[🐞] basicts.utils.adjacent_matrix_norm 中函数返回类型与签名不一致

[Les1ie]

Is there an existing issue / discussion for this? | 是否已有关于该错误的issue或讨论？

• I have searched the existing issues / discussions | 我已经搜索过已有的issues和讨论

Is there an existing answer for this in tutorial? | 该问题是否在教程中有解答？

• I have searched tutorial | 我已经搜索过tutorial

Current Behavior | 当前行为

• 实际返回类型为 scipy.sparse._csc.csc_matrix
• 签名为 np.matrix

---

以 calculate_symmetric_message_passing_adj 为例

def calculate_symmetric_message_passing_adj(adj: np.ndarray) -> np.matrix:
    """
    Calculate the renormalized message-passing adjacency matrix as proposed in GCN.

    The message-passing adjacency matrix is defined as A' = D^{-1/2} (A + I) D^{-1/2}.

    Args:
        adj (np.ndarray): Adjacency matrix A.

    Returns:
        np.matrix: Renormalized message-passing adjacency matrix.
    """

    adj = adj + np.eye(adj.shape[0], dtype=np.float32)
    adj = sp.coo_matrix(adj)

    row_sum = np.array(adj.sum(1)).flatten()
    d_inv_sqrt = np.power(row_sum, -0.5)
    d_inv_sqrt[np.isinf(d_inv_sqrt)] = 0.0

    d_mat_inv_sqrt = sp.diags(d_inv_sqrt)
    mp_adj = d_mat_inv_sqrt.dot(adj).transpose().dot(d_mat_inv_sqrt).astype(np.float32)

    return mp_adj

Expected Behavior | 期望行为

实际类型与签名类型保持一致

Environment | 运行环境

- BasicTS: 1.1.0

BasicTS logs | BasicTS日志

No response

Steps To Reproduce | 复现方法

No response

Anything else? | 备注

No response