3-2. 子空間

Define Subspace

(V V , +, ⋅ \cdot ): vector space,

  • W⊆V W \subseteq V
  • (W W , +, ⋅ \cdot ): vector space

則稱 W W 為 V V 之子空間(subspace)

Theorem of 子空間的充要條件

已知 V V : vector space, W⊆V W \subseteq V , W≠∅ W \ne \emptyset , 則下列敘述等價:

  1. W W 為 V V 的子空間
  2. ∀u⃗ \forall \vec{u} , v⃗∈W \vec{v} \in W , u⃗ \vec{u} + v⃗∈W \vec{v} \in W
  3. ∀c∈F \forall c \in F , ∀v⃗∈W \forall \vec{v} \in W , cv⃗∈W c\vec{v} \in W
  4. ∀c \forall c , d∈F d \in F , ∀u⃗ \forall \vec{u} , v⃗∈W \vec{v} \in W , cu⃗ c\vec{u} + dv⃗∈W d\vec{v} \in W

    Corollary of 2. & 3.

  5. ∀ci∈F \forall c_i \in F , vi∈W v_i \in W , i i = 1 1 , 2 2 , ..., k k , ∑i=1kcivi∈W \displaystyle\sum_{i=1}^k c_i v_i \in W

    Corollary of 4.

→ \rightarrow 在 W⊆V W \subseteq V 和 W≠∅ W \ne \emptyset 的條件下,只需驗證向量加法及純量積的封閉性即可證明 W W 為 V V 的子空間

Square of R 的三型子空間

  1. 原點: {(0 0 , 0 0 )} = { 0⃗ \vec{0} }: zero space → \rightarrow 點

    ∀ \forall vector spaces, ∃ \exists { 0⃗ \vec{0} }

  2. 過原點直線 → \rightarrow 線

    若非直線,則不具純量積封閉性

  3. x−y x-y 平面: R2 R^2 → \rightarrow 面

    V V 為 V V 的子空間

Theorem of 兩子空間交集

W1 W_1 , W2 W_2 are subspaces of V V , 則 W1∩W2 W_1 \cap W_2 is a subspace of V V

  1. ⊆ \subseteq
  2. ≠∅ \ne \emptyset
  3. 封閉性
proof:
  1. W1⊆V W_1 \subseteq V 且 W2⊆V W_2 \subseteq V , 則 W1∩W2⊆V W_1 \cap W_2 \subseteq V
  2. ∵0⃗∈W1∩W2 \because \vec{0} \in W_1 \cap W_2 , ∴W1∩W2≠∅ \therefore W_1 \cap W_2 \ne \emptyset
  3. ∀c \forall c , d∈F d \in F , ∀u⃗ \forall \vec{u} , v⃗∈W1∩W2 \vec{v} \in W_1 \cap W_2 ,
    ∵u⃗ \because \vec{u} , v⃗∈W1 \vec{v} \in W_1 且 W1 W_1 is a subspace of V V , ∴cu⃗ \therefore c\vec{u} + dv⃗ d\vec{v} = W1 W_1
    ∵u⃗ \because \vec{u} , v⃗∈W2 \vec{v} \in W_2 且 W2 W_2 is a subspace of V V , ∴cu⃗ \therefore c\vec{u} + dv⃗ d\vec{v} = W2 W_2
    ∴cu⃗ \therefore c\vec{u} + dv⃗ d\vec{v} = W1∩W2 W_1 \cap W_2

兩子空間聯集

W1 W_1 , W2 W_2 are subspaces of V V , W1∩W2 W_1 \cap W_2 不一定為 V V 的 subspace

反例:

V V = R2 R^2 has two subsapces W1 W_1 and W2 W_2 ,

  • W1 W_1 = { (x x , 0 0 ) | x∈R x \in R },

    x x 軸

  • W2 W_2 = { (0 0 , y y ) | y∈R y \in R },

    y y 軸

but W1∪W2 W_1 \cup W_2 is not a subspace of V V .

ex. ^{ex.} (1 1 , 0 0 ), (0 0 , 1 1 ) ∈W1∪W2 \in W_1 \cup W_2 , but (1 1 , 0 0 ) + (0 0 , 1 1 ) = (1 1 , 1 1 ) ∉W1∪W2 \notin W_1 \cup W_2

Theorem of 兩子空間聯集

W1 W_1 , W2 W_2 are subspaces of V V , W1∩W2⇔W1⊆W2 W_1 \cap W_2 \Leftrightarrow W_1 \subseteq W_2 or W2⊆W1 W_2 \subseteq W_1

Define 和空間 (Sum Space)

W1 W_1 , W2 W_2 are subspaces of V V , W1 W_1 + W2 W_2 = { w1⃗ \vec{w_1} + w2⃗ \vec{w_2} | w1⃗∈W1 \vec{w_1} \in W_1 , w2⃗∈W2 \vec{w_2} \in W_2 }, 稱為 W1 W_1 , W2 W_2 之和空間(sum space)

Theorem of 和空間為 V 的子空間

W1 W_1 , W2 W_2 are subspaces of V V , 則 W1 W_1 + W2 W_2 is a subspace of V V

proof:
  1. W1⊆V W_1 \subseteq V 且 W2⊆V W_2 \subseteq V , 則 W1 W_1 + W2⊆V W_2 \subseteq V
  2. ∵0⃗ \because \vec{0} = 0⃗ \vec{0} + 0⃗∈W1 \vec{0} \in W_1 + W2 W_2 , ∴W1 \therefore W_1 + W2≠∅ W_2 \ne \emptyset
  3. ∀c \forall c , d∈F d \in F , ∀u⃗ \forall \vec{u} , v⃗∈W1 \vec{v} \in W_1 + W2 W_2 ,
    u⃗ \vec{u} = u1⃗ \vec{u_1} + u2⃗ \vec{u_2} , u1⃗∈W1 \vec{u_1} \in W_1 , u2⃗∈W2 \vec{u_2} \in W_2
    v⃗ \vec{v} = v1⃗ \vec{v_1} + v2⃗ \vec{v_2} , v1⃗∈W1 \vec{v_1} \in W_1 , v2⃗∈W2 \vec{v_2} \in W_2
    ⇒cu⃗ \Rightarrow c\vec{u} + dv⃗ d\vec{v} = c c (u1⃗ \vec{u_1} + u2⃗ \vec{u_2} ) + d d (v1⃗ \vec{v_1} + v2⃗ \vec{v_2} ) = (cu1⃗ c\vec{u_1} + dv1⃗ d\vec{v_1} ) + (cu2⃗ c\vec{u_2} + dv2⃗ d\vec{v_2} ) ∈W1 \in W_1 + W2 W_2

Define 四個基本子空間

A∈Fm×n A \in F^{m \times n} ,

  • 核空間(kernel of A A , nullspace, N N (A A )):
    • ker ker (A A ) = { x⃗ \vec{x} : n×1 n \times 1 | Ax⃗ A\vec{x} = 0⃗ \vec{0} }
    • 齊次解集,收集 Ax⃗ A\vec{x} = 0⃗ \vec{0} 之 x⃗ \vec{x}
  • 行空間(column space):
    • CS CS (A A ) = { Ax⃗ A\vec{x} : m×1 m \times 1 | x⃗ \vec{x} : n×1 n \times 1 }
    • 收集 Ax⃗ A\vec{x} = y⃗ \vec{y} 之 y⃗ \vec{y}
  • 左核空間(left kernel of A A , left nullspace):
    • Lker Lker (A A ) = { x⃗ \vec{x} : 1×m 1 \times m | x⃗A \vec{x}A = 0⃗ \vec{0} }
    • 收集 x⃗A \vec{x}A = 0⃗ \vec{0} 之 x⃗ \vec{x}
  • 列空間(row space):
    • RS RS (A A ) = { x⃗A \vec{x}A : 1×n 1 \times n | x⃗ \vec{x} : 1×m 1 \times m }
    • 收集 x⃗A \vec{x}A = y⃗ \vec{y} 之 y⃗ \vec{y}

ex.A=[123112][???]=[00] ^{ex.} A = \begin{bmatrix} 1 & 2 & 3 \\ 1 & 1 & 2 \end{bmatrix} \begin{bmatrix} ? \\ ? \\ ? \end{bmatrix} = \begin{bmatrix} 0 \\ 0 \end{bmatrix} ⇒ \Rightarrow ker ker (A A ) ={[aa−a]∣a∈R} = \begin{Bmatrix} \begin{bmatrix} a \\ a \\ -a \end{bmatrix} | a \in R \end{Bmatrix} ;

ex.B=[120120][???]=[bb] ^{ex.} B = \begin{bmatrix} 1 & 2 & 0 \\ 1 & 2 & 0 \end{bmatrix} \begin{bmatrix} ? \\ ? \\ ? \end{bmatrix} = \begin{bmatrix} b \\ b \end{bmatrix} ⇒ \Rightarrow CS CS (B B ) ={[bb]∣b∈R} = \begin{Bmatrix} \begin{bmatrix} b \\ b \end{bmatrix} | b \in R \end{Bmatrix}

Theorem of 四個基本子空間

A∈Fm×n A \in F^{m \times n} ,

  1. ker ker (A A ) is a subspace of Fn×1 F^{n \times 1}
  2. CS CS (A A ) is a subspace of Fm×1 F^{m \times 1}

    ∀y⃗∈CS \forall \vec{y} \in CS (A A ) ⇔y⃗ \Leftrightarrow \vec{y} = Ax⃗ A\vec{x} , for some x⃗ \vec{x}

  3. Lker Lker (A A ) is a subspace of F1×m F^{1 \times m}
  4. RS RS (A A ) is a subspace of F1×n F^{1 \times n}
proof:
  1. ∵A⋅0⃗ \because A \cdot \vec{0} = 0⃗ \vec{0} , ∴0⃗∈ker \therefore \vec{0} \in ker (A A ) ⇒ker \Rightarrow ker (A A ) ≠∅ \ne \emptyset
    ∀c \forall c , d∈F d \in F , ∀x1⃗ \forall \vec{x_1} , x2⃗∈ker \vec{x_2} \in ker (A A ) ⇒Ax1⃗ \Rightarrow A\vec{x_1} = Ax2⃗ A\vec{x_2} = 0⃗ \vec{0}
    ⇒A \Rightarrow A (cx1⃗ c\vec{x_1} + dx2⃗ d\vec{x_2} ) = cAx1⃗ cA\vec{x_1} + dAx2⃗ dA\vec{x_2} = c⋅0⃗ c \cdot \vec{0} + d⋅0⃗ d \cdot \vec{0} = 0⃗ \vec{0}
    ⇒cx1⃗ \Rightarrow c\vec{x_1} + dx2⃗∈ker d\vec{x_2} \in ker (A A )

    3 同理

  2. 0⃗ \vec{0} = A⋅0⃗∈CS A \cdot \vec{0} \in CS (A A ) ⇒CS \Rightarrow CS (A A ) ≠∅ \ne \emptyset
    ∀c \forall c , d∈F d \in F , ∀y1⃗ \forall \vec{y_1} , y2⃗∈CS \vec{y_2} \in CS (A A ) ⇒y1⃗ \Rightarrow \vec{y_1} = Ax1⃗ A\vec{x_1} , y2⃗ \vec{y_2} = Ax2⃗ A\vec{x_2} , x1⃗ \vec{x_1} , x2⃗ \vec{x_2} : n×1 n \times 1
    ⇒cy1⃗ \Rightarrow c\vec{y_1} + dy2⃗ d\vec{y_2} = cAx1⃗ cA\vec{x_1} + dAx2⃗ dA\vec{x_2} = A A (cx1⃗ c\vec{x_1} + dx2⃗ d\vec{x_2} ) ∈CS \in CS (A A )

    4 同理

四個基本子空間 with Invertible Matrix

  1. ker ker (B B ) ⊆ker \subseteq ker (AB AB ), 當 A A is nonsingular 時, ker ker (B B ) = ker ker (AB AB )

    ABx⃗ AB\vec{x} = 0⃗→Bx⃗ \vec{0} \rightarrow B\vec{x} = 0⃗ \vec{0}

  2. Lker Lker (A A ) ⊆Lker \subseteq Lker (AB AB ), 當 B B is nonsingular 時, Lker Lker (A A ) = Lker Lker (AB AB )
  3. CS CS (AB AB ) ⊆CS \subseteq CS (A A ), 當 B B 為可逆時, CS CS (AB AB ) = CS CS (A A )
  4. RS RS (AB AB ) = RS RS (B B )

四個基本子空間 with Row and Column Equivalence

A A , B∈Fm×n B \in F^{m \times n} ,

  1. A A 列等價於 B B , 則
    • ker ker (A A ) = ker ker (B B )
    • RS RS (A A ) = RS RS (B B )
  2. A A 行等價於 B B , 則
    • Lker Lker (A A ) = Lker Lker (B B )
    • CS CS (A A ) = CS CS (B B )
Copyright© saberLiou all rights reserved.            last updated at 2019-09-04 17:09:58

results matching ""

    No results matching ""