ENGSCI 211

QUESTION / ANSWER BOOKLET

THE UNIVERSITY OF AUCKLAND

SEMESTER ONE 2021
Campus: City

MATHEMATICAL MODELLING 2

Test 2

(Time allowed: 60 minutes plus 3 minutes reading time)

University email: @aucklanduni.ac.nz

NOTE: Answer all questions. Show all your working.

Answers for each question should be written in the space provided directly after the
question. Do NOT write in the QR code area.
Write in dark ink suitable for reproduction. Ensure ink does not bleed through the paper.
Calculators are not permitted.
A separate sheet contains the Formula Sheet. This will not be collected, so blank space
can be used for rough working.
Overflow pages are provided at the end of this booklet. Please indicate clearly in the
regular answer space if your answer is continued in these pages. Additional paper will
only be provided when you have exhausted all overflow pages.
DO NOT REMOVE ANY PAGES FROM THIS BOOKLET.

Page 1 of 12
 ENGSCI 211
QUESTION / ANSWER BOOKLET

========== DO NOT WRITE IN THIS MARGIN ==========

Question 1 (5 marks)
RR
(a) Consider evaluating a double integral R f (x, y) dA, where the integration region R is
shown in the figure below with bounding functions labelled.
y
4 x2
y= 2
+2

2 R
y =4−x
x2
y= 2
1 y =2−x

x
1 2 3 4
Identify a suitable change of variables for evaluating this integral, and state the integral
bounds in the new coordinate system. (3 marks)

empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty

Page 2 of 12
 ENGSCI 211
========== DO NOT WRITE IN THIS MARGIN ========== QUESTION / ANSWER BOOKLET

RR
(b) A second, separate, double integral T g(x, y) dA has an integration region T that
requires a change of variables to the coordinate system u = xy, v = xy 2 .
Evaluate the Jacobian determinant for this change of variables. (2 marks)

empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty

Page 3 of 12
 ENGSCI 211
QUESTION / ANSWER BOOKLET

========== DO NOT WRITE IN THIS MARGIN ==========

Question 2 (7 marks)
(a) Consider the following scenario:
A wood processing plant is currently producing planks that are 2m in length. To upkeep
quality standards of the final product, the length of planks produced are regularly
monitored by measuring a sample of 30 planks every hour. Over the length of a day’s
production, the p-value of one-sample t-tests show a clear decreasing trend, from 0.8 to
0.02.
Explain what might be happening. Should the processing plant be concerned about this
observation? (2 marks)

(b) For a simple linear regression model of the form y = β0 + β1 x, a t-test on the regression
coefficient of β1 resulted in a p-value of 0.62.
(i) With respect to the relevant null hypothesis, interpret the p-value of this t-test.
(2 marks)

(ii) Give an interpretation of this result in terms of the effect on y of a 1-unit change
in x. (1 mark)

Page 4 of 12
 ENGSCI 211
========== DO NOT WRITE IN THIS MARGIN ========== QUESTION / ANSWER BOOKLET

(c) In the context of machine learning, explain briefly what is meant by a model that is
overfitted. What might we expect to observe when we use such a model for prediction?
(2 marks)

Page 5 of 12
 ENGSCI 211
QUESTION / ANSWER BOOKLET

========== DO NOT WRITE IN THIS MARGIN ==========

Question 3 (7 marks)
The following questions relate to the analysis given in
Appendix A: Concrete Compression.

(a) Comment on the exploratory analysis conducted on this dataset. (3 marks)

(b) With reference to an appropriate plot, comment on whether compression.lm satisfies

the Normality assumption. (1 mark)

WARNING: Marks will be deducted for answers that are not specifically asked
for in the questions. Please keep answers concise. Do NOT do brain dumps.
(c) Assuming that the relevant assumptions are satisfied, write a brief Executive Sum-
mary of the main conclusions you would draw from this analysis. (2 marks)

Page 6 of 12
 ENGSCI 211
========== DO NOT WRITE IN THIS MARGIN ========== QUESTION / ANSWER BOOKLET

(d) State another method by which this analysis could have been conducted. (1 mark)

Page 7 of 12
 ENGSCI 211
QUESTION / ANSWER BOOKLET

========== DO NOT WRITE IN THIS MARGIN ==========

Question 4 (9 marks)
The following questions relate to the analysis given in Appendix B: Planets.

(a) What is the difference between the models planets.lm2 and planets.lm3?
Why was the model planets.lm3 rejected? (2 marks)

(b) Write down the equation of the model fitted in planets.lm2. (2 marks)

(c) Give a prediction for the possible range of locations of the asteroid belt (planet order 5)
in another solar system with the same size sun as ours, as given by the model we have
fitted to the observed data. (1 mark)

Page 8 of 12
 ENGSCI 211
========== DO NOT WRITE IN THIS MARGIN ========== QUESTION / ANSWER BOOKLET

WARNING: Marks will be deducted for answers that are not specifically asked
for in the questions. Please keep answers concise. Do NOT do brain dumps.
(d) Assuming that the relevant assumptions are satisfied, write a brief Executive Sum-
mary of the main conclusions you would draw from this analysis. (4 marks)

Page 9 of 12
 ENGSCI 211
QUESTION / ANSWER BOOKLET

========== DO NOT WRITE IN THIS MARGIN ==========

Question 5 (7 marks)
(a) Write down the permutation matrix P that performs the following transformation:
   
1 3 1 2 3 1
 2 3 1  7→  5 5 1 
5 5 1 1 3 1
(2 marks)

empty
empty
empty
empty
empty
empty
empty
empty
empty
empty

(b) Find the PLU factorisation of  

2 −1 4
 −6 3 1 
4 −3 3 (5 marks)

empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty

Page 10 of 12
 ENGSCI 211
========== DO NOT WRITE IN THIS MARGIN ========== QUESTION / ANSWER BOOKLET

empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty

Page 11 of 12
 ENGSCI 211
QUESTION / ANSWER BOOKLET

========== DO NOT WRITE IN THIS MARGIN ==========

Spare answer page (this will be marked)
Please use the blank space below if you run out of space inside the answer booklet.
Clearly state the number of the question(s) to which your work here relates.

empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty
empty

Page 12 of 12
APPENDIX ENGSCI 211

THE UNIVERSITY OF AUCKLAND

SEMESTER ONE 2021
Campus: City

ENGINEERING SCIENCE

Mathematical Modelling 2

TEST 2 APPENDICES

Contents
Page

APPENDIX A: Concrete Compression 2

APPENDIX B: Planets 4

Formula Sheet 9

Page 1 of 10
APPENDIX ENGSCI 211

APPENDIX A: Concrete Compression

Large, slender, high-strength reinforced concrete column are loaded eccentrically (i.e. off-center).
It is of interest to determine whether the compressive strength of columns loaded in such a
manner differed between moist conditions and lab-dry conditions, and to quantify that difference
if one existed.

The following variables were measured:

strength compressive strength (MPa)
condition condition experiment was conducted in, either Moist or LabDry

References: Withheld (Copyright Act 1994 s 49)

library(s20x)
compression.df = read.table("RCcompression.txt", header = TRUE)
compression.df$condition = factor(compression.df$condition)

boxplot(strength ~ condition, data = compression.df)

95
strength

90
85
80

LabDry Moist

condition

onewayPlot(strength ~ condition, data = compression.df)

Plot of `strength' by levels of `condition',
with TUKEY intervals (95%, pooled SDs)
95
strength

90
85
80

LabDry Moist

condition

Page 2 of 10
APPENDIX ENGSCI 211

compression.lm = lm(strength ~ condition, data = compression.df)

modcheck(compression.lm)

Sample Quantiles
Residuals

Fitted values Theoretical

Cook's Quantiles
Distance plot

0.30
6
Fitted values

Cook's distance
0.20
9

0.10
0.00

Residuals from lm(strength ~ condition)

summary(compression.lm)

##
## Call:
## lm(formula = strength ~ condition, data = compression.df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -9.658 -2.658 0.825 2.133 8.142
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 90.3917 1.1670 77.459 <2e-16 ***
## conditionMoist -0.7333 1.6503 -0.444 0.661
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 4.042 on 22 degrees of freedom
## Multiple R-squared: 0.008895, Adjusted R-squared: -0.03616
## F-statistic: 0.1974 on 1 and 22 DF, p-value: 0.6611

confint(compression.lm)

## 2.5 % 97.5 %
## (Intercept) 87.97153 92.811808
## conditionMoist -4.15593 2.689263

Page 3 of 10
APPENDIX ENGSCI 211

APPENDIX B: Planets
It is theorised that there is a mathematical relationship between how far a planet is away from
its local star (i.e. the sun in our solar system) and the order of the planet in that solar system
(e.g. Earth is order 3 in our solar system). It is of interest to determine this relationship for our
solar system, in order to help find exoplanets in other solar systems with similar stars to ours.

The following variables were recorded:

Name name of planet
Order order of planet (including the asteroid belt and Pluto)
Distance distance of planet from sun, in astronomical units (AU)
(Note 1 AU is the distance between the Sun and Earth)
planets.df = read.table("planets.txt", header = TRUE)
planets.df

## Name Order Distance

## 1 Mercury 1 0.387
## 2 Venus 2 0.723
## 3 Earth 3 1.000
## 4 Mars 4 1.524
## 5 AsteroidBelt 5 2.900
## 6 Jupiter 6 5.203
## 7 Saturn 7 9.546
## 8 Uranus 8 19.200
## 9 Neptune 9 30.090
## 10 Pluto 10 39.500

plot(Distance ~ Order, data = planets.df)

40
30
Distance

20
10
0

2 4 6 8 10

Order

Page 4 of 10
APPENDIX ENGSCI 211

planets.lm = lm(Distance ~ Order, data = planets.df)

modcheck(planets.lm)

Sample Quantiles
Residuals

Fitted values Theoretical

Cook's Quantiles
Distance plot
10
Fitted values

0.8
Cook's distance
0.6
1

0.4
0.2

9
0.0

Residuals from lm(Distance ~ Order)

planets.lm2 = lm(log(Distance) ~ Order, data = planets.df)

modcheck(planets.lm2)
Sample Quantiles
Residuals

Fitted values Theoretical

Cook's Quantiles
Distance plot
10
0.5

Fitted values
0.4
Cook's distance
0.3

8
2
0.2
0.1
0.0

Residuals from lm(log(Distance) ~ Order)

Page 5 of 10
APPENDIX ENGSCI 211

summary(planets.lm2)

##
## Call:
## lm(formula = log(Distance) ~ Order, data = planets.df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.188576 -0.079848 0.000433 0.095100 0.197265
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -1.53782 0.09268 -16.59 1.76e-07 ***
## Order 0.53693 0.01494 35.95 3.93e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1357 on 8 degrees of freedom
## Multiple R-squared: 0.9938, Adjusted R-squared: 0.9931
## F-statistic: 1292 on 1 and 8 DF, p-value: 3.928e-10

planets.lm3 = lm(log(Distance) ~ Order, data = planets.df[-10,])

summary(planets.lm3)

##
## Call:
## lm(formula = log(Distance) ~ Order, data = planets.df[-10, ])
##
## Residuals:
## Min 1Q Median 3Q Max
## -0.192887 -0.064724 -0.007704 0.086044 0.161162
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -1.58524 0.09125 -17.37 5.15e-07 ***
## Order 0.54987 0.01622 33.91 5.03e-09 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.1256 on 7 degrees of freedom
## Multiple R-squared: 0.9939, Adjusted R-squared: 0.9931
## F-statistic: 1150 on 1 and 7 DF, p-value: 5.028e-09

Page 6 of 10
APPENDIX ENGSCI 211

confint(planets.lm2)

## 2.5 % 97.5 %
## (Intercept) -1.7515344 -1.3240976
## Order 0.5024888 0.5713764

exp(confint(planets.lm2))

## 2.5 % 97.5 %
## (Intercept) 0.1735075 0.2660429
## Order 1.6528297 1.7707027

pred.df = data.frame(Order = 5)
predict(planets.lm2, pred.df, interval = "confidence")

## fit lwr upr

## 1 1.146847 1.046427 1.247267

exp(predict(planets.lm2, pred.df, interval = "confidence"))

## fit lwr upr

## 1 3.148251 2.847459 3.480818

predict(planets.lm2, pred.df, interval = "prediction")

## fit lwr upr

## 1 1.146847 0.8182739 1.47542

exp(predict(planets.lm2, pred.df, interval = "prediction"))

## fit lwr upr

## 1 3.148251 2.266584 4.372873

Page 7 of 10
APPENDIX ENGSCI 211

This page has been intentionally left blank for your rough working. It will not be marked.

Page 8 of 10
 ENGSCI 211 Formula Sheet Version: Semester One 2021

Trigonometric identities Univariate calculus

A+B A−B
sin2 A + cos2 A = 1 sin A + sin B = 2 sin cos function derivative function antiderivative
2 2 0 0 n xn+1
2 2 f (x)g(x) f (x)g(x) + f (x)g (x) x 6= −1
sec A = 1 + tan A A+B A−B n+1 , n
2 2
sin A − sin B = 2 cos sin f (x) f 0 (x)g(x)−f (x)g 0 (x) 1
cosec A = 1 + cot A 2 2 g(x) g(x)2 , g 6= 0 x ln |x|
A+B A−B x n
nxn−1 e cx 1 cx
cos A + cos B = 2 cos cos ce
sin(A ± B) = sin A cos B ± cos A sin B 2 2 1 f 0 (x)
A+B B−A x − x12 f (x) ln |f (x)|
cos(A ± B) = cos A cos B ∓ sin A sin B cos A − cos B = 2 sin sin ln |x| 1
sin x − cos x
2 2 x
tan A ± tan B loga x 1
cos x sin x
tan(A ± B) = x ln a
1 ∓ tan A tan B cx cx
2 sin A cos B = sin(A + B) + sin(A − B) e ce tan x ln |sec x|
x x
sin 2A = 2 sin A cos A 2 cos A cos B = cos(A + B) + cos(A − B) a a ln a, a > 0 cosec x ln |cosec x − cot x|
cos 2A = cos2 A − sin2 A 2 sin A sin B = cos(A − B) − cos(A + B) sin x cos x = ln tan x2

= 2 cos2 A − 1 cos x − sin x sec x ln |sec x + tan x|

2
tan x sec x = ln tan( π4 + x2 )
= 1 − 2 sin2 A
2 tan A eix + e−ix eix − e−ix cosec x − cosec x cot x cot x ln |sin x|
tan 2A = cos x = sin x = 1 1 a+x
1 − tan2 A 2 2i sec x sec x tan x a2 −x2 2a ln a−x , |x| < a
eix = cos x + i sin x e−ix = cos x − i sin x arcsin x √ 1 1 1 x−a

Page 9 of 10
1−x2 x2 −a2 2a ln x+a , |x| > a
√ 1 1 1 x
arccos x − 1−x 2 a2 +x2 a arctan a
Partial fractions 1 √ 1
arctan x 1+x2 a2 −x2
arcsin xa , |x| < a
√
√ 1 ln(x + x2 + a2 )
p(x) A B C a2 +x2
= + + + ... Integration by Parts: √
√ 1 ln x + x2 − a2 , |x| > a
(x − a)(x − b)(x − c) . . . x−a x−b x−c Z Z x2 −a2
dv du
p(x) A B1 B2 B3 u dx = u v − v dx
= + + + dx dx
(x − a)(x − b)3 x − a x − b (x − b)2 (x − b)3
p(x) A Bx + C Numerical methods for ODEs
= +
(x − a)(x2 + bx + c) x − a x2 + bx + c
E
yn+1 = yn + ∆t k1
Geometric series k1 + k2
IE
yn+1 = yn + ∆t
2
n−1 ∞
X 1 − rn X 1 k1 = f (tn , yn )
ark = a ark = a , |r| < 1 E


1−r 1−r k2 = f tn+1 , yn+1
k=0 k=0
ENGSCI 211
 ENGSCI 211 Formula Sheet Version: Semester One 2021

Laplace transforms Vector Calculus

 
i j k
 
grad f = ∇f div v = ∇ · v ∂ ∂ ∂
Z ∞ curl v = ∇ × v = det 
 ∂x ∂y ∂z


L {f (t)} = F (s) = e−st f (t) dt vx vy vz
0
( " #
ZZ ZZ ∂x ∂x
0 t<a ∂s ∂t
u(t − a) = f (x, y) dx dy = f (x(s, t), y(s, t)) |J(s, t)| ds dt J(s, t) = det ∂y ∂y
1 t≥a ∂s ∂t
( Z ∞
0 t 6= a
δ(t − a) = and δ(t − a) dt = 1 Polynomial approximation
undefined t = a −∞
( R∞
f (τ )g(t − τ ) dτ general case (x − a)2 00 (x − a)k (k)
(f ∗ g)(t) = t
R−∞ f (x) = f (a) + (x − a)f 0 (a) + f (a) + . . . + f (a) + . . .
f (τ )g(t − τ ) dτ if f (t) = g(t) = 0 for t < 0 2! k!
0

leading terms k th term

2 3
x x xk
ex 1+x+ 2! + 3! + . . . k! all x
function transform condition function transform condition 3 5 7 (−1)k x2k+1
1 1
sin x x − x3! + x5! − x7! + . . . (2k+1)! all x
1 s t s2 2 4 6 (−1)k x2k
n! 1 cos x 1 − x2! + x4! − x6! + . . . (2k)! all x
tn sn+1 n ≥ 0, integer eat s−a s>a 3 5 7
x2k+1
e−as −as sinh x x + x3! + x5! + x7! + . . . (2k+1)! all x
u(t − a) s a≥0 δ(t − a) e a≥0 2 4 6

Page 10 of 10
x2k
ω s cosh x 1 + x2! + x4! + x6! + . . . (2k)! all x
sin ωt 2
s +ω 2 cos ωt s2 +ω 2 n n(n−1)(n−2)...(n−k+1) k
ω s (1 + x) 1 + nx + n(n−1)2! x2 + . . . k! x |x| < 1
sinh ωt s2 −ω 2 s > |ω| cosh ωt s2 −ω 2 s > |ω|
2ωs s2 −ω 2
t sin ωt (s2 +ω 2 )2 t cos ωt (s2 +ω 2 )2 Fourier series

∞     
operation function transform X 2nπx 2nπx
f (x) = a0 + an cos + bn sin , f (x + T ) = f (x)
shift in s-domain eat f (t) F (s − a), a < s n=1
T T
shift in t-domain u(t − a)f (t − a) e−as F (s)
Z T /2
1
a0 = f (x) dx
1st t-derivative y 0 (t) sY − y(0) T −T /2

2nd t-derivative y 00 (t) s2 Y − sy(0) − y 0 (0)

Z T /2  
2 2nπx
an = f (x) cos dx
nth t-derivative y (n) (t) sn Y − sn−1 y(0) − · · · − y (n−1) (0) T −T /2 T
R τ =t Z T /2  
time integral τ =0
y(τ ) dτ s−1 Y 2 2nπx
R t=T bn = f (x) sin dx
periodic function f (t) = f (t + T ) (1 − e−sT )−1 t=0 e−st f (t) dt T −T /2 T
convolution f (t) ∗ g(t) F (s) G(s)
ENGSCI 211