Statistical Theory of the Energy Levels of Complex Systems. IV
Статистическая теория энергетических уровней сложных систем. IV
1963-05-01
SCID: 54.1/amy673tk
Discuss with AI
eigenvalue statisticsenergy level spacinglong-range orderneutron capture levelsrandom matrix theory
Figures from the paper
Abstract (AI)
In previous papers of this series, a theory was constructed for the description of the statistical properties of the eigenvalues of a random matrix of high order. This paper deduces from the same theoretical model the statistical behavior to be expected for a finite stretch of observed eigenvalues chosen out of a much longer stretch of unobserved ones. In comparing the model with experimental data, we have always to deal with such a finite stretch of observed levels. Three ``statistics'' are investigated; these are quantities which can be computed directly from the observed data, and for which the expectation values and statistical variances can be calculated from the theory. One statistic gives only a precise way of measuring the average level spacing. One provides a test of the model by measuring the degree of long-range order in the level series. The third provides an independent test of the model by measuring the degree of short-range order. In Sec. V these methods are applied to a preliminary analysis of some experimental data on neutron capture levels in heavy nuclei. The results are inconclusive. Discrepancies between theory and observation are large, but the discrepancies might be produced either by incompleteness of the data or by incorrectness of the theoretical model.
Key Findings
1
A preliminary application to neutron-capture levels in heavy nuclei produced inconclusive results, with substantial theory–experiment discrepancies.
2
It defines three directly computable statistics with theoretically calculable expectations and variances: mean spacing, long-range order, and short-range order.
3
The long-range and short-range statistics provide independent tests of the random-matrix model’s spectral correlations.
4
The observed discrepancies could reflect incomplete data or inadequacy of the theoretical model, preventing a definitive conclusion.
5
The paper derives statistical predictions for finite observed stretches of eigenvalues sampled from much longer unobserved random-matrix spectra.
Research Object
finite stretches of observed eigenvalues (energy levels) from complex systems, modeled as eigenvalues of high-order random matrices
Research Subject
statistical behavior and long- and short-range level-order properties, including average spacing, of the observed eigenvalue series
Publication Details
Publication Date
1963-05-01
Journal
Publisher
ISSN
Cited by
638
Access Type
Author Information
Download PDF
Subscribe to digest