移至主內容

QWQ. 週一, 2026-03-23 12:25

(1) 若 \(\mathbf{v}_1, \dots, \mathbf{v}_n\) 為 linearly dependent,  
表示 \(\exists c_1, \dots, c_n\) 不全為 \(0\),s.t.
\[
c_1 \mathbf{v}_1 + \dots + c_n \mathbf{v}_n = \mathbf{0}_V
\]

則  
\[
T(c_1 \mathbf{v}_1 + \dots + c_n \mathbf{v}_n) = T(\mathbf{0}_V)
\]  
因 \(T\) 為 linear transformation,  
\[
c_1 T(\mathbf{v}_1) + \dots + c_n T(\mathbf{v}_n) = \mathbf{0}_W
\]  
∴ \(T(\mathbf{v}_1), \dots, T(\mathbf{v}_n)\) 為 linearly dependent.

故由 contrapositive method(反證法),  
知若 \(T(\mathbf{v}_1), \dots, T(\mathbf{v}_n)\) 為 linearly independent,
則 \(\mathbf{v}_1, \dots, \mathbf{v}_n\) 為 linearly independent.


(2) 若 \(T(\mathbf{v}_1), \dots, T(\mathbf{v}_n) \in W\) 為 linearly dependent,  
表示 \(\exists c_1, \dots, c_n\) 不全為 \(0\),s.t.
\[
c_1 T(\mathbf{v}_1) + \dots + c_n T(\mathbf{v}_n) = \mathbf{0}_W
\]

因 \(T\) 為 linear transformation,  
\[
T(c_1 \mathbf{v}_1 + \dots + c_n \mathbf{v}_n) = \mathbf{0}_W
\]  
∴ \(c_1 \mathbf{v}_1 + \dots + c_n \mathbf{v}_n \in N(T)\)
此時由假設 \(\mathbf{v}_1, \dots, \mathbf{v}_n\) 為 linearly independent,且 \(c_1, \dots, c_n\) 不全為 \(0\),知 \(c_1 \mathbf{v}_1 + \dots + c_n \mathbf{v}_n \neq \mathbf{0}_V\)
又 \(c_1 \mathbf{v}_1 + \dots + c_n \mathbf{v}_n \in \operatorname{Span}(\mathbf{v}_1, \dots, \mathbf{v}_n)\),  
∴ \(\operatorname{Span}(\mathbf{v}_1, \dots, \mathbf{v}_n) \cap N(T) \neq \{\mathbf{0}_V\}\)

故由 contrapositive method(反證法),  
知若 \(\operatorname{Span}(\mathbf{v}_1, \dots, \mathbf{v}_n) \cap N(T) = \{\mathbf{0}_V\}\),
則 \(T(\mathbf{v}_1), \dots, T(\mathbf{v}_n) \in W\) 為 linearly independent.