Python 多个矩阵的加权和

我有一个m n x n矩阵的列表和一个m实值(alphas)的列表.n和m的值可能相当大.我试图用alphas计算矩阵的加权和.

我想知道是否有一个numpy函数(或任何其他库)可以比manual for loop方法更快地完成这项工作.

我在下面列出了我当前的功能.

def calculate_matrix_sums(mats, alphas):
    """
        Calculate the weighted sum of matrices in mats with weights alpha
    """
    k_mults = [np.multiply(mats[i], alphas[i]) for i in range(len(alphas))]
    k_sums1 = np.matrix(k_mults[0]) + np.matrix(k_mults[1])
    for i in range(2, len(k_mults)):
        k_sums1 = k_sums1 + np.asmatrix(k_mults[i])
    k_sums2 = np.asarray(k_sums1).astype(float)
    k_sums2 = k_sums2.reshape(len(mats[0]), len(mats[0]))
    return k_sums2

和示例代码:

matrices = np.asarray([np.array([[1., 0.77841638, 0.53239253, 0.9444068, 0.93024477],
                                 [0.77841638, 1., 0.7221497, 0.5805838, 0.68501944],
                                 [0.53239253, 0.7221497, 1., 0.36986265, 0.62792847],
                                 [0.9444068, 0.5805838, 0.36986265, 1., 0.88303226],
                                 [0.93024477, 0.68501944, 0.62792847, 0.88303226, 1.]]),
                       np.array([[1., 0.45650032, 0.13898701, 0.83605729, 0.79743304],
                                 [0.45650032, 1., 0.36094014, 0.18229867, 0.30596445],
                                 [0.13898701, 0.36094014, 1., 0.04443844, 0.23300302],
                                 [0.83605729, 0.18229867, 0.04443844, 1., 0.67745532],
                                 [0.79743304, 0.30596445, 0.23300302, 0.67745532, 1.]])])
alpha_vals = [0.47547796, 0.52452204]

print(calculate_matrix_sums(matrices, alpha_vals))

欢迎提出任何建议.