Getting Started | Documentation | Glish | Learn More | Programming | Contact Us |
Version 1.9 Build 1367 |
|
This interpretation of Imaging Model is perhaps the one which many of us at Green Bank had in mind but did not have time to write down completely. I hope that the emphasis given here to linear algebra takes some of the mystery out of inverse methods and emphasizes the distinction between the linear part of the problem AI = D and the non-linear part encoded in algorithms (or Imagers) like MEM and CLEAN. The key advantage of this abstraction is that in AIPS++ we should be able to mix and match deconvolution algorithms and Imaging Models as required.
I therefore propose that AIPS++ adopt the definition of Imaging Model as a class with the limited, fixed set of services summarized in Appendix A. I'm sure that someone will be able to think of instruments and deconvolution algorithms for which this abstraction does not work. However, the fact that it does work in so many different situations means that we have much to gain by adopting it as the definition of an Imaging Model.