3-3. 生成與線性獨立

Define Linear Combination

V V 為一佈於 F F 的向量空間, S⊆V S \subseteq V , S S 中任意有限個向量 v1⃗ \vec{v_1} , ..., vk⃗ \vec{v_k} , F F 中任意有限個純量 α1 \alpha_1 , ..., αk \alpha_k , 稱 v⃗ \vec{v} = α1v1⃗ \alpha_1 \vec{v_1} + ... + αkvk⃗ \alpha_k \vec{v_k} 為 S S 之一線性組合(linear combination, LC)

Theorem of 行向量線性組合關係式

A A , B∈Fm×n B \in F^{m \times n} , A A 列等價於 B B , 則 A A 與 B B 之行向量線性組合關係式相同

∵A \because A 列等價於 B⇒ker B \Rightarrow ker (A A ) = ker ker (B B )
ex.A=[123413401455] ^{ex.} A = \begin{bmatrix} 1 & 2 & 3 & 4 \\ 1 & 3 & 4 & 0 \\ 1 & 4 & 5 & 5 \end{bmatrix} 列運算至 rref U=[101001100001] U = \begin{bmatrix} \color{red}{1} & 0 & 1 & 0 \\ 0 & \color{red}{1} & 1 & 0 \\ 0 & 0 & 0 & \color{red}{1} \end{bmatrix} , 其中 U U 的每個非 pivot 行皆可寫成 pivot 行(column ↓ \downarrow with 1 \color{red}{1} )的 LC

Define Span and Spanning Set

V V 為一佈於 F F 的向量空間, S⊆V S \subseteq V , span span (S S ) = { v⃗ \vec{v} | v⃗ \vec{v} 為 S S 的一組 LC }, 稱為 S S 的生成空間(span);
若 span span (S S ) = V V , 則稱 S S 生成 V V (S S spans V V )或 S S 為 V V 的生成集(spanning set)

ex.R2 ^{ex.} R^2 :

  • span span {(1 1 , 0 0 )} = { x x (1 1 , 0 0 ) | x∈R x \in R } = { (x x , 0 0 ) | x∈R x \in R } = x x 軸
  • span span {(0 0 , 1 1 )} = { y y (0 0 , 1 1 ) | y∈R y \in R } = { (0 0 , y y ) | y∈R y \in R } = y y 軸
  • span span {(1 1 , 0 0 ), (0 0 , 1 1 )} = { x x (1 1 , 0 0 ) + y y (0 0 , 1 1 ) | x x , y∈R y \in R } = { (x x , y y ) | x x , y∈R y \in R } = R2 R^2

→span \rightarrow span (S S ): set of all LC in S S

Properties of Span

  • span span (∅ \emptyset ) = { 0⃗ \vec{0} }; span span ({ 0⃗ \vec{0} }) = { 0⃗ \vec{0} } as well
  • A∈Fm×n A \in F^{m \times n} :
    • CS CS (A A ) = span span { a1⃗ \vec{a_1} , ..., an⃗ \vec{a_n} }, ax⃗ \vec{a_x} : A A 的行向量
    • RS RS (A A ) = span span { A1 A_1 , ..., Am A_m }, Ax A_x : A A 的列向量
  • span span (S S ) 為 V V 的子空間

    ∵span \because span (S S ) 補齊 S S 之封閉性

  • span span (S S ) 為包含 S S 之最小子空間 ⇒ \Rightarrow 所有包含 S S 的子空間之交集
  • S S is a subspace of V⇔span V \Leftrightarrow span (S S ) = S S
  • S1⊆S2⇒span S_1 \subseteq S_2 \Rightarrow span (S1 S_1 ) ⊆span \subseteq span (S2 S_2 )

    but ⇍ \nLeftarrow :
    ex.S1 ^{ex.} S_1 = {(1 1 , 0 0 )} ⊈S2 \nsubseteq S_2 = {(2 2 , 0 0 ), (0 0 , 1 1 )} ⇒span \Rightarrow span (S1 S_1 ) = x x 軸 ⊆span \subseteq span (S2 S_2 ) = R2 R^2

  • span span (S1∩S2 S_1 \cap S_2 ) ⊆span \subseteq span (S1 S_1 ) ∩ \cap span span (S2 S_2 )
  • span span (S1∪S2 S_1 \cup S_2 ) ⊇span \supseteq span (S1 S_1 ) ∪ \cup span span (S2 S_2 )

Theorem of Sum of Spans

V V 為一佈於 F F 的向量空間, S1 S_1 , S2⊆V S_2 \subseteq V , 若 W1 W_1 = span span (S1 S_1 ), W2 W_2 = span span (S2 S_2 ), 則 W1 W_1 + W2 W_2 = span span (S1∪S2 S_1 \cup S_2 )

ex.S1 ^{ex.} S_1 = {(1 1 , 0 0 )} ⇒W1 \Rightarrow W_1 = span span {(1 1 , 0 0 )} = x x 軸; S2 S_2 = {(0 0 , 1 1 )} ⇒W2 \Rightarrow W_2 = span span {(0 0 , 1 1 )} = y y 軸
⇒W1 \Rightarrow W_1 + W2 W_2 = span span {(1 1 , 0 0 ), (0 0 , 1 1 )} = R2 R^2

Corollary of Sum of Spans

V V 為一佈於 F F 的向量空間, ∀i \forall i = 1 1 , ..., k k , Si⊆V S_i \subseteq V , 若 Wi W_i = span span (Si S_i ), 則 ∑i=1kWi \displaystyle\sum_{i=1}^k W_i = span span (S1∪...∪Sk S_1 \cup ... \cup S_k )

Define Linear Dependent and Independent Sets

V V 為一佈於 F F 的向量空間, S⊆V S \subseteq V , ∃ \exists 有限個 vi⃗⊆S \vec{v_i} \subseteq S , i i = 1 1 , ..., k k

  • S S 為線性相依集(linear dependent set, LD) ⇔ \Leftrightarrow
    ∃ \exists 有限個不全為 0 0 的 ci⊆F c_i \subseteq F s.t. c1v1⃗ c_1 \vec{v_1} + ... + ckvk⃗ c_k \vec{v_k} = 0⃗ \vec{0}

    ∃vi⃗∈S \exists \vec{v_i} \in S s.t. 該 vi⃗ \vec{v_i} 可寫成其他 ∈S \in S 的向量之 LC

  • S S 為線性獨立集(linear independent set, LI) ⇔ \Leftrightarrow
    c1v1⃗ c_1 \vec{v_1} + ... + ckvk⃗ c_k \vec{v_k} = 0⃗⇒c1 \vec{0} \Rightarrow c_1 = c2 c_2 = ... = ck c_k = 00

    ∀vi⃗∈S \forall \vec{v_i} \in S 皆無法寫成其他 ∈S \in S 的向量之 LC

Properties of LD and LI

  • S1⊆S2 S_1 \subseteq S_2 ,
    • S2 S_2 : LI ⇒S1 \Rightarrow S_1 : LI

      S2 S_2 : LI ⇍S1 \nLeftarrow S_1 : LI

    • S1 S_1 : LD ⇒S2 \Rightarrow S_2 : LD
  • 0⃗∈S⇒S \vec{0} \in S \Rightarrow S : LD

    ∵c0⃗ \because c \vec{0} = 0⃗ \vec{0} , c∈F c \in F

  • v⃗≠0⃗⇒ \vec{v} \ne \vec{0} \Rightarrow { v⃗ \vec{v} }: LI

    ∵cv⃗ \because c \vec{v} = 0⃗⇒c \vec{0} \Rightarrow c = 0 0

  • ∅ \emptyset : LI
  • { u⃗ \vec{u} , v⃗ \vec{v} }: LD ⇔u⃗ \Leftrightarrow \vec{u} = cv⃗ c \vec{v} , or v⃗ \vec{v} = cu⃗ c \vec{u} , c∈F c \in F

Wronskian

f1 f_1 , f2 f_2 , ..., fn∈C(n−1) f_n \in C^{(n-1)} [a a , b b ]: 在 [a a , b b ] 上的 n−1 n - 1 次可微分函數所成的集合(即函數向量空間)
定義 W W (x x ) = ∣f1(x)f2(x)…fn(x)f1′(x)f2′(x)…fn′(x)⋮⋮⋱⋮f1(n−1)(x)f2(n−1)(x)…fn(n−1)(x)∣ \begin{vmatrix} f_1(x) & f_2(x) & \ldots & f_n(x) \\ f_1'(x) & f_2'(x) & \ldots & f_n'(x) \\ \vdots & \vdots & \ddots & \vdots \\ f_1^{(n-1)}(x) & f_2^{(n-1)}(x) & \ldots & f_n^{(n-1)}(x) \end{vmatrix}
稱為 f1 f_1 , f2 f_2 , ..., fn f_n 之 Wronskian

ex.f1 ^{ex.} f_1 (x x ) = 2x2 2x^2 , f2 f_2 (x x ) = 3x3 3x^3 , W W (x x ) = ∣2x23x34x9x∣ \begin{vmatrix} 2x^2 & 3x^3 \\ 4x & 9x \end{vmatrix} = 6x4 6x^4

Theorem of Wronskian

f1 f_1 , f2 f_2 , ..., fn∈C(n−1) f_n \in C^{(n-1)} [a a , b b ], ∃x0∈ \exists x_0 \in [a a , b b ] s.t. W W (x0 x_0 ) ≠0⇒f1 \ne 0 \Rightarrow f_1 , f2 f_2 , ..., fn f_n : LI

→ \rightarrow 若 W W (x x ) = 0 0 , 未必保證 f1 f_1 , f2 f_2 , ..., fn f_n : LD

ex.f1 ^{ex.} f_1 (x x ) = x2 x^2 , f2 f_2 (x x ) = x∣x∣ x|x| on [−1 -1 , 1 1 ] ⇒W \Rightarrow W (x x ) = ∣x2x∣x∣2x2∣x∣∣ \begin{vmatrix} x^2 & x|x| \\ 2x & 2|x| \end{vmatrix} = 0 0
但假設 c1x2 c_1 x^2 + c2x∣x∣ c_2 x|x| = 0 0 {c1+c2=0,if x=1c1−c2=0,if x=−1⇒c1 \begin{cases} c_1 + c_2 = 0, & \text{if } x = 1 \\ c_1 - c_2 = 0, & \text{if } x = -1 \end{cases} \Rightarrow c_1 = c2 c_2 = 0 0 , f1 f_1 , f2 f_2 : LI
→ \rightarrow 若定義域改成[0 0 , 1 1 ], 則 f1 f_1 , f2 f_2 : LD

Copyright© saberLiou all rights reserved.            last updated at 2019-09-04 13:27:20

results matching ""

    No results matching ""