Solmidterm

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Solutions to the Midterm Exam, Math 214,

Spring 2020

Question 1. True or false. Give a reason or a counter-example

(a) If an R-vector space has a finite generating set, then it is finite dimensional.

(b) A generating subset in a finite dimensional R-vector space must consist of finitely many
vectors.

(c) If S is a finite set, and K is a field, then the vector space Fun(S, K) of all functions
from S to K is finite dimensional.

Answer 1. Statement (a) is True because every finite generating set contains a maximal
linearly independent subset and hence contains a basis.

Statement (b) is False since the set of all vectors in a vectors space is a spanning set. For
instance if we view V = R as an R-vector space, then V contains infinitely many elements
and they trivially genererate V .

Statement (c) is True since the collection of delta functions {δs }s∈S is a basis of Fun(S, K).
2

Question 2. Let V be a vector space over a field K, and let x, y ∈ V be two vectors, and
a, b ∈ K be two scalars. Show that

ax + by = bx + ay

if and only if a = b and/or x = y.

1
Answer 2. Since
ax + by = bx + ay
the existence of additive inverses for vector addition gives

ax + by − bx − ay = 0.

Commutativity of addition and distributivity of scaling and addition then give

(a − b)(x − y) = 0.

If a − b 6= 0 we can multiply both sides of the last identity by 1/(a − b) which gives x − y = 0.
2

Question 3. Which of the following subsets of vectors are vector subspaces. In each case
either check the subspace properties or point out a property that fails and explain why.

(a) In the real 2-space R2 the subset S ⊂ R2 of all vectors with integral coordinates:
  
a 2
S= ∈R a, b ∈ Z .
b

(b) in the complex space C∞ of all sequences (a1 , a2 , . . . , an , . . .) of complex numbers (with
the term-by-by term addition and scaling) the subset B ⊂ C∞ of all bounded sequences:
 
∞ ∞ there exists a positive real constant
B = (ai )i=1 ∈ C
c > 0 so that |ai | < c for all i

Answer 3. In part (a) S is not a subspace. It is closed under addition but it is not closed
under scaling. Specifically if we scale a vector with integral coordinates by a general real
number we will get a vector with non-integral coordinates. For instance
√ 1
  √ 
2
2 = .
0 0

In part (b) S is a subspace. To check this suppose a = (ai ) and b = (bi ) are two bounded
sequences of complex numbers and α is a real number.

• Since (ai ) is bounded we can find a positive real constant A so that |ai | < A for all
i = 1, 2, . . .. Similarly since (bi ) is bounded we can find a positive real constant B so
that |bi | < B for all i = 1, 2, . . ..

2
Consider the sum a + b. Since the sum of sequences is defined term by term it follows
that
a + b = (ai + bi )∞
i=1 .

But by the triangle inequality for the absolute value we have

|ai + bi | ≤ |ai | + |bi | < A + B,

for all i = 1, 2, . . .. Therefore a + b is a bounded sequence as well. This shows that


the sum in R∞ preserves the condition of being bounded.

• Since the scaling of a sequence is defined term by term we have that

αa = (α · ai )∞
i=1 .

Then by the multiplicativity of the absolute value we have

|α · ai | = |α| · |ai | < |α| · A

for all i = 1, 2, . . .. This shows that scaling in R∞ preserves the condition of being
bounded.

Question 4. Let Pol be the vector space of all polynomials with real coefficients in one
variable. Suppose that V ⊂ Pol is a vector subspace such that:
• For every k = 0, 1, 2, . . . , n the subspace V contains a polynomial of degree exactly k.
In other words for every k = 0, 1, 2, . . . , n we have a polynomial pk (x) ∈ V such that
pk (x) = ck xk + lower degree terms, and ck 6= 0.

• V does not contain any polynomials of degree > n.


Show that V nust be equal to the subspace Poln ⊂ Pol of polynomials of degree at most n.

Answer 4. By assumption V does not contain any polynomials of degree > n. Therefore
V ⊂ Poln . To show that V = Poln it suffices to check that V contains a set of polynomials
that spans Poln .
We are given polynomials p0 (x), p1 (x), . . . , pn (x) in V such that for every k = 0, 1, . . . , n
we have
pk (x) = ck xk + lower degree terms, and ck 6= 0.
We can use these polynomials to argue that V contains all monomials 1, x, x2 , . . . , xn .
We will argue by induction on n.

3
Base: n = 0. We need to show that 1 ∈ V . By assumption we know that we have a
polynomial p0 (x) ∈ V where
p0 (x) = c0 , and c0 6= 0.
1 1
Since V is a vector subspace we will have that p (x)
c0 0
∈ V But p (x)
c0 0
= 1 hence 1 ∈ V .

Step: Suppose that we know that if V contains polynomials p0 (x), . . . pn−1 (x) satisfying

pk (x) = ck xk + lower degree terms, with ck 6= 0.

for k = 1, . . . , n − 1, then V contains the monomials 1, x, . . . , xn−1 . Suppose in addition V


contains a polynomial pn (x) such that

pn (x) = cn xn + lower degree terms, with cn 6= 0.

We need to show that V contains the monomial xn .


Explicitly
pn (x) = cn xn + an−1 xn−1 + an−2 xn−2 · · · a1 x + a0 ,
and so
1 an−1 n−1 a1 a0
xn = pn (x) − x − ··· − x − .
cn cn cn cn
Since pn (x) ∈ V and by the inductive assumption 1, x, . . . , xn−1 it follows that the right
hand side is a linear combination of polynomials in V . Since V is a vector space this implies
xn ∈ V and completes the check.
2

Question 5. Let U ⊂ Mat2×2 (R) be the subspace of all symmetric matrices and V ⊂
Mat2×2 (R) be the subspace of all strictly upper triangular matrices:
  
a b
U= a, b, c ∈ R ,
b c
  
0 d
V = d∈R .
0 0

(a) Show that U ⊕ V = Mat2×2 (R).


 
1 1
(b) Decompose the matrix E = into a sum E = A + B
2 −1
with A ∈ U and B ∈ U

4
Answer 5. For part (a) consider the subspace W := U +V ⊂ Mat2×2 (R). Note that U ∩V = {0}.
Indeed, if X ∈ U ∩ V is a matrix which is both in U and V , then on one hand we have
 
a b
X= ,
b c

and on the other  


0 d
X= .
0 0
Therefore we must have b = d, and a = 0, b = 0, and c = 0. This shows that W = U ⊕ V . But
every matrix in U can be written uniquely as
       
a b 1 0 0 1 0 0
=a· +b· +c· .
b c 0 0 1 0 0 1

Therefore      
1 0 0 1 0 0
, ,
0 0 1 0 0 1
 
0 1
is a basis of U and so dim U = 3. Similarly, note that every matrix in V is a scaling of
0 0
and thus dim V = 1. Since W = U ⊕ V this imples that dim W = dim U + dim V = 3 + 1 = 4. But
dim Mat2×2 (R) is also equal to 4 and since W is a subspace we must have W = Mat2×2 (R). This
proves part (a).

For part (b) we need to solve the equation


     
1 1 a b 0 d
= + .
2 −1 b c 0 0

This is equivalent to 1 = a, 1 = b + d, 2 = b, and −1 = c, and so we get


     
1 1 1 2 0 −1
= + .
2 −1 2 −1 0 0

Question 6. Let S be a finite set and let V = (P(S), +, ·) be the power set of S considered as
a vector space over F2 where for A, B ⊂ S, and α ∈ F2 we have

A + B = A∆B = A ∪ B − A ∩ B
(
A, if α = 1,
α·A=
∅, if α = 0,

Suppose that X, Y , Z are subsets in S such that X 6⊂ Y ∪ Z, Y 6⊂ X ∪ Z, and Z 6⊂ X ∪ Y . Show


that X, Y , and Z are linearly independent when viewed as vectors in V .

5
Answer 6. We have to check that there is no non-trivial linear combination of X, Y , and Z which
is equal to 0 ∈ V . Since the coefficients in any linear combination can be equal to either 0 or 1 the
non-trivial linear combinations are
1 · X + 0 · Y + 0 · Z = X,
0 · X + 1 · Y + 0 · Z = Y,
0 · X + 0 · Y + 1 · Z = Z,
1 · X + 1 · Y + 0 · Z = X + Y, (1)
1 · X + 0 · Y + 1 · Z = X + Z,
0 · X + 1 · Y + 1 · Z = Y + Z,
1 · X + 1 · Y + 1 · Z = X + Y + Z.

Since in V the zero vector corresponds to the empty subset ∅ ⊂ S, we need to show that the
subsets in the (1) are never empty.
First note that ∅ is contained in every subset, and so the conditions X 6⊂ Y ∪ Z, Y 6⊂ X ∪ Z,
and Z 6⊂ X ∪ Y imply that none of X, Y , and Z can be empty.
Let us examine X + Y next. By definition X + Y = (X ∪ Y ) − (X ∩ Y ) consists of all points
in the unon of X and Y which do not belong simultaneoously in X and Y . But we know that
X 6⊂ Y ∪ Z so we know that there is a point x ∈ X which does not belong to Y and does not belong
to Z. Hence x 6∈ X ∩ Y and so x ∈ (X ∪ Y ) − (X ∩ Y ). This shows that (X ∪ Y ) − (X ∩ Y ) is
not empty or equivalently that X + Y 6= 0. The same reasoning shows that X + Z 6= 0 and that
Y + Z 6= 0.
Finally note that we chose x ∈ X such that x 6∈ Y and x 6∈ Z. Thus x ∈ X +Y = (X ∪Y )−(X ∩
Y ) but x 6∈ (X ∪ Y ) ∩ Z ⊃ (X + Y ) ∩ Z. Therefore x ∈ X + Y + Z = ((X + Y ) ∪ Z) − (X + Y ) ∩ Z).
This shows that X + Y + Z 6= ∅ or equivalently X + Y + Z 6= 0. 2

Question 7. Let V and W be real vector spaces with bases E= {e1 , e2, e3 } and F = {f1 , f2 }
0 1 2
respectively. Suppose that the linear map T : V → W has matrix . Find the matrix of
3 4 6
T in the bases E0 = {e1 , e1 + e2 , e1 + e2 + e3 } and F0 = {f1 , f1 + f2 }.

Answer 7. Write e01 , e02 , e03 for the elements of the basis E0 and f10 , f20 for the elements of the
elements of the basis F0 . To compute the matrix of T in these bases we need to compute the
coordinates of the vectors in the collection T (E0 ) in the basis F0 . Using the matrix of T in the bases

6
E and F we compute
T (e01 ) = T (e1 ) = 0 · f1 + 3 · f2
= 3f2 ,
T (e02 ) = T (e1 + e2 ) = T (e1 ) + T (e2 ) = (3f2 ) + (1 · f1 + 4 · f2 )
= f1 + 7f2 ,
T (e03 ) = T (e1 + e2 + e3 ) = T (e1 ) + T (e2 ) + T (e3 ) =
= (3f2 ) + (1 · f1 + 4 · f2 ) + (2 · f1 + 6 · f6 )
= 3f1 + 13f2 .

This gives the vectors T (E0 ) in terms of the basis F. To get expressions for these vectors in terms of
the basis F0 we need to solve for the vectors in F in terms of the vectors in F0 . This is straightforward:
since f10 = f1 and f20 = f1 + f2 we get f1 = f10 and f2 = −f10 + f20 . Substituting these expression in the
previous formulas gives
T (e01 ) = 3f2 = −3f10 + 3f20 ,
T (e02 ) = f1 + 7f2 = −6f10 + 7f20 ,
T (e03 ) = 3f1 + 13f2 = −10f10 + 13f20 .
Hence the matrix of T in the bases E0 and F0 is
 
−3 −6 −10
.
3 7 13
2

Question 8. Let V be a vector space over a field K and let f : V → K be a linear function
which is not identically zero. Consider the subspace U = {x ∈ V | f (x) = 0} and let a ∈ V be any
vector that does not belong to U .

(a) Show that for every vector v ∈ V the vector


f (v)
x=v− a
f (a)
is well defined and belongs to U .
(b) Show that U ⊕ span(a) = V .

Answer 8. For part (a) note that a 6∈ U means f (a) 6= 0 in K. Therefore we can divide by f (a)
in K and so the vector
f (v)
x=v− a
f (a)
is well defined. To check that this vector belongs to U we evaluate f on x:
 
f (v) f (v)
f (x) = f v − a = f (v) − f (a) = f (v) − f (v) = 0.
f (a) f (a)

7
This shows that x ∈ U .

For part (b) note that part (a) implies that any vector v ∈ V is equal to the sum

f (v)
v =x+ a,
f (a)

and that x ∈ U . Since (f (v)/f (a))·a is a scaling of a it belongs to span(a) and so V = U +span(a).
To check that this is a direct sum we need to check that U ∩ span(a) = {0}.
Suppose x ∈ U ∩ span(a). Then f (x) = 0 and x = αa for some α ∈ K. But then 0 = f (x) =
f (αa) = αf (a). Since f (a) 6= 0 it follows that we must have α = 0. This implies that x = 0 · a = 0
and so U ∩ span(a) = {0}.
2

You might also like