Naneh Apkarian

Math 20F
Winter 2013

Midterm #2 Solutions
Here is a solution key for the second midterm. The solutions presented here are more
complete and thorough than your responses needed to be - in order to help you better
understand the answers.

You have two scores on your exam. The raw score, and the adjusted one. The scores
were adjusted by the addition of 5 points.
unadjusted average: 18/30
adjusted average: 23/30

1. Find the inverse of the following matrix:

 
0 1 2
A = 1 0 3
4 −3 8

Setup the augmented matrix:  

A | I
And row reduce until the left side is the identity. Then you will have
I | A −1
 

Final solution:
− 29 − 23
 
7
 
−1
 
A  −2
= 4 −1 

 
3 1
2 −2 2


2.

(a) Let   
x
W= : x ≤ 0 and y ≥ 0 .
y
Determine whether or not W is a subspace of R2 .

(b) Let U and W be two subspaces of a vector space V. Prove that the intersection of
U and W, written U ∩ W, is also a subspace of V. Find an example in R2 which
shows that the union U ∪ W is not, in general, a subspace.

1
(a) Since W is a subset of R2 , there are only three criteria we need concern ourselves with:
 
0
(a) Is in W?
0

(b) Is W closed under addition? i.e. if u, v ∈ W, is (u + v) ∈ W?

(c) Is W closed under scalar multiplication? i.e. if u ∈ W, is cu ∈ W for all c ∈ R?
If all three are true, then W is indeed a subspace of R2 . If any one of them is false, then W
is not a subspace of R2 . As it happens, the third condition is violated. Here is an example
of why:  
−1
−1 ≤ 0, and 1 ≥ 0 ⇒ u = ∈W
1
But −1 ∈ R, so what about (−1)u?

  
−1 1
(−1)u = −1 = ∈
/ W.
1 −1
Thus W is not a subspace. 

(b) This problem was in your homework. To refresh your memory:

intersection of sets: U ∩ W = {v : v ∈ U AND v ∈ W }

union of sets: U ∪ W = {v : v ∈ U OR v ∈ W }
As in part (a), we need to check three conditions. Both U, W are subspaces of V, which
tells us that 0 ∈ U and 0 ∈ W, which means 0 ∈ U ∩ W. If u, w ∈ U ∩ W, then by
definition u, w ∈ U and u, w ∈ W. Since U, W are subspaces, they are closed under
addition - meaning u + w ∈ U and u + w ∈ W, implying u + w ∈ U ∩ W. Any single
element u ∈ U ∩ W is in U and in W, so that cu ∈ U and cu ∈ W for any scalar c, meaning
of course that cu ∈ U ∩ W. Thus theintersection
 U ∩ W is a subspace
  of V.
x 0
For a counterexample, let U = { : x ∈ R} and W = { : y ∈ R}. Then
0 y
   
1 0
u= ∈ U ⊂ U∪W w= ∈ W ⊂ U∪W
0 1
But  
1
u+w = ∈
/ U or W ⇒ (u + w) ∈
/ U ∪ W.
1


3. Let W = span{v1 , v2 , v3 , v4 , v5 }. Find a subset of of vectors to be a basis of W.

Row reduce the augmented matrix [v1 |v2 |v3 |v4 |v5 ] to determine which columns are lin-
early independent. There is more than one correct answer here, but you needed three
linearly independent vectors. One answer is {v1 , v2 , v3 }. 

2
4. Let A be a 4 × 5 matrix and let B denote its echelon form.

(a) Find bases for Col ( A) and Row( A) using the labeling given.

(b) Find Rank( A).

(c) Find a basis for Null ( A).

(a) We can clearly see that B has three pivots, in columns r10 , r20 , r40 . By theorem, the corre-
sponding columns of A form a basis for the columnspace, that is

{r1 , r2 , r3 } is a basis for Col ( A).

Those pivots are in rows c10 , c20 , c30 . Similarly, the corresponding rows of A form a basis for
the rowspace. That is,
{c1 , c2 , c3 } is a basis for Row( A).


(b) There are a few ways to see this.

rank( A) = dim(Col ( A)) = # vectors in basis for Col ( A) = 3

rank( A) = dim(Col ( A)) = dim( Row( A)) = # vectors in basis for Row( A) = 3
rank( A) = dim(Col ( A)) = # pivot columns of A = 3
The answer was 3. 

(c) To find a basis for Null ( A), we must first establish Null ( A) = { x ∈ R4 : Ax = 0},
and discover what it looks like. Since B is already in reduced echelon form, and we are
concerned with the homogenous equation, we may jump straight in. A solution x to
Ax = 0 is of the form
       
x1 x1 + x3 + x5 = 0 − x3 − x5 −1 −1
 x2 
  x2 − 2x3 + 3x5 = 0 2x3 − 3x5 
 
 2 
 
 −3
 
x =  x3  with
  x3 free ⇒x=  x3  = x3  1  + x5  0 
    
 x4  x4 − 5x5 = 0  5x5   0   5 
x5 x5 free x5 0 1

Thus    

 −1 −1 

  2   −3

    


 1  ,  0  is a basis for Null ( A).
   


  0   5  
 
0 1
 


5. (v.1) The set B = {1 + t, 1 + t2 , t + t2 } is a basis for P2 . Find the B -coordinate vector

of p(t) = 6 + 3t − t2 .

3
The coordinate vector of p(t) with respect to the basis B is the vector
 
c1
[ p(t)]B = c2  such that p ( t ) = c1 (1 + t ) + c2 (1 + t2 ) + c3 ( t + t2 ).
c3
In our case, that means
6 + 3t − t2 = c1 (1 + t) + c2 (1 + t2 ) + c3 (t + t2 )
There are a few ways of seeing this, but the net result must be that
c1 + c2 = 6 c1 + c3 = 3 c2 + c3 = −1.
This can be rewritten     
1 1 0 c1 6
1 0 1  c2  =  3 
0 1 1 c3 −1
Whichever way you attack this problem, the correct answer is
 
5
[ p(t)]B =  1 
−2


5. (v.2) The set B = {1 + t, 1 + t2 , t + t2 } is a basis for P2 . Find the B -coordinate vector

of p(t) = 6 − t + t2 .

The coordinate vector of p(t) with respect to the basis B is the vector
 
c1
[ p(t)]B = c2 
 such that p ( t ) = c1 (1 + t ) + c2 (1 + t2 ) + c3 ( t + t2 ).
c3
In our case, that means
6 − t + t2 = c1 (1 + t ) + c2 (1 + t2 ) + c3 ( t + t2 )
There are a few ways of seeing this, but the net result must be that
c1 + c2 = 6 c1 + c3 = −1 c2 + c3 = 1.
This can be rewritten     
1 1 0 c1 6
1 0 1  c2  =  −1
0 1 1 c3 1
Whichever way you attack this problem, the correct answer is
 
2
[ p(t)]B = 4 

−3


4
6. Define a linear transformation T : P3 → R3 by
 
p (0)
T ( p(t)) =  p(0)
p (0)

Find a basis for the kernel of T, and describe the range of T.

This is almost identical to a homework problem we had. A polynomial p ∈ P2 has the
form
p ( t ) = a0 + a1 t + a2 t2 + a3 t3
where the coefficients ai are real numbers. Thus
p (0) = a0 + a1 (0) + a2 (0)2 + a3 ( t )3 = a0 .
We see that an equivalent definition of the transformation T would be
   
a0 1
T ( p(t)) = a0 = a0 1 (since a0 is a scalar).
  
a0 1
Now we can begin to discuss the kernel and the range of T.
kernel ( T ) = { p ∈ P3 : T ( p) = 0}
    
 1 0 
= a0 + a1 t + a2 t2 + a3 t3 : a0 1 = 0
1 0
 
   
1 0
Clearly we see that a0 must be zero for a0 1 = 0 . Thus
  
1 0
kernel ( T ) = { a1 t + a2 t2 + a3 t3 : ai ∈ R}.
A basis of this set would be {t, t2 , t3 }, as these are three linearly independent polynomials
and they span the kernel. As for the range,
range( T ) = {v ∈ R3 : T ( p) = v for some p ∈ P3 }
  
 p (0) 
=  p(0) : p ∈ P3
p (0)
 
   
 1 
= a0 1 : a0 ∈ R
1
 
 
 1 
= span 1
1
 

Since the range of T is the span of a single vector, it is actually a line in R3 .