📚 2 Rules of Matrix Algebra | 矩阵代数的两条法则
When you first meet matrices, they appear as neat grids of numbers that follow a powerful set of algebraic rules. While matrix addition behaves just like ordinary addition of numbers – it is commutative and associative – matrix multiplication introduces surprises that are central to understanding linear algebra. Among the many properties of matrix operations, two rules stand out as essential for IB Mathematics students: the fact that matrix multiplication is not commutative, and the fact that it is associative. Mastering these two ideas will help you avoid careless mistakes and build deeper insight into why matrices are so useful in transformations, systems of equations and beyond.
当你第一次接触矩阵时，它们看起来只是排列整齐的数字，但背后却有一套强大的代数规则。矩阵加法与普通数字加法一样，满足交换律和结合律，然而矩阵乘法却带来了一些出人意料的特性，而这些正是理解线性代数的关键。在所有矩阵运算性质中，有两条规则对 IB 数学的学生尤为重要：矩阵乘法不可交换，以及矩阵乘法满足结合律。深入掌握这两个概念，不仅能帮你避免许多常见错误，还能让你理解为何矩阵在变换、方程组等众多领域如此强大。
1. The Basics of Matrix Operations | 矩阵运算基础
matrix multiplication. For matrices of the same dimensions, addition is performed element‑ tivity: kA = Ak. The interesting behaviour appears when we multiply two matrices together.
在我们重点讨论两条核心规则之前，先回顾一下基本运算：加法、标量乘法和矩阵乘法。对于维度相同的矩阵，加法是按对应元素进行的。如果 A 和 B 都是 m × n 矩阵，那么 A + B 也是 m × n 矩阵，并且 A + B = B + A。这与实数的加法完全一致。标量乘法——即每个元素都乘以常数 k——也满足交换性：kA = Ak。真正有趣的现象出现在两个矩阵相乘的时候。
2. Rule 1: Matrix Multiplication is Not Commutative | 规则一：矩阵乘法不可交换
The first rule – and perhaps the most important one to remember – is that, in general, AB ≠ BA. Unlike multiplication of ordinary numbers, the order in which you multiply matrices matters greatly. Changing the order often results in a completely different product, and in many cases one of the products may not even be defined because the dimensions do not match.
第一条规则——或许也是最需要记住的一条——就是一般情况下 AB ≠ BA。与普通数字的乘法不同，矩阵相乘的顺序至关重要。调换顺序通常会得到一个完全不同的结果；在很多情况下，甚至可能因为维度不兼容而导致其中一个乘积根本无法定义。
3. Why AB = BA Fails: Dimension Mismatch | 为何 AB = BA 不成立：维度不匹配
To see why non‑ matching inner dimensions. This means that for rectangular matrices, even the very possibility of commuting requires very specific shapes – e same size. Even when both A and B are n × n square matrices, AB usually differs from BA.
要理解为什么矩阵乘法普遍不可交换，可以从维度入手。假设 A 是 m × n 矩阵，B 是 n × p 矩阵，那么乘积 AB 有定义，结果是一个 m × p 矩阵。而反向乘积 BA 要有定义，就必须满足 p = m，这样 B (n × p) 和 A (m × n) 的内层维度才能匹配。这意味着对于非方阵，可交换性几乎无法实现——通常只有两个矩阵都是同阶方阵时，AB 和 BA 才都有定义。即便如此，对于方阵，AB 一般也仍然不等于 BA。
4. Example: Non‑Commutativity in 2 × 2 Matrices | 例子：2 × 2 矩阵的不可交换
Let us take two simple 2 × 2 matrices and compute both products. Set
我们取两个简单的 2 × 2 矩阵，并计算两个乘积。设
| A = |
|
, B = |
|
Then AB =
=
(1×0 + 2×-1) (1×1 + 2×0) (3×0 + 4×-1) (3×1 + 4×0)
,
while BA =
-2 1 -4 3
=
(0×1 + 1×3) (0×2 + 1×4) (-1×1 + 0×3) (-1×2 + 0×4)
.
Clearly AB ≠ BA. The entries are completely different, demonstrating that even for tiny square matrices, order matters.3 4 -1 -2
计算得 AB =
，
而 BA =
-2 1 -4 3
。
显然 AB ≠ BA，两者元素完全不同。这个例子说明，即便对最简单的方阵，相乘的顺序也极为关键。3 4 -1 -2
5. Special Cases When Matrices Do Commute | 矩阵可交换的特例
you should never assume commutativity unless you have verified it or the matrices belong to a family where commuting is guaranteed.
确实有一些矩阵是可交换的，但这些都属于特殊情况。同阶的对角矩阵彼此总是可交换的。单位矩阵 I 与任何同阶方阵都可交换：AI = IA = A。此外，矩阵与其自身的幂总是可交换的，例如 A A² = A² A。然而，除非经过验证或矩阵属于那些必定可交换的特殊类型，否则绝不能随意假定交换律成立。
6. Rule 2: Matrix Multiplication is Associative | 规则二：矩阵乘法满足结合律
triple product ABC without any parentheses, just as we do for numbers. Associativity is a powerful property that lies at the heart of many matrix applications, including transformations in computer graphics and the solving of linear systems. It guarantees that however we group the multiplication, the final outcome is the same.
第二条关键规则是结合律：对于任何维度相容的矩阵 A、B 和 C，总有 (AB)C = A(BC)。这使得我们可以像写数字乘法一样，省略括号直接写出乘积 ABC。结合律是矩阵应用中的一项强大性质，在计算机图形学中的变换、线性方程组的求解等众多场合都起着核心作用。它确保了无论我们如何分组进行乘法运算，最终结果都不会改变。
7. Proof Outline for Associativity | 结合律的证明概要
ding the idea builds confidence. Suppose A is m × n, B is n × p and C is p × q. The (i, j)‑entry of (AB)C is the sum over k of (sum over l aᵢₗ bₗₖ) cₖⱼ. Re‑grouping the sums shows this equals the sum over l of aᵢₗ (sum over k bₗₖ cₖⱼ), which is exactly the (i, j)‑ ums. This proof confirms that parentheses are unnecessary in matrix multiplication chains.
虽然 IB 考试中通常不要求书写完整证明，但理解其思路能帮助你建立信心。假设 A 是 m × n 矩阵，B 是 n × p 矩阵，C 是 p × q 矩阵。(AB)C 的第 (i, j) 元等于先对 l 求和再对 k 求和：∑ₖ (∑ₗ aᵢₗ bₗₖ) cₖⱼ。重新组合求和顺序后，该表达式等于 ∑ₗ aᵢₗ (∑ₖ bₗₖ cₖⱼ)，这正是 A(BC) 的第 (i, j) 元。这一等式之所以成立，源于普通数的结合律以及我们可以交换有限求和顺序的事实。该证明说明，在多个矩阵连乘时括号完全是多余的。
8. Why Associativity Matters in Practice | 结合律在实际中的意义
Associativity is not just a theoretical curiosity – 0×100×50 = 25000 + 50000 = 75000 multiplications. A smart choice of brackets can make a dramatic difference. Moreover, associativity allows us to compose a sequence of linear transformations – rotation, scaling, shearing – into a single transformation matrix before applying it to vectors.
结合律不仅是理论上的奇妙性质，它在实践中也非常重要。连乘多个矩阵时，你可以选择一种分组方式来大幅降低计算量。例如，若 A 是 10 × 100 矩阵，B 是 100 × 5 矩阵，C 是 5 × 50 矩阵，那么计算 (AB)C 需要 10×100×5 + 10×5×50 = 5000 + 2500 = 7500 次乘法，而计算 A(BC) 则需要 100×5×50 + 10×100×50 = 25000 + 50000 = 75000 次乘法。合理的括号安排能带来惊人的效率差异。此外，结合律让我们能将一系列线性变换——旋转、缩放、剪切——预先合并成一个变换矩阵，然后再作用于向量，从而提升图形渲染等应用的速度。
9. Additional Rule: The Distributive Law | 附加规则：分配律
the dimensions are compatible. It is important to note that because multiplication is not commutative, left‑distributivity and right‑ with associativity, the distributive law makes matrix algebra a rich and flexible system.
除两条核心规则之外，另一个基本性质是分配律。矩阵乘法对加法满足分配律：只要维度相容，就有 A(B + C) = AB + AC 以及 (A + B)C = AC + BC。值得留意的是，正因乘法不可交换，左分配律和右分配律是两条独立的规则，而它们都成立。分配律与结合律一起，使矩阵代数成为一个既丰富又灵活的系统。
10. Common Pitfalls and How to Avoid Them | 常见误区与对策
e. Always check dimensions before multiplying and never reorder products carelessly. Write out the steps explicitly, and when in doubt, test with small numerical matrices to see whether a property holds. Keep a checklist: (1) Are the dimensions compatible? (2) Did I respect the order? (3) Have I used associativity or distributivity correctly? By internalising the two fundamental rules – non‑commutativity and associativity – you will develop a robust foundation for all subsequent topics in linear algebra.
许多同学失分，往往是因为错误地将矩阵乘法当作可交换运算，随意“消去”项，或误以为 AB = 0 就意味着 A = 0 或 B = 0。请记住：若 AB = AC，不能直接推出 B = C，除非 A 可逆。相乘之前务必检查维度，切莫随意调换乘积顺序。写出清晰的运算步骤，感到不确定时，可以用小规模数字矩阵试验某一性质是否成立。建立一份检查清单：(1) 维度是否相容？(2) 顺序是否被正确保持？(3) 结合律与分配律的使用是否得当？真正内化这两条基本规则——不可交换性与结合律，你就能为后续所有线性代数内容打下坚实基础。
11. Summary Table of Key Rules | 关键规则总结表
| Property / 性质 | Expression / 表达式 | Holds? / 是否成立 |
|---|---|---|
| Commutativity of addition / 加法交换律 | A + B = B + A | Yes / 是 |
| Associativity of multiplication / 乘法结合律 | (AB)C = A(BC) | Yes / 是 |
| Commutativity of multiplication / 乘法交换律 | AB = BA | No (generally) / 否 (一般) |
| Distributive law / 分配律 | A(B+C) = AB + AC, (A+B)C = AC + BC | Yes / 是 |
This table highlights the contrast: matrix addition follows the familiar rules of arithmetic, but matrix multiplication introduces a critical twist – non‑commutativity – while still preserving associativity and distributivity.
上表突显了对比：矩阵加法遵循熟悉的算术规则，而矩阵乘法则引入了一个关键转折——不可交换性——同时仍保留了结合律和分配律。
Published by TutorHao | Mathematics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图