def test_bpmf(self):
Y = scipy.sparse.rand(10, 20, 0.2)
Y, Ytest = macau.make_train_test(Y, 0.5)
results = macau.bpmf(Y, Ytest = Ytest, num_latent = 4,
verbose = False, burnin = 50, nsamples = 50,
univariate = False)
self.assertEqual(Ytest.nnz, results.prediction.shape[0])
self.assertTrue( (results.prediction.columns[0:2] == ["row", "col"]).all() )
def test_macau_univariate(self):
Y = scipy.sparse.rand(10, 20, 0.2)
Y, Ytest = macau.make_train_test(Y, 0.5)
side1 = scipy.sparse.coo_matrix( np.random.rand(10, 2) )
side2 = scipy.sparse.coo_matrix( np.random.rand(20, 3) )
results = macau.bpmf(Y, Ytest = Ytest, side = [side1, side2], num_latent = 4,
verbose = False, burnin = 50, nsamples = 50,
univariate = True)
self.assertEqual(Ytest.nnz, results.prediction.shape[0])
def test_make_train_test(self):
X = scipy.sparse.rand(15, 10, 0.2)
Xtr, Xte = macau.make_train_test(X, 0.5)
self.assertEqual(X.nnz, Xtr.nnz + Xte.nnz)
diff = np.linalg.norm( (X - Xtr - Xte).todense() )
self.assertEqual(diff, 0.0)
#print(Y)
grid = np.indices((drug_num, assay_num))
rows_idx = grid[0].flatten()
col_idx = grid[1].flatten()
import random
idx_sel = random.sample(range(len(rows_idx)), (nums_0))
row_sel = rows_idx[idx_sel]
col_sel = col_idx[idx_sel]
# Set some samples to 0
Y[row_sel[0:nums_0], col_sel[0:nums_0]] = 0
#Training
D = scipy.sparse.coo_matrix(Y)
#Test
D, Dtest = macau.make_train_test(D, test_per)
#Censor below censoring threshold:
a = D.data
cens = [1] * sum(D.data < cens_threshold)
C_mat = scipy.sparse.coo_matrix(
(cens,
(D.row[D.data < cens_threshold], D.col[D.data < cens_threshold])),
shape=D.shape)
D.data[a < cens_threshold] = cens_threshold
cens_numbers[idx_i, loop] = len(C_mat.data) / len(D.data)
result = macau.macau(Y=D,
Ytest=Dtest,
side=[None, None],
num_latent=latents,
