Alessandro Cheli
Alessandro Cheli
> note i've gone ahead and wrapped that rtmidi library here https://github.com/anandijain/RtMIDI.jl and put the jll in yggdrasil > > however, I haven't done any work to integrate it with...
> why not make egraphs faster :p > > I'll look into building it I have v2 of Metatheory in progress, do you want to help out? I need a...
I guess the failures are not related to this MR, but I still have to check manually, I'll merge after checking. Thanks a lot
I've managed to implement a trie tree of chars in gobba. It generates completions and hints from a list of strings (the keywords in the language). Tell me if you...
It would be nice to match on vectors. Other solutions such as https://github.com/kmsquire/Match.jl and https://github.com/RelationalAI-oss/Rematch.jl allow matching on arrays. This feature has to be carefully designed. Should iterators be allowed?...
A slowdown could be caused by the ematcher iterating through ALL eclasses in the egraph in every match trial. An optimization could be keeping a `Dict{Any, Vector{Int64}}` telling us in...
This optimization was done.
Got a greate speedup close to 2x and halved memory usage. Still not close to `egg`.
@MrVPlusOne could you help out with this issue? Your PRs were really useful
The ematch algorithm is detailed here. https://leodemoura.github.io/files/ematching.pdf I mostly based my implementation from egg. I was thinking today that instead of a virtual machine or staging through metaprogramming the ematcher...