Jeffrey Brown - FGL, Digraphs With Text
969 بار بازدید -
9 سال پیش
-
Not everything that tries to
Not everything that tries to be a tree is a tree, but everything can be represented as a graph*.
In this talk first we'll look at the Functional Graph Library.
https://hackage.haskell.org/package/fgl
I won't present his introduction, but Tikhon Jelvis provides a short, beautiful introduction to the FGL (and the ViewPatterns pragma, and randomness):
http://jelv.is/blog/Generating-Mazes-...
After describing FGL, I will present my library Digraphs With Text:
https://github.com/JeffreyBenjaminBro...
which allows relationships of arity greater than two to be represented in FGL, without exposing (all of) the details of the underlying arity-two graph. There are two files in the src/Dwt/ directory, but ParseMm.hs is for our purposes ignorable; the other one, Graph.hs, is under 200 lines.
* So I believe.
references
* https://hackage.haskell.org/package/fgl
* https://github.com/JeffreyBenjaminBro...
thanks
* to Elliot Cameron, whose idea it was to use an underlying Graph in FGL, in which nodes and edges are very different, to represent a Mindmap in DWT, in which relationships can involve other relationships.
* to Tikhon Jelvis, whose Generating Mazes with Inductive Graphs I used some of the images from.
This was recorded at Brainium in Santa Monica during a Santa Monica Haskell meetup.
http://www.meetup.com/santa-monica-ha...
In this talk first we'll look at the Functional Graph Library.
https://hackage.haskell.org/package/fgl
I won't present his introduction, but Tikhon Jelvis provides a short, beautiful introduction to the FGL (and the ViewPatterns pragma, and randomness):
http://jelv.is/blog/Generating-Mazes-...
After describing FGL, I will present my library Digraphs With Text:
https://github.com/JeffreyBenjaminBro...
which allows relationships of arity greater than two to be represented in FGL, without exposing (all of) the details of the underlying arity-two graph. There are two files in the src/Dwt/ directory, but ParseMm.hs is for our purposes ignorable; the other one, Graph.hs, is under 200 lines.
* So I believe.
references
* https://hackage.haskell.org/package/fgl
* https://github.com/JeffreyBenjaminBro...
thanks
* to Elliot Cameron, whose idea it was to use an underlying Graph in FGL, in which nodes and edges are very different, to represent a Mindmap in DWT, in which relationships can involve other relationships.
* to Tikhon Jelvis, whose Generating Mazes with Inductive Graphs I used some of the images from.
This was recorded at Brainium in Santa Monica during a Santa Monica Haskell meetup.
http://www.meetup.com/santa-monica-ha...
9 سال پیش
در تاریخ 1394/10/10 منتشر شده
است.
969
بـار بازدید شده