Fun Mathematical Problem in Stochastic Geometry: Random Triangles

This article is part of a sequence about gratifying points, supplied with choices. The earlier one might be found proper right here. This new draw back falls in a category known as stochastic geometry. We define a random triangle as a triangle inscribed in a circle of radius ρ, with its three vertices uniformly distributed on the circle. Without lack of generality, we are going to assume that the circle is centered on the origin, its radius is ρ = 1, and one vertex is positioned at (1, 0). The metric of curiosity proper right here, does not depend upon the scaling concern ρ.

The widespread question is to hunt out the distribution of the random variable R = SQRT(S) / L, the place S is the world of the triangle, and L its perimeter. The first step is to point that R does not depend upon ρ, after which uncover the utmost potential value for R, the minimal being 0. The probability distribution of R might be approximated using Monte-Carlo simulations. Note that R will be unbiased of the unit used for the measurements, because of using SQRT(S) reasonably than S. 

1. The draw back, and backbone

Let us assume that the three vertices of the triangle are (ρ cos θ0, ρ sin θ0), (ρ cos θ1, ρ sin θ1), (ρ cos θ2, ρ sin θ2), with θ0 = 0, and (θ1, θ2) uniformly distributed on [0, 2π] x [0, 2π]. The following outcomes are easy to amass:

Here χ is the indicator carry out, see proper right here. Note that R = R(θ1, θ2) does not depend upon ρ. The most house S is achieved for the equilateral triangle, that is, when θ1 = 2π/3 and θ2 = 4π/3. However this moreover corresponds to the utmost perimeter L. So it is not optimistic that the equilateral triangle achieves the utmost value for R. One technique to affirm that’s to hunt out the utmost of R(θ1, θ2) by differentiating its expression with respect to θ1 and θ2. But there is a quite a bit easier decision: take into consideration a triangle of fixed perimeter: its house is most if the triangle is equilateral. See the detailed decision proper right here. 

The picture underneath represents the possibility distribution for R, with R considered a random variable. The X-axis represents r, and the Y-axis represents P(R  <  r). It was produced using 100,000 simulated random triangles.

Note that the utmost value for R is about 0.219.

2. Generalizations, and related points

For the equilateral triangle, sq., circle and deltoid curve, R is respectively equal to (roughly) 0.219, 0.250, 0.282, and 0.257. The precise value is easy to amass in each case, see desk underneath. It does not depend upon the size. Note that the deltoid (see proper right here and in the picture underneath) is non-convex, thus you may anticipate a lower R. Nothing can beat the circle!

The precise values are as follows:

An fascinating MIT article about random triangles, specializing in the hypothesis of shapes, might be found proper right here. The picture underneath choices 1,000 random triangles from that article, generated using a Gaussian distribution. The draw back might be generalized to random polygons, random polyhedrons (that is, in 3 dimensions, see proper right here) or to random convex models. Other fascinating points in stochastic geometry embrace Buffon’s needle (see proper right here) and partial overlaying of the airplane by infinitely many random circles (see proper right here).

Applications of stochastic geometry (along with stereology and spatial statistics) are described in this e-book, revealed in 2013. A recent e-book (2019) might be found proper right here. 

To receive a weekly digest of our new articles, subscribe to our e-newsletter, proper right here.

About the creator:  Vincent Granville is a data science pioneer, mathematician, e-book creator (Wiley), patent proprietor, former post-doc at Cambridge University, former VC-funded authorities, with 20+ years of firm experience along with CNET, NBC, Visa, Wells Fargo, Microsoft, eBay. Vincent is usually a self-publisher at DataShaping.com, and primarily based and co-founded a variety of start-ups, along with one with a worthwhile exit (Data Science Central acquired by Tech Target). You can entry Vincent’s articles and books, proper right here. Plenty of the most recent ones might be found on vgranville.com.