For three irreducible characters $\phi,\psi,\rho$ of a finite group $G$ define the Kronecker multiplicities as: $$g(\phi,\psi,\rho) = \langle \phi,\psi\cdot\rho\rangle $$ where $$\langle \chi,\eta\rangle = \frac{1}{|G|}\sum_{x\in G} \chi(x)\, \overline{\eta(x)} $$ and $[\psi\cdot\rho] (x) = \psi(x) \rho(x)$ is the usual product.

I am interested in Kronecker multiplicities for the Monster group $M$. While the group is large, there are only 194 conjugacy classes.

$$(1) \qquad \max_{\phi,\psi,\rho} g(\phi,\psi,\rho)$$

$$(2) \qquad \sum_{\phi,\psi,\rho} g(\phi,\psi,\rho)$$ $$(3) \qquad \sum_{\phi,\psi,\rho} g(\phi,\psi,\rho)^2$$

These sums are over all triples of irreducible characters, but because of the symmetries only about 1/6 of them need to be computed to get the answer. If you can do this, I would also be curious about the specific characters maximizing (1).

The computation is beyond my computer algebra skills, but I know that GAP has the whole character table of $M$ ready to use.

  • 2
    $\begingroup$ One can try to ping @user:2820 (Derek Holt), he might know. $\endgroup$ Feb 27 '18 at 0:55

In GAP, you could simply iterate in a triple loop over $\phi,\psi,\rho$, calculate the $g$-values and find maximum and sum values:


m:=0; s:=0;q:=0; # max, sum, sumsquare

for rho in irrs do
  for psi in irrs do
    ten:=rho*psi; # tensor product
    for phi in irrs do
      if g>m then m:=g;fi;

Afterwards look at the values of m, s, and q. Unless I have mistyped something, the results I get are (for the monster group):

  1. Maximum is $21458051228477513179513856=2^{10}\cdot281\cdot443\cdot599\cdot6571\cdot42768299767$

  2. Sum is $247017097351847432984363535932$ (Thank you, @James for the correction)

  3. Sum-Squares is $808017424794512875894769468067441075690144312450960558$ (ditto corrected, also typo fixed)

  • 3
    $\begingroup$ I computed the same maximum, but got different values for the sum and sum of squares. My value for the sum is $247017097351847432984363535932$ and, for the sum of squares is $808017424794512875894769468067441075690144312450960558$. If I'm not mistaken, I think your code should have added your "g" and "g^2" rather than "m" and "m^2$ to "s" and "q" respectively. $\endgroup$
    – James
    Feb 28 '18 at 16:26
  • 2
    $\begingroup$ @James, Oops, you're completely right. I'll fix it. $\endgroup$
    – ahulpke
    Feb 28 '18 at 17:18
  • 1
    $\begingroup$ I was just about to report that I modified your GAP code and ran it and obtained the same results that I had seen from my own code. It is perhaps useful to know that the results can be replicated independently in two different CASs. $\endgroup$
    – James
    Feb 28 '18 at 17:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.