https://github.com/chrispsn/kgen

Experimenting with code to generate and optimise K/APL code.
https://github.com/chrispsn/kgen

apl code-generation k optimization synthesis

Last synced: 11 months ago
JSON representation

Experimenting with code to generate and optimise K/APL code.

• Host: GitHub
• URL: https://github.com/chrispsn/kgen
• Owner: chrispsn
• License: gpl-3.0
• Created: 2018-09-02T12:45:52.000Z (almost 8 years ago)
• Default Branch: master
• Last Pushed: 2019-04-04T13:48:16.000Z (about 7 years ago)
• Last Synced: 2025-03-31T06:51:11.622Z (about 1 year ago)
• Topics: apl, code-generation, k, optimization, synthesis
• Language: JavaScript
• Homepage:
• Size: 30.3 KB
• Stars: 5
• Watchers: 4
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# kGen

Experimenting with code to generate and optimise K/APL code.

Motivation: can programs generate:

- simple K/APL programs, like those here? http://kparc.com/z/fun.k

- magic functions (implementations of new primitives, or faster implemetations of common recognised idioms)? https://www.dyalog.com/blog/2015/06/in-praise-of-magic-functions-part-one/

- C/asm code for old primitives?

- deduce the optimal abstractions to assign to n available 'symbol slots' (say the ASCII character set)?

Other resources to be fully grokked:

- https://dyalog.tv/Dyalog17/?v=Wff2InJn3p0

- https://github.com/e9gille/gpapl

- STOKE http://stoke.stanford.edu/ (especially Stochastic Program Optimization)

- Souper https://arxiv.org/abs/1711.04422

Putting this up so I can track progress and compare before/after results, but no promises this will go anywhere...