You are on page 1of 2

Tutorial 5 Regression and Interpolation

1.

Use linear and quadratic regression to find the value of y(18). Find the value of
correlation coefficient (r) in both cases.
X
5
10
15
20
25
30
35
40
45
50
Y
17
24
31
33
37
37
40
40
42
41
Answer: Linear Regression A = 20.6, B = 0.4945, y(18) = 29.501, r = 0.9157
Quadratic Regression A = 11.7667, B = 1.3779, C = - 0.0161, y(18) = 31.3525, r = 0.9899

2.

Bessel functions, J, are used extensively in engineering and are usually compiled in
standard mathematical tables. For example:
x
1.8
2
2.2
2.4
2.6
J(x) 0.5815 0.5767 0.556 0.5202 0.4708
Using linear OR quadratic regression, whichever is more accurate, find the value of
J(2.1).
Answer: Linear Regression - A = 0.8467, B = - 0.1390, r = -0.9532
Quadratic Regression A = - 0.0398, B = 0.6806, C = - 0.1863, r = 0.9999
J(2.1) = 0.5679

3.

Given the following data:


x 2
3
4
5
y 5.1 7.6 10.4 12.0
a)

b)

Fit a straight line to the x and y values of the table using the linear regression
method. Find the standard estimate of error, Sy/x, the coefficient of
determination, r2 and the value of y when x = 3.5. Show all your workings.
Redo the regression and force the intercept to zero. Recalculate the value of y
when x = 3.5. (Hint: the linear best fit will be given by the equation y = ax).

Answer: (a) A = 0.55, B = 2.35, Sy/x = 0.3969, r2 = 0.9887, y(3.5) = 8.775


(b) a = 2.4926, y(3.5) = 8.7241

4.

Generate a Newton polynomial of order 3 and Lagrange polynomial of order 3 to


approximate f(2.5) for the following data. Given that the true value is 0.86207,
compute the true error in each case.
x
f(x)

1
0.5

2
0.8

3
0.9

4
0.941176

Answer: Newton Polynomial f(2.5) = 0.8662, true error = 0.48%


Lagrange Polynomial f(2.5) = 0.8662, true error = 0.48%

5
0.96154

5.

The following data were obtained from an experiment on the non-linearity effect of
Ohms Law:
i
-2
-1
-0.5
0.5
1
2
V -637 -96.5 -20.5 20.5
96.5
637
From the last four points of the data, find the value of V(0.1) using a third order
Newton Polynomial . Answer: f(0.1) = 2.324

6.

The followings are data of temperature vs. depth from a nuclear reactor.
Depth (m)
0.0 1.0 2.0 3.0
o
Temperature ( C) 70 55 13 10
Using a 3rd order Lagrange Polynomial method,
a) Obtain an equation for temperature as a function of depth in the form of
T = Ad3+Bd2+Cd+E where A, B, C and E are constant numbers.
b) Find the value of the temperature at depth of 2.5 m.
c) Find the depth where the temperature is 40 oC.
Answer: a) T = 11d3 46.5d 2+20.5d+70, b) T(2.5) = 2.5 oC,
c) when T = 40 oC , d = 1.3499m

7.

As an engineering student, you are required to do an experiment which involves


measuring the electric current I at certain times t, by specifying different values of
voltage V across an inductor. The lab manual states that you should measure the
current six times, at t = 0, 0.125, 0.25, 0.3, 0.375 and 0.5. At the end of the lab session,
however, you realized that you only took five readings, as tabulated below:
Time, t (s)
Voltage, V (Volt)
Current, I (Ampere)

0
0
0

0.125
2.5
6

0.25
3.2
8

0.375
2
5

0.5
0
0

a. Use quadratic regression to calculate the current I at t = 0.3s and determine how
accurate is the result.
b. Given that V = 3 Volts at time t = 0.3s, apply the well-known Faradays Law,
V = L dI dt and the equation obtained from part (a) to find the inductance L (in
unit Henrys).
Answer: a) A = 0.1429, B = 60.9143, C = - 123.4286, I(0.3) = 7.3086, r = 0.9940
b) L = 0.2283 Henrys

You might also like