Assignment 1

ENGM 2041 - Linear Algebra

September 28, 2010

1
1 T/F/N Questions
(a) T. They both have the same solution. This is the definition of a consistent
system.
(b) T. The three elementary row ops are:
• R1 ↔ R2
• R1 → 61 R1
• R1 → −5R2 + R1
(c) F. Creating a zero is the same as eliminating a variable.
(d) T. See the definition of REF in the class notes.
(e) F. The number of free parameters is given by n − rank(A). FOr this matrix,
that is 2.
(f) N. All that matters is the rank of the matrix and the number of variables. For
example, the following augmented matrix has inf many solutions:
 
1 2 3 4
(1.1)
0 1 0 2

(g) N. The system may be consistent or inconsistent.

(h) F. Gaussian elimination is more efficient for inconsistent systems.
(i) T. See class notes.
(j) T.

2
2 Supplementary Exercises 3a
Find the conditions on a such that the system has zero, one or infinitely
many solutions.
x + 2y − 4z = 4
3x − y + 13z = 2
4x + y + a2 z = a + 3

Since we are only interested in determining the type of solution for this system, it
is more efficient for us to use Gaussian elimination. The augmented matrix for this
system is:  
1 2 −4 4
3 −1 13 2 
2
4 1 a a+3

Row reduction will give:

 
1 2 −4 4
0 1 −25/7 10/7 
0 0 a2 − 9 a − 3

As long as a2 − 9 6= 0, the system will have a unique solution, since we would be able
to create a leading 1 in postion 3,3 meaning that the matrix would then have a rank
of 3.
If a = 3 then R3 will be a row of zeros, reducing both the rank of the augmented
and oefficient matrices to 2. Under this condition, there are infinitely many solu-
tions.
 
Finally, if a = −3, there is no solution, since R3 would be 0 0 0 −6 and the
rank of the augmented matrix would differ from that of the coefficient matrix.
Summary:

no solution if
 a = −3
number of solutions: unique solution if a=6 ±3

infinitely many solutions if a = 3


3
3 Exercises 1.5, Question 4
Solve the circuit of question 4.

From Kirchoff’s Current Law:

0 = I1 − I3 − I4
0 = I1 − I5 − I6
0 = I2 − I4 + I6
0 = I2 + I3 − I5

From Kirchoff’s Voltage Law:

10 = −10I3 + 10I4
20 = +10I3 + 10I5
10 = +10I5 − 10I6

We can place these into an augmented matrix and perform G-J elimination to
solve:
   
1 0 −1 −1 0 0 0 1 0 0 0 0 0 2
1 0 0
 0 −1 −1 0 

0 1 0 0 0 0
 1 

0 1 0
 −1 0 1 0 

0 0 0 0 0 0
 1/2

0 1 1
 0 −1 0 0  → 0 0 1 0 0 0
  3/2

0 0 −10 +10 0 0 10 0 0 0 1 0 0 3/2
   
0 0 +10 0 10 0 20 0 0 0 0 1 0 1/2
0 0 0 0 +10 −10 10 0 0 0 0 0 0 0

The currents in the circuit are:

I1 = 2A I2 = 1A
I3 = 1/2A I4 = 3/2A
I5 = 3/2A I6 = 1/2A

4
4 Exercises 1.6, Question 3
Balance the chemical equation:

k1 CO2 + k2 H2 O → k3 C6 H12 O6 + k4 O2

We know that atoms are neither created, descroyed, nor changed in a chemical reac-
tion. The number of each type of atom must therefore be the same on each side of
this stoichiometric equation.

C: k1 = 6k3
O: 2k1 + k2 = 6k3 + 2k4
H: 2k2 = 12k3

This can be re-written in an augmented matrix and solved with Gaussian elimina-
tion:    
1 0 −6 0 0 1 0 0 −1 0
2 1 −6 −2 0 → 0 1 0 −1 0
0 2 −12 0 0 0 0 1 −1/6 0

Since the rank of this matrix (3) is less than the number of variables (4), there must
be infinitely many solutions. The equations represented by this matrix can be solved
by assigning a free parameter:

k1 =t
k1 − k4 =0
k2 =t
k2 − k4 =0 →
k3 = 1/6t
k3 − 1/6k4 = 0
k4 =t

We can obtain a particular solution by selecting a value of t to that makes the

chemical coefficients all natural (positive, non-fraction) numbers. Choose t = 6 so
that the balanced equation is:

6CO2 + 6H2 O → 1C6 H12 O6 + 6O2

5
5 Curve Fitting
Find a polynomial of degree 3 that passes through y = sin(x), at x = π/4,
3π/4, 5π/4, 7π/4.
a) Write the augmented matrix that corresponds to the problem.
b) Use MATLAB to find coefficients of the polynomial.
c) Use MATLAB the graph bot sin(x) and the polynomial.

Part a)

y = a0 + a1 x + a2 x2 + a3 x3 (5.1)
√ √
The polynomial
√ must pass through
√ the ordered pairs: (π/4, 2/2), (3π/4, 2/2),
(5π/4, − 2/2) and 7π/4, − 2/2). This suggests that we can solve this as a system
of equations, since it requires that:
√
2/2 = a0 + a1 π/4 + a2 (π/4)2 + a3 (π/4)3
√
2/2 = a0 + a1 3π/4 + a2 (3π/4)2 + a3 (3π/4)3
√
− 2/2 = a0 + a1 5π/4 + a2 (5π/4)2 + a3 (5π/4)3
√
− 2/2 = a0 + a1 7π/4 + a2 (7π/4)2 + a3 (7π/4)3

We can then write this system of equations in an augmented matrix, thus satisfying
the requirement for part a of this question:
√
1 π/4 (π/4)2 (π/4)3
 
√ 2/2
1 3π/4 (3π/4)2 (3π/4)3
√2/2 


2 3
1 5π/4 (5π/4) (5π/4) − 2/2
√
1 9π/4 (7π/4)2 (7π/4)3 − 2/2

6
Part b)

rref(A) in MATLAB produces the output:

 
1 0 0 0 −.707
0 1 0 0 2.63 
  (5.2)
0 0 1 0 −1.14
0 0 0 1 1.22

We can now write out the polynomial:

y = −.707 + 2.63x − 1.14x2 + 1.22x3 (5.3)

Part c)

1 S a mp l e p oi nts
y = si n (x)
y = p ol yn omi al

0.5
y−axis

−0.5

−1
0 1 2 3 4 5 6
x−axis

Figure 1: The two functions, y = sin x (black line) and y = −.707 + 2.63x − 1.14x2 +
1.22x3 (red line) plotted. The polynomial fits through the four sample points (black
dots) as was required for the assignment question.