Generalizations of theorems of Wilson, Fermat and Euler (Q1166551): Difference between revisions

The author uses Pólya-Bruijn enumeration theory to give some generalizations of the theorems in the title. So, for instance, Wilson's theorem is a consequence of the following congruence \[ \sum_{d\mid n} (\varphi(d))^2 \cdot d^{n/d-1} \cdot \left(\frac{n}{d} - 1\right)! \equiv 0 \bmod n \] being true for every positive integer \(n\). The proofs in this part of the paper are based on the explicit knowledge of the cycle index of the cyclic permutation group \(C_n = <(12\cdots n)>\) of order \(n\) generated by the cycle \((12\cdots n)\) acting on the set \(D= \{1,2,\ldots,n\}\). Then an algorithm for the determination of equivalence classes of functions \(f\in R^n\) where \(D= \{1,2,\ldots,n\}\), \(R= \{1,2,\ldots,t\}\) is given. The algorithm uses the incidence matrix of the function which class is determined and the permutation matrix groups corresponding to the groups acting on \(D\) and \(R\). In the final section of the paper the author presents some interpretations for Wilson's and Fermat's quotients using some properties of this algorithm. Thus \(\frac1{p}(a^p - a)\) is the number of equivalence classes each containing \(p\) functions in \(R^n\) \((n=p,\ t=a)\) if \(C_p\) acts on \(D\) and the identity group on \(R\).