Numerical examples are given to illustrate that the norm errors of the approximate algorithm is generally small, while the smaller the absolute value is, the greater the component errors will be.

• 数值算例表明：该近似算法的范数误差一般较小，而绝对量较小的分量误差较大。