The Setting
For Fortuin-Kasteleyn planar maps with Potts parameter , set and . The inventory coding represents map pieces by words over an alphabet of two burger types and three order types (two fixed and one flexible), reduced via burger and order stacks. On the event of a left or right cut at the origin in a stationary bi-infinite sequence of random pieces, the question is what limiting proportions of burger types and order types survive at the bottom of long reduced intervals, and whether the flexible-order weight is positive with equal weights on the two fixed orders.
Inventory Evaluation and Piece Laws
Let denote the two burger types, let denote the inventory orders consisting of the fixed orders for (with for types ) and the flexible order ✦, and let be the alphabet consisting of a burger letter for each and an order letter for each . An inventory stack evaluation state is a pair consisting of a list of pending orders in and a list of unmatched burger types in stored freshest-first.✦
The evaluation of a finite word over on an evaluation state is the evaluation state obtained by updating letter by letter from left to right via the single-step stack transition (returning on the empty word), in which each order consumes the freshest eligible burger in .✦ For a finite word over , the word reduction is the evaluation state obtained by evaluating from empty pending orders and an empty burger stack.✦ For a finite list of length and an integer , the bottom segment is the suffix of obtained by dropping its first entries (leaving the final entries, or all of when ).✦
For , the letter mass function is the function assigning weight to each burger letter, to the flexible order letter , and to each fixed order letter and ✦, and the letter distribution is the probability distribution on with weights given by .✦ For and any index set , the independent and identically distributed (i.i.d.) inventory stream measure on is the infinite product measure of copies of , with .✦ Given any probability measure on a measurable space , the i.i.d. bi-infinite sequence law on is the infinite product measure of copies of at every coordinate.✦
The letter increment is the map assigning to each burger letter and to each order letter.✦ For a one-sided sequence in and , the letter increment sum is ✦, and for an initial level the downward inventory path is .✦ For an integer-valued sequence , the first hitting time of zero is the least with , or if no such exists.✦ For and , the first downward hitting time from level is .✦ For and , the revealed word of length is (empty for )✦, and for a map the stopped word is , where when and otherwise.✦ For and , the reversed right piece word is ✦, the reversed right piece law is the pushforward of under ✦, and the right piece law is the pushforward of under word reversal.✦
At step for , the survivor indicator is if the burger stack of is empty and otherwise✦, and the first survivor time is .✦ For , set if and otherwise; the unmatched order count is if and otherwise equals the number of pending orders in ✦, the first survivor type is the freshest burger type in the burger stack of (defaulting to type if that stack is empty)✦, and the first survivor data is the pair of the unmatched order count and first survivor type.✦ For and , the shifted stream is defined by for ✦, and the first survivor tail is .✦ For and , the parsing step survivor data at step is the pair , and the parsing step defect count is .✦
For a sequence of nonnegative integer defect counts on a space and a point , the defect walk at step is (1)✦ and the defect walk survival event at step is .✦ The defect stopping index of (1) is the least with (the first step where ), or if for all .✦ For and , the left parsing step count is .✦ At each parsing step for , a finite word over (the -th raw chunk of ) reduces to an evaluation state with an assigned normal word over ; for , the short piece prefix is the empty word at , and at step is obtained by prepending the -th normal word to .✦ The short left piece word of is ✦, and for the shortened left piece law is the pushforward of under .✦
Cut Events, Outputs, and Counts
Given a sequence of finite words over and , the concatenated piece prefix is the empty word at and at .✦ For a bi-infinite word sequence , an integer , and , the concatenated interval word is ✦, and its reduction gives the reduced interval order stack and reduced interval burger stack . A bi-infinite word sequence has a left cut at , written , if is empty for all ✦, and has a right cut at , written , if is empty for all .✦
For a bi-infinite word sequence , integers , a burger type , and an order , writing for the number of entries equal to in a list , the bottom burger count and the terminal order count are (2)✦ (3)✦
For a predicate on and , the left cut distance and right cut distance are the least satisfying and , respectively, whenever such a exists (and nonnegative integers otherwise). Given a predicate on , a list family indexed by and , a default value , and , the cut-interval output is the entry at index of (indexing from ), or when that list has length at most .✦ For a bi-infinite word sequence and , the right-cut output order is (4)✦ For and a bi-infinite word sequence , the output order indicator is the indicator that equals if and otherwise.✦ For and , the output order mass is (5)✦
The Claim
Theorem 1. Let . For each burger type and -almost every sequence satisfying , (6)✦ and for each order and -almost every sequence satisfying , (7)✦ The output order masses satisfy (8)✦ (9)✦ and for each , (10)✦
In the paper, the result reads: almost surely, each burger type has frequency in the stationary L-output, the stationary R-output has almost-sure frequencies , , and for the two fixed orders and the flexible order , respectively, for a deterministic weight , and for a random sequence and each , on the event and on the event (where , , , , and stand for the paper’s , , , being an L-cut, and being an R-cut).
What Is Checked
The claims are stated for any , each burger type , and each order over the stationary bi-infinite sequence laws and built from the shortened left piece law and the right piece law , on the left-cut event and right-cut event for the bottom burger counts (2), the terminal order counts (3), and the output order mass (5) defined via (4).