AlgebraDifficulty 5.8AIME, harderProve itUnited States
Problem:
Denote {1,2,…,n} by [n], and let S be the set of all permutations of [n]. Call a subset T of S good if every permutation σ in S may be written as t1t2 for elements t1 and t2 of T, where the product of two permutations is defined to be their composition. Call a subset U of S extremely good if every permutation σ in S may be written as s−1us for elements s of S and u of U. Let τ be the smallest value of ∣T∣/∣S∣ for all good subsets T, and let v be the smallest value of ∣U∣/∣S∣ for all extremely good subsets U. Prove that v≥τ.
Solution
Solution:
Call an element t∈S an involution if and only if t2 is the identity permutation. We claim that the set of all involutions in S constitutes a good subset of S. The proof is simple. Let s be an arbitrary permutation in S. Note that s may be decomposed into a product of disjoint cycles of length at least 2, and suppose that there are m such cycles in its decomposition. For 1≤i≤m, let li denote the length of the ith cycle, so that s may be written as (a1,1a1,2…a1,l1)(a2,1a2,2…a2,l2)…(am,1am,2…am,lm) for some pairwise distinct elements a1,1,a1,2,…,a1,l1,a2,1,a2,2,…,a2,l2,…,am,1,am,2,…,am,lm of [n]. Consider the permutations q and r defined by q(ai,j)=ai,li+1−j for all 1≤j≤li and r(ai,1)=ai,1 and r(ai,k)=ai,li+2−k for all 2≤k≤li, for all 1≤i≤m, and by q(x)=r(x)=x for all x∈[n] otherwise. Since q,r∈S, q2=r2=1, and rq=s, it follows that the set of all involutions in S is indeed good, as desired.
For all integer partitions λ of n, let fλ denote the number of standard Young tableaux of shape λ. By the Robinson-Schensted-Knuth correspondence, the permutations of [n] are in bijection with pairs of standard Young tableaux of the same shape in such a way that the involutions of [n] are in bijection with pairs of identical standard Young tableaux. In other words, the number of permutations of [n] is equal to the number of pairs of identically shaped standard Young tableaux whose shape is a partition of n, and the number of involutions of [n] is equal to the number of standard Young tableaux whose shape is a partition of n. Hence n!=∑λ(fλ)2 and τ≤∑λn!fλ, where both sums range over all partitions λ of n.
For all elements u∈S, define the conjugacy class of u to be the set of elements that may be written in the form s−1us for some s∈S. It is easy to see that for all u,u′∈S, the conjugacy classes of u and u′ are either identical or disjoint. It follows that S may be partitioned into disjoint conjugacy classes and that any extremely good subset of S must contain at least one element from each distinct conjugacy class. We claim that the number of distinct conjugacy classes of S is at least the number of integer partitions of n. It turns out that the two numbers are in fact equal, but such a result is not necessary for the purposes of this problem. Let u be an arbitrary permutation in S. Recall that u may be written as (a1,1a1,2…a1,l1)(a2,1a2,2…a2,l2)…(am,1am,2…am,lm) for some pairwise distinct elements a1,1,a1,2,…,a1,l1,a2,1,a2,2,…,a2,l2,…,am,1,am,2,…,am,lm of [n]. Associate to the conjugacy class of u the partition n=lw1+lw2+…+lwm+1+…+1, where w1,w2,…,wm is a partition of 1,2,…,m such that w1≥w2≥…≥wm and the 1's represent all the fixed points of s. Now note that if s is any permutation in S, s−1us may be written in the form s−1(a1,1a1,2…a1,l1)ss−1(a2,1a2,2…a2,l2)s…s−1(am,1am,2…am,lm)s which may be written as (b1,1b1,2…b1,l1)(b2,1b2,2…b2,l2)…(bm,1bm,2…bm,lm) for some pairwise distinct elements b1,1,b1,2,…,b1,l1,b2,1,b2,2,…,b2,l2,…,bm,1,bm,2,…,bm,lm of [n] because multiplying by s−1 on the left and by s on the right is equivalent to re-indexing the letters 1,2,…,n. Hence if partitions of n are associated to all conjugacy classes of S analogously, the same partition that is associated to u is associated to s−1us for all s∈S. Since exactly one partition is associated to each conjugacy class, it follows that the number of conjugacy classes cannot exceed the number of partitions, as desired.
To conclude, we need only observe that v≥∑λn!1 by our above claim, for then n!v=n!λ∑1=(λ∑(fλ)2)(λ∑1)≥λ∑fλ≥n!τ by the Cauchy-Schwarz inequality, and this completes the proof.
Want a route through all this instead of an archive? The track
puts 2,000 problems in a working order, from AMC 10 level to the IMO shortlist.
Source: MathNet,
licensed CC-BY-4.0.
Statement reproduced verbatim; metadata (topic, difficulty) added by this project.