This preview shows pages 1–3. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: TODAY: Short recurrence for counting lattice paths by qweight Number partitions Short recurrence for counting lattice paths by qweight Warmup: combinatorial interpretation of (a choose b) b = a (a1 choose b1) Consequence: (a choose b) = (a/b) (a1 choose b1) Put a=m+n, b=m (motivation: counting lattice paths from (0,0) to (m,n)) Let P(m,n;1) = the polynomial P(m,n) from the homework due today, evaluated at q=1; i.e. P(m,n;1) = the number of lattice paths from (0,0) to (m,n). P(m,n;1) = (m+n)/(m) P(m1,n;1) This is a special case of a qversion you proved in HW #10: P(m,n) = ((1q^(m+n))/(1q^m)) P(m1,n) = ((1+q+...+q^(m+n1))/(1+q+...+q^(m1))) P(m1,n) (just send q 1) Definition: [n] = 1+q+...+q^(n1). We call this the qanalogue of n. (Some authors prefer to define [n] as 1q^n.) [1], [2], [3], ... are the qintegers. (Compare with an earlier definition of [n] as {1,...,n}. Ill continue to use both, but it should always be clear which one I have in mind.) Digress for a moment: We defined P(m,n) as a sum of weights of lattice paths, where the weight of a lattice path was defined as q to the power of the sum of the heights of the horizontal edges. (Give example: compute P(2,2).) Is there another way to think about the weight? ...: q to the power of the area under the path. Lets find a direct proof of (1+q+...+ q^(m1)) P(m,n) = (1+q+...+q^(m+n1)) P(m1,n): Motivation: nonq version: m P(m,n) = (m+n1) P(m1,n) LHS: Counts lattice paths from (0,0) to (m,n) with a marked horizontal edge, where the area associated with such a marked path is defined as its ordinary area plus the xcoordinate of the left endpoint of the marked edge. RHS: Counts lattice paths from (0,0) to (m1,n) with a marked vertex, where the area associated with such a marked path is defined as its ordinary area plus the xcoordinate of the vertex plus the ycoordinate of the vertex. Weightpreserving bijection from LHSobjects to RHSobjects: ... Replace each marked dot by a marked horizontal edge, shifting everything after it to the right. Do example with (m1,n) = (1,1), (m,n) = (2,1). Discuss why its a bijection. Analysis: Consider a latticepath from (0,0) to (m,n) with weight x^A. When you mark a vertex (i,j) on a latticepath from (0,0) to (m1,n), you get a vertexmarked path of weight x^(A+i+j). On the other hand, if you replace the marked vertex by a marked horizontal edge, you get an edgemarked path of weight x^(A+i+j). path of weight x^(A+i+j)....
View
Full
Document
This note was uploaded on 02/13/2012 for the course MATH 491 taught by Professor Propp during the Fall '08 term at UMass Lowell.
 Fall '08
 Propp
 Algebra, Combinatorics, Counting

Click to edit the document details