A gardener collected 17 apples...

Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
32 replies · 4K views
PeroK said:
I've learned something new. The permament of a matrix:

https://en.wikipedia.org/wiki/Permanent_(mathematics)

This simplifies the counting, since the number of terms in the determinant for ##M## is the permanent of the matrix ##M'##, where all the ##\pm 1## entries are ##1##. And, the recurrence relation for the permanent of ##M'## is easy to show:
$$a(n) = (n-1)[a(n-1) + a(n-2)] = na(n-1) + (-1)^n$$This is clearly odd when ##n## is even.
Yes, new to me and good to know, especially for problems involving counting.

It explicitly shows the formula I've arrived to:

1700133862843.png


I suspect that because permanent of a matrix is basis dependent, its use in physics is limited.
 
Reply
  • Like
Likes   Reactions: PeroK
on Phys.org
Let ##X## be an ##n\times n## matrix with even numbers on the diagonal, and odd numbers everywhere else. Then ##\det X = n+1\pmod 2##.

Pf. Consider ##X## modulo ##2##. Then its diagonal is zero and there are ones everywhere else. Let ##S## be an ##n\times n## matrix whose every entry is ##1##. The eigenvalues of ##S## are precisely ##n## and ##0##. Therefore, ##S-I_n## has eigenvalues ##n-1## and ##-1##. Hence ##\det (S-I_n) = (n-1)^a(-1)^b##. Thus, the parity of ##\det X## is ##n+1\pmod{2}##. ##_{\blacksquare}##

Now, let ##n+1\geqslant 3## be odd and ##A## a square matrix of dimension ##n+1## such that its diagonal is zero and there are equal number of ##\pm 1## on each row. Since ##A(1,1,\ldots, 1)^t=0##, we have that ##\mathrm{rank}A \leqslant n##. Consider the ##n\times n## submatrix obtained from first ##n## rows and columns. By the previous claim, its determinant is an odd number, hence ##\mathrm{rank}A = n##.

Thus, in the initial problem, the vector ##(1,1,\ldots ,1)## generates the kernel of ##A##. Hence, every apple must have the same weight.
 
Last edited:
Reply
  • Like
Likes   Reactions: PeroK, WWGD and Hill
nuuskur said:
Let ##X## be an ##n\times n## matrix with even numbers on the diagonal, and odd numbers everywhere else. Then ##\det X = n+1\pmod 2##.

Pf. Consider ##X## modulo ##2##. Then its diagonal is zero and there are ones everywhere else. Let ##S## be a matrix whose every entry is ##1##. The eigenvalues of ##S## are precisely ##n## and ##0##. Therefore, ##S-I_n## has eigenvalues ##n-1## and ##-1##. Hence ##\det (S-I_n) = (n-1)^a(-1)^b##. Thus, the parity of ##\det X## is ##n+1\pmod{2}##. ##_{\blacksquare}##

Now, let ##n+1\geqslant 3## be odd and ##A## a square matrix of dimension ##n+1## such that its diagonal is zero and there are equal number of ##\pm 1## on each row. Since ##A(1,1,\ldots, 1)=0##, we have that ##\mathrm{rank}A \leqslant n##. Consider the ##n\times n## submatrix obtained from first ##n## rows and columns. By the previous claim, its determinant is an odd number, hence ##\mathrm{rank}A = n##.

Thus, in the initial problem, the vector ##(1,1,\ldots ,1)## generates the kernel of ##A##. Hence, every apple must have the same weight.
Thank you. This is much shorter and more elegant than my solution. No explicit counting needed. Beautiful!