Solution of equations

Welcome to the World of Numerical Solutions!

In your earlier math studies, you learned to solve equations like \(x^2 - 5x + 6 = 0\) using factoring or the quadratic formula. But what happens when you meet a "monster" equation like \(x - \cos(x) = 0\)? There isn't a simple algebraic trick to solve that!

This is where Numerical Methods come to the rescue. Instead of finding an exact answer, we use clever "guessing" techniques to get closer and closer to the true value. In this chapter, we’ll explore five main methods to hunt down these elusive roots. Don't worry if it seems a bit abstract at first—think of it like a game of "hot or cold" with numbers!

1. The Bisection Method: The "Higher or Lower" Game

The Bisection Method is the simplest and most reliable way to find a root. It relies on the Change of Sign rule: if a continuous function moves from negative to positive between two points, it must have crossed zero somewhere in between!

How it works (Step-by-Step):

1. Find two numbers, \(a\) and \(b\), where \(f(a)\) and \(f(b)\) have different signs (one is positive, one is negative).
2. Find the midpoint: \(m = \frac{a + b}{2}\).
3. Check the sign of \(f(m)\).
4. Replace either \(a\) or \(b\) with \(m\) so that you still have a sign change.
5. Repeat until your interval is small enough for the accuracy you need!

Analogy: Imagine you are looking for a specific page in a book. You open it exactly in the middle. Is your page number higher or lower? You rip the "wrong" half away and repeat. You’ll find your page eventually!

Quick Review: The Bisection Method is slow but steady. It always works as long as the function is continuous and you start with a sign change.

2. False Position and the Secant Method

While Bisection always goes to the middle, False Position (Linear Interpolation) tries to be smarter. It connects the two points with a straight line and sees where that line hits the x-axis.

False Position vs. Secant Method

• False Position: Always keeps the root "trapped" between two points with different signs. It is very safe.
• Secant Method: Similar to False Position, but it doesn't care about keeping the root trapped. It just uses the two most recent guesses to draw the line.

Common Mistake: Because the Secant Method doesn't "trap" the root, it can sometimes fly off into space and fail to find the answer (it diverges).

Did you know? "Linear Interpolation" is just a fancy way of saying "assuming the graph is a straight line between two points."

3. Fixed Point Iteration: \(x_{n+1} = g(x_n)\)

This method is like a mathematical echo. We rearrange our equation \(f(x) = 0\) into the form \(x = g(x)\). We pick a starting number, plug it in, get an answer, and then plug that answer back into the same formula.

Visualizing the Hunt: Cobwebs and Staircases

When we plot these on a graph, we see two distinct patterns:
• Cobweb Diagrams: These happen when the sequence oscillates (jumps back and forth) around the root. It looks like a spider web spiraling in (or out!).
• Staircase Diagrams: These happen when the sequence approaches the root from one side, looking like a set of steps leading to the answer.

Memory Aid: If the gradient (steepness) of \(g(x)\) is too steep at the root—specifically if \(|g'(x)| > 1\)—the method will fail. The "web" will spiral away instead of in!

4. The Newton-Raphson Method

This is the "Ferrari" of numerical methods. It is incredibly fast because it uses the tangent line (the gradient) at your current guess to find the next one.

The Formula:

\(x_{n+1} = x_n - \frac{f(x_n)}{f'(x_n)}\)

Why it's great: It usually has second-order convergence. This means the number of correct decimal places roughly doubles with every step!

When it fails:
1. If your guess is near a stationary point (where the gradient is zero), the formula involves dividing by zero. Disaster!
2. If the curve is very complex, it might jump to a different root entirely.

Key Takeaway: Fixed point iteration is "first-order" (slow), but Newton-Raphson is "second-order" (fast).

5. Improving Success with Relaxation

Sometimes, a Fixed Point Iteration sequence is frustratingly slow, or even worse, it diverges (goes away from the root). We can use a trick called Relaxation to fix this.

The Relaxed Formula:

Instead of just using \(x_{n+1} = g(x_n)\), we use:
\(x_{n+1} = (1 - \lambda)x_n + \lambda g(x_n)\)

Here, \(\lambda\) (lambda) is a "weighting" factor.
• If the method is diverging, we use a small \(\lambda\) to "pull it back" and force it to converge.
• If it's too slow, we can pick a \(\lambda\) that speeds it up.

Don't worry if this seems tricky: In the exam, the formula for the relaxed iteration is usually given to you. Your job is to plug in the numbers and observe if it's getting closer to the root!

Summary: Choosing Your Weapon

• Need a guaranteed answer? Use Bisection.
• Equation has a clear derivative? Use Newton-Raphson for speed.
• Iteration failing? Try Relaxation or check the gradient of \(g(x)\).
• Using a spreadsheet? These methods are perfect for dragging down formulas to see the "convergence" in real-time!

Common Exam Question: You might be asked to "justify the accuracy" of your answer. To do this, check the sign of the function just above and just below your rounded answer. If the sign changes, the root must be within that tiny gap!