<numSemantics):
print("cant report top semantics")
sys.exit()
idfMovArr = idf.idfMovieTag()
for i in range(movieLen):
mat[i] = idf.tfIdfMovieTag(movies[i][0], idfMovArr)
U, s, V = np.linalg.svd(mat,full_matrices=False)
movieFacts = np.zeros((movieLen, numSemantics))
for i in range(movieLen):
for j in range(numSemantics):
movieFacts[i][j] = U[i][j]
return movieFacts
def loadDataset(filename, trainingSet=[], testSet=[]):
with open(filename, 'r') as csvfile:
lines = csv.reader(csvfile)
dataset = list(lines)
labels = ['' for i in range(len(dataset))]
movies = di.getAllMovies()
tagIds = di.getAllTags()
allTagLen = len(tagIds)
dataset_copy = [['' for i in range(allTagLen)]
for j in range(len(dataset))]
#dataset_copy = numpy.zeros((len(movies),allTagLen+1))
#dataset_copy = [[0 for i in range(allTagLen+1)] for j in range(len(movies))]
idfMovArr = idf.idfMovieTag()
#print(idfMovArr)
for i in range(len(dataset)):
idfVect = idf.tfIdfMovieTag(dataset[i][0], idfMovArr)
for j in range(len(idfVect)):
dataset_copy[i][j] = idfVect[j]
#dataset_copy[i][allTagLen]=dataset[i][1]
labels[i] = dataset[i][1]
trainingSet.append(dataset_copy[i])
train = [0 for i in range(len(dataset))]
target = ['' for i in range(len(movies))]
for i in range(len(dataset)):
train[i] = int(dataset[i][0])
k = 0
test = []
label = ['0', '1']
testset_copy = [['' for i in range(allTagLen)] for j in range(len(movies))]
for i in range(len(movies)):
if (int(movies[i][0]) in train):
pass
else:
test.append(movies[i][0])
idfVect1 = idf.tfIdfMovieTag(movies[i][0], idfMovArr)
for j in range(len(idfVect1)):
testset_copy[k][j] = idfVect1[j]
#testset_copy[k][allTagLen]=di.getMovieGenre(movies[i][0])[0]
#testset_copy[k][allTagLen]=random.choice(labels)
target[k] = random.choice(label)
testSet.append(testset_copy[k])
k = k + 1
#print("train data =",trainingSet)
#print("\n\n test data =",testSet)
return trainingSet, testSet, labels, target, test
def formPPRMatrix(): mat = np.zeros((movieLen,allTagLen)) idfMovArr = idf.idfMovieTag() for i in range(len(movies)): mat[i] = idf.tfIdfMovieTag(movies[i][0], idfMovArr) return mat
if __name__ == "__main__":
trainingSet = []
movies = loadDataset('foo.csv')
numsemantics = 500
allMovies = di.getAllMovies()
tagIds = di.getAllTags()
allTagLen = len(tagIds)
movNames = di.getAllMovieNames()
movieNames = np.array(['' for i in range(len(movies))])
for i in range(len(allMovies)):
mov = allMovies[i]
if(mov[0] in movies):
idx = movies.index(mov[0])
movieNames[idx] = movNames[i][0]
mat = np.zeros((len(movies),allTagLen))
idfMovArr = idf.idfMovieTag()
for i in range(len(movies)):
mat[i] = idf.tfIdfMovieTag(movies[i], idfMovArr)
a = svd.svdUout(mat,numsemantics)
d = 5
xmax = 20
points = np.array(a)
l = int(input("Enter number of layers:"))
k = int(input("Enter number of hashes per layer:"))
r = int(input("Enter number of neighbours:"))
q = input("Enter query movieid:")
ks = []
ls = []
ks.append(k)
ls.append(l)
query=[]