mirror of
https://github.com/szdytom/LADRSolutions.git
synced 2025-10-19 16:30:16 +00:00
3B 22-23
Signed-off-by: szdytom <szdytom@qq.com>
This commit is contained in:
parent
8627c454df
commit
1e8fa70977
@ -1,4 +1,4 @@
|
||||
#import "../styles.typ": exercise_sol, tab, exercise_ref
|
||||
#import "../styles.typ": exercise_sol, tab, exercise_ref, math_numbering
|
||||
#import "../math.typ": null, range, LinearMap, span, restricted
|
||||
|
||||
#exercise_sol(type: "answer")[
|
||||
@ -427,3 +427,49 @@
|
||||
|
||||
#tab 代入上面的结果,得到 $dim V_U = dim null T + dim (U inter range T)$。
|
||||
]
|
||||
|
||||
#exercise_sol(type: "proof")[
|
||||
设 $U$ 和 $V$ 都是有限维向量空间,$S in LinearMap(V, W)$,$T in LinearMap(U, V)$. 证明:
|
||||
|
||||
$ dim null S T <= dim null S + dim null T $
|
||||
][
|
||||
#let TN = $restricted(T, N)$
|
||||
#show: math_numbering(true)
|
||||
令 $N = null S T$。由于 $S T in LinearMap(U, W)$,故 $N$ 是 $U$ 的子空间。设 $u_1, dots, u_m$ 是 $N$ 的一组基。根据线性映射引理(原书3.4),存在 $TN in LinearMap(N, V)$,使得 $TN u_i = T u_i$。设 $u in N$,根据零空间的定义(原书3.11),有 $S T u = 0$,故 $range TN subset.eq null S$,即
|
||||
|
||||
$ dim range TN <= dim null S $ <3B-c-range-TN-leq-null-S>
|
||||
|
||||
#tab 根据线性映射基本定理(原书3.21),有
|
||||
|
||||
$ dim N = dim null TN + dim range TN $ <3B-c-dim-N-eq-null-TN-plus-range-TN>
|
||||
|
||||
#tab 注意到
|
||||
|
||||
$ null TN = {u in N : TN u = 0} = {u in N : T u = 0} = N inter null T $ <3B-c-null-TN-eq-N-inter-null-T>
|
||||
|
||||
#tab 将@3B-c-range-TN-leq-null-S 和@3B-c-null-TN-eq-N-inter-null-T 代入@3B-c-dim-N-eq-null-TN-plus-range-TN,得到
|
||||
#show: math_numbering(false)
|
||||
|
||||
$ dim N <= dim (N inter null T) + dim null S $
|
||||
|
||||
#tab 由于 $N inter null T subset.eq null T$,根据子空间的维数(原书2.37),$dim (N inter null T) <= dim null T$,因此 $dim null S T <= dim null T + dim null S$。
|
||||
]
|
||||
|
||||
#exercise_sol(type: "proof")[
|
||||
设 $U$ 和 $V$ 都是有限维向量空间,$S in LinearMap(V, W)$,$T in LinearMap(U, V)$. 证明:
|
||||
|
||||
$ dim range S T <= min{dim range S, dim range T} $
|
||||
][
|
||||
首先证明 $dim range S T <= dim range S$。设 $u in U$,则 $S T u = S (T u) in range S$,故 $range S T subset.eq range S$,即$dim range S T <= dim range S$。
|
||||
|
||||
#let SI = $restricted(S, I)$
|
||||
#tab 现在证明 $dim range S T <= dim range T$。令 $I = range T$,则 $I$ 是 $V$ 的子空间。根据线性映射引理(原书3.4),存在 $SI in LinearMap(I, W)$,使得对于任意 $v in I$,有 $SI v = S v$。设 $u in U$,则 $T u in I$,因此 $S T u = SI (T u)$。故 $range S T = range SI T subset.eq range SI$,即 $dim range S T <= dim range SI$。
|
||||
|
||||
#tab 由于 $SI$ 是从 $I$ 到 $W$ 的线性映射,根据线性映射基本定理(原书3.21),有
|
||||
|
||||
$ dim I = dim null SI + dim range SI $
|
||||
|
||||
#tab 即 $dim I >= dim range SI$,又因为 $dim range S T <= dim range SI$,故 $dim range S T <= dim range T$。
|
||||
|
||||
#tab 综上所述,$dim range S T <= min{dim range S, dim range T}$。
|
||||
]
|
||||
|
Loading…
x
Reference in New Issue
Block a user