pomaps: Maps and sets of partial orders

[ data-structures, library, mit ] [ Propose Tags ]

Maps (and sets) indexed by keys satisfying PartialOrd.

The goal is to provide asymptotically better data structures than simple association lists or lookup tables. Asymptotics depend on the partial order used as keys, its width \(w\) specifically (the size of the biggest anti-chain).

For partial orders with great width, this package won't provide any benefit over using association lists, so benchmark for your use-case!

[Skip to Readme]
Change log CHANGELOG.md
Dependencies base (>= && <4.12), containers (>= && <=, deepseq (>=1.1 && <1.5), ghc-prim (>=0.4 && <0.6), lattices (>=1.7 && <2) [details]
License MIT
Maintainer Sebastian Graf <sgraf1337@gmail.com>
Category Data Structures
Home page https://github.com/sgraf812/pomaps#readme
Bug tracker https://github.com/sgraf812/pomaps/issues
Source repo head: git clone https://github.com/sgraf812/pomaps
Uploaded by sgraf812 at Sun Apr 8 19:07:23 UTC 2018
Distributions LTSHaskell:, NixOS:, openSUSE:
Downloads 314 total (25 in the last 30 days)
Rating (no votes yet) [estimated by rule of succession]
Your Rating
  • λ
  • λ
  • λ
Status Docs available [build log]
Last success reported on 2018-04-13 [all 1 reports]
Hackage Matrix CI





Don't depend on the lattices package and extract the PartialOrd class.


Use -f <flag> to enable a flag, or -f -<flag> to disable that flag. More info


Maintainer's Corner

For package maintainers and hackage trustees

Readme for pomaps-

[back to package description]

[pomaps][pomaps] Build Status Hackage

Reasonably fast maps (and possibly sets) based on keys satisfying PartialOrd. This package tries to load off as much work as possible to the excellent containers library, in order to achieve acceptable performance. The interface is kept as similar to Data.Map.{Strict,Lazy} as possible, which is an excuse for somewhat lacking documentation. POMaps basically store a decomposition of totally ordered chains (e.g. something Maps can handle). Functionality and strictness properties should be pretty much covered by the testsuite. But it's not battle-tested yet, so if you encounter space leaks in the implementation, let me know. A rather naive implementation leads to O(w*n*log n) lookups, where w is the width of the decomposition (which should be the size of the biggest anti-chain). This is enough for me at the moment to get things going, but there is room for improvement (Sorting and Selection in Posets). Let me know if things are too slow and I'll see what I can do! [pomaps]: https://github.com/sgraf812/pomaps