摘要
A Bayesian statistics approach for subtraction of incoherent scattering from neutron total-scattering data has been developed and implemented in a public domain software package. In this approach, the estimated background signal associated with incoherent scattering maximizes the posterior probability, which combines the likelihood of this signal in reciprocal and real spaces with the prior that favors smooth lines. The probability distributions are constructed according to the principle of maximum entropy. The method enables robust subtraction of incoherent-scattering backgrounds while providing estimated uncertainties for recovered signals. The developed procedure was first tested using simulated data and then demonstrated using three representative experimental data sets, collected on bulk materials and nanoparticles, featuring distinct ratios of coherent to incoherent scattering.
- 出版日期2014-12