Session-1

Introduction
Exact numbers:
Numbers that are free from uncertainty or approximation are known as exact numbers.
For example: 2, 1/3, 2 , π, e, …

Approximate numbers: Some numbers are not exact. For example:

1/3 = 0.33333…
2  1.414213...
π = 3.1415926…

These are not exact numbers because they contain infinitely many non-recurring digits. Numbers
obtained by retaining only a few digits are called approximate numbers.

Significant figures (digits):

All digits from 1 to 9 are significant figures. Zero, however, may or may not be significant,
depending on its context. When zero is used to position the decimal point or to fill the places of
omitted digits, it is not considered a significant figure.

Examples :
 The no. 320 has 3 significant digits 0, 2, and 3.
 The no. 0.0305 has 3 significant digits since the digits 3, 0, and 5 are significant because they
represent measured or meaningful values.
 The nos. 3.1416, 0.66667, 4.0687 contain five significant digits.
 The no. 1.016 has 4 significant digits
 The no. 0.05790 has 3 significant digits
 The no. 16.00 has 2 significant digits
 The number 0.00023 has 2 significant figures, not 5. The leading zeros are not significant;
they only serve to position the decimal point.

To represent any number based on the desired number of significant digits, we use either
chopping or rounding off.

Chopping :
It is a number chopped after required number of significant digits.

Example :
The number 0.859378 is chopped to 0. 8593 by keeping four significant digits
The number 0.5789 is chopped to 0. 57 by keeping two significant digits
Rounding:
Rounding of a number is the process of adjusting the digits of a number to make it simpler, often
by eliminating some digits after a certain point, based on a specified level of precision. The
number is rounded to the nearest specified place value, such as the nearest integer, tenth,
hundredth, or significant digit.

To round off a number to n significant digits, discard all digits to the right of the nth digit, and
follow these rules for the discarded number:
1. If the discarded digits represent less than half a unit in the nth place, leave the nth digit
unchanged.
2. If the discarded digits represent greater than half a unit in the nth place, increase the nth
digit by 1.
3. If the discarded digits represent exactly half a unit in the nth place, increase the nth digit
by 1 if it is odd, otherwise leave it unchanged.
The number thus rounded off is said to be correct to n significant figures.

Example :
 2.6483 is rounded to 2.65 by keeping 2 decimal places.
 10.1427 is rounded to 10.14 by keeping 2 decimal places.
 0.0358 is rounded to 0.04 by keeping 2 decimal places.
 1.24567 is rounded to 1.24 by keeping 2 decimal places.

 0.859378 is rounded to 0. 8594 by keeping 4 significant digits

 1.6583 is rounded to 1.658 by keeping 4 significant digits.
 30.0567 is rounded to 30.06 by keeping 4 significant digits.
 0.859458 is rounded to 0.8594 by keeping 4 significant digits.
 3.14159 is rounded to 3.142 by keeping 4 significant digits.

What is Error?

Error : Error = True Value – Approximate Value

Absolute value : |Error|

Note : In case of rounding, the error is bounded by 10-(n+1) where ‘n’ is the number of decimal
places to be maintained for accuracy.

Very often these absolute errors may not be sufficient to analyse the problem. Because if the
solution of the problem is very small/large, the absolute error is going to be very small/large. But
if the order of magnitude of the solution is 10-5 (for example) and if we get absolute error is 10-6,
we may think that it is a very good solution. But, it is not so. Because, we have to talk about
magnitude of the solution only nut not error. In view of these observations, in order to determine
the accuracy in an approximate solution to a problem, either we find the bounds of the
Relative Error (it is a good estimator):
Error
Relative Error = (it will take case of the small/large solutions)
True value

Percentage Error :
Error
Percentage Error =100 X
True value

1
Example: If 0.333 is the approximate value of , then
3
 Absolute Error
1 1 333 1
=  0.333     0.00033
3 3 1000 3000
Error 0.00033
 Relative Error = =  0.00099
True value 1/ 3
Error
 Percentage Error =100  = 0.00099 100  0.099
True value

Errors in Computation:
In numerical methods, we typically encounter the following types of errors due to blunders or
mistakes:

Inherent error (machine epsilon):

The errors in problems caused by the use of calculators, mathematical tables, computers,
physical measurements of parameters, etc., are called inherent errors.

Truncation error :
The errors caused by expanding functions like sin(x), cos(x), exp(x), etc., or by using familiar
series like Taylor's series and Maclaurin's series, considering only a finite number of terms and
neglecting all higher-order terms, are called truncation errors.