Section 4.13

Geometrical probability

[18/100%]
Show
LaTeX
Problem 4.13.1

Pick two points A and B independently at random on the circumference of a circle CC with centre OO and unit radius. Let Π\Pi be the length of the perpendicular from OO to the line ABA B, and let Θ\Theta be the angle ABA B makes with the horizontal. Show that (Π,Θ)(\Pi , \Theta ) has joint density

f(p,θ)=1π21−p2,0≤p≤1,0≤θ<2π f(p, \theta ) = \frac{1}{\pi ^{2} \sqrt{1-p^{2}}}, \quad 0 \leq p \leq 1,0 \leq \theta < 2 \pi
?
Problem 4.13.2

Let S1S_{1} and S2S_{2} be disjoint convex shapes with boundaries of length b(S1),b(S2)b\left(S_{1}\right), b\left(S_{2}\right), as illustrated in the figure beneath. Let b(H)b(H) be the length of the boundary of the convex hull of S1S_{1} and S2S_{2}, incorporating their exterior tangents, and b(X)b(X) the length of the crossing curve using the interior tangents to loop round S1S_{1} and S2S_{2}. Show that the probability that a random line crossing S1S_{1} also crosses S2S_{2} is {b(X)−b(H)}/b(S1)\left\{ b(X)-b(H)\right\} / b\left(S_{1}\right). (See Example (4.13.2) for an explanation of the term 'random line'.) How is this altered if S1S_{1} and S2S_{2} are not disjoint?

The circles are the shapes S1S_{1} and S2S_{2}. The shaded regions are denoted AA and BB, and b(X)b(X) is the sum of the perimeter lengths of AA and BB.

?
Problem 4.13.3

Let S1S_{1} and S2S_{2} be convex figures such that S2⊆S1S_{2} \subseteq S_{1}. Show that the probability that two independent random lines λ1\lambda_{1} and λ2\lambda_{2}, crossing S1S_{1}, meet within S2S_{2} is 2π∣S2∣/b(S1)22 \pi \left|S_{2}\right| / b\left(S_{1}\right)^{2}, where ∣S2∣\left|S_{2}\right| is the area of S2S_{2} and b(S1)b\left(S_{1}\right) is the length of the boundary of S1S_{1}. (See Example (4.13.2) for an explanation of the term 'random line'.)

?
Problem 4.13.4

Let ZZ be the distance between two points picked independently at random in a disk of radius aa. Show that E[Z]=128a/(45π)\mathbb {E}\left[Z\right] = 128 a /(45 \pi ), and E[Z2]=a2\mathbb {E}\left[Z^{2}\right] = a^{2}.

?
Problem 4.13.5

Pick two points AA and BB independently at random in a ball with centre OO. Show that the probability that the angle AOB^\widehat{\mathrm{AOB}} is obtuse is 58\frac{5}{8}. Compare this with the corresponding result for two points picked at random in a circle.

?
Problem 4.13.6

A triangle is formed by A,B\mathrm{A}, \mathrm{B}, and a point P picked at random in a set SS with centre of gravity G. Show that E[∣ABP∣]=∣ABG∣\mathbb {E}\left[\left|\mathrm{ABP}\right|\right] = \left|\mathrm{ABG}\right|.

?
Problem 4.13.7

A point DD is fixed on the side BCB C of the triangle ABCA B C. Two points PP and QQ are picked independently at random in ABDA B D and ADCA D C respectively. Show that E[∣APQ∣]=∣AG1G2∣=29∣ABC∣\mathbb {E}\left[\left|A P Q\right|\right] = \left|\mathrm{AG}_{1} \mathrm{G}_{2}\right| = \frac{2}{9}\left|A B C\right|, where G1G_{1} and G2\mathrm{G}_{2} are the centres of gravity of ABD and ADC.

?
Problem 4.13.8

From the set of all triangles that are similar to the triangle ABCA B C, similarly oriented, and inside ABC, one is selected uniformly at random. Show that its mean area is 110∣ABC∣\frac{1}{10}\left|\mathrm{ABC}\right|.

?
Problem 4.13.9

Two points XX and YY are picked independently at random in the interval (0,a)(0, a). By varying aa, show that F(z,a)=P(∣X−Y∣≤z)F(z, a) = \mathbb {P}\left(|X-Y| \leq z\right) satisfies

∂F∂a+2aF=2za2,0≤z≤a \frac{\partial F}{\partial a}+\frac{2}{a} F = \frac{2 z}{a^{2}}, \quad 0 \leq z \leq a

and hence find F(z,a)F(z, a). Let r≥1r \geq 1, and show that mr(a)=E[∣X−Y∣r]m_{r}(a) = \mathbb {E}\left[|X-Y|^{r}\right] satisfies

admrda=2{arr+1−mr} a \frac{d m_{r}}{d a} = 2\left\{ \frac{a^{r}}{r+1}-m_{r}\right\}

Hence find mr(a)m_{r}(a).

?
Problem 4.13.10

Lines are laid down independently at random on the plane, dividing it into polygons. Show that the average number of sides of this set of polygons is 4. [Hint: Consider nn random great circles of a sphere of radius RR; then let RR and nn increase.]

?
Problem 4.13.11

A point PP is picked at random in the triangle ABCA B C. The lines AP,BP,CPA P, B P, C P, produced, meet BCB C, AC,ABA C, A B respectively at L,M,NL, M, N. Show that E[∣LMN∣]=(10−π2)∣ABC∣\mathbb {E}\left[\left|L M N\right|\right] = \left(10-\pi^{2}\right)\left|A B C\right|.

?
Problem 4.13.12

If four points are picked independently at random inside the triangle ABCA B C, show that the probability that no one of them lies inside the triangle formed by the other three is 23\frac{2}{3}.

?
Problem 4.13.13

If three points P,Q,RP, Q, R are picked independently at random in a disk of radius aa, show that E[∣PQR∣]=35a2/(48π)\mathbb {E}\left[\left|\mathrm{PQR}\right|\right] = 35 a^{2} /(48 \pi ). [You may find it useful that ∫0π∫0πsin⁡3xsin⁡3ysin⁡∣x−y∣dxdy=35π/128\int_{0}^{\pi } \int_{0}^{\pi } \sin^{3} x \sin^{3} y \sin \left|x-y\right| d x d y = 35 \pi / 128.]

?
Problem 4.13.14

Two points AA and BB are picked independently at random inside a disk CC. Show that the probability that the circle having centre A and radius ∣AB∣\left|\mathrm{AB}\right| lies inside CC is 16\frac{1}{6}.

?
Problem 4.13.15

Two points A and B are picked independently at random inside a ball SS. Show that the probability that the sphere having centre A and radius ∣AB∣\left|\mathrm{AB}\right| lies inside SS is 120\frac{1}{20}.

?
Problem 4.13.16

Pick two points independently and uniformly at random on the surface of the sphere of R3\mathbb {R}^{3} with radius 1. Show that the density of the Euclidean distance DD between them is f(d)=12df(d) = \frac{1}{2} d for d∈(0,2)d \in (0,2).

?
Problem 4.13.17

Two points are chosen independently and uniformly at random on the perimeter (including the diameter) of a semicircle with unit radius. What is the probability of the event DD that exactly one of them lies on the diameter, and of the event NN that neither lies on the diameter?

Let AA be the area of the triangle formed by the two points and the midpoint of the diameter. Show that E[A∣D]=1/(2π)\mathbb {E}\left[A \mid D\right] = 1 /(2 \pi ), and E[A∣N]=1/π\mathbb {E}\left[A \mid N\right] = 1 / \pi. Hence or otherwise show that E[A]=1/(2+π)\mathbb {E}\left[A\right] = 1 /(2+\pi ).

?
Problem 4.13.18

Stevens's solution of Jeffreys's bicycle wheel problem. After passing over a stretch of road strewn with tacks, a cyclist looks repeatedly at the front wheel to check whether the tyre has picked up a tack. One inspection of the wheel covers a fraction xx of the tyre. After nn independent such inspections, each uniformly positioned on the wheel, show that the probability of having inspected the entire tyre is 1−(n1)(1−x)n−1+(n2)(1−2x)n−1−⋯1-\binom {n}{1}(1-x)^{n-1}+\binom {n}{2}(1-2 x)^{n-1}-\cdots, where the series terminates at the term in 1−kx1-k x where k=⌊1/x⌋k = \lfloor 1 / x\rfloor is the integer part of 1/x1 / x.

?