Skip to content
All apps

Konvexe Mengen, Hülle und Ecken

About this app

Die Grundbegriffe hinter der linearen Optimierung in 2D und 3D: konvexe Menge (Sehnenregen findet Gegenbeispiele, Ring und Stern gegen Dreieck und Halbebene), konvexe Hülle (Gummiband und Schrumpffolie), konvexe Linearkombination (Gewichte als Massen, alle Kombinationen füllen die Hülle, negative Gewichte führen hinaus), konvexes Polyeder als Schnitt von Halbräumen (Produktionsplanung, Werkstatt-LP) und Ecke als Extrempunkt (Speichen durch einen Probepunkt; Kreis und Kugel mit unendlich vielen Extrempunkten). Dazu die Zielgerade, die in einer Ecke oder auf einer ganzen Kante hängen bleibt.

With explain mode: an AI tutor walks you through the app, points out what matters in the scene and answers your questions.

What it is about

A set is convex if, together with two points, it contains the whole segment between them. Convex linear combinations are weighted averages with non-negative weights summing to one; together they form the convex hull. A convex polyhedron is the intersection of finitely many half-spaces, which is exactly the feasible region of a linear programme. Its corners or extreme points lie in the interior of no segment of the set. This yields the fundamental theorem of linear optimisation: a linear objective attains its optimum at a corner.

What the app shows

Six terms, each in the plane and in space. A chord rain tests shapes such as an ellipse, a star, a ring, a torus or two balls for convexity. A rubber band and shrink wrap close around points as their convex hull. Weights appear as masses whose centre of mass is the combination. Half-planes and planes cut out the feasible region of a production planning LP and a workshop LP. Spokes through a probe point show whether it is a corner. Finally the objective line slides to the optimal corner.

Try it yourself

Drag points A and B into the star and find a segment that leads outside. Move an interior point outwards until it changes the hull. Allow negative weights and watch x leave the hull. Switch off constraints and see when the polyhedron becomes unbounded. Move the probe point into a corner, onto an edge or onto the boundary of the disc. Turn the objective until two corners are equally good and a whole edge becomes optimal.

Frequently asked questions

What is a convex set?
A set X is convex if for any two points x and y in X every point λx + (1 − λ)y with 0 ≤ λ ≤ 1 also lies in X, that is the whole connecting segment. Discs, triangles and half-planes are convex, a ring or a star is not. The intersection of convex sets is convex again.
What is an extreme point or corner?
A point x of a convex set X is an extreme point if it cannot be written as λy + (1 − λ)z with two different points y, z in X and 0 < λ < 1. For a polyhedron these are exactly the corners. For a disc every boundary point is extreme, while a half-plane has no extreme point at all.
Why does the optimum of a linear programme lie at a corner?
The feasible region of an LP is a convex polyhedron, and every point of a bounded polyhedron is a convex combination of its corners. A linear function cannot be larger at a convex combination than at the best corner involved. So it suffices to check the finitely many corners; in a tie a whole edge or face is optimal.
Can I switch between 2D and 3D in the app?
Yes. Each of the six terms has a view in the plane and one in space, for example rubber band and shrink wrap, triangle and tetrahedron, or objective line and objective plane. The magnifier explains every part of the scene at three levels.

Subject: Operations research

More from Operations research

A VisuApp by heyprof: interactive, free in your browser, no sign-up. What is a VisuApp? · All apps