Linear Algebra Exercises-n-Answers.pdf
Linear Algebra Exercises-n-Answers.pdf
Linear Algebra Exercises-n-Answers.pdf
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
142 <strong>Linear</strong> <strong>Algebra</strong>, by Hefferon<br />
and this second description exactly says this.<br />
⎛<br />
⎞<br />
N (f) ⊥ = [{ ⎝ 1 2⎠}]<br />
3<br />
(b) The generalization is that for any f : R n → R there is a vector ⃗ h so that<br />
⎛ ⎞<br />
v 1<br />
⎜<br />
⎝<br />
⎟<br />
. ⎠ ↦−→ f<br />
h 1 v 1 + · · · + h n v n<br />
v n<br />
and ⃗ h ∈ N (f) ⊥ . We can prove this by, as in the prior item, representing f with respect to the<br />
standard bases and taking ⃗ h to be the column vector gotten by transposing the one row of that<br />
matrix representation.<br />
(c) Of course,<br />
( ) 1 2 3<br />
Rep E3 ,E 2<br />
(f) =<br />
4 5 6<br />
and so the nullspace is this set.<br />
⎛<br />
N (f){ ⎝ v ⎞<br />
1<br />
v 2<br />
⎠ ∣ v 3<br />
That description makes clear that<br />
⎛<br />
⎞<br />
⎝ 1 2⎠ ,<br />
3<br />
( ) ⎛<br />
1 2 3 ⎝ v ⎞<br />
(<br />
1<br />
v<br />
4 5 6 2<br />
⎠ 0<br />
= }<br />
0)<br />
v 3<br />
⎛<br />
⎝ 4 5<br />
6<br />
⎞<br />
⎠ ∈ N (f) ⊥<br />
and since N (f) ⊥ is a subspace of R n , the span of the two vectors is a subspace of the perp of the<br />
nullspace. To see that this containment is an equality, take<br />
⎛ ⎞<br />
⎛ ⎞<br />
M = [{ ⎝ 1 2⎠}] N = [{ ⎝ 4 5⎠}]<br />
3<br />
6<br />
in the third item of Exercise 23, as suggested in the hint.<br />
(d) As above, generalizing from the specific case is easy: for any f : R n → R m the matrix H representing<br />
the map with respect to the standard bases describes the action<br />
⎛ ⎛<br />
⎜<br />
⎝<br />
⎞<br />
v 1<br />
. ⎟<br />
. ⎠<br />
v n<br />
f<br />
↦−→<br />
⎜<br />
⎝<br />
⎞<br />
h 1,1 v 1 + h 1,2 v 2 + · · · + h 1,n v n<br />
.<br />
⎟<br />
.<br />
⎠<br />
h m,1 v 1 + h m,2 v 2 + · · · + h m,n v n<br />
and the description of the nullspace gives that on transposing the m rows of H<br />
⎛ ⎞ ⎛ ⎞<br />
h 1,1<br />
h m,1<br />
⃗ h 1,2<br />
h1 = ⎜ ⎟<br />
⎝ . ⎠ , . . .⃗ h m,2<br />
h m = ⎜ ⎟<br />
⎝ . ⎠<br />
h 1,n h m,n<br />
we have N (f) ⊥ = [{ ⃗ h 1 , . . . , ⃗ h m }]. (In [Strang 93], this space is described as the transpose of the<br />
row space of H.)<br />
Three.VI.3.25 (a) First note that if a vector ⃗v is already in the line then the orthogonal projection<br />
gives ⃗v itself. One way to verify this is to apply the formula for projection into the line spanned by<br />
a vector ⃗s, namely (⃗v ⃗s/⃗s ⃗s) · ⃗s. Taking the line as {k · ⃗v ∣ k ∈ R} (the ⃗v = ⃗0 case is separate but<br />
easy) gives (⃗v ⃗v/⃗v ⃗v) · ⃗v, which simplifies to ⃗v, as required.<br />
Now, that answers the question because after once projecting into the line, the result proj l (⃗v) is<br />
in that line. The prior paragraph says that projecting into the same line again will have no effect.<br />
(b) The argument here is similar to the one in the prior item. With V = M ⊕ N, the projection of<br />
⃗v = ⃗m+⃗n is proj M,N (⃗v ) = ⃗m. Now repeating the projection will give proj M,N (⃗m) = ⃗m, as required,<br />
because the decomposition of a member of M into the sum of a member of M and a member of N<br />
is ⃗m = ⃗m + ⃗0. Thus, projecting twice into M along N has the same effect as projecting once.