SciML / SciML/ComponentArrays.jl

Slow construction

Open
#107 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Julia
Stars
375
Forks
42
Avg merge
7h 25m
Merged PRs (30d)
17

Description

  1. Construction of this ComponentArray is very slow
    n=3000; a=rand(n,n);b=rand(n)
    @btime C=ComponentArray(a=a,b=b);
    This takes more than 9 seconds even after the timing stablizes. Meanwhile, the following will be very fast
    @time C=ComponentArray(b=b); C=ComponentArray(C,b=b,a=a) #--> 0.002 seconds
    If you switch the above around it will be slow
    @time C=ComponentArray(a=a); C=ComponentArray(C,a=a,b=b) #--> 11 seconds

  2. It will also be nice to have a constructor in this format
    n = 3000;
    symbol_list = [:a, :b, :c]; size_list = [(n,n),(n,),(n,n)]
    C=ComponentArray(data, symbol_list, size_list); # Data may be pre-allocated, or it could be omitted, in which case the code calculates how much total is needed from size_list.
    Or it could be like this:
    ax=ConstructAxis(symbol_list, size_list);
    C=ComponentArray(data, ax)

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reproducing the two ComponentArray construction timings in the issue with n=3000, then trace the constructor path used for multiple named components. Compare the fast incremental construction with the slow combined construction; done means the reported slowdown is addressed and the requested constructor form is clarified or implemented.

Written by the indexing model from the issue text.

Assessment

Tech stack
julia
Domain
performance
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.