3

I have seen in other questions (Painting the faces of a cube with distinct colours) , and found for myself there are 30 colorings of the cube with exactly 6 colors. My issue is when I use Polya enumeration theory I take the number of colorings up to 6 to be $\frac{1}{24}(6^6+3*6^4+12*6^3+8*6^2)$ (https://en.wikipedia.org/wiki/P%C3%B3lya_enumeration_theorem) and subtract the number of colorings up to 5, $\frac{1}{24}(5^6+3*5^4+12*5^3+8*5^2)$ to get 1426. What is wrong with my application of Polya enumeration theory? How could I fix it?

Mars
  • 822

2 Answers2

3

Let $f(m) = \frac{1}{24} (m^6 + 3m^4 + 12m^3 + 8m^2)$.

The problem is that you want to subtract the number of colorings that use 5 out of 6 colors, not the number that use 5 particular colors.

The fix to this approach would be to start by computing $f(6)-6f(5)$, but of course this overcounts some 5-colorings. So we proceed by inclusion-exclusion and find that the answer is $\sum_{i=0}^6 (-1)^i{m\choose i} f(m-i) = f(6) - 6f(5) + 15f(4) - 20f(3) + 15f(2) - 6f(1) = 30$.


Another equivalent approach is to compute the exponential generating function $F(X) = \sum_{m=0}^\infty f(m) X^n$.

It turns out that $F(X) = e^X p(X)$, where $p(X)$ is a degree $6$ polynomial. Then $G(X) = e^{-X} F(X) = p(X)$ is the exponential generating function for $g(m)$, the number of colorings with exactly $m$ colors.

We can compute $p$ explicitly, but it is straightforward to see that it has the same leading term as $f$, namely $\frac{1}{24} X^6$.

Then we can compute $g(6) = 6! \cdot [X^6] p(X) = \frac{6!}{24} = 30$.

Andrew Dudzik
  • 30,074
  • You may have forgot the $(-1)^i$ in your summation notation. – Mars Sep 18 '21 at 07:55
  • Great explanation, I'll look into the inclusion-exclusion principle for smaller numbers, and hopefully gain some sense about the general formula which you used. – Mars Sep 18 '21 at 08:00
  • @Mars Of course, this isn't a very efficient method for this exact problem. I've included a second approach using generating functions that avoids this tedious computation and only uses information about the leading coefficient of $f$. The two techniques are equivalent, since inclusion-exclusion is just convolution with the power series for $e^{-x}$. – Andrew Dudzik Sep 18 '21 at 12:51
  • I defined a function $e(m)$ to be the number of exactly $m$ colors and used induction to prove the first result. The relationship I used in the proof was $e(m)=f(m)-\Sigma_{i=0}^{m-1}\binom{m}{i}e(i)$. I'll try and digest the additional content you provided. Are you're suggesting is that the computation complexity of the original formula you posted is larger than the formula in the second approach? – Mars Sep 18 '21 at 22:16
  • 1
    A convolution over a degree 6 polynomial is certainly more complicated than the expression $6!/24$, I'm just saying why the two are equivalent. – Andrew Dudzik Sep 18 '21 at 23:19
2

To find the number of ways to color the faces of a cube with exactly $6$ colors using Polya's Enumeration Theorem:

First, as you already know, the cycle index of the symmetries of the face permutations of a cube is $$Z = \frac{1}{24}(x_1^6 + 6x_1^2x_4+3x_1^2x_2^2 + 8x_3^2 + 6x_2^3)$$ The figure inventory for $6$ colors represented by $a,b,c,d,e,f$ is $$a+b+c+d+e+f$$ "Substituting" the figure inventory into the cycle index, we have $$\frac{1}{24} \left((a+b+c+d+e+f)^6 \\ \\+6 (a+b+c+d+e+f)^2 \left(a^4+b^4+c^4+d^4+e^4+f^4\right) \\+3 (a+b+c+d+e+f)^2 \left(a^2+b^2+c^2+d^2+e^2+f^2\right)^2\\+8 \left(a^3+b^3+c^3+d^3+e^3+f^3\right)^2 \\ +6 \left(a^2+b^2+c^2+d^2+e^2+f^2\right)^3 \right) $$ The number of ways to color the cube using each of the six colors exactly once is the coefficient of $abcdef$ when this polynomial is expanded. In difficult cases I would use a computer algebra to find the coefficient, but in this example it's easy to see that what we want is $1/24$ the coefficient of $abcdef$ in $(a+b+c+d+e+f)^6$, i.e. $$\frac{6!}{24} = 30$$

awkward
  • 14,736
  • Thank you very much, this solution was unexpected to me, but very useful. It will help me better understand the "figure index" part of Polya enumeration theory. I think of "+" as "or" and the choice of "$x_i$" to mean that the orbit of period $i$ must be monochromatic. But I am very unsure, just trying to make sense of things. – Mars Sep 18 '21 at 22:28