Problem Set

1. Calculate the value of the definite integral

Z 2
1
dx
1 x
correct to five decimal places using the following numerical integration methods:

• Trapezoidal Rule,
• Simpson’s One-Third Rule,
• Simpson’s Three-Eighths Rule,
• Weddle’s Rule.

Also, compute the error of approximation for each method by comparing the results
with the exact value of the integral.

2. Consider a probability density function (PDF) defined as:

1 2
f (x) = √ e−x /2 .
2π
Write a C program to approximate the integral:
Z 2
f (x) dx
−2

using the trapezoidal rule with different values of h = 0.1, 0.01, 0.001. Compare the
accuracy of the results with the known probability from the standard normal table.

3. Given the following conditions:

f (0) = 0
f (1) + f (2) = 10
f (3) + f (4) + f (5) = 65

find the value of f (4), assuming that the function f (x) is of the form

f (x) = a + bx + cx2 .

1
4. Find the smallest positive real root of the equation

xex − 2 = 0,

correct to four significant figures, using the method of false position (also known
as the regula falsi method).

5. Using the Newton–Raphson method, find the root of the equation

x+1
sin x − = 0,
x−1
correct to five significant figures. (Hints. An approximate value of the root, based on
the graphs of y = sin x and y = x+1 x−1
, is known to be −0.4.)

6. Find the positive root of the equation

xx + 5x − 1000 = 0,

correct to four significant digits.

7. Consider the equation

x5 − 3x4 + 2x3 − x2 + x = 3.

(a) Solve the equation using the Bisection method.

(b) Solve the equation using the Newton–Raphson method.

Also, determine how many real solutions the equation has.

8. Solve the equation

x = cos x
using the Bisection method and the Newton–Raphson method. Also, determine
how many real solutions the equation has.

Next, solve the equation

sin x = cos x
using the Bisection method and the Newton–Raphson method. Again, deter-
mine how many real solutions the equation has.

9. Consider the linear system:

x + 31 y = 1.33,
3x + y = 4.

Find the solutions assuming the following approximations of 31 :

1
3
= 0.3, 0.33, 0.333, 0.3333.

2
 For each approximation, compute the condition number
λmax
κ= ,
λmin
where λ denotes the eigenvalues of the matrix
 1
1 3
.
3 1

10. Consider the matrix  

1 112
1
 222 
A = 1
 332 .
 .. .... 
. . . 
1 T T2

a) Find the inverse of D = A′ A, when T = 5.

b) Find the eigenvalues of D for T = 50, 100, 200.
c) Find the determinant of D for T = 50, 100, 200.
λmax
d) Compute the condition number κ = for T = 50, 100, 200, where λ denotes
λmin
the eigenvalues of D.

11. Write a C program to numerically approximate the first derivative of the function

f (x) = e−x · sin(x2 ) + ln(1 + x2 )

at a given value of x, using the central difference formula:

f (x + h) − f (x − h)
f ′ (x) ≈ ,
2h
where h = 0.001.
Evaluate the numerical derivative for different values of h to examine the effect of step
size on accuracy. Use:
h = 10−2 , 10−4 , 10−6
Compare the outputs and comment on the numerical stability and convergence of the
derivative approximation.
Hints:

a) Define the function f (x) in C using the math library.

b) Prompt the user to enter a value of x.
c) Approximate the first derivative f ′ (x) using the central difference method.
d) Display the result with at least 6 decimal places of precision.

3
12. Consider the system of linear equations:
2x + 3y − z = 5,
4x + y + 2z = 9,
−6x − 2y + 3z = −17.

Write a C program that performs the following tasks:

a) Represent the system in matrix form AX = B, where:
   
2 3 −1 5
A= 4 1 2 , B =  9 .
−6 −2 3 −17
 
x
b) Solve for the vector X = y  using:

z
• Gaussian elimination,
• Matrix inversion (i.e., compute X = A−1 B if A−1 exists).
c) Output the values of x, y, and z correct to at least four decimal places.
13. Given the OLS estimator β̂ = (X ′ X)−1 X ′ Y and estimated residual variance:
1
σ̂ 2 = (Y − X β̂)′ (Y − X β̂),
n−k
write a C program to compute β̂ and the variance-covariance matrix:
Var(β̂) = σ̂ 2 (X ′ X)−1 .
where    
1 4 2.2
1 5 2.8
X=
1
, Y =
3.6 ,

6
1 7 4.5
14. Consider a logistic model:
1
P (y = 1|x) =
1+ e−(β0 +β1 x)
Given a small dataset of binary outcomes and a single predictor, write a C program
that:
a) Initializes β0 , β1 to 0.
b) Iteratively updates the parameters using the Newton–Raphson algorithm:
−1
β (t+1) = β (t) − H(β (t) ) ∇ℓ(β (t) ),


where ∇ℓ is the gradient and H is the Hessian.

4
 c) Stops when the change in parameters is below a tolerance level (e.g., 10−6 ).

15. Suppose you are given the following dataset, which represents the number of customer
service calls received per minute over a 30-minute period:

Data: 1, 2, 2, 1, 3, 2, 0, 4, 1, 3, 2, 1, 2, 3, 1, 0, 1, 2, 4, 3, 2, 3, 2, 1, 0, 1, 2, 2, 3, 1

a) Store the data in an array in C.

b) Count the frequency of each unique value of the random variable X (i.e., number
of calls per minute).
c) Estimate the empirical PMF using the formula:

frequency of xi
P̂ (X = xi ) = .
30

d) Display a table showing each unique value xi and its estimated probability P̂ (X =
xi ) with at least 4 decimal places.
e) Plot the PMF as a text-based histogram.
f) Compute the empirical mean and variance:
X X
x̄ = xi P̂ (X = xi ), Var(X) = (xi − x̄)2 P̂ (X = xi ).

References
1. S. D. Conte and C. de Boor: Elementary Numerical Analysis: An Algorithmic Ap-
proach

2. D. K. Faddeev and V. H. Faddeeva: Computational Methods in Linear Algebra.

3. G. E. Forsythe and G. B. Moler: Computer Solution of Linear Algebraic Systems.

4. W. H. Press, S. A. Teukolsky, W. T. Vetterling, B. P. Flannery: Numerical Recipes in

C.