在python中使用稀疏矩阵时如何避免负值？

我正在使用稀疏矩阵编写以下代码

def Transition_Matrix(adj,beta):
    Iden = sp.eye(adj.shape[0])
    adj = sp.coo_matrix(adj)
    rowsum = np.array(adj.sum(1))
    degree_mat_inv_sqrt = sp.diags(np.power(rowsum, -0.5).flatten()
    adj_normalized = adj.dot(degree_mat_inv_sqrt).tocoo()
    matrix_=Iden-(beta * adj_normalized)
    ****print(matrix_)**** this is the result
here beta=0.9


1.         -0.3        -0.28460499 ... -0.36742346  0.
   0.        ]
 [-0.225       1.         -0.28460499 ...  0.          0.
   0.        ]
 [-0.225      -0.3         1.         ...  0.         -0.25980762
   0.        ]
 ...
 [-0.225       0.          0.         ...  1.         -0.25980762
  -0.21828206]
 [ 0.          0.         -0.28460499 ... -0.36742346  1.
  -0.21828206]
 [ 0.          0.          0.         ... -0.36742346 -0.25980762
   1.        ]]


matrixinv=np.linalg.pinv(matrix_)
Transition=(1-beta)*matrixinv
return Transition

输入是稀疏矩阵：

adj=nx.to\u scipy\u sparse\u矩阵（G）

如何避免这种负值？任何帮助都将不胜感激