the principle of overcountingfaculty.cord.edu/ahendric/2010fall325/notburnside.pdf · the...

68
The Principle of Overcounting In combinatorics, we often have to count in a two-step procedure: 1 Use a “wrong” procedure to count, which yields a number that’s too high. 2 Then figure out how wrong our answer was, and reduce it to the correct number. Example How many subsets of {A, B, C, D, E} have exactly 3 elements? We start by counting the ways to choose three elements: · ·

Upload: others

Post on 15-Aug-2020

2 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

The Principle of Overcounting

In combinatorics, we often have to count in a two-step procedure:

1 Use a “wrong” procedure to count, which yields a numberthat’s too high.

2 Then figure out how wrong our answer was, and reduce it tothe correct number.

Example

How many subsets of {A,B,C,D,E} have exactly 3 elements?We start by counting the ways to choose three elements:

5

·

4

·

3 = 60

Page 2: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

The Principle of Overcounting

In combinatorics, we often have to count in a two-step procedure:

1 Use a “wrong” procedure to count, which yields a numberthat’s too high.

2 Then figure out how wrong our answer was, and reduce it tothe correct number.

Example

How many subsets of {A,B,C,D,E} have exactly 3 elements?We start by counting the ways to choose three elements:

5

·

4

·

3 = 60

Page 3: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

The Principle of Overcounting

In combinatorics, we often have to count in a two-step procedure:

1 Use a “wrong” procedure to count, which yields a numberthat’s too high.

2 Then figure out how wrong our answer was, and reduce it tothe correct number.

Example

How many subsets of {A,B,C,D,E} have exactly 3 elements?We start by counting the ways to choose three elements:

5

·

4

·

3 = 60

Page 4: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

The Principle of Overcounting

In combinatorics, we often have to count in a two-step procedure:

1 Use a “wrong” procedure to count, which yields a numberthat’s too high.

2 Then figure out how wrong our answer was, and reduce it tothe correct number.

Example

How many subsets of {A,B,C,D,E} have exactly 3 elements?We start by counting the ways to choose three elements:

5 ·

4

·

3 = 60

Page 5: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

The Principle of Overcounting

In combinatorics, we often have to count in a two-step procedure:

1 Use a “wrong” procedure to count, which yields a numberthat’s too high.

2 Then figure out how wrong our answer was, and reduce it tothe correct number.

Example

How many subsets of {A,B,C,D,E} have exactly 3 elements?We start by counting the ways to choose three elements:

5 · 4 ·

3 = 60

Page 6: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

The Principle of Overcounting

In combinatorics, we often have to count in a two-step procedure:

1 Use a “wrong” procedure to count, which yields a numberthat’s too high.

2 Then figure out how wrong our answer was, and reduce it tothe correct number.

Example

How many subsets of {A,B,C,D,E} have exactly 3 elements?We start by counting the ways to choose three elements:

5 · 4 · 3

= 60

Page 7: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

The Principle of Overcounting

In combinatorics, we often have to count in a two-step procedure:

1 Use a “wrong” procedure to count, which yields a numberthat’s too high.

2 Then figure out how wrong our answer was, and reduce it tothe correct number.

Example

How many subsets of {A,B,C,D,E} have exactly 3 elements?We start by counting the ways to choose three elements:

5 · 4 · 3 = 60

Page 8: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

The Principle of Overcounting

ABC

ABD

ABE

ACB

ACD

ACE

ADB

ADC

ADE

AEB

AEC

AED

BAC

BAD

BAE

BCA

BCD

BCE

BDA

BDC

BDE

BEA

BEC

BED

CAB

CAD

CAE

CBA

CBD

CBE

CDA

CDB

CDE

CEA

CEB

CED

DAB

DAC

DAE

DBA

DBC

DBE

DCA

DCB

DCE

DEA

DEB

DEC

EAB

EAC

EAD

EBA

EBC

EBD

ECA

ECB

ECD

EDA

EDB

EDC

Page 9: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

The Principle of Overcounting

Page 10: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

The Principle of Overcounting

ABC

ABD

ABE

ACB

ACD

ACE

ADB

ADC

ADE

AEB

AEC

AED

BAC

BAD

BAE

BCA

BCD

BCE

BDA

BDC

BDE

BEA

BEC

BED

CAB

CAD

CAE

CBA

CBD

CBE

CDA

CDB

CDE

CEA

CEB

CED

DAB

DAC

DAE

DBA

DBC

DBE

DCA

DCB

DCE

DEA

DEB

DEC

EAB

EAC

EAD

EBA

EBC

EBD

ECA

ECB

ECD

EDA

EDB

EDC

{A,B,C}{A,B,D}{A,B,E}{A,C ,D}{A,C ,E}{A,D,E}{B,C ,D}{B,C ,E}{B,D,E}{C ,D,E}

Page 11: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

The Principle of Overcounting

ABC

ABD

ABE

ACB

ACD

ACE

ADB

ADC

ADE

AEB

AEC

AED

BAC

BAD

BAE

BCA

BCD

BCE

BDA

BDC

BDE

BEA

BEC

BED

CAB

CAD

CAE

CBA

CBD

CBE

CDA

CDB

CDE

CEA

CEB

CED

DAB

DAC

DAE

DBA

DBC

DBE

DCA

DCB

DCE

DEA

DEB

DEC

EAB

EAC

EAD

EBA

EBC

EBD

ECA

ECB

ECD

EDA

EDB

EDC

{A,B,C}{A,B,D}{A,B,E}{A,C ,D}{A,C ,E}{A,D,E}{B,C ,D}{B,C ,E}{B,D,E}{C ,D,E}

60

6= 10 =

(53

)

Page 12: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Another Example

Another Example

How many ways to seat three people around a round table?

Initial (wrong) answer: 3! = 6.

Page 13: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Another Example

Another Example

How many ways to seat three people around a round table?

3

Initial (wrong) answer: 3! = 6.

Page 14: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Another Example

Another Example

How many ways to seat three people around a round table?

3

2

Initial (wrong) answer: 3! = 6.

Page 15: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Another Example

Another Example

How many ways to seat three people around a round table?

3

21

Initial (wrong) answer: 3! = 6.

Page 16: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Another Example

Another Example

How many ways to seat three people around a round table?

3

21

Initial (wrong) answer: 3! = 6.

Page 17: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Another Example

Page 18: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Another Example

Page 19: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Another Example

Page 20: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Another Example

Correct answer:3!

3= 2.

Page 21: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

A More Complicated Example

Example

You have two colors, red and blue. How many ways are there topaint the corners of a scalene triangle?

Initial answer: 2 · 2 · 2 = 8.

Page 22: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

A More Complicated Example

Example

You have two colors, red and blue. How many ways are there topaint the corners of a scalene triangle?

2

Initial answer: 2 · 2 · 2 = 8.

Page 23: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

A More Complicated Example

Example

You have two colors, red and blue. How many ways are there topaint the corners of a scalene triangle?

2

2

Initial answer: 2 · 2 · 2 = 8.

Page 24: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

A More Complicated Example

Example

You have two colors, red and blue. How many ways are there topaint the corners of a scalene triangle?

2

2

2

Initial answer: 2 · 2 · 2 = 8.

Page 25: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

A More Complicated Example

Example

You have two colors, red and blue. How many ways are there topaint the corners of a scalene triangle?

2

2

2

Initial answer: 2 · 2 · 2 = 8.

Page 26: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

A More Complicated Example

Example

You have two colors, red and blue. How many ways are there topaint the corners of an equilateral triangle?

Initial (wrong) answer: 2 · 2 · 2 = 8.

Page 27: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

A More Complicated Example

Page 28: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

A More Complicated Example

Page 29: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

A More Complicated Example

Page 30: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

A More Complicated Example

Correct answer: 4.

Page 31: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

An Even More Complicated Example

An Even More Complicated Example

You have two colors, red and blue. How many ways are there tocolor the corners of a square?

Page 32: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

An Even More Complicated Example

An Even More Complicated Example

You have two colors, red and blue. How many ways are there tocolor the corners of a square?

22

2 2

Wrong answer: 24 = 16

Page 33: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

An Even More Complicated Example

Page 34: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

An Even More Complicated Example

Page 35: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

An Even More Complicated Example

Page 36: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

An Even More Complicated Example

Correct answer: 6.

Page 37: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Absurdly Complicated Example

Absurdly Complicated Example

How many ways are there to paint the corners of a square withfour colors? Count colorings as the same if one can be rotated orflipped to look like the other.

The initial (wrong) answer: 44 = 256.

I don’t want to draw all those out,let alone figure out where the repeats are.

We need a faster way to predict the answer, withoutexamining all the possibilities by hand.

The answer: group actions!

Page 38: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

R90

Page 39: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

R90

Page 40: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j
Page 41: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Observations

In our example:

Every coloring we wanted was in the whole list of 24

possibilities.

Some of those were equivalent and shouldn’t be countedseparately.

We considered two colorings equivalent when one could betransformed into the other by a rotation.

That is, we were looking at the orbits of the colorings underthe action of the group of rotations {R0,R90,R180,R270}.Our problem: Counting the orbits.

Page 42: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

The Set-Up

G acts on S where

S = all coloringsG = symmetry group of the figure

“distinguishable” = in different orbits

Page 43: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Combinatorics Strategy

1 Think of all possible ways (including duplicates).

2 Decide what makes two ways count as equivalent.

3 Decide what group action has those equivalence classes as itsorbits.

4 Use the not-Burnside Theorem to count the orbits.

Page 44: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Combinatorics Strategy

1 Think of all possible ways (including duplicates).

2 Decide what makes two ways count as equivalent.

3 Decide what group action has those equivalence classes as itsorbits.

4 Use the not-Burnside Theorem to count the orbits.

Page 45: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Combinatorics Strategy

1 Think of all possible ways (including duplicates).

2 Decide what makes two ways count as equivalent.

3 Decide what group action has those equivalence classes as itsorbits.

4 Use the not-Burnside Theorem to count the orbits.

Page 46: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Combinatorics Strategy

1 Think of all possible ways (including duplicates).

2 Decide what makes two ways count as equivalent.

3 Decide what group action has those equivalence classes as itsorbits.

4 Use the not-Burnside Theorem to count the orbits.

Page 47: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

The not-Burnside Theorem

Theorem (not-Burnside)

Let G act on S. For each g ∈ G, let ψ(g) = |Fix(g)|. Then thenumber of orbits of S under the action of G is

1

|G |∑g∈G

ψ(g).

Proof Sketch:

Count {(g , s) ∈ G × S : s · g = s} in two different ways. �

Page 48: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

The not-Burnside Theorem

Theorem (not-Burnside)

Let G act on S. For each g ∈ G, let ψ(g) = |Fix(g)|. Then thenumber of orbits of S under the action of G is

1

|G |∑g∈G

ψ(g).

Proof Sketch:

Count {(g , s) ∈ G × S : s · g = s} in two different ways. �

Page 49: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j
Page 50: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

Even More Complicated Example, revisited

You have two colors, red and blue. How many ways are there tocolor the corners of a square? Count two colorings as the same ifthey can be rotated to resemble one another.

Our symmetry group is {R0,R90,R180,R270} ≤ D8.

g ψ(g)

R0

16

R90

2

R180

4

R270

224

Thus the number of distinguishable colorings is

1

|G |∑g∈G

ψ(g) =

1

4· 24 = 6.

Page 51: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

Even More Complicated Example, revisited

You have two colors, red and blue. How many ways are there tocolor the corners of a square? Count two colorings as the same ifthey can be rotated to resemble one another.

Our symmetry group is {R0,R90,R180,R270} ≤ D8.

g ψ(g)

R0 16R90

2

R180

4

R270

224

Thus the number of distinguishable colorings is

1

|G |∑g∈G

ψ(g) =

1

4· 24 = 6.

Page 52: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

Even More Complicated Example, revisited

You have two colors, red and blue. How many ways are there tocolor the corners of a square? Count two colorings as the same ifthey can be rotated to resemble one another.

Our symmetry group is {R0,R90,R180,R270} ≤ D8.

g ψ(g)

R0 16R90 2R180

4

R270

224

Thus the number of distinguishable colorings is

1

|G |∑g∈G

ψ(g) =

1

4· 24 = 6.

Page 53: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

Even More Complicated Example, revisited

You have two colors, red and blue. How many ways are there tocolor the corners of a square? Count two colorings as the same ifthey can be rotated to resemble one another.

Our symmetry group is {R0,R90,R180,R270} ≤ D8.

g ψ(g)

R0 16R90 2R180 4R270

224

Thus the number of distinguishable colorings is

1

|G |∑g∈G

ψ(g) =

1

4· 24 = 6.

Page 54: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

Even More Complicated Example, revisited

You have two colors, red and blue. How many ways are there tocolor the corners of a square? Count two colorings as the same ifthey can be rotated to resemble one another.

Our symmetry group is {R0,R90,R180,R270} ≤ D8.

g ψ(g)

R0 16R90 2R180 4R270 2

24

Thus the number of distinguishable colorings is

1

|G |∑g∈G

ψ(g) =

1

4· 24 = 6.

Page 55: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

Even More Complicated Example, revisited

You have two colors, red and blue. How many ways are there tocolor the corners of a square? Count two colorings as the same ifthey can be rotated to resemble one another.

Our symmetry group is {R0,R90,R180,R270} ≤ D8.

g ψ(g)

R0 16R90 2R180 4R270 2

24

Thus the number of distinguishable colorings is

1

|G |∑g∈G

ψ(g) =

1

4· 24 = 6.

Page 56: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

Even More Complicated Example, revisited

You have two colors, red and blue. How many ways are there tocolor the corners of a square? Count two colorings as the same ifthey can be rotated to resemble one another.

Our symmetry group is {R0,R90,R180,R270} ≤ D8.

g ψ(g)

R0 16R90 2R180 4R270 2

24

Thus the number of distinguishable colorings is

1

|G |∑g∈G

ψ(g) =1

4· 24 = 6.

Page 57: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

Absurdly Complicated Example, revisited

How many ways are there to paint the corners of a square withfour colors? Count colorings as the same if one can be rotated orflipped to look like the other.

Note that our symmetry group is D8.

Page 58: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

g ψ(g)

R0

256

R90

4

R180

16

R270

4

H

16

V

16

D1

64

D2

64440

Thus the number of distinguishable colorings is

1

|D8|∑g∈D8

ψ(g) =

1

8· 440 = 55.

Page 59: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

g ψ(g)

R0 256R90

4

R180

16

R270

4

H

16

V

16

D1

64

D2

64440

Thus the number of distinguishable colorings is

1

|D8|∑g∈D8

ψ(g) =

1

8· 440 = 55.

Page 60: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

g ψ(g)

R0 256R90 4R180

16

R270

4

H

16

V

16

D1

64

D2

64440

Thus the number of distinguishable colorings is

1

|D8|∑g∈D8

ψ(g) =

1

8· 440 = 55.

Page 61: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

g ψ(g)

R0 256R90 4R180 16R270

4

H

16

V

16

D1

64

D2

64440

Thus the number of distinguishable colorings is

1

|D8|∑g∈D8

ψ(g) =

1

8· 440 = 55.

Page 62: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

g ψ(g)

R0 256R90 4R180 16R270 4H

16

V

16

D1

64

D2

64440

Thus the number of distinguishable colorings is

1

|D8|∑g∈D8

ψ(g) =

1

8· 440 = 55.

Page 63: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

g ψ(g)

R0 256R90 4R180 16R270 4H 16V

16

D1

64

D2

64440

Thus the number of distinguishable colorings is

1

|D8|∑g∈D8

ψ(g) =

1

8· 440 = 55.

Page 64: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

g ψ(g)

R0 256R90 4R180 16R270 4H 16V 16D1

64

D2

64440

Thus the number of distinguishable colorings is

1

|D8|∑g∈D8

ψ(g) =

1

8· 440 = 55.

Page 65: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

g ψ(g)

R0 256R90 4R180 16R270 4H 16V 16D1 64D2

64440

Thus the number of distinguishable colorings is

1

|D8|∑g∈D8

ψ(g) =

1

8· 440 = 55.

Page 66: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

g ψ(g)

R0 256R90 4R180 16R270 4H 16V 16D1 64D2 64

440

Thus the number of distinguishable colorings is

1

|D8|∑g∈D8

ψ(g) =

1

8· 440 = 55.

Page 67: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

g ψ(g)

R0 256R90 4R180 16R270 4H 16V 16D1 64D2 64

440

Thus the number of distinguishable colorings is

1

|D8|∑g∈D8

ψ(g) =

1

8· 440 = 55.

Page 68: The Principle of Overcountingfaculty.cord.edu/ahendric/2010Fall325/notBurnside.pdf · The not-Burnside Theorem. Theorem (not-Burnside) Let G act on S. For each g 2G, let (g) = jFix(g)j

Applying not-Burnside

g ψ(g)

R0 256R90 4R180 16R270 4H 16V 16D1 64D2 64

440

Thus the number of distinguishable colorings is

1

|D8|∑g∈D8

ψ(g) =1

8· 440 = 55.