Matrix prime number theorems (Q913848): Difference between revisions

Let \(N\in {\mathbb{N}}\) and \(\alpha =N^{A/N}\), where \(A>1\) is constant, and define \[ f(\vartheta)=\sum^{N}_{j=1}\{\psi (\alpha^ j)-\psi (\alpha^{j-1})-\alpha^{j-1}(\alpha -1)\}e^{ij\vartheta}. \] Then it is shown that \(f(\vartheta)=o_ A(\alpha^ N)\) uniformly in \(\vartheta\). The proof uses standard methods. If one defines \[ c_ n=\int^{2\pi}_{0}e^{in\vartheta}(f(\vartheta)+f(-\vartheta)) d\vartheta \] one may deduce a non-trivial bound for the spectral radius of a family of Toeplitz matrices \((c_{j-k})_{jk}\) whose entries are formed from the coefficients of f(\(\vartheta\)), and thus reflect the distribution of primes in short intervals. Various similar deductions are mentioned.