Recursive domain equations
Please expand.
Solving recursive domain equations
The 'standard' way of solving recursive domain equations is through domain theory.
Broadly speaking, there are three methods for solving them:
- The colimit method. This is the original method developed by Scott in the 1960s, and it involves taking a colimit of embedding-projection pairs. It is explained in detail in the classic paper by Smyth and Plotkin 1982. It yields a domain
D and an isomorphismD \cong F(D) . - The information system method. This method is based on Scott's alternative presentation of domains as information systems. It has the advantage of giving a solution
D that satisfiesD = F(D) on the nose (!). It is described in Larsen and Winskel 1991. - Through universal domains. This uses least fixed points of continuous operations on universal domains. It is described in Gunter and Scott 1990.
References
@article{smyth_1982,
title = {The {Category}-{Theoretic} {Solution} of {Recursive} {Domain} {Equations}},
volume = {11},
doi = {10.1137/0211062},
number = {4},
journal = {SIAM Journal on Computing},
author = {Smyth, M. B. and Plotkin, G. D.},
year = {1982},
pages = {761--783}
}
@article{larsen1991,
title = {Using information systems to solve recursive domain equations},
volume = {91},
doi = {10.1016/0890-5401(91)90068-D},
pages = {232--258},
number = {2},
journaltitle = {Information and Computation},
author = {Larsen, Kim Guldstrand and Winskel, Glynn},
date = {1991},
}
@incollection{gunter_1990,
title = {Semantic Domains},
volume = {B},
pages = {633--674},
booktitle = {Handbook of Theoretical Comptuer Science},
publisher = {Elsevier},
author = {Gunter, D. and Scott, D. S.},
editor = {van Leeuwen, J.},
date = {1990},
doi = {10.1016/B978-0-444-88074-1.50017-2}
}