Equidistributions of Mahonian statistics over pattern ...

55
Equidistributions of Mahonian statistics over pattern avoiding permutations Nima Amini (KTH, Stockholm) June 26, 2017 Permutation Patterns, Reykjavik

Transcript of Equidistributions of Mahonian statistics over pattern ...

Page 1: Equidistributions of Mahonian statistics over pattern ...

Equidistributions of Mahonian statistics overpattern avoiding permutations

Nima Amini (KTH, Stockholm)

June 26, 2017

Permutation Patterns, Reykjavik

Page 2: Equidistributions of Mahonian statistics over pattern ...

Overall Point

I Combine the theory of pattern avoidance with the theory ofstatistics.

I Study the generating function

F (q) =∑

σ∈Avn(Π)

qstat(σ).

I Example questions: Equidistribution? Recurrence relation?Unimodality/Log-concavity/real-rootedness? etc..

Page 3: Equidistributions of Mahonian statistics over pattern ...

Overall Point

I Combine the theory of pattern avoidance with the theory ofstatistics.

I Study the generating function

F (q) =∑

σ∈Avn(Π)

qstat(σ).

I Example questions: Equidistribution? Recurrence relation?Unimodality/Log-concavity/real-rootedness? etc..

Page 4: Equidistributions of Mahonian statistics over pattern ...

Overall Point

I Combine the theory of pattern avoidance with the theory ofstatistics.

I Study the generating function

F (q) =∑

σ∈Avn(Π)

qstat(σ).

I Example questions: Equidistribution? Recurrence relation?Unimodality/Log-concavity/real-rootedness? etc..

Page 5: Equidistributions of Mahonian statistics over pattern ...

Overall Point

I Combine the theory of pattern avoidance with the theory ofstatistics.

I Study the generating function

F (q) =∑

σ∈Avn(Π)

qstat(σ).

I Example questions: Equidistribution? Recurrence relation?Unimodality/Log-concavity/real-rootedness? etc..

Page 6: Equidistributions of Mahonian statistics over pattern ...

Our focus

I We study equidistributions of the form∑σ∈Avn(π1)

qstat1(σ) =∑

σ∈Avn(π2)

qstat2(σ)

where π1, π2 ∈ S3 are (classical) patterns and stat1, stat2 are(Mahonian) permutation statistics.

Page 7: Equidistributions of Mahonian statistics over pattern ...

Our focus

I We study equidistributions of the form∑σ∈Avn(π1)

qstat1(σ) =∑

σ∈Avn(π2)

qstat2(σ)

where π1, π2 ∈ S3 are (classical) patterns and stat1, stat2 are(Mahonian) permutation statistics.

Page 8: Equidistributions of Mahonian statistics over pattern ...

Mahonian d-functions

I Let Π denote the set of vincular patterns of length at most d .A d-function is a statistic of the form

stat =∑π∈Π

απ(π),

where απ ∈ N and (π) counts occurrences of the pattern π.

I Theorem (Babson-Steingrımsson ’00)

For each d ≥ 0 there is a finite number of Mahonian d-functions.

I Babson-Steingrımsson classify all Mahonian 3-functions.

Page 9: Equidistributions of Mahonian statistics over pattern ...

Mahonian d-functions

I Let Π denote the set of vincular patterns of length at most d .A d-function is a statistic of the form

stat =∑π∈Π

απ(π),

where απ ∈ N and (π) counts occurrences of the pattern π.

I Theorem (Babson-Steingrımsson ’00)

For each d ≥ 0 there is a finite number of Mahonian d-functions.

I Babson-Steingrımsson classify all Mahonian 3-functions.

Page 10: Equidistributions of Mahonian statistics over pattern ...

Mahonian d-functions

I Let Π denote the set of vincular patterns of length at most d .A d-function is a statistic of the form

stat =∑π∈Π

απ(π),

where απ ∈ N and (π) counts occurrences of the pattern π.

I Theorem (Babson-Steingrımsson ’00)

For each d ≥ 0 there is a finite number of Mahonian d-functions.

I Babson-Steingrımsson classify all Mahonian 3-functions.

Page 11: Equidistributions of Mahonian statistics over pattern ...

Mahonian d-functions

I Let Π denote the set of vincular patterns of length at most d .A d-function is a statistic of the form

stat =∑π∈Π

απ(π),

where απ ∈ N and (π) counts occurrences of the pattern π.

I Theorem (Babson-Steingrımsson ’00)

For each d ≥ 0 there is a finite number of Mahonian d-functions.

I Babson-Steingrımsson classify all Mahonian 3-functions.

Page 12: Equidistributions of Mahonian statistics over pattern ...

Name Vincular pattern statistic Original reference

maj (132) + (231) + (321) + (21) MacMahon

inv (231) + (312) + (321) + (21) MacMahon

mak (132) + (312) + (321) + (21) Foata-Zeilberger

makl (132) + (231) + (321) + (21) Clarke-Steingrımsson-Zeng

mad (231) + (231) + (312) + (21) Clarke-Steingrımsson-Zeng

bast (132) + (213) + (321) + (21) Babson-Steingrımsson

bast′ (132) + (312) + (321) + (21) Babson-Steingrımsson

bast′′ (132) + (312) + (321) + (21) Babson-Steingrımsson

foze (213) + (321) + (132) + (21) Foata-Zeilberger

foze′ (132) + (231) + (231) + (21) Foata-Zeilberger

foze′′ (231) + (312) + (312) + (21) Foata-Zeilberger

sist (132) + (132) + (213) + (21) Simion-Stanton

sist′ (132) + (132) + (231) + (21) Simion-Stanton

sist′′ (132) + (231) + (231) + (21) Simion-Stanton

Table: Mahonian 3-functions.

Page 13: Equidistributions of Mahonian statistics over pattern ...

Question

I Question: What equidistributions exist between Mahonian3-functions over pattern avoiding sets?

I Some motivation:I Existing bijections for proving Mahonity usually do not restrict

to bijections over Avn(π). A priori no reason for suchequidistributions to exist.

I |Avn(π)| = 1n+1

(2nn

)for π ∈ S3. Get induced equidistributions

between statistics on other Catalan structures underappropriate bijections (and vice versa).

Page 14: Equidistributions of Mahonian statistics over pattern ...

Question

I Question: What equidistributions exist between Mahonian3-functions over pattern avoiding sets?

I Some motivation:I Existing bijections for proving Mahonity usually do not restrict

to bijections over Avn(π). A priori no reason for suchequidistributions to exist.

I |Avn(π)| = 1n+1

(2nn

)for π ∈ S3. Get induced equidistributions

between statistics on other Catalan structures underappropriate bijections (and vice versa).

Page 15: Equidistributions of Mahonian statistics over pattern ...

Question

I Question: What equidistributions exist between Mahonian3-functions over pattern avoiding sets?

I Some motivation:I Existing bijections for proving Mahonity usually do not restrict

to bijections over Avn(π). A priori no reason for suchequidistributions to exist.

I |Avn(π)| = 1n+1

(2nn

)for π ∈ S3. Get induced equidistributions

between statistics on other Catalan structures underappropriate bijections (and vice versa).

Page 16: Equidistributions of Mahonian statistics over pattern ...

Question

I Question: What equidistributions exist between Mahonian3-functions over pattern avoiding sets?

I Some motivation:I Existing bijections for proving Mahonity usually do not restrict

to bijections over Avn(π). A priori no reason for suchequidistributions to exist.

I |Avn(π)| = 1n+1

(2nn

)for π ∈ S3. Get induced equidistributions

between statistics on other Catalan structures underappropriate bijections (and vice versa).

Page 17: Equidistributions of Mahonian statistics over pattern ...

I Proposition (A. ’17)

Let σ ∈ Avn(π) where π ∈ {132, 213, 231, 312}. Then

mak(σ) = imaj(σ).

Moreover for any n ≥ 1,∑σ∈Avn(π)

qmaj(σ)tdes(σ) =∑

σ∈Avn(π−1)

qmak(σ)tdes(σ).

I Remark: By above proposition and a result of Stump we have∑σ∈Avn(231)

qmaj(σ)+mak(σ) =1

[n + 1]q

[2n

n

]q

(MacMahon’s q-analogue of the Catalan numbers)

Page 18: Equidistributions of Mahonian statistics over pattern ...

I Proposition (A. ’17)

Let σ ∈ Avn(π) where π ∈ {132, 213, 231, 312}. Then

mak(σ) = imaj(σ).

Moreover for any n ≥ 1,∑σ∈Avn(π)

qmaj(σ)tdes(σ) =∑

σ∈Avn(π−1)

qmak(σ)tdes(σ).

I Remark: By above proposition and a result of Stump we have∑σ∈Avn(231)

qmaj(σ)+mak(σ) =1

[n + 1]q

[2n

n

]q

(MacMahon’s q-analogue of the Catalan numbers)

Page 19: Equidistributions of Mahonian statistics over pattern ...

I Theorem (Dokos-Dwyer-Johnson-Sagan-Selsor ’12)∑σ∈Avn(231)

qinv(σ) = Cn(q),

where

Cn(q) =n−1∑k=0

qk Ck(q)Cn−k−1(q)

(Carlitz-Riordan’s q-analogue of the Catalan numbers)

Page 20: Equidistributions of Mahonian statistics over pattern ...

I Theorem (A. ’17)

For any n ≥ 1,∑σ∈Avn(321)

qmaj(σ)xDB(σ)yDT(σ) =∑

σ∈Avn(321)

qmak(σ)xDB(σ)yDT(σ),

∑σ∈Avn(123)

qmaj(σ)xAB(σ)yAT(σ) =∑

σ∈Avn(123)

qmak(σ)xAB(σ)yAT(σ).

where DB(σ) = {σi+1 : σi > σi+1}, DT(σ) = {σi : σi > σi+1} andAB(σ),AT(σ) defined similarly.

I Proved via an involution φ : Avn(321)→ Avn(321).I Let 5612379468 ∈ Av9(321)

Page 21: Equidistributions of Mahonian statistics over pattern ...

I Theorem (A. ’17)

For any n ≥ 1,∑σ∈Avn(321)

qmaj(σ)xDB(σ)yDT(σ) =∑

σ∈Avn(321)

qmak(σ)xDB(σ)yDT(σ),

∑σ∈Avn(123)

qmaj(σ)xAB(σ)yAT(σ) =∑

σ∈Avn(123)

qmak(σ)xAB(σ)yAT(σ).

where DB(σ) = {σi+1 : σi > σi+1}, DT(σ) = {σi : σi > σi+1} andAB(σ),AT(σ) defined similarly.

I Proved via an involution φ : Avn(321)→ Avn(321).I Let 5612379468 ∈ Av9(321)

Page 22: Equidistributions of Mahonian statistics over pattern ...

I Theorem (A. ’17)

For any n ≥ 1,∑σ∈Avn(321)

qmaj(σ)xDB(σ)yDT(σ) =∑

σ∈Avn(321)

qmak(σ)xDB(σ)yDT(σ),

∑σ∈Avn(123)

qmaj(σ)xAB(σ)yAT(σ) =∑

σ∈Avn(123)

qmak(σ)xAB(σ)yAT(σ).

where DB(σ) = {σi+1 : σi > σi+1}, DT(σ) = {σi : σi > σi+1} andAB(σ),AT(σ) defined similarly.

I Proved via an involution φ : Avn(321)→ Avn(321).I Let 561237948∈ Av9(321)I Red letters are left-to-right maxima and Blue letters are

non-left-to-right maxima.

Page 23: Equidistributions of Mahonian statistics over pattern ...

I Theorem (A. ’17)

For any n ≥ 1,∑σ∈Avn(321)

qmaj(σ)xDB(σ)yDT(σ) =∑

σ∈Avn(321)

qmak(σ)xDB(σ)yDT(σ),

∑σ∈Avn(123)

qmaj(σ)xAB(σ)yAT(σ) =∑

σ∈Avn(123)

qmak(σ)xAB(σ)yAT(σ).

where DB(σ) = {σi+1 : σi > σi+1}, DT(σ) = {σi : σi > σi+1} andAB(σ),AT(σ) defined similarly.

I Proved via an involution φ : Avn(321)→ Avn(321).I Let 561237948∈ S9(321)I Green letters are descent tops and descent bottoms.

Page 24: Equidistributions of Mahonian statistics over pattern ...

I Theorem (A. ’17)

For any n ≥ 1,∑σ∈Avn(321)

qmaj(σ)xDB(σ)yDT(σ) =∑

σ∈Avn(321)

qmak(σ)xDB(σ)yDT(σ),

∑σ∈Avn(123)

qmaj(σ)xAB(σ)yAT(σ) =∑

σ∈Avn(123)

qmak(σ)xAB(σ)yAT(σ).

where DB(σ) = {σi+1 : σi > σi+1}, DT(σ) = {σi : σi > σi+1} andAB(σ),AT(σ) defined similarly.

I Proved via an involution φ : Avn(321)→ Avn(321).I Let 561237948∈ S9(321)I The involution preserves the relative order of Green letters

(descent pairs) and swaps the role of Red (LRMax) and Blue(non-LRMax) letters.

I We get 561237948 7→ 236189457.

Page 25: Equidistributions of Mahonian statistics over pattern ...

An induced equidistribution

I The bijection φ : Avn(321)→ Avn(321) induces anequidistribution on statistics associated with shortenedpolyominoes (another Catalan structure).

I A shortened polyomino is a pair (P,Q) of N (north), E (east)lattice paths P = (Pi )

ni=1 and Q = (Qi )

ni=1 satisfying

1. P and Q begin at the same vertex and end at the same vertex.2. P stays weakly above Q and the two paths can share E -steps

but not N-steps.

Denote the set of shortened polyominoes with |P| = |Q| = nby Hn.

I |Hn| = 1n+1

(2nn

).

Page 26: Equidistributions of Mahonian statistics over pattern ...

An induced equidistribution

I The bijection φ : Avn(321)→ Avn(321) induces anequidistribution on statistics associated with shortenedpolyominoes (another Catalan structure).

I A shortened polyomino is a pair (P,Q) of N (north), E (east)lattice paths P = (Pi )

ni=1 and Q = (Qi )

ni=1 satisfying

1. P and Q begin at the same vertex and end at the same vertex.2. P stays weakly above Q and the two paths can share E -steps

but not N-steps.

Denote the set of shortened polyominoes with |P| = |Q| = nby Hn.

I |Hn| = 1n+1

(2nn

).

Page 27: Equidistributions of Mahonian statistics over pattern ...

An induced equidistribution

I The bijection φ : Avn(321)→ Avn(321) induces anequidistribution on statistics associated with shortenedpolyominoes (another Catalan structure).

I A shortened polyomino is a pair (P,Q) of N (north), E (east)lattice paths P = (Pi )

ni=1 and Q = (Qi )

ni=1 satisfying

1. P and Q begin at the same vertex and end at the same vertex.2. P stays weakly above Q and the two paths can share E -steps

but not N-steps.

Denote the set of shortened polyominoes with |P| = |Q| = nby Hn.

I |Hn| = 1n+1

(2nn

).

Page 28: Equidistributions of Mahonian statistics over pattern ...

An induced equidistribution

I The bijection φ : Avn(321)→ Avn(321) induces anequidistribution on statistics associated with shortenedpolyominoes (another Catalan structure).

I A shortened polyomino is a pair (P,Q) of N (north), E (east)lattice paths P = (Pi )

ni=1 and Q = (Qi )

ni=1 satisfying

1. P and Q begin at the same vertex and end at the same vertex.2. P stays weakly above Q and the two paths can share E -steps

but not N-steps.

Denote the set of shortened polyominoes with |P| = |Q| = nby Hn.

I |Hn| = 1n+1

(2nn

).

Page 29: Equidistributions of Mahonian statistics over pattern ...

A bijection Ψ : Hn → Avn(321)

Q

P

Page 30: Equidistributions of Mahonian statistics over pattern ...

A bijection Ψ : Hn → Avn(321)

1

2

3 45

67

8

9

Q

P

Page 31: Equidistributions of Mahonian statistics over pattern ...

A bijection Ψ : Hn → Avn(321)

1

2

3 45

67

8

9

3

Q

P

Ψ(P,Q) = 3

Page 32: Equidistributions of Mahonian statistics over pattern ...

A bijection Ψ : Hn → Avn(321)

1

2

3 45

67

8

9

3 4

Q

P

Ψ(P,Q) = 34

Page 33: Equidistributions of Mahonian statistics over pattern ...

A bijection Ψ : Hn → Avn(321)

1

2

3 45

67

8

9

3 4

1

Q

P

Ψ(P,Q) = 341

Page 34: Equidistributions of Mahonian statistics over pattern ...

A bijection Ψ : Hn → Avn(321)

1

2

3 45

67

8

9

3 4

16

Q

P

Ψ(P,Q) = 3416

Page 35: Equidistributions of Mahonian statistics over pattern ...

A bijection Ψ : Hn → Avn(321)

1

2

3 45

67

8

9

3 4

16

2

59

7

8

Q

P

Ψ(P,Q) = 341625978

Page 36: Equidistributions of Mahonian statistics over pattern ...

A bijection Ψ : Hn → Avn(321)

1

2

3 45

67

8

9

3 4

16

2

59

7

8

Q

P

Ψ(P,Q) = 341625978 ∈ Av9(321)

Page 37: Equidistributions of Mahonian statistics over pattern ...

Two statistics on polyominoes

I Let Valley(Q) = {i : QiQi+1 = EN} denote the set of indicesof the valleys in Q and let val(Q) = |Valley(Q)|.

I Define the statistics valley-column area, vcarea(P,Q), andvalley-row area, vrarea(P,Q).

Q

P

(a) vcarea(P,Q) = 2 + 3 + 2 = 7

Q

P

(b) vrarea(P,Q) = 2 + 4 + 3 = 9

Page 38: Equidistributions of Mahonian statistics over pattern ...

Two statistics on polyominoes

I Let Valley(Q) = {i : QiQi+1 = EN} denote the set of indicesof the valleys in Q and let val(Q) = |Valley(Q)|.

I Define the statistics valley-column area, vcarea(P,Q), andvalley-row area, vrarea(P,Q).

Q

P

(a) vcarea(P,Q) = 2 + 3 + 2 = 7

Q

P

(b) vrarea(P,Q) = 2 + 4 + 3 = 9

Page 39: Equidistributions of Mahonian statistics over pattern ...

Two statistics on polyominoes

I Let Valley(Q) = {i : QiQi+1 = EN} denote the set of indicesof the valleys in Q and let val(Q) = |Valley(Q)|.

I Define the statistics valley-column area, vcarea(P,Q), andvalley-row area, vrarea(P,Q).

Q

P

(a) vcarea(P,Q) = 2 + 3 + 2 = 7

Q

P

(b) vrarea(P,Q) = 2 + 4 + 3 = 9

Page 40: Equidistributions of Mahonian statistics over pattern ...

An induced equidistribution

I Theorem (A. ’17)

For any n ≥ 1,∑(P,Q)∈Hn

qvcarea(P,Q)tval(Q) =∑

(P,Q)∈Hn

qvrarea(P,Q)tval(Q).

I vcarea(P,Q) = ((21) + (312))Ψ(P,Q)

I vrarea(P,Q) = ((21) + (231))Ψ(P,Q)

I ((21) + (312))φ(σ) = ((21) + (231))σ.

I Equidistribution follows via bijection Ψ−1 ◦ φ ◦Ψ.

Page 41: Equidistributions of Mahonian statistics over pattern ...

An induced equidistribution

I Theorem (A. ’17)

For any n ≥ 1,∑(P,Q)∈Hn

qvcarea(P,Q)tval(Q) =∑

(P,Q)∈Hn

qvrarea(P,Q)tval(Q).

I vcarea(P,Q) = ((21) + (312))Ψ(P,Q)

I vrarea(P,Q) = ((21) + (231))Ψ(P,Q)

I ((21) + (312))φ(σ) = ((21) + (231))σ.

I Equidistribution follows via bijection Ψ−1 ◦ φ ◦Ψ.

Page 42: Equidistributions of Mahonian statistics over pattern ...

An induced equidistribution

I Theorem (A. ’17)

For any n ≥ 1,∑(P,Q)∈Hn

qvcarea(P,Q)tval(Q) =∑

(P,Q)∈Hn

qvrarea(P,Q)tval(Q).

I vcarea(P,Q) = ((21) + (312))Ψ(P,Q)

I vrarea(P,Q) = ((21) + (231))Ψ(P,Q)

I ((21) + (312))φ(σ) = ((21) + (231))σ.

I Equidistribution follows via bijection Ψ−1 ◦ φ ◦Ψ.

Page 43: Equidistributions of Mahonian statistics over pattern ...

An induced equidistribution

I Theorem (A. ’17)

For any n ≥ 1,∑(P,Q)∈Hn

qvcarea(P,Q)tval(Q) =∑

(P,Q)∈Hn

qvrarea(P,Q)tval(Q).

I vcarea(P,Q) = ((21) + (312))Ψ(P,Q)

I vrarea(P,Q) = ((21) + (231))Ψ(P,Q)

I ((21) + (312))φ(σ) = ((21) + (231))σ.

I Equidistribution follows via bijection Ψ−1 ◦ φ ◦Ψ.

Page 44: Equidistributions of Mahonian statistics over pattern ...

Another example

∑σ∈Avn(321)

qinv(σ)

Page 45: Equidistributions of Mahonian statistics over pattern ...

Another example

∑σ∈Avn(321)

qinv(σ) =(∗)∑P∈Dn

qspea(P)

(∗) Cheng-Elizalde-Kasraoui-Sagan ’13

spea(P) =∑

p∈Peak(P)

(htP(p)− 1).

Figure: The Dyck path corresponding to σ = 341625978 underKrattenthaler’s bijection. Maps inv to spea.

Page 46: Equidistributions of Mahonian statistics over pattern ...

Another example

∑σ∈Avn(321)

qinv(σ) =∑P∈Dn

qspea(P) =(∗)∑P∈Dn

qstun(P)

(∗) Cheng-Elizalde-Kasraoui-Sagan ’13

stun(P) =∑

(i ,j)∈Tunnel(P)

(j − i)/2

Figure: The tunnel lengths of a Dyck path (indicated with dashes).

Page 47: Equidistributions of Mahonian statistics over pattern ...

Another example

∑σ∈Avn(321)

qinv(σ) =∑P∈Dn

qspea(P) =∑P∈Dn

qstun(P) =(∗)∑P∈Dn

qmass(P)+dr(P)

(∗) A. ’17

The mass corresponding to two consecutive U-steps, is half thenumber of steps between their matching D-steps(i.e. if P = UUP ′DP ′′D, then mass of the pair UU is |P ′′|/2).

mass(P) = sum of masses over all occurrences of UU

dr(P) = number of double rises in P.

Figure: The mass associated with the first double rise is highlighted inred.

Page 48: Equidistributions of Mahonian statistics over pattern ...

Another example

∑σ∈Avn(321)

qinv(σ) =∑P∈Dn

qspea(P) =∑P∈Dn

qstun(P) =∑P∈Dn

qmass(P)+dr(P)

=(∗)∑

σ∈Avn(231)

qmad(σ)

(∗) A. ’17

Follows via Knuth’s ‘standard’ bijection

f : Avn(231)→ Dn

213[1, σ1, σ2] 7→ Uf (σ1)Df (σ2).

Page 49: Equidistributions of Mahonian statistics over pattern ...

Another example

∑σ∈Avn(321)

qinv(σ) =∑P∈Dn

qspea(P) =∑P∈Dn

qstun(P) =∑P∈Dn

qmass(P)+dr(P)

=∑

σ∈Avn(231)

qmad(σ) =(∗)∑

σ∈Avn(132)

qinc(σ)

(∗) A. ’17

inc = ι1 +∞∑k=2

(−1)k−12k−2ιk

where ιk−1 = (12 . . . k) is the statistic that counts the number ofincreasing subsequences of length k in a permutation.

Uses the Catalan continued fraction framework ofBranden-Claesson-Steingrımsson.

Page 50: Equidistributions of Mahonian statistics over pattern ...

Another example

∑σ∈Avn(321)

qinv(σ) =∑P∈Dn

qspea(P) =∑P∈Dn

qstun(P) =∑P∈Dn

qmass(P)+dr(P)

=∑

σ∈Avn(231)

qmad(σ) =∑

σ∈Avn(132)

qinc(σ)

Aside ∑σ∈Avn(321)

qinv(σ) =(∗)∑

(P,Q)∈Hn

qarea(P,Q).

(∗) Cheng-Eu-Fu ’07

Q

P

Page 51: Equidistributions of Mahonian statistics over pattern ...

Another example

∑σ∈Avn(321)

qinv(σ) =∑P∈Dn

qspea(P) =∑P∈Dn

qstun(P) =∑P∈Dn

qmass(P)+dr(P)

=∑

σ∈Avn(231)

qmad(σ) =∑

σ∈Avn(132)

qinc(σ)

Aside ∑σ∈Avn(321)

qinv(σ) =∑

(P,Q)∈Hn

qarea(P,Q)

∑σ∈Avn(321)

qinv(σ) =(∗)∑P∈Dn

qsups(P).

(∗) A. ’17

sups(P) =∑

i∈Up(P)

dhtP(i)/2e.

Page 52: Equidistributions of Mahonian statistics over pattern ...

I Several more equidistributions hold between Mahonian3-functions over Avn(π)!

I The following table includes all established equidistributions(in black) and all conjectured equidistributions (in red).

Page 53: Equidistributions of Mahonian statistics over pattern ...

I Several more equidistributions hold between Mahonian3-functions over Avn(π)!

I The following table includes all established equidistributions(in black) and all conjectured equidistributions (in red).

Page 54: Equidistributions of Mahonian statistics over pattern ...

maj inv mak makl mad bast bast′ bast′′ foze foze′ foze′′ sist sist′ sist′′

maj132, 231213, 312

123, 123132, 132132, 312213, 213213, 231231, 132231, 312312, 213312, 231321, 321

132, 231213, 312231, 231312, 312321, 321

132, 213213, 231231, 213312, 231

132, 132231, 132

213, 231312, 231

132, 132213, 231231, 132312, 231

inv • 132, 213231, 312

231, 312312, 312321, 231

231, 132312, 132321, 213

231, 132312, 132321, 213

231, 213312, 213321, 132

231, 231312, 231321, 132

231, 132312, 132321, 231

mak • • 132, 312213, 231

132, 231213, 312231, 312312, 231321, 321

132, 213213, 231231, 231312, 213

132, 132312, 132

213, 231231, 231

132, 132213, 231231, 231312, 132

makl • • •132, 132231, 213312, 231

231, 132 312, 231132, 213231, 132312, 231

mad • • • • 231, 213312, 132

231, 213312, 132

231, 132312, 213

132, 213231, 132312, 231

213, 213231, 231312, 132

bast • • • • • 213, 132 231, 231

123, 123213, 132132, 213231, 231312, 312321, 321

bast′ • • • • • • 132, 132bast′′ • • • • • • • 231, 231

foze • • • • • • • •

foze′ • • • • • • • • • 132, 132213, 213

132, 213213, 132

132, 231213, 132

132, 132213, 231

foze′′ • • • • • • • • • • 213, 132132, 213

213, 132132, 231

132, 132213, 231

sist • • • • • • • • • • •132, 132213, 231312, 312

132, 231213, 132231, 312

sist′ • • • • • • • • • • • • 132, 231231, 132

sist′′ • • • • • • • • • • • • •

Table: Established equidistributions in black and conjecturedequidistributions in red.

Page 55: Equidistributions of Mahonian statistics over pattern ...

Thank you for listening!