fixed point theorem notes

Author: Daniel-42-z

images/math-talk/fixed-point-theorem/background-infinity.png

@@ -0,0 +1,354 @@
+---
+layout: default
+title: Fixed Point Theorem
+permalink: /math-talks/fixed-point-theorem/
+toc: true
+---
+
+# Fixed Point Theorem: Why Must Something Stay Unchanged?
+
+Mathematical Analysis • Mathematical Foundations • Analysis & Applications
+
+---
+
+## Building Intuition: The Crumpled Map Problem
+
+A thought experiment that reveals a profound mathematical truth.
+
+### Imagine This Scenario
+
+You take a map, **crumple it up**, and place it back on the table. The map is now folded, wrinkled, and distorted in countless ways.
+
+### The Central Question
+
+Is there a point on the map that remains in the **exact same location** as before?
+
+### Mathematical Insight
+
+- **Intuition says:** Probably not — everything moved!
+- **Mathematics says:** Absolutely **YES**!
+
+There must exist at least one point that hasn't moved.
+
+---
+
+## Formal Definition: What is a Fixed Point?
+
+### The Mathematical Definition
+
+A point $x$ is called a **Fixed Point** if it satisfies: $f(x) = x$
+
+In other words: The point remains completely unchanged when the function $f$ is applied to it.
+
+- **Input equals Output**
+- **An "Anchor" Point**
+- **Invariant Under $f$**
+
+---
+
+## Concrete Example: $f(x) = x^2$
+
+Let's find the fixed points of a basic quadratic function.
+
+### Step-by-Step Solution
+
+1. **Set up the equation:** $f(x) = x$ $x^2 = x$
+2. **Rearrange:** $x^2 - x = 0$
+3. **Factor:** $x(x - 1) = 0$
+4. **Solution:** $x = 0$ or $x = 1$
+
+### Graphical Visualization
+
+![](/images/math-talk/fixed-point-theorem/y-equals-x-squared-fixed-points.png){: width="75%"}
+
+**Key Insight:** Fixed points occur where the curve $y = x^2$ intersects the line $y = x$ (shown in gold).
+
+---
+
+## The Challenge: The Key Question
+
+### What if the function is more complicated?
+
+- **Simple Cases:** For $f(x) = x^2$, we can solve algebraically. But what about functions without closed-form solutions?
+- **The Real Challenge:** Can we still **guarantee** that a fixed point exists, even when we can't find it explicitly?
+
+**We need a general theorem — not just examples!** (Existence Proof • General Guarantee • Universal Result)
+
+---
+
+## Foundation: Review: Intermediate Value Theorem (IVT)
+
+A powerful tool from calculus that we'll use as our foundation.
+
+### The IVT Statement
+
+If $f$ is **continuous** on $[a, b]$ and $f(a) \cdot f(b) < 0$ then there exists $c \in (a, b)$ such that $f(c) = 0$
+
+**Core Insight:** A sign change guarantees crossing zero.
+
+### Visual Intuition
+
+![](/images/math-talk/fixed-point-theorem/ivt-visual-intuition.png){: width="75%"}
+
+- Continuous curve from negative to positive.
+- Must cross the $x$-axis somewhere.
+
+---
+
+## Strategy: Reformulating the Problem
+
+The crucial transformation that connects fixed points to IVT.
+
+### The Key Transformation
+
+- **Original Problem:** $f(x) = x$
+- **Reformulated:** $f(x) - x = 0$
+
+### Why This Works
+
+Finding where $f(x) = x$ is equivalent to finding where $f(x) - x = 0$. This transforms a fixed point problem into a **root-finding problem**!
+
+### The Connection
+
+Now we can apply IVT! If we can show that $f(x) - x$ changes sign, a root (and thus a fixed point) must exist.
+
+---
+
+## Construction: Constructing a New Function
+
+Define the auxiliary function: $g(x) = f(x) - x$
+
+- **$g(x)$ Definition:** A new function that measures the **difference** between $f(x)$ and $x$.
+- **Root of $g(x)$:** If $g(c) = 0$, then $f(c) = c$.
+- **Fixed Point Found!** Finding a root of $g$ is equivalent to finding a fixed point of $f$.
+
+**Key Insight:** We've transformed the problem from "find where $f(x) = x$" to "find where $g(x) = 0$" — a classic root-finding problem solvable with IVT!
+
+---
+
+## Conditions: Applying the Conditions
+
+Setting up the boundary conditions for IVT.
+
+### Our Assumptions
+
+1. **Assumption 1:** $f(0) \geq 0$ (The function at 0 is non-negative)
+2. **Assumption 2:** $f(1) \leq 1$ (The function at 1 doesn't exceed 1)
+
+_Note: These are natural conditions for functions mapping $[0,1]$ into itself._
+
+### What This Means for $g(x)$
+
+- At $x = 0$: $g(0) = f(0) - 0 = f(0) \geq 0$
+- At $x = 1$: $g(1) = f(1) - 1 \leq 0$ (since $f(1) \leq 1$)
+
+**$g(x)$ changes sign from $\geq 0$ to $\leq 0$!**
+
+---
+
+## Application: Applying IVT to $g(x)$
+
+### By the Intermediate Value Theorem
+
+**Conditions Met:**
+
+- $g$ is continuous ($f$ is continuous)
+- $g(0) \geq 0$
+- $g(1) \leq 0$
+
+**IVT Conclusion:** Therefore, there exists some point $c \in (0, 1)$ such that: $g(c) = 0$
+
+We've proven that $g(c) = 0$ for some $c$ in $(0,1)$. But what does this mean for $f$?
+
+---
+
+## Conclusion: Fixed Point Exists!
+
+The final step that completes the proof.
+
+### The Final Deduction
+
+We know: $g(c) = 0$
+
+By definition of $g$: $f(c) - c = 0$
+
+Therefore: $f(c) = c$
+
+### Visual Proof
+
+![](/images/math-talk/fixed-point-theorem/g-c-equals-zero-fixed-point-proof.png){: width="75%"}
+
+**A FIXED POINT EXISTS!** The point $c$ where $g(c) = 0$ is exactly where $f(c) = c$.
+
+---
+
+## Interactive Example: $f(x) = \cos(x)$
+
+Does the cosine function have a fixed point?
+
+- **Does $f(x) = \cos(x)$ have a fixed point?** **Yes!**
+- **Reasoning:**
+  1. **Continuity:** $\cos(x)$ is continuous everywhere.
+  2. **Interval $[0, 1]$:** $\cos(0) = 1 \geq 0$, $\cos(1) \approx 0.54 \leq 1$.
+  3. **Apply Theorem:** All conditions satisfied!
+
+**Fixed point exists in $[0, 1]$!**
+
+---
+
+## Interactive Example: $f(x) = 2x$
+
+A case where conditions matter!
+
+- **Does $f(x) = 2x$ have a fixed point in $(0,1)$?** **No!**
+- **Answer:** Only $x = 0$ is a fixed point, but it's not in $(0,1)$.
+- **Why Theorem Doesn't Apply:**
+  - $f(0) = 0$ (satisfies $f(0) \geq 0$)
+  - $f(1) = 2$ (**violates** $f(1) \leq 1$)
+- **The function maps $[0,1]$ outside itself!** At $x = 1$, $f(1) = 2$, which is outside the interval.
+
+**The conditions are crucial!**
+
+---
+
+## Critical Insight: Continuity is Crucial
+
+Without continuity, the guarantee disappears.
+
+### Why Continuity Matters
+
+If $f$ is **not continuous**, it can "jump" over the fixed point without ever hitting it.
+
+### Counterexample:
+
+$f(x) = x + 0.5$ for $x < 0.5$ $f(x) = x - 0.5$ for $x \geq 0.5$ This function has **no fixed point** because of the discontinuity at $x = 0.5$.
+
+### Visual Explanation
+
+![](/images/math-talk/fixed-point-theorem/discontinuity-no-fixed-point.png){: width="75%"}
+
+- **Continuous function:** must cross $y = x$.
+- **Discontinuous:** can jump over the line.
+
+---
+
+## Extension: Brouwer Fixed Point Theorem
+
+From 1D to 2D: A profound generalization.
+
+### The 2D Version
+
+- **The Setup:** Consider a **closed disk** (or any convex, compact set) in 2D. Let $f$ be a **continuous function** that maps the disk into itself.
+- **The Theorem:** Brouwer's Theorem states: $\exists p \text{ such that } f(p) = p$ At least one fixed point **must exist!**
+
+**Significance:** This extends to any finite dimension! It's one of the most important theorems in topology.
+
+---
+
+## Visualization: 2D Visualization: The Disk Transformation
+
+Any continuous transformation of a disk into itself has a fixed point.
+
+### The Crumpled Map Analogy (2D)
+
+Imagine taking a **circular disk**, crumpling it, stretching it, twisting it, and placing it back on top of itself.
+
+**The Question:** Is there a point that remains in exactly the same position? **Answer: YES!** Brouwer guarantees it.
+
+### Visual Intuition
+
+![](/images/math-talk/fixed-point-theorem/disk-transformation-visual.png){: width="50%"}
+
+No matter how you transform the disk, at least one point must stay fixed.
+
+---
+
+## Philosophy: The Beautiful Intuition
+
+### The Elegant Insight
+
+> "You cannot 'move everything' without leaving at least one point unchanged."
+
+- **Counterintuitive:** It seems like you should be able to move everything, but mathematics says otherwise.
+- **Universal:** This applies to ANY continuous transformation, no matter how complex.
+- **Profound:** This simple idea has applications across mathematics, economics, and physics.
+
+---
+
+## Application: Fixed Point Iteration
+
+A practical numerical method for solving equations.
+
+### The Iteration Method
+
+**THE RECURRENCE** $x_{n+1} = f(x_n)$
+
+- **How it works:** Start with an initial guess $x_0$, then repeatedly apply $f$. If the sequence converges, it converges to a fixed point!
+- **Visual:** The "staircase" or "cobweb" diagram shows how iteration converges to the fixed point where $y = f(x)$ intersects $y = x$.
+
+![](/images/math-talk/fixed-point-theorem/fixed-point-iteration-staircase-cobweb.jpg){: width="75%"}
+
+This turns equation-solving into simple iteration!
+
+---
+
+## Worked Example: Solving $x = \cos(x)$
+
+Using iteration to find the fixed point.
+
+### The Iteration Process
+
+- **Initial Value:** $x_0 = 0.5$
+- **Iteration Formula:** $x_{n+1} = \cos(x_n)$
+- **First few iterations:**
+  - $x_1 = \cos(0.5) \approx 0.8776$
+  - $x_2 = \cos(0.8776) \approx 0.6390$
+  - $x_3 = \cos(0.6390) \approx 0.8027$
+  - $x_4 = \cos(0.8027) \approx 0.6948$
+
+### Convergence Visualization
+
+![](/images/math-talk/fixed-point-theorem/cosine-fixed-point-convergence.png){: width="75%"}
+
+**Converges to $x \approx 0.7391$** This is the unique fixed point of $\cos(x)$!
+
+---
+
+## Theory: Why Iteration Works
+
+### Convergence Conditions
+
+- **When It Works:** Iteration converges to a fixed point under certain conditions: $\vert f'(x)\vert < 1$ near the fixed point
+- **The Intuition:** If the function is "flat enough" near the fixed point, each iteration gets closer:
+  - Small derivative = small steps
+  - Sequence approaches fixed point
+
+**Practical Value:** Fixed point iteration provides a simple, robust numerical method for solving equations that might be difficult to solve algebraically.
+
+---
+
+## Profound Application: Game Theory & Nash Equilibrium
+
+Where fixed point theorems changed economics forever.
+
+### The Nash Equilibrium
+
+- **Definition:** A set of strategies where no player can benefit by changing their strategy unilaterally.
+- **John Nash's Proof (1950):** Nash used **Brouwer's Fixed Point Theorem** to prove that **every finite game has at least one equilibrium!**
+- **Nobel Prize in Economics (1994)**
+- **Example:** Prisoner's Dilemma — both players confessing is the Nash Equilibrium, even though both would be better off cooperating.
+
+Fixed point theorems provide the mathematical foundation for modern game theory!
+
+---
+
+## Summary: Final Summary
+
+### Key Takeaways
+
+1. **Core Idea: Continuity $\Rightarrow$ Existence** A continuous function mapping $[0,1]$ into itself **must** have a fixed point.
+2. **The Strategy: Transform problem $\rightarrow$ Apply IVT** Rewrite $f(x) = x$ as $g(x) = f(x) - x = 0$, then use the Intermediate Value Theorem.
+3. **Generalization: Brouwer's Theorem** Extends to any dimension: Any continuous transformation of a disk into itself has a fixed point.
+4. **Applications: Wide-ranging impact** From numerical methods (fixed point iteration) to economics (Nash Equilibrium).
+
+**In a changing world, something always stays the same.**