By: Habtamu Garoma 3/27/2017

NUMERICAL ANALYSIS I
(MATH 2061)

For: Second Year Mathematics Student

By:Habtamu Garoma
Email address: habte200@[Link]

1
 By: Habtamu Garoma 3/27/2017

Numerical analysis

o is the branch of mathematics that is used to

find approximations to difficult problems
such as:

 finding the roots of non−linear equations

 integration involving complex expressions

 solving differential equations for which

analytical solutions do not exist
2
 3/27/2017

o It is applied to a wide variety of disciplines

such as :
-business
-all fields of engineering
-computer science
-education
-geology
-meteorology and others.
o It is the area of mathematics and computer
science that creates, analyzes, and
implements algorithms for solving numerically
the problems of continuous mathematics.
3
By: Habtamu Garoma
 By: Habtamu Garoma 3/27/2017

Chapter 1
Basic concepts in error
estimation

4
 By: Habtamu Garoma 3/27/2017

Source of Error
 Modeling Error
-Blunders
-Formulation Error
-Data uncertainty
 Numerical Error
-Rounding Error
-Truncation Error

5
 By: Habtamu Garoma 3/27/2017

Round off Error

 which result when numbers having
limited significant figures are used to
represent exact numbers.
 Caused by representing a number
approximately.

Example:

6
 By: Habtamu Garoma 3/27/2017

TRUNCATION ERROR
 Error caused by truncating or
approximating a mathematical
procedure.
Example of Truncation Error
1. Taking only a few terms of a
Maclaurin series to

7
 3/27/2017

If only 3 terms are used,

2. Using a finite to approximate

8
By: Habtamu Garoma
 By: Habtamu Garoma 3/27/2017

 Using finite rectangles to approximate

an integral.

9
 By: Habtamu Garoma 3/27/2017

Example 1: Maclaurin series

Calculate the value of with an absolute
relative approximate error of less than 1%.

6 terms are required. How many are required

to get at least 1 significant digit correct in
your answer?
10
 By: Habtamu Garoma 3/27/2017

Example 2: Diffrentiation
Find for using
and

The actual value is

Truncation error is then,

Can you find the truncation error with
?
11
 By: Habtamu Garoma 3/27/2017

Example 3: Integrations
Use two rectangles of equal width to
approximate the area under the curve
for f ( x)  x 2 over the interval

12
 By: Habtamu Garoma 3/27/2017

INTEGRATION EXAMPLE (CONT’D…)

 Choosing a width of 3, we have

 Actual value is given by

Truncation error is then

Can you find the truncation error with 4

rectangles?
13
 By: Habtamu Garoma 3/27/2017

APPROXIMATIONS AND ROUND-OFF ERRORS

 For many engineering problems, we cannot

obtain analytical solutions.
 Numerical methods yield approximate results,
results that are close to the exact analytical
solution. We cannot exactly compute the errors
associated with numerical methods.
 Only rarely given data are exact, since they
originate from measurements. Therefore
there is probably error in the input
information.
14
 By: Habtamu Garoma 3/27/2017

CONT’D
o Algorithm itself usually introduces errors as
well, e.g., unavoidable round-offs, etc
o The output information will then contain
error from both of these sources.

 How confident we are in our approximate

result?
 The question is “how much error is
present in our calculation and is it
tolerable?”
15
 By: Habtamu Garoma 3/27/2017

 Accuracy: How close is a computed or

measured value to the true value
 Precision (or reproducibility): How close is a
computed or measured value to previously
computed or measured values.
 Inaccuracy (or bias): A systematic deviation
from the actual value.
 Imprecision (or uncertainty): Magnitude of
scatter.
16
By: Habtamu Garoma 3/27/2017

17
 By: Habtamu Garoma 3/27/2017

SIGNIFICANT FIGURES
 Number of significant figures indicates
precision. Significant digits of a number are
those that can be used with confidence.
e.g., the number of certain digits plus one
estimated digit.
 53,800 How many significant figures?

5.38 x 104 3
5.380 x 104 4
5.3800 x 104 5

18
 By: Habtamu Garoma 3/27/2017

Zeros are sometimes used to locate the

decimal point not significant figures.

0.00001753 4
0.0001753 4
0.001753 4
Error Definitions
True Value = Approximation + Error
Et = True value – Approximation (+/-)
True error

19
 By: Habtamu Garoma 3/27/2017

• For numerical methods, the true value

will be known only when we deal with
functions that can be solved analytically
(simple systems).

• In real world applications, we usually

not know the answer a priori. Then

20
 By: Habtamu Garoma 3/27/2017

CON’D

Iterative Approach, example Newton's

method

 Use absolute value.

 Computations are repeated until
stopping criterion is satisfied

21
 By: Habtamu Garoma 3/27/2017

CONT’D…

 If the following criterion is met

you can be sure that the result is

correct to at least n significant
figures.
a) Inaccurate and imprecise; (b) accurate and
imprecise; (c) inaccurate and precise; (d) accurate
and precise

22
 3/27/2017

Absolute and Relative Errors

Absolute Error ( Ea )

Absolute error= Exact value  Approximate value

Relative Errors ( Er )

Exact value  Approximate value

Er  x 100%
Exact value

current approximation  previous approximation

a  X 100%
current approximation

23
 By: Habtamu Garoma 3/27/2017

Round of Errors
 Round-off errors: originate from the fact that
computers retain only a fixed number of
significant figures during a calculation.
 Numbers such as π, e, or 7 cannot be
expressed by a fixed number of significant
figures.
 Therefore, they cannot be represented
exactly by the computer.

24
FIGURE
How the (a) decimal (base 10) and the (b) binary (base 2) systems work. In (b), the
binary number 10101101 is equivalent to the decimal number 173.
FIGURE
The representation of the decimal integer −173 on a 16-bit computer using
the signed magnitude method.

FIGURE
The manner in which a floating-point number is stored in a word
CHOPPING
Example:
p=3.14159265358 to be stored on a base-10
system carrying 7 significant digits.
p=3.141592 chopping error t=0.00000065
If rounded
p=3.141593 t=0.00000035

 Some machines use chopping, because rounding

adds to the computational overhead. Since
number of significant figures is large enough,
resulting chopping error is negligible.
 3/27/2017

Propagation of Error
 The purpose of this section is to study how
errors in numbers can propagate through
mathematical functions.
 If we multiply two numbers that have errors,
we would like to estimate the error in the
product.
* Functions of a Single Variable
* Functions of More than One Variable
28
 By: Habtamu Garoma 3/27/2017

 Suppose that we have a function f (x) that is

dependent on a single independent variable x.

 Assume that x is an approximation of x.

 to assess the effect of the discrepancy between

x and on the value of the function.

 We would like to estimate by

f ( x)  f ( x)  f ( x)

By expansion of Taylor’s series, we obtain:

f ( x)  f ' ( x) x, where x  x  x
29
 By: Habtamu Garoma 3/27/2017

FIGURE 4.7
Graphical depiction of first order error propagation

30
 By: Habtamu Garoma 3/27/2017

Example: Given a value of x = 2.5 with an

error of x= 0.01, estimate the resulting error

in the function f ( x)  x3 .
n

Ans: f(2.5) = 15.625 ± 0.1875

Functions of More than One Variable

For n independent variables x1 , x2 ,...., xn
having errors  x ,  x ,...,  x the following
1 2 n

general relationship holds:

f f f
f ( x1 , x2 ,..., xn )   x1   x2  ...   xn
x1 x2 xn

31
 Habtamu Garoma 3/27/2017

EXAMPLE 1
Find the bounds for the propagation in adding two
numbers. For example if one is calculating X +Y where
X = 1.5 ± 0.05
Y = 3.4 ± 0.04
Solution
Maximum possible value of X = 1.55 and Y = 3.44
Maximum possible value of X + Y = 1.55 + 3.44 = 4.99
Minimum possible value of X = 1.45 and Y = 3.36.
Minimum possible value of X + Y = 1.45 + 3.36 = 4.81
Hence
4.81 ≤ X + Y ≤ 4.99.
32
 3/27/2017

EXAMPLE
The strain in an axial member of a
square cross-section is given by

Given

Find the maximum possible error in the

measured strain.

33
 3/27/2017

Solution:

34
 3/27/2017

CONT’D…
Thus

Hence

35
 3/27/2017

EXAMPLE 3
Subtraction of numbers that are nearly equal
can create unwanted inaccuracies. Using the
formula for error propagation, show that this
is true.
Solution: Let
Then

So the relative change is

36
 3/27/2017

CONT’D
For example if

= 0.6667
= 66.67%

37
 3/27/2017

Numerical stability

 refers to the accuracy of an algorithm

in the presence of rounding errors

• an algorithm is unstable if rounding

errors cause large errors in the result

 rigorous definition depends on what

‘accurate’ and ‘large error’ mean

38
3/27/2017

39