Winding numbers and domain coloring

This video presents an elegant, visually driven algorithm for solving two-dimensional equations, using domain coloring to visualize complex functions and extending the one-dimensional bisection method into higher dimensions. The key insight is that a loop with a non-zero winding number guarantees a zero inside, so the solver recursively splits such regions to isolate solutions. This approach not only yields a practical numerical tool for finding polynomial zeros but also leads to a beautiful proof of the fundamental theorem of algebra, since a polynomial's leading term forces a winding number equal to its degree. Beyond the specific technique, the deeper meta-lesson is the value of designing mathematical constructs that "compose nicely," allowing winding numbers to add along paths and cancel on loops.

• ▶ 0:03 The video's main topic is a visually appealing algorithm for solving two-dimensional equations, with the deeper meta-lesson of designing constructs that "compose nicely."

• ▶ 1:29 The bisection method for zero-finding is demonstrated: repeatedly checking midpoints and keeping the half where the function changes sign, which narrows in on a zero.

• ▶ 2:46 The core heuristic is that solving equations equals finding zeros, and because a smooth function that is positive at one point and negative at another must cross zero, the same idea extends to two-dimensional equations.

• ▶ 3:38 Visualizing 2D functions requires alternatives since graphing both 2D input and output needs 4D; options include side-by-side input/output spaces, vector fields, and domain coloring.
• ▶ 4:38 Domain coloring maps output direction to hue and distance from origin to brightness, making the color of each input point show where the function maps it, with zeros appearing as black dots.
• ▶ 6:50 The central challenge is creating an efficient algorithm to find inputs that map to zero, avoiding computing every pixel, plus a theoretical construct for determining if a zero exists inside a region.
• ▶ 7:39 In 2D, zero-finding generalizes from sign changes in 1D to checking whether a region’s boundary loop covers every possible output direction/color.
• ▶ 8:34 The naive hypothesis: if a boundary covers all directions/colors, then a zero must exist inside that region.
• ▶ 11:24 This naive solver fails: after splitting a colorful region, neither subregion’s boundary covers all colors, so the search stops without finding a zero—unlike 1D bisection, where opposite-sign endpoints always survive.
• ▶ 14:31 Winding along connected paths is additive: a second short path adds to the first, so total winding simply adds up.
• ▶ 15:26 On a closed loop, the winding number is always a whole number (e.g., 3 for three full clockwise turns), and counterclockwise winding counts as negative, allowing cancellation.
• ▶ 18:32 A loop with non-zero winding number guarantees a zero inside: repeatedly splitting it homes in on a point, giving a two-dimensional equation solver for equations like f(x)=g(x).
• ▶ 20:05 The algorithm recursively splits loops with non-zero winding numbers to isolate polynomial zeros, converging to all solutions in a complex plane example.
• ▶ 21:26 Because a polynomial’s leading term forces a total winding number equal to its degree, winding-number reasoning proves that every complex polynomial has a zero—the fundamental theorem of algebra.
• ▶ 22:14 Winding numbers offer both practical value (numerically solving equations) and theoretical value (guaranteeing existence of solutions), with a follow-up promised on topological applications.

• ▶ 0:03 Introduction and One-Dimensional Motivation (0:03 - 3:22) - Summary: Introduces the main and meta topics, motivates zero-finding with real-valued functions and 1D bisection, and extends the viewpoint to 2D equations.
• ▶ 3:18 Visualizing 2D Functions and the Zero-Finding Challenge (3:18 - 7:39) - Summary: Explains color-coding the input/output spaces to visualize 2D functions and poses the efficiency/existence challenge for 2D zero-finding.
• ▶ 7:39 Boundary Arguments and the Failure of a Naive Solver (7:39 - 14:10) - Summary: Generalizes the 1D sign argument to boundary directions, tries a 2D solver that can miss zeros, and motivates tracking color order as a better property.
• ▶ 14:10 Winding Numbers and the 2D Equation Solver (14:10 - 20:05) - Summary: Introduces winding of color paths/loops, uses additivity/Stokes' theorem to detect zeros, and completes an efficient 2D equation solver.
• ▶ 20:05 Examples, Implications, and Conclusion (20:05 - 23:10) - Summary: Shows the algorithm on complex examples, connects total winding to the fundamental theorem of algebra, and closes with practical/theoretical value and follow-up.