Symmetry still for the win!

A completely random walk (infinitesimal steps with equal probability in all directions) starts at ( cot ( π ( 1 3 1 2 + 1 2 ) ) , cot ( π ( 1 4 1 3 + 1 2 ) ) , cot ( π ( 1 5 1 4 + 1 2 ) ) , , cot ( π ( 1 n + 2 1 n + 1 + 1 2 ) ) , 1 ) \left(\cot\left(\pi\left(\frac13-\frac12+\frac12\right)\right),\cot\left(\pi\left(\frac14-\frac13+\frac12\right)\right),\cot\left(\pi\left(\frac15-\frac14+\frac12\right)\right),\dots,\cot\left(\pi\left(\frac1{n+2}-\frac1{n+1}+\frac12\right)\right),1\right) and ends when the last coordinate is no longer positive. As n n goes to infinity, what is the reciprocal of the probability of the walker stopping with no positive coordinates, given that there is additionally some unknown force that doubles this probability with each dimension?

Inspired by Symmetry for the win! <3


The answer is 3.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

1 solution

Ryan S
Jun 7, 2020

For the axis of the dimension of any coordinate that is not the last, let P k P_k be the probability that the walker touches the non-positive part of this axis before the positive when they start at an angle of θ 0 = k n π \theta_0=\frac kn\pi radians from the axis, that angle which we will name θ \theta , of the dimension of the last coordinate, with position restricted to these two dimensions and ignoring the probability doubling force. k n \frac kn is rational and can approximate any value to an arbitrary accuracy. The question's condition is met when the event associated with P n P_n occurs, and thus we label P n = 1 P_n=1 , and is not when that with P 0 P_0 does, and so we let P 0 = 0 P_0=0 . There is an equal probability at any time, that the walk will change θ \theta positively by an amount as negatively by the same amount, given the walk does not end as a result, and so P k = 1 2 P k + 1 + 1 2 P k 1 P_k=\frac12P_{k+1}+\frac12P_{k-1} . We can rearrange this to P k + 1 = P k + ( P k P k 1 ) P_{k+1}=P_k+(P_k-P_{k-1}) and see that P P undergoes an arithmetic progression. Should it exist (we may make it), P n 2 P_{\frac n2} is 1 2 \frac12 , as its representative event has the point rotationally half-way between the positive and negative parts of the axis. Thus, P k = k n = θ 0 π P_k=\frac kn=\frac{\theta_0}\pi .

The question's condition occurs when this happens for all non-last dimensions with the last dimension. Thus we must multiply, still ignoring the force, the θ 0 π \frac{\theta_0}\pi s of each dimension. A rearrangement of the fractions of each non-last initial coordinate yields values of the form cot ( π ( q ( q + 3 ) 2 ( q + 1 ) ( q + 2 ) ) ) \cot\left(\pi\left(\frac{q(q+3)}{2(q+1)(q+2)}\right)\right) , for q q from 1 1 to the n n as defined by the question. Height of 1 1 , θ 0 \theta_0 is arctan 1 cot ( π ( q ( q + 3 ) 2 ( q + 1 ) ( q + 2 ) ) ) = π q ( q + 3 ) 2 ( q + 1 ) ( q + 2 ) \arctan\frac1{\cot\left(\pi\left(\frac{q(q+3)}{2(q+1)(q+2)}\right)\right)}=\pi\frac{q(q+3)}{2(q+1)(q+2)} absolutely. Our resulting product is, dividing each term by π \pi and considering the force and limit, lim n q = 1 n q ( q + 3 ) ( q + 1 ) ( q + 2 ) = 1 3 \lim\limits_{n\to\infty}\prod\limits_{q=1}^n\frac{q(q+3)}{(q+1)(q+2)}=\frac13

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...