π Lecture 1 β Numerical Analysis
π Overview: This introductory lecture covers the fundamental concepts of number systems used in computing and the representation of numbers on computers. It is essential because numerical analysis relies on understanding how numbers are stored and manipulated in binary form, and introduces the course contents including solution of equations, interpolation, differentiation, integration, and differential equations.
ποΈ Topics Covered
The lecture begins with an introduction to numerical analysis and problem solving using computers, then covers number systems including decimal, binary, octal, and hexadecimal bases. It explains how real numbers are represented in different bases, provides examples of converting decimal numbers to binary (both integers and fractions), and demonstrates conversion from binary to octal. The course contents are outlined at the start.
π Lecture Summary
Introduction
This chapter begins with basic concepts of representation of numbers on computers and errors introduced during computation. Problem solving using computers and the steps involved are also discussed in brief.
Number (s) System (s)
In our daily life, we use numbers based on the decimal system, which uses ten symbols 0, 1,...,9 and the number 10 is called the base of the system. When a base N is given, we need N different symbols 0, 1, 2, ..., (N β 1) to represent an arbitrary number.
The number systems commonly used in computers are:
| Base, N | Number |
|---|---|
| 2 | Binary |
| 8 | Octal |
| 10 | Decimal |
| 16 | Hexadecimal |
An arbitrary real number, a, can be written as: a = aβNα΅ + aβββNα΅β»ΒΉ + ... + aβNΒΉ + aβ + aββNβ»ΒΉ + ... + aββNβ»α΅
In the binary system, it has the form: a = aβ2α΅ + aβββ2α΅β»ΒΉ + ... + aβ2ΒΉ + aβ + aββ2β»ΒΉ + ... + aββ2β»α΅
The decimal number 1729 is represented and calculated as: (1729)ββ = 1Γ10Β³ + 7Γ10Β² + 2Γ10ΒΉ + 9Γ10β°
The decimal equivalent of binary number 10011001 is: 1Γ2β° + 0Γ2β»ΒΉ + 0Γ2β»Β² + 1Γ2β»Β³ + 1Γ2β»β΄ + 0Γ2β»β΅ + 0Γ2β»βΆ + 1Γ2β»β· = 1 + 1/8 + 1/16 + 1/128 = (1.1953125)ββ
Electronic computers use the binary system whose base is 2. The two symbols used are 0 and 1, which are called binary digits or simply bits. The internal representation of any data within a computer is in binary form. However, data input and output of numerical results are preferred in decimal system. Within the computer, arithmetic is carried out in binary form.
π Definition β Base: The number of distinct symbols used in a number system; for decimal it is 10, for binary it is 2. π Definition β Bit: A binary digit, either 0 or 1, the fundamental unit of data in computers.
Conversion of decimal number 47 into its binary equivalent
The conversion is done by repeated division by 2:
| Division | Quotient | Remainder |
|---|---|---|
| 47 Γ· 2 | 23 | 1 (LSB) |
| 23 Γ· 2 | 11 | 1 |
| 11 Γ· 2 | 5 | 1 |
| 5 Γ· 2 | 2 | 1 |
| 2 Γ· 2 | 1 | 0 |
| 1 Γ· 2 | 0 | 1 (MSB) |
(47)ββ = (101111)β
π Formula: Decimal to binary integer conversion β Repeatedly divide by 2, reading remainders from last (most significant bit) to first (least significant bit). π Example: (47)ββ β Divide by 2 repeatedly; remainders from last to first give 101111.
Binary equivalent of the decimal fraction 0.7625
The conversion is done by repeated multiplication by 2:
| Multiplication | Product | Integer Part |
|---|---|---|
| 0.7625 Γ 2 | 1.5250 | 1 |
| 0.5250 Γ 2 | 1.0500 | 1 |
| 0.05 Γ 2 | 0.1 | 0 |
| 0.1 Γ 2 | 0.2 | 0 |
| 0.2 Γ 2 | 0.4 | 0 |
| 0.4 Γ 2 | 0.8 | 0 |
| 0.8 Γ 2 | 1.6 | 1 |
| 0.6 Γ 2 | 1.2 | 1 |
| 0.2 Γ 2 | 0.4 | 0 |
Reading the integer parts from top to bottom: (0.7625)ββ = (0.110000110...)β
π Formula: Decimal to binary fraction conversion β Repeatedly multiply by 2, recording integer parts in order. π Example: (0.7625)ββ β Multiply by 2 repeatedly; integer parts in order give 0.110000110...
Conversion (59)ββ into binary and then into octal
First, convert 59 to binary:
| Division | Quotient | Remainder |
|---|---|---|
| 59 Γ· 2 | 29 | 1 |
| 29 Γ· 2 | 14 | 1 |
| 14 Γ· 2 | 7 | 0 |
| 7 Γ· 2 | 3 | 1 |
| 3 Γ· 2 | 1 | 1 |
| 1 Γ· 2 | 0 | 1 |
(59)ββ = (111011)β
Then convert binary to octal by grouping bits in threes from right: (111011)β = 111 011 = (73)β
π Formula: Binary to octal conversion β Group binary digits in threes from right, convert each group to its octal equivalent. π Example: (111011)β β Group as 111 and 011 β (7)(3) = (73)β
π‘ Why this matters: Understanding number system conversions is critical for numerical analysis because all computer arithmetic is performed in binary, and errors can arise during conversion between bases.
β Key Takeaways
The lecture establishes that numerical analysis involves representing and solving mathematical problems on computers, where numbers are stored in binary form. Four number systems are commonly used in computing: decimal (base 10), binary (base 2), octal (base 8), and hexadecimal (base 16). Converting integers from decimal to binary requires repeated division by 2, reading remainders from last to first. Converting decimal fractions to binary requires repeated multiplication by 2, recording integer parts in order. Binary numbers can be converted to octal by grouping bits in threes, and understanding these conversions is fundamental to managing computational errors in numerical methods.
π§ Quick Revision Questions
- What is the binary equivalent of the decimal number 47?
- How do you convert a decimal fraction to binary, and what is the binary equivalent of 0.7625?
- What is the octal representation of the binary number 111011?
- Why is the binary system used in electronic computers instead of the decimal system?
- What are bits and why are they important in numerical computing?
π Lecture 2 β Errors in Computations
π Overview: This lecture introduces the fundamental types of errors that arise in numerical computations. Understanding these errors is critical because numerically computed solutions are never exact, and knowing the sources and growth of errors helps in selecting appropriate algorithms and interpreting results with confidence.
ποΈ Topics Covered
The lecture classifies errors into three main categories: inherent errors, which exist in the problem statement before any computation; local round-off errors, which arise due to finite computer word length and binary-decimal conversion; and local truncation errors, which result from approximating infinite series by a finite number of terms. Each error type is defined, explained with examples, and illustrated with practical significance.
π Lecture Summary
Errors in Computations
Numerically computed solutions are subject to certain errors. It may be fruitful to identify the error sources and their growth while classifying the errors in numerical computation. These are inherent errors, local round-off errors, and local truncation errors.
Inherent Errors
It is that quantity of error which is present in the statement of the problem itself, before finding its solution. It arises due to the simplified assumptions made in the mathematical modeling of a problem. It can also arise when the data is obtained from certain physical measurements of the parameters of the problem.
π‘ Why this matters: Inherent errors are unavoidable since they exist before any computation begins β they set a lower bound on the achievable accuracy.
Local Round-off Errors
Every computer has a finite word length and therefore it is possible to store only a fixed number of digits of a given input number. Since computers store information in binary form, storing an exact decimal number in its binary form into the computer memory gives an error. This error is computer dependent.
At the end of computation of a particular problem, the final results in the computer, which is obviously in binary form, should be converted into decimal form β a form understandable to the user β before their print out. Therefore, an additional error is committed at this stage too. This error is called local round-off error.
π Definition β Error: Error = True value β Computed value
π Definition β Absolute error: Absolute error = |Error|
π Definition β Relative error: Relative error = Error / True value
π Example: The decimal number (0.7625)ββ equals (0.110000110011)β in binary. If a computer system has a word length of 12 bits only, then the decimal number 0.7625 is stored in the computer memory in binary form as 0.110000110011. However, it is equivalent to 0.76245. Thus, in storing the number 0.7625, we have committed an error equal to 0.00005, which is the round-off error; inherent with the computer system considered.
Local Truncation Error
It is generally easier to expand a function into a power series using Taylor series expansion and evaluate it by retaining the first few terms. For example, we may approximate the function f(x) = cos x by the series: cos x = 1 β xΒ²/2! + xβ΄/4! β β― + (β1)βΏ xΒ²βΏ/(2n)! + β―
If we use only the first three terms to compute cos x for a given x, we get an approximate answer. Here, the error is due to truncating the series. Suppose, we retain the first n terms, the truncation error (TE) is given by: TE β€ xΒ²βΏβΊΒ² / (2n + 2)!
The TE is independent of the computer used.
If we wish to compute cos x accurate with five significant digits, the question is, how many terms in the expansion are to be included? In this situation: xΒ²βΏβΊΒ² / (2n + 2)! < 0.5 Γ 10β»β΅ = 5 Γ 10β»βΆ
Taking logarithm on both sides, we get: (2n + 2) log x β log[(2n + 2)!] < logββ 5 β 6 logββ 10 = 0.699 β 6 = β5.3 or log[(2n + 2)!] β (2n + 2) log x > 5.3
We can observe that the above inequality is satisfied for n = 7. Hence, seven terms in the expansion are required to get the value of cos x, with the prescribed accuracy.
The truncation error is given by: TE β€ xΒΉβΆ / 16!
π Example: To compute cos x with five significant digits accuracy, one must retain seven terms in the Taylor series expansion because the inequality log[(2n+2)!] β (2n+2) log x > 5.3 is satisfied only when n=7.
π‘ Why this matters: Truncation error is analyst-controlled β by choosing how many terms to include, we can achieve any desired accuracy, though at the cost of more computation.
β Key Takeaways
The three fundamental error types in numerical computation are inherent errors (present in the problem statement or data), local round-off errors (due to finite computer word length and binary-decimal conversion), and local truncation errors (due to approximating infinite processes by finite ones). The error is defined as True value minus Computed value, with absolute error being its magnitude and relative error being the error divided by the true value. Round-off errors are computer-dependent, while truncation errors are independent of the computer and can be controlled by retaining more terms. To achieve five significant digits of accuracy when computing cos x using its Taylor series, seven terms are required, as determined by solving the inequality involving the truncation error bound.
π§ Quick Revision Questions
-
What are the three types of errors discussed in numerical computations, and what is the source of each?
-
Define error, absolute error, and relative error mathematically.
-
Why does the decimal number 0.7625 become 0.76245 in a 12-bit computer system? What type of error does this represent?
-
What is the bound on the truncation error when approximating cos x by the first n terms of its Taylor series?
-
How many terms of the Taylor series for cos x are needed to guarantee five significant digits of accuracy? Show the reasoning.
π Lecture 3 β Polynomial
π Overview: This lecture introduces the fundamental concepts of polynomial and algebraic equations, along with transcendental equations. It establishes the foundational definitions and key facts necessary for understanding root-finding methods in numerical analysis, explaining why these classification and properties matter for solving equations.
ποΈ Topics Covered
The lecture covers the definition of a polynomial expression of nth degree, the concept of an algebraic equation f(x)=0, and key facts about the roots of such equationsβincluding the number and nature of roots based on degree and sign of the last term. It then defines transcendental equations as those involving logarithmic, trigonometric, or exponential functions, and finally defines the root of an equation as a value 'a' such that f(a)=0.
π Lecture Summary
Polynomial
An expression of the form ( f(x) = a_0 x^n + a_1 x^{n-1} + a_2 x^{n-2} + ... + a_{n-1} x + a_n ), where n is a positive integer and ( a_0, a_1, a_2, ..., a_n ) are real constants, is called an nth degree polynomial in x if ( a_0 \neq 0 ). This is the basic building block for algebraic equations in numerical analysis.
π Definition β Polynomial: An expression of the form ( f(x) = a_0 x^n + a_1 x^{n-1} + ... + a_n ) with ( a_0 \neq 0 ), where n is a positive integer and all coefficients are real constants.
Algebraic equation
An equation f(x)=0 is said to be an algebraic equation in x if it is purely a polynomial in x. For example, ( x^5 + x^4 + 3x^2 + x - 6 = 0 ) is a fifth order polynomial and thus an algebraic equation. Other examples include ( x^3 - 6 = 0 ), ( x^6 - 7x = 0 ), ( y^4 - 4y^3 + 3y^2 - y - 2 = 0 ) (polynomial in y), and ( t^4 - 6t^2 - 21 = 0 ) (polynomial in t).
Some facts
- Every equation of the form f(x)=0 has at least one root; it may be real or complex.
- Every polynomial of nth degree has n and only n roots.
- If f(x)=0 is an equation of odd degree, then it has at least one real root whose sign is opposite to that of the last term.
- If f(x)=0 is an equation of even degree whose last term is negative, then it has at least one positive and at least one negative root.
π‘ Why this matters: These facts give a quick way to predict the existence and sign of real roots without solving the equation, which is essential for selecting appropriate numerical methods.
Transcendental equation
An equation is said to be a transcendental equation if it involves logarithmic, trigonometric, and exponential functions, or a combination of all three. For example, ( e^x - 5x - 3 = 0 ), ( e^x - \sin x = 0 ), ( \ln x - \sin x = 0 ), and ( 2 \sec^2 x - \tan x - e^x = 0 ) are all transcendental equations. These are more complex than algebraic equations and often require iterative numerical methods for solution.
π Definition β Transcendental equation: An equation that contains logarithmic, trigonometric, or exponential functions (or any combination thereof).
Root of an equation
For an equation f(x)=0, to find the solution we find such a value which satisfies f(x)=0; these values are known as the roots of the equation. A value 'a' is known as the root of an equation f(x)=0 if and only if f(a)=0.
π Definition β Root of an equation: A value 'a' such that f(a)=0; it satisfies the equation f(x)=0.
π Example: For the polynomial ( x^3 - 6 = 0 ), if we find a number 'a' such that ( a^3 - 6 = 0 ), then 'a' is a root of this algebraic equation.
β Key Takeaways
You must distinguish between algebraic (purely polynomial) and transcendental (involving log, trig, or exponential functions) equations. Every nth degree polynomial has exactly n roots (counting multiplicity and complex roots). The sign of the last term and the degree (odd/even) provide critical information about the existence and sign of real rootsβodd degree equations always have at least one real root, and even degree equations with a negative last term have both positive and negative real roots. Finally, the root of any equation f(x)=0 is defined precisely by the condition f(a)=0, which is the foundation for all root-finding methods.
π§ Quick Revision Questions
- What is the definition of an nth degree polynomial in x?
- State the four facts about the roots of polynomial equations.
- What distinguishes a transcendental equation from an algebraic equation?
- Give three examples of transcendental equations.
- What is a root of an equation f(x)=0?
π Lecture 4 β Properties of an Algebraic Equation
π Overview: This lecture introduces fundamental properties of algebraic equations that are essential for root-finding in numerical analysis. It covers the nature of roots, sign rules for determining root signs, and the classification of numerical methods into direct and iterative approaches.
ποΈ Topics Covered
The lecture covers properties of algebraic equations including complex roots occurring in conjugate pairs and factor properties. It explains Descartes' rule of signs for determining positive and negative roots, the intermediate value property for root existence in intervals, and classifies numerical methods into direct methods (like Graefee method) and iterative methods.
π Lecture Summary
Properties of an Algebraic equation
An algebraic equation of degree n has specific properties regarding its roots. First, complex roots occur in conjugate pairs β if (a+ib) is a root of f(x)=0, then (a-ib) is also a root. Second, if x=a is a root of the nth-degree polynomial f(x)=0, then (x-a) is a factor of f(x), and dividing f(x) by (x-a) yields a polynomial of degree n-1.
π Definition β Complex conjugate roots: If a polynomial equation has real coefficients and a complex root (a+ib), its complex conjugate (a-ib) is also a root.
Descartes rule of signs
Descartes' rule of signs establishes the relationship between the signs of coefficients and the roots of an equation. The rule states: "The number of positive roots of an algebraic equation f(x)=0 with real coefficients cannot exceed the number of changes in the signs of the coefficients in f(x)=0. Similarly, the number of negative roots cannot exceed the number of changes in the sign of coefficients of f(-x)=0."
π Definition β Descartes' rule of signs: A rule that determines the maximum possible number of positive and negative real roots of a polynomial by counting sign changes in coefficients.
π Example: Consider the equation xΒ³ β 3xΒ² + 4x β 5 = 0. In f(x), the coefficient signs are +, β, +, β β there are three sign changes (first + to β, second β to +, third + to β), so at most three positive roots exist. For f(βx) = βxΒ³ β 3xΒ² β 4x β 5, the signs are β, β, β, β β no sign change, so there are no negative roots.
Intermediate value property
The intermediate value property states: If f(x) is a real-valued continuous function in the closed interval a β€ x β€ b, and if f(a) and f(b) have opposite signs, then f(x)=0 has at least one root Ξ² such that a β€ Ξ² β€ b. Simply put, if f(a) and f(b) have different signs, the equation must have at least one real root between a and b.
π Definition β Intermediate value property: For a continuous function, if function values at interval endpoints have opposite signs, at least one root exists within that interval.
π‘ Why this matters: This property provides the theoretical foundation for bracketing methods like the bisection method, guaranteeing root existence before starting numerical computation.
Numerical methods for solving equations
Numerical methods for solving algebraic or transcendental equations are classified into two groups: Direct methods and Iterative methods.
Direct methods are those that do not require any information about the initial approximation of the root to start the solution. Examples include Graefee root squaring method, Gauss elimination method, and Gauss Jordan method. All these methods do not require any type of initial approximation.
Iterative methods require an initial approximation to start the solution process.
π Definition β Direct methods: Numerical methods that find the solution without needing an initial guess for the root.
π Definition β Iterative methods: Numerical methods that begin with an initial approximation and progressively refine it to approach the actual root.
β Key Takeaways
Complex roots of polynomial equations always occur in conjugate pairs, and if a value is a root, the corresponding linear factor divides the polynomial. Descartes' rule of signs provides a quick way to estimate the maximum number of positive and negative real roots by counting sign changes in f(x) and f(-x). The intermediate value property guarantees at least one root exists in an interval where the function changes sign, forming the basis for bracketing methods. Numerical methods split into direct methods (no initial guess needed, like Graefee and Gauss methods) and iterative methods (require initial approximation). Understanding these properties is essential before applying any root-finding algorithm.
π§ Quick Revision Questions
- If a polynomial equation has a complex root (3+2i), what other root must it have?
- How many sign changes occur in the polynomial xβ΄ β 2xΒ³ + 3xΒ² β x + 1 = 0, and what does this indicate about positive roots?
- For the polynomial f(x) = xΒ³ + 2xΒ² + x + 1, use Descartes' rule to determine the maximum number of negative roots.
- If f(2) = 4 and f(5) = β3, what does the intermediate value property tell us about roots in the interval [2,5]?
- Name two direct methods for solving algebraic equations and state what characteristic they share.
π Lecture 5 β Bisection Method, Initial Approximations, and Iterative Methods
π Overview: This lecture introduces the concept of iterative methods for finding roots of equations and focuses on obtaining initial approximations through graphical and analytical methods. It then provides a detailed explanation and application of the Bisection Method (Bolzano), including procedural steps, examples, and criteria for stopping iterations.
ποΈ Topics Covered
The lecture covers iterative methods, methods for obtaining initial approximations (graphical and analytical), the Bisection method (Bolzano) with its procedure and step-by-step examples, and the criteria for stopping the iterative process to achieve a desired accuracy.
π Lecture Summary
Bisection method, Newton raphson method, secant method, jacobi method are all examples of iterative methods. How to get an initial approximation?
The initial approximation for finding a root can be found using two methods: the graphical method and the analytical method.
π Definition β Graphical Method: The equation f(x)=0 can be rewritten as f1(x)=f2(x). The initial approximation is taken as the abscissa (x-coordinate) of the point of intersection of the graphs of y=f1(x) and y=f2(x).
π Example: For the equation f(x)=x β sin x β 1 = 0, it is rewritten as y=x-1 and y=sin x. The graphs of these two functions intersect at approximately x=1.9, so the initial approximation is 1.9.
π Definition β Analytical Method: This method is based on the intermediate value property. It involves locating two values, a and b, such that f(a) and f(b) have opposite signs. The root is then known to lie between these two points.
π Example: Consider the equation f(x)=3x β 1 + sin x = 0. f(0) = -1 and f(1) = 3 β 1 + sin(1 Γ 180/Ο) = 2 + 0.84147 = 1.64299. Since f(0) and f(1) have opposite signs, one root lies between 0 and 1.
π‘ Why this matters: The analytical method ensures that a root exists in a given interval by checking the sign change of the function at its endpoints.
Bisection method (Bolzano)
The Bisection method is used to locate the root of an equation f(x)=0 within an interval (x0, x1) where f(x0) and f(x1) have opposite signs, i.e., f(x0)f(x1) < 0. This sign change indicates that the graph crosses the x-axis, guaranteeing at least one root in the interval.
The procedure starts by defining the midpoint x2 = (x0 + x1)/2. If f(x2) = 0, then x2 is the root. Otherwise, the root lies either between x0 and x2 or x1 and x2, depending on which pair produces a sign change. The process of halving the interval is repeated until the root is found to the desired accuracy.
π Formula: Midpoint: x_new = (x_left + x_right) / 2 β The next approximation is the average of the two interval endpoints that bracket the root.
π Example: Solve xΒ³ β 9x + 1 = 0 for the root between x=2 and x=4.
- f(2) = 8 β 18 + 1 = -9; f(4) = 64 β 36 + 1 = 29. Since f(2)f(4) < 0, root lies between 2 and 4.
- x2 = (2+4)/2 = 3; f(3) = 27 β 27 + 1 = 1. Since f(2)f(3) < 0, root lies between 2 and 3.
- x3 = (2+3)/2 = 2.5; f(2.5) = 15.625 β 22.5 + 1 = -5.875.
- x4 = (2.5+3)/2 = 2.75; f(2.75) = -2.9531.
- x5 = (2.75+3)/2 = 2.875; f(2.875) = -1.1113.
- x6 = (2.875+3)/2 = 2.9375; f(2.9375) = -0.0901.
- The process is continued until the desired accuracy is obtained.
π Example: Carry out five iterations for f(x)=2x cos(2x) β (x+1)Β².
- f(-1) = 0.8322 > 0, f(0) = -1 < 0. Root lies between -1 and 0.
- x2 = (0-1)/2 = -0.5
- f(-0.5) = -0.7903 < 0. Root lies between -1 and -0.5.
- x3 = (-0.5-1)/2 = -0.75
- f(-0.75) = -0.1686 < 0. Root lies between -1 and -0.75.
- x4 = (-0.75-1)/2 = -0.875
- f(-0.875) = 0.296275 > 0. Root lies between -0.875 and -0.75.
- x5 = (-0.75-0.875)/2 = -0.8125
- f(-0.8125) = 0.052970 > 0. Root lies between -0.8125 and -0.75.
- x6 = (-0.75-0.8125)/2 = -0.78125
π Example: Carry out the first five iterations for f(x)=x cosx β 2xΒ² + 3x β 1, for 1.2 β€ x β€ 1.3.
- f(1.2) = 0.1548 > 0, f(1.3) = -0.1322 < 0. Root lies between 1.2 and 1.3.
- x2 = (1.2+1.3)/2 = 1.25; f(1.25) = 0.0191 > 0. Root lies between 1.25 and 1.3.
- x3 = (1.25+1.3)/2 = 1.275; f(1.275) = -0.0545 < 0. Root lies between 1.25 and 1.275.
- x4 = (1.25+1.275)/2 = 1.2625; f(1.2625) = -0.0172 < 0. Root lies between 1.25 and 1.2625.
- x5 = (1.25+1.2625)/2 = 1.25625; f(1.25625) = 0.00108 > 0. Root lies between 1.25625 and 1.2625.
- x6 = (1.25625+1.2625)/2 = 1.259375
When to stop the process of iteration?
The iteration process stops when a specified condition is met. This condition can be either a fixed number of iterations or a desired accuracy.
π Definition β Accuracy: To achieve an accuracy of 10β»Β³, the absolute difference between two consecutive approximations is calculated. If this difference is less than or equal to the required accuracy, the process stops.
π Example: If two consecutive roots are 2.135769 and 2.135648, the difference is 2.135769 β 2.135648 = 0.000121. Since 0.000121 is less than 0.001 (10β»Β³), the accuracy of 10β»Β³ is achieved and the process stops.
β Key Takeaways
The most critical points to remember from this lecture are that iterative methods, such as the Bisection method, require an initial approximation, which can be found graphically or analytically by ensuring function values at interval endpoints have opposite signs. The core procedure of the Bisection method is to repeatedly halve the interval bracketing the root by calculating the midpoint and selecting the subinterval where a sign change occurs. All calculations for trigonometric functions in numerical methods are performed in radians mode with Ο approximated as 3.14. Finally, the iteration process is terminated either after a predetermined number of steps or when the absolute difference between two successive approximations is less than a specified tolerance or accuracy.
π§ Quick Revision Questions
- What are the two methods to obtain an initial approximation for a root?
- State the fundamental condition that must be satisfied for the analytical method to guarantee a root in an interval.
- In the Bisection method, if f(xβ) and f(xβ) have opposite signs, what is the formula for the next approximation, xβ?
- In which mode (degrees or radians) are all calculations in numerical analysis performed?
- How do you determine when to stop the Bisection method if you are asked to find the root with an accuracy of 10β»β΄?
π Lecture 6 β Regula-Falsi Method (Method of False Position) & Method of Iteration
π Overview: This lecture covers two root-finding techniques: the Regula-Falsi (false position) method, which uses a chord between two points with opposite signs to approximate a root, and the Iteration method, which rewrites the equation as (x = \phi(x)) for successive approximations. These methods are essential for solving algebraic and transcendental equations numerically.
ποΈ Topics Covered
Regula-Falsi method derivation and formula; convergence observation about initial interval size; worked examples on cubic, transcendental, and mixed exponential-trigonometric equations; Method of Iteration definition; convergence theorem; examples using cubic and trigonometric functions with high accuracy.
π Lecture Summary
Regula-Falsi method (Method of false position)
Here we choose two points (x_n) and (x_{n-1}) such that (f(x_n)) and (f(x_{n-1})) have opposite signs. Intermediate value property suggests that the graph of (y=f(x)) crosses the x-axis between these two points and therefore, a root lies between these two points. Thus to find the real root of (f(x)=0) using Regula-Falsi method, we replace the part of the curve between the points (A(x_n, f(x_n))) and (B(x_{n-1}, f(x_{n-1}))) by a chord in the interval and we take the point of intersection of this chord with x-axis as initial approximation.
The equation of the chord joining points A and B is: [\frac{y - f(x_n)}{x - x_n} = \frac{f(x_{n-1}) - f(x_n)}{x_{n-1} - x_n}]
Setting (y=0) gives the first approximation to the root: [x_{n+1} = x_n - \frac{x_n - x_{n-1}}{f(x_n) - f(x_{n-1})} f(x_n)]
We observe that (f(x_{n-1})) and (f(x_{n+1})) are of opposite signs, thus it is possible to apply the above procedure to determine the line through B and Aβ and so on.
π Definition β Regula-Falsi formula: (x_{n+1} = x_n - \frac{x_n - x_{n-1}}{f(x_n) - f(x_{n-1})} f(x_n)) β finds the x-intercept of the chord joining two points with opposite function signs.
π Example: Use Regula-Falsi to compute a real root of (x^3 - 9x + 1 = 0): (i) between 2 and 4, (ii) between 2 and 3.
Let (f(x) = x^3 - 9x + 1).
(f(2) = 8 - 18 + 1 = -9), (f(4) = 64 - 36 + 1 = 29). Opposite signs, root lies between 2 and 4.
First approximation ((x_1=2, x_2=4)): [x_3 = 4 - \frac{4-2}{29-(-9)}(29) = 4 - \frac{2(29)}{38} = 4 - 1.5263 = 2.4736] (f(x_3) = (2.4736)^3 - 9(2.4736) + 1 = 15.13520 - 22.2624 + 1 = -6.12644)
Second approximation: [x_4 = 2.4736 - \frac{2.4736-4}{-6.12644-29}(-6.12644) = 2.4736 + 0.26619 = 2.73989] (f(x_4) = 20.5683 - 24.65901 + 1 = -3.090707)
Third approximation: [x_5 = 2.73989 - \frac{2.73989-4}{-3.090707-29}(-3.090707) = 2.86125] (f(x_5) = 23.42434 - 25.75125 + 1 = -1.326868)
(ii) Between 2 and 3:
(f(2) = -9), (f(3) = 27 - 27 + 1 = 1). Opposite signs, root lies between 2 and 3.
First approximation ((x_1=2, x_2=3)): [x_3 = 3 - \frac{3-2}{1-(-9)}(1) = 3 - \frac{1}{10} = 2.9] (f(x_3) = 24.389 - 26.1 + 1 = -0.711)
Second approximation: [x_4 = 2.9 - \frac{2.9-3}{-0.711-1}(-0.711) = 2.9 + 0.04156 = 2.94156] (f(x_4) = 25.45265 - 26.47404 + 1 = -0.0207)
Third approximation: [x_5 = 2.94156 - \frac{2.94156-3}{-0.0207-1}(-0.0207) = 2.94275] (f(x_5) = 25.48356 - 26.48475 + 1 = -0.0011896)
π‘ Why this matters: The value when the interval is (2,3) converges faster and is closer to the root. Important observation: The initial interval ((x_1, x_2)) in which the root lies should be sufficiently small.
Example β Transcendental Equation
Use Regula-Falsi to find a real root of (\ln x - \cos x = 0) accurate to four decimal places after three successive approximations. (Note: all calculations in radians mode.)
(f(x) = \ln x - \cos x)
(f(1) = 0 - 0.540302 = -0.540302 < 0)
(f(2) = 0.69315 - 0.41615 = 1.109)
(f(1)f(2) < 0), root lies between 1 and 2.
First approximation ((x_1=1, x_2=2)): [x_3 = 2 - \frac{2-1}{1.109 - (-0.540302)}(1.109) = 2 - \frac{1.1093}{1.6496} = 1.3275] (f(x_3) = \ln 1.3275 - \cos 1.3275 = 0.2833 - 0.2409 = 0.0424)
Second approximation (using (x_1) and (x_3)): [x_4 = 1.3275 - \frac{(0.3275)(0.0424)}{0.0424 + 0.5403} = 1.3037] (f(x_4) = 1.24816 \times 10^{-3})
Third approximation:
[x_5 = 1.3037 - \frac{(0.3037)(0.001248)}{0.001248 + 0.5403} = 1.3030]
(f(x_5) = 0.6245 \times 10^{-4})
The required real root is 1.3030.
Example β Mixed Equation
Use method of false position to solve (e^{-x} + 2^{-x} + 2\cos x - 6 = 0) for (1 \le x \le 2).
(f(x) = e^{-x} + 2^{-x} + 2\cos x - 6) with (x_0=1, x_1=2).
(f(1) = 2.7182 + 0.5 + 2(0.5403) - 6 = -1.7011)
(f(2) = 7.3886 + 0.25 + 2(-0.4161) - 6 = 0.8068)
For (n=1): [x_2 = 2 - \frac{2-1}{0.8068 + 1.7011}(0.8068) = 2 - \frac{1}{2.5079}(0.8068) = 1.6783] (f(1.6783) = -0.5457)
For (n=2): [x_3 = 1.6783 - \frac{1.6783-2}{-0.5457-0.8068}(-0.5457) = 1.6783 + 0.12979 = 1.8081] (f(1.8081) = -0.08575)
For (n=3): [x_4 = 1.8081 - \frac{1.8081-1.6783}{-0.08575+0.5457}(-0.08575) = 1.8323] (f(1.8323) = 0.01199)
For (n=4): [x_5 = 1.8323 - \frac{1.8323-1.8081}{0.01199+0.08575}(0.01199) = 1.8323 - 0.00296 = 1.8293] (f(1.8293) = -0.000343)
For (n=5): [x_6 = 1.8293 - \frac{1.8293-1.8323}{-0.000343-0.01199}(-0.000343) = 1.8293]
Example β Trigonometric Equation
Solve (2x\cos 2x - (x-2)^2 = 0) for (2 \le x \le 3), perform three iterations.
(f(x) = 2x\cos 2x - (x-2)^2) with (x_0=2, x_1=3).
(f(2) = 4\cos 4 - 0 = -2.6146)
(f(3) = 6\cos 6 - 1 = 4.7610)
For (n=1): [x_2 = 3 - \frac{3-2}{4.7610 - (-2.6146)}(4.7610) = 3 - \frac{4.7610}{7.3756} = 3 - 0.6455 = 2.3545] (f(2.3545) = 4.709\cos 4.709 - 0.1257 = -0.1416)
For (n=2): [x_3 = 2.3545 - \frac{2.3545-3}{-0.1416-4.7610}(-0.1416) = 2.3731] (f(2.3731) = 4.7462\cos 4.7462 - 0.1392 = -0.1392)
For (n=3): [x_4 = 2.3731 - \frac{2.3731-2.3545}{-0.1392+0.1416}(-0.1392) = 2.3707] (f(2.3707) = 4.7414\cos 4.7412 - 0.1374 = 0.00013)
For (n=4): [x_5 = 2.3707 - \frac{2.3707-2.3731}{0.00013-0.0212}(0.00013) = 2.3707]
Example β Logarithmic Equation
Using Regula-Falsi, find the real root of (x\log_{10}x = 1.2) correct to three decimal places.
(f(x) = x\log_{10}x - 1.2)
(f(2) = 2(0.3010) - 1.2 = -0.5979)
(f(3) = 3(0.4771) - 1.2 = 0.2314)
Opposite signs, root lies between (x_1=2, x_2=3).
First approximation: [x_3 = 3 - \frac{3-2}{0.2314+0.5979}(0.2314) = 3 - \frac{0.2314}{0.8293} = 2.72097] (f(x_3) = 2.72097\log_{10} 2.72097 - 1.2 = -0.01713)
Second approximation: [x_4 = 2.72097 - \frac{2.72097-3}{-0.01713-0.2314}(-0.01713) = 2.7402] (f(x_4) = 2.7402\log_{10} 2.7402 - 1.2 = -3.8905 \times 10^{-4})
Thus, the root correct to three decimal places is 2.740.
π‘ Why this matters: If TOL is given, we can find its value by subtracting consecutive roots and writing in exponential notation. Stop when required TOL is obtained.
Method of Iteration
Method of iterations can be applied to find a real root of (f(x)=0) by rewriting it in the form: [x = \phi(x)]
Let (x = x_0) be the initial approximation to the actual root (\alpha). Then: [x_1 = \phi(x_0), \quad x_2 = \phi(x_1), \quad x_3 = \phi(x_2), \quad \ldots, \quad x_n = \phi(x_{n-1})] If the sequence of approximate roots converges to (\alpha), it is taken as the root of (f(x)=0).
π Definition β Iteration method: (x_{n+1} = \phi(x_n)) where (x = \phi(x)) is obtained by rewriting (f(x)=0).
Theorem
If (\alpha) is a root of (f(x)=0) which is equivalent to (x = \phi(x)), and I is any interval containing (x = \alpha) with (|\phi'(x)| < 1) for all (x \in I), then the sequence of approximations (x_1, x_2, x_3, \ldots, x_n) will converge to the root (\alpha) provided that the initial approximation (x_0) is chosen in I.
Example β Trigonometric Iteration
(f(x) = \cos x - 2x + 3 = 0) rewritten as:
[x = \frac{1}{2}(\cos x + 3) = \phi(x)]
(\phi'(x) = -\frac{1}{2}\sin x)
(f(1) = 1.54030 > 0), (f(2) = -1.41614 < 0), root lies between 1 and 2.
(|\phi'(1)|) and (|\phi'(2)| < 1), so iteration method can be applied.
Let (x_0 = 1.5):
(x_1 = \frac{1}{2}(\cos 1.5 + 3) = 1.999825)
(x_2 = \frac{1}{2}(\cos 1.999825 + 3) = 1.999695)
(x_3 = \frac{1}{2}(\cos 1.999695 + 3) = 1.999695)
This is the required root correct up to 5 decimal places.
Example β Cubic Iteration
Find the real root of (x^3 + x^2 - 1 = 0) by iteration.
(f(x) = x^3 + x^2 - 1)
(f(0) = -1 < 0), (f(1) = 1 > 0), root lies between 0 and 1.
Rewrite: (x^2(x+1) = 1 \implies x = \frac{1}{\sqrt{x+1}} = \phi(x))
(\phi'(x) = -\frac{1}{2}(x+1)^{-3/2})
(\phi'(0) = -\frac{1}{2} < 1), (\phi'(1) = -\frac{1}{2\sqrt{8}} < 1), so (|\phi'(x)| < 1) for all values in the interval.
Let (x_0 = 0.65):
(x_1 = \frac{1}{\sqrt{1.65}} = 0.7784989)
(x_2 = \frac{1}{\sqrt{1.7784989}} = 0.7498479)
(x_3 = \frac{1}{\sqrt{1.7498479}} = 0.7559617)
(x_4 = \frac{1}{\sqrt{1.7559617}} = 0.7546446)
(x_5 = \frac{1}{\sqrt{1.7546446}} = 0.7549278)
(x_6 = \frac{1}{\sqrt{1.7549278}} = 0.7548668)
(x_7 = \frac{1}{\sqrt{1.7548668}} = 0.7548799)
(x_8 = \frac{1}{\sqrt{1.7548799}} = 0.7548771)
(x_9 = \frac{1}{\sqrt{1.7548771}} = 0.7548777)
(x_{10} = \frac{1}{\sqrt{1.7548777}} = 0.7548776)
(x_{11} = \frac{1}{\sqrt{1.7548776}} = 0.7548776)
Hence root is 0.7548776.
Example β Cosine Iteration
Find a real root of (\cos x = 3x - 1) correct to seven decimal places.
(f(x) = \cos x - 3x + 1)
(f(0) = 1 - 0 + 1 = 1 > 0)
(f(\pi/2) = \cos 1.57 - 4.71 + 1 = 0.0007963 - 4.71 + 1 = -3.7092037 < 0)
Root lies between 0 and (\pi/2).
(\phi(x) = \frac{1}{3}(\cos x + 1)), (\phi'(x) = -\frac{1}{3}\sin x)
Since (\sin x) is bounded between -1 and 1, (|\phi'(x)| < 1), so iteration method can be applied.
Let (x_0 = 0.5):
(x_1 = \frac{1}{3}[\cos 0.5 + 1] = 0.6258608)
(x_2 = \frac{1}{3}[\cos 0.6258608 + 1] = 0.6034863)
(x_3 = \frac{1}{3}[\cos 0.6034863 + 1] = 0.6077873)
(x_4 = \frac{1}{3}[\cos 0.6077873 + 1] = 0.6069711)
(x_5 = \frac{1}{3}[\cos 0.6069711 + 1] = 0.6071264)
(x_6 = \frac{1}{3}[\cos 0.6071264 + 1] = 0.6070969)
(x_7 = \frac{1}{3}[\cos 0.6070969 + 1] = 0.6071025)
(x_8 = \frac{1}{3}[\cos 0.6071025 + 1] = 0.6071014)
(x_9 = \frac{1}{3}[\cos 0.6071014 + 1] = 0.6071016)
(x_{10} = \frac{1}{3}[\cos 0.6071016 + 1] = 0.6071016)
Hence root is 0.6071016.
β Key Takeaways
The Regula-Falsi method uses a chord between two points with opposite signs to approximate the root, and its formula is derived from the equation of a line. The initial interval must be sufficiently small for faster convergence. The Method of Iteration rewrites (f(x)=0) as (x=\phi(x)) and generates successive approximations; it converges if (|\phi'(x)|<1) in the interval containing the root. Both methods are applied to algebraic, transcendental, and mixed equations, with iteration commonly requiring more steps but offering clear convergence criteria.
π§ Quick Revision Questions
- Derive the Regula-Falsi formula from the equation of the chord joining two points on the curve.
- Why must the initial interval in Regula-Falsi be sufficiently small?
- State the convergence condition for the Method of Iteration and explain its significance.
- Rewrite (x^3 - 4x + 1 = 0) in the form (x = \phi(x)) suitable for iteration, and verify the convergence condition near the root in (0,1).
- How do you determine when to stop iterating in both methods when a tolerance (TOL) is given?
π Lecture 7 β Newton-Raphson Method
π Overview: This lecture introduces the Newton-Raphson method, a powerful and widely used technique for finding roots of nonlinear equations. It derives the iterative formula from Taylor series, provides a geometric interpretation using tangents, and discusses convergence properties, drawbacks, and practical applications through solved examples.
ποΈ Topics Covered
The lecture covers the derivation of the Newton-Raphson formula from Taylor series expansion, its geometrical interpretation using tangents to the curve, important notes on its application and limitations, step-by-step solved examples for finding roots of polynomial and transcendental equations, drawbacks including divergence at inflection points and division by zero, and a detailed analysis of its quadratic convergence rate.
π Lecture Summary
Newton-Raphson Method
This method is used for finding a root of f(x)=0. The formula is derived by using the first two terms in Taylor's series expansion. Setting f(x_{n+1}) = 0 and simplifying gives the Newton-Raphson iterative formula.
π Definition β Newton-Raphson Formula: x_{n+1} = x_n - f(x_n) / f'(x_n) for n=0,1,2,β¦ This gives successive approximations to the root.
Geometrical Interpretation
If the curve f(x)=0 meets the x-axis at x=Ξ± (the root), and x_0 is a point near Ξ±, then the tangent at P_0[x_0, f(x_0)] is: y - f(x_0) = f'(x_0)(x - x_0). This tangent cuts the x-axis at x_1 = x_0 - f(x_0)/f'(x_0), which is the first approximation. Repeating this process gives better approximations rapidly.
Important Notes
- When f'(x) is very large (slope is large), h will be small, and the root can be found quickly.
- If the initial approximation x_0 is close to the root, convergence is very fast.
- The method fails if f'(x) = 0 in the neighborhood of the root. In such cases, the Regula Falsi method should be used.
- If the initial approximation is not given, choose two points a and b such that f(a) and f(b) have opposite signs. If |f(a)| < |f(b)|, use a as the initial approximation.
- Newton-Raphson is also called the method of tangents.
Example 1: Solve x^3 - x - 1 = 0
Find a real root correct to four decimal places.
Given f(x) = x^3 - x - 1
- f(1) = -1 < 0, f(2) = 5 > 0, so root lies between 1 and 2.
- f'(x) = 3x^2 - 1
- f''(x) = 6x
- f(2) and f''(2) have the same sign (+), so x_0 = 2
π Iteration 1: x_1 = 2 - 5/11 = 1.54545 π Iteration 2: f(1.54545) = 1.14576, f'(1.54545) = 6.16487, x_2 = 1.54545 - 1.14576/6.16525 = 1.35961 π Iteration 3: f(1.35961) = 0.15369, f'(1.35961) = 4.54562, x_3 = 1.35961 - 0.15369/4.54562 = 1.32579 π Iteration 4: f(1.32579) = 4.60959Γ10^-3, f'(1.32579) = 4.27316, x_4 = 1.32579 - 0.00460959/4.27316 = 1.32471 π Iteration 5: f(1.32471) = -3.39345Γ10^-5, f'(1.32471) = 4.26457, x_5 = 1.32471 + 0.0000339345/4.26457 = 1.324718 Since f(x_5) = 1.823Γ10^-7, the required root is 1.3247.
Comparison with Bracketing Methods
Methods like Bisection and False Position require bracketing the root and are always convergent. The Newton-Raphson method is an open method β it needs only one initial guess and is not guaranteed to converge, but when it does, it converges very quickly.
Drawbacks of Newton-Raphson Method
- Divergence at inflection points: If a guess is near an inflection point (where f''(x) = 0), the iterations may diverge.
- Division by zero or near-zero: If f'(x_i) is very small, x_{i+1} can become very large.
- Root jumping: For oscillating functions with multiple roots, an initial guess close to one root may converge to a different root.
- Oscillations near local max/min: Iterations may oscillate about a local maximum or minimum without converging to a root, eventually causing division by near-zero.
Convergence of Newton-Raphson Method
Comparing the N-R formula with the general iteration formula x_{n+1} = Ο(x_n), we have Ο(x) = x - f(x)/f'(x). The method converges if |Ο'(x)| < 1, which requires f(x) f''(x) < [f'(x)]^2 in the interval of interest.
Definition: Let x_n = Ξ± + Ξ΅_n and x_{n+1} = Ξ± + Ξ΅_{n+1}, where Ξ± is the root. If Ξ΅_{n+1} = K Ξ΅_n^p, then p is the rate of convergence.
The N-R method has quadratic convergence (p=2). Using Taylor expansion and the fact that f(Ξ±)=0, we derive: Ξ΅_{n+1} = [Ξ΅_n^2 f''(Ξ±)] / [2 f'(Ξ±)]
So Ξ΅_{n+1} = K Ξ΅_n^2 where K = f''(Ξ±) / [2 f'(Ξ±)]. This shows second-order (quadratic) convergence.
Example: Square Root of a Positive Number N
To find βN, solve f(x) = x^2 - N = 0. With f'(x) = 2x, the N-R iteration is: x_{n+1} = x_n - (x_n^2 - N)/(2x_n) = (1/2)(x_n + N/x_n)
π Example: Find β12
- Initial guess: x_0 = (3 + 4)/2 = 3.5
- x_1 = (1/2)(3.5 + 12/3.5) = 3.4643
- x_2 = (1/2)(3.4643 + 12/3.4643) = 3.4641
- x_3 = (1/2)(3.4641 + 12/3.4641) = 3.4641 Hence, β12 = 3.4641
Example: Solve f(x) = x - 0.8 - 0.2 sin x in [0, Ο/2]
Given f(x) = x - 0.8 - 0.2 sin x
- f(0) = -0.8, f(1.57) β 0.570, so root exists.
- f'(x) = 1 - 0.2 cos x
- Since |f(0)| is less, take x_0 = 1.57? No β the note says if |f(a)| < |f(b)|, take a. Here we start with x_0 = 0 based on |f(0)| = 0.8 vs |f(1.57)|=0.57; 0.57 < 0.8, so x_0 = 1.57? The text uses x_0 = 0 differently. Letβs follow the worked solution.
π Iteration 1: x_1 = 0 - (-0.8)/0.8 = 1 π Iteration 2: f(1)=0.0317, f'(1)=0.8919, x_2 = 1 - 0.0317/0.8919 = 0.9645 π Iteration 3: f(0.9645)=0.0002, f'(0.9645)=0.88604, x_3 = 0.9645 - 0.0002/0.88604 = 0.9643
Example: Solve f(x) = 4x cos x - (x-2)^2 in [0, 8]
Given f(x) = 4x cos x - (x-2)^2
- f(0) = -4, f(8) β -66.645, so both negative β no sign change.
- f'(x) = 4 cos 2x - 8x sin 2x - 2(x-2)
- Since |f(8)| is larger, x_0 = 8
π Iteration 1: x_1 = 8 - (-66.645)/2.5952 = 33.6801 π Iteration 2: x_2 = 33.6801 + 1028.25/200.79 = 38.8011 π Iteration 3: x_3 = 38.8011 + 1446.14/(-326.205) = 43.2343 (Note: This example shows divergence β the iterates are moving away from the [0,8] interval.)
Example: Solve ln(x-1) + cos(x-1) = 0 in [1.2, 2]
Given f(x) = ln(x-1) + cos(x-1)
- f(1.2) = -0.6293, f(2) = 0.5403
- f'(x) = 1/(x-1) - sin(x-1)
- f'(1.2) = 5 - 0.1986 = 4.8014, so x_0 = 1.2
π Iteration 1: x_1 = 1.2 - (-0.6293)/4.8014 = 1.3311 π Iteration 2: f(1.3311)=-0.1596, f'(1.3311)=2.6951, x_2 = 1.3311 - (-0.1596)/2.6951 = 1.3903 π Iteration 3: f(1.3903)=-0.016, f'(1.3903)=2.1816, x_3 = 1.3903 - (-0.016)/2.1816 = 1.3976
π‘ Why this matters: The Newton-Raphson method is essential for quickly finding roots of complex equations where analytical solutions are impossible, such as in engineering, physics, and economics. Its quadratic convergence makes it highly efficient when a good initial guess is provided.
β Key Takeaways
The Newton-Raphson method is a powerful open iterative method for finding roots of f(x)=0, using the formula x_{n+1} = x_n - f(x_n)/f'(x_n), derived from Taylor series expansion. It has quadratic convergence, meaning the error reduces as the square of the previous error, making it very fast when the initial guess is close to the root. However, the method has critical limitations: it fails when f'(x)=0 near the root, may diverge at inflection points, and can exhibit root jumping or oscillations near local maxima/minima. For optimal results, the initial approximation should be chosen carefully β if bracketing is possible, choose the endpoint where f(x) and f''(x) have the same sign. The method is particularly useful for extracting square roots and solving transcendental equations.
π§ Quick Revision Questions
-
Derive the Newton-Raphson formula using the first two terms of Taylor's series expansion. What assumption is made about f(x_{n+1})?
-
Explain the geometrical interpretation of the Newton-Raphson method. How does the tangent line at each iteration lead to the improved estimate?
-
What are the four main drawbacks of the Newton-Raphson method? Under what conditions does each problem occur?
-
Show that Newton-Raphson method has quadratic convergence. Starting from x_n = Ξ± + Ξ΅_n, derive the relationship Ξ΅_{n+1} = K Ξ΅_n^2 and find the expression for K.
-
Using Newton's method, derive the iterative scheme for finding the square root of a positive number N. Then find β12 correct to four decimal places using this scheme.