Switch from internal 5-arg searchsorted* methods to views
#440
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Codecov Report
@@ Coverage Diff @@
## main #440 +/- ##
==========================================
+ Coverage 92.42% 93.20% +0.78%
==========================================
Files 12 12
Lines 7667 7668 +1
==========================================
+ Hits 7086 7147 +61
+ Misses 581 521 -60
... and 1 file with indirect coverage changes 📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more |
|
The benchmark in the OP is not benchmarking the direct replacement. Is there a benchmark for e.g. |
|
Adding the drop-in replacement to the OP doesn't change much julia> @btime searchsortedfirst($x, 250, 400, 600, Base.Forward)
18.597 ns (0 allocations: 0 bytes)
498
julia> @btime searchsortedfirst(view($x, 400:600), 250) + 400 - 1
18.806 ns (0 allocations: 0 bytes)
498
julia> @btime searchsortedfirst(view($x, 400:600), 250)
18.806 ns (0 allocations: 0 bytes)
99I expect the additional addition and decrement operations to be negligible in a larger program because the decrement can probably be eliminated through some means or another and the subtraction might result in increased or decreased work, depending on if it cancels out another operation. For end to end, I'm getting noisy and unreliable benchmarking results :( however, they do not show a regression. I'd be somewhat surprised if this caused a noticeable regression. Before julia> versioninfo()
Julia Version 1.11.0-DEV.471
Commit 153b538ff0* (2023-09-05 21:23 UTC)
Platform Info:
OS: macOS (arm64-apple-darwin22.4.0)
CPU: 8 × Apple M2
WORD_SIZE: 64
LLVM: libLLVM-15.0.7 (ORCJIT, apple-m1)
Threads: 1 on 4 virtual cores
Environment:
JULIA_EDITOR = code
julia> using SparseArrays, BenchmarkTools, Random
julia> Random.seed!(1729)
TaskLocalRNG()
julia> x = sprand(1000, 1000, .001)
1000×1000 SparseMatrixCSC{Float64, Int64} with 1057 stored entries:
⎡⢉⠀⠐⠀⠁⠁⠂⠀⠁⠄⠀⠠⠀⠀⠈⠁⠈⠀⡁⡁⠀⢀⠀⠠⠃⠉⠆⡀⠢⠠⢀⠂⠂⠀⠠⠀⠄⠄⠄⠀⎤
⎢⠰⡗⠈⠥⠁⠄⠠⢀⠡⠂⠀⠀⠄⠄⠤⡅⢀⠂⡄⡀⠀⠂⠁⠔⡀⠘⠀⢈⠀⡰⠌⠀⢰⠀⠒⠂⠀⠒⢂⠀⎥
⎢⢤⡀⠀⠈⠐⠀⠈⠄⠀⠀⠀⡤⡀⠀⠄⡐⠌⢂⠔⠀⢐⠖⢀⠥⠀⡀⠐⡠⠄⠐⡐⠀⢠⠀⡀⠒⠘⡘⠒⠀⎥
⎢⠐⠄⠀⠅⠀⠀⠤⠀⢀⠀⠀⠀⠀⡀⡈⠰⠡⠀⠄⡀⡀⠀⠈⠢⠈⡄⠀⠀⠊⠜⠀⠘⠠⠠⠁⠊⠅⠒⠎⢀⎥
⎢⠀⠂⡐⠀⠌⠁⢡⢐⢄⠈⠂⠀⠀⠐⠒⢄⠠⠄⠁⠁⠀⢀⣀⠀⠀⠃⠂⠀⡐⠄⠄⠀⠀⢀⠀⠍⠠⠄⡠⠌⎥
⎢⠀⠐⠐⠁⠀⠰⢨⢀⠂⠀⠀⡠⡂⠐⡒⠄⠀⠆⠘⢘⢂⠀⠀⠈⠈⡀⠆⡘⠌⠃⠀⠂⡄⠀⠈⠀⢂⠐⠁⠀⎥
⎢⠀⠁⠑⠀⢀⡄⠂⠀⠂⢀⡁⡀⠲⢈⡠⠀⢀⠀⠀⠆⠀⢂⢔⠀⠀⠡⠄⠠⠠⢠⠄⠈⢈⠐⠋⡨⡀⡀⠌⠀⎥
⎢⠀⠠⠀⡈⠐⡀⠂⠀⠠⠀⠈⢅⠁⠀⠀⠀⠀⠀⡄⢉⠀⠅⠁⠠⠁⢐⠀⠀⡀⠈⠀⠀⠀⠠⠠⠄⠁⠀⢀⢠⎥
⎢⠐⣠⢂⠊⡠⡲⠘⠀⠀⠁⠐⠄⠍⠁⢄⠀⠠⠀⠀⠄⠐⠀⡀⢋⠀⠠⠄⠡⠀⠁⠐⡀⠀⠀⠈⡩⠀⡀⠠⠐⎥
⎢⠀⠉⠁⠀⠈⢐⠰⠂⠀⢄⠀⠠⠆⡌⢅⠀⣑⠅⠓⠐⡄⢀⠅⡀⠠⠁⡀⠡⠀⠠⠠⡐⠈⡄⠔⠴⠤⡡⡀⠁⎥
⎢⠁⡁⠀⠀⠝⠃⠉⠀⠰⡀⠀⠈⠐⡌⠂⠀⠠⢡⠈⡀⠱⠐⠀⢀⢀⠵⠀⠒⡂⠐⠈⣤⣀⡀⡌⠀⠠⠈⠀⢀⎥
⎢⠂⢀⠰⠀⡀⠁⠈⢉⢑⡂⢀⡈⣀⠂⢁⠀⠊⢀⡀⢀⡀⠄⡂⠈⠀⠁⡀⠁⠂⠀⡄⠀⠀⠀⠐⠄⡜⠈⠀⠀⎥
⎢⠠⡀⠂⠀⢄⠄⠀⠠⡀⢁⠂⢈⢰⠉⠂⡀⠂⠠⠀⠐⡀⠠⠂⡀⡀⢀⢑⢀⠐⠐⠠⡄⠀⠀⠐⠀⢒⠄⢁⠌⎥
⎢⠀⢀⠌⠐⠂⢀⠉⠌⠄⢁⠰⠁⡀⠀⡀⠄⠀⠈⡀⠀⢄⢀⠀⢢⠆⠊⡀⠤⡀⠀⡈⠐⠀⠅⠀⠌⢎⠐⠁⠅⎥
⎢⠊⠀⠀⠀⠰⡰⡨⡀⠀⢀⡄⠀⡀⡠⠰⠀⠀⠀⠀⣒⢆⣁⠂⠈⠀⠐⠐⠂⡈⠀⠃⠂⠠⠠⠌⠠⠨⠂⠢⡀⎥
⎢⠁⢀⡒⠂⠈⢩⢀⠀⢨⡄⡀⢄⠈⡂⠐⠀⡀⠠⠤⡰⢀⠀⠀⠀⠀⡄⠀⠪⠀⠁⠈⠀⡄⠠⠃⠀⠈⢠⡡⠄⎥
⎢⠀⡂⠕⠡⠀⠆⠠⠅⠁⠀⡑⠀⢌⠀⡂⠂⠅⠩⠁⢀⠑⠀⢠⠄⠠⡆⠂⠣⠀⠄⠂⠢⠂⢰⠄⠖⠥⠐⠂⠈⎥
⎢⠐⠀⠀⡄⠀⠌⠠⢄⢀⠌⡀⠎⠐⡁⠐⠁⡀⠣⠁⠂⠃⠁⢴⠃⠀⡁⠉⠊⠀⠒⠁⠀⠀⠨⠈⠀⠠⠒⡄⠀⎥
⎢⡂⠀⠠⠍⠀⠀⠄⡀⠔⢀⠁⠈⢄⣅⢄⡂⢈⠰⠠⠃⡁⠱⠒⢀⡀⠀⠀⠙⠂⠈⠁⠀⠘⠔⠄⠀⠒⡀⠂⣀⎥
⎣⠈⠀⠈⢀⠉⠀⠀⠑⠀⠀⡐⠀⠆⠀⢈⠀⠀⡄⠂⠉⠈⠀⡘⠀⠀⠈⠒⠐⢒⢉⠈⠁⠁⠀⠂⠀⠠⠂⢀⠄⎦
julia> @benchmark getindex($x, i, j) setup=(i=rand(1:1000); j=rand(1:1000)) gctrial=false gcsample=false evals=1_000 samples=10_000
BenchmarkTools.Trial: 10000 samples with 1000 evaluations.
Range (min … max): 2.125 ns … 22.875 ns ┊ GC (min … max): 0.00% … 0.00%
Time (median): 2.542 ns ┊ GC (median): 0.00%
Time (mean ± σ): 2.616 ns ± 0.469 ns ┊ GC (mean ± σ): 0.00% ± 0.00%
█▁ ▃ ▄
▂▃██▃▁▄▃▂▂▁▇█▃▅▁█▇▂▂▃▂▃▂▂▂▁▂▂▅█▁▃▂▂▂▁▂▂▂▂▂▁▂▂▂▂▁▁▁▁▁▁▁▂▂▂▂ ▃
2.12 ns Histogram: frequency by time 4.08 ns <
Memory estimate: 0 bytes, allocs estimate: 0.
julia> nonzeros = findall(x .!= 0);
julia> @benchmark getindex($x, i, j) setup=((i, j)=Tuple(rand(nonzeros))) gctrial=false gcsample=false evals=1_000 samples=10_000
BenchmarkTools.Trial: 10000 samples with 1000 evaluations.
Range (min … max): 20.292 ns … 14.414 μs ┊ GC (min … max): 0.00% … 99.58%
Time (median): 22.375 ns ┊ GC (median): 0.00%
Time (mean ± σ): 26.562 ns ± 168.124 ns ┊ GC (mean ± σ): 12.13% ± 2.60%
▂▁ ▁▃▃█
▁▂▂▂▃▅▇██▆█████▄▃▂▂▃▂▂▂▂▂▂▃▄▃▄▄▃▄▅▄▅▄▃▃▃▃▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁ ▃
20.3 ns Histogram: frequency by time 29.5 ns <
Memory estimate: 64 bytes, allocs estimate: 2.
julia> @benchmark getindex($x, i, j) setup=((i, j)=rand((rand(1:1000), rand(1:1000)), Tuple(rand(nonzeros)))) gctrial=false gcsample=false evals=1_000 samples=10_000 seconds=10
BenchmarkTools.Trial: 10000 samples with 1000 evaluations.
Range (min … max): 18.917 ns … 58.875 ns ┊ GC (min … max): 0.00% … 0.00%
Time (median): 20.375 ns ┊ GC (median): 0.00%
Time (mean ± σ): 20.537 ns ± 1.161 ns ┊ GC (mean ± σ): 0.00% ± 0.00%
▁█▅▆█▃▆▁ ▂
▂▁▂▂▂▃▃▄▇▇██████████▆▅▆▄▃▄▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▁▂▁▂▂▂▂▂▁▂▂▁▂▂▂▂▂▂ ▃
18.9 ns Histogram: frequency by time 24.7 ns <
Memory estimate: 64 bytes, allocs estimate: 2.
julia> @less getindex(x, 1, 1)
@RCI @propagate_inbounds function getindex(A::AbstractSparseMatrixCSC{T}, i0::Integer, i1::Integer) where T
@boundscheck checkbounds(A, i0, i1)
r1 = Int(@inbounds getcolptr(A)[i1])
r2 = Int(@inbounds getcolptr(A)[i1+1]-1)
(r1 > r2) && return zero(T)
r1 = searchsortedfirst(rowvals(A), i0, r1, r2, Forward)
((r1 > r2) || (rowvals(A)[r1] != i0)) ? zero(T) : nonzeros(A)[r1]
endAfter (@v1.11) pkg> activate --temp
Activating new project at `/var/folders/hc/fn82kz1j5vl8w7lwd4l079y80000gn/T/jl_08EiCA`
(jl_08EiCA) pkg> dev SparseArrays
Resolving package versions...
Updating `/private/var/folders/hc/fn82kz1j5vl8w7lwd4l079y80000gn/T/jl_08EiCA/Project.toml`
[2f01184e] + SparseArrays v1.11.0 `~/.julia/dev/SparseArrays`
Updating `/private/var/folders/hc/fn82kz1j5vl8w7lwd4l079y80000gn/T/jl_08EiCA/Manifest.toml`
[0dad84c5] + ArgTools v1.1.1
[56f22d72] + Artifacts
[2a0f44e3] + Base64
[ade2ca70] + Dates
[f43a241f] + Downloads v1.6.0
[7b1f6079] + FileWatching
[b77e0a4c] + InteractiveUtils
[b27032c2] + LibCURL v0.6.4
[76f85450] + LibGit2
[8f399da3] + Libdl
[37e2e46d] + LinearAlgebra
[56ddb016] + Logging
[d6f4376e] + Markdown
[ca575930] + NetworkOptions v1.2.0
[44cfe95a] + Pkg v1.11.0
[de0858da] + Printf
[3fa0cd96] + REPL
[9a3f8284] + Random
[ea8e919c] + SHA v0.7.0
[9e88b42a] + Serialization
[6462fe0b] + Sockets
[2f01184e] + SparseArrays v1.11.0 `~/.julia/dev/SparseArrays`
[fa267f1f] + TOML v1.0.3
[a4e569a6] + Tar v1.10.0
[cf7118a7] + UUIDs
[4ec0a83e] + Unicode
[e66e0078] + CompilerSupportLibraries_jll v1.0.5+1
[deac9b47] + LibCURL_jll v8.0.1+1
[29816b5a] + LibSSH2_jll v1.11.0+1
[c8ffd9c3] + MbedTLS_jll v2.28.2+1
[14a3606d] + MozillaCACerts_jll v2023.1.10
[4536629a] + OpenBLAS_jll v0.3.23+2
[bea87d4a] + SuiteSparse_jll v7.2.0+1
[83775a58] + Zlib_jll v1.2.13+1
[8e850b90] + libblastrampoline_jll v5.8.0+1
[8e850ede] + nghttp2_jll v1.52.0+1
[3f19e933] + p7zip_jll v17.4.0+2
julia> versioninfo()
Julia Version 1.11.0-DEV.471
Commit 153b538ff0* (2023-09-05 21:23 UTC)
Platform Info:
OS: macOS (arm64-apple-darwin22.4.0)
CPU: 8 × Apple M2
WORD_SIZE: 64
LLVM: libLLVM-15.0.7 (ORCJIT, apple-m1)
Threads: 1 on 4 virtual cores
Environment:
JULIA_EDITOR = code
julia> using SparseArrays, BenchmarkTools, Random
julia> Random.seed!(1729)
TaskLocalRNG()
julia> x = sprand(1000, 1000, .001)
1000×1000 SparseMatrixCSC{Float64, Int64} with 1057 stored entries:
⎡⢉⠀⠐⠀⠁⠁⠂⠀⠁⠄⠀⠠⠀⠀⠈⠁⠈⠀⡁⡁⠀⢀⠀⠠⠃⠉⠆⡀⠢⠠⢀⠂⠂⠀⠠⠀⠄⠄⠄⠀⎤
⎢⠰⡗⠈⠥⠁⠄⠠⢀⠡⠂⠀⠀⠄⠄⠤⡅⢀⠂⡄⡀⠀⠂⠁⠔⡀⠘⠀⢈⠀⡰⠌⠀⢰⠀⠒⠂⠀⠒⢂⠀⎥
⎢⢤⡀⠀⠈⠐⠀⠈⠄⠀⠀⠀⡤⡀⠀⠄⡐⠌⢂⠔⠀⢐⠖⢀⠥⠀⡀⠐⡠⠄⠐⡐⠀⢠⠀⡀⠒⠘⡘⠒⠀⎥
⎢⠐⠄⠀⠅⠀⠀⠤⠀⢀⠀⠀⠀⠀⡀⡈⠰⠡⠀⠄⡀⡀⠀⠈⠢⠈⡄⠀⠀⠊⠜⠀⠘⠠⠠⠁⠊⠅⠒⠎⢀⎥
⎢⠀⠂⡐⠀⠌⠁⢡⢐⢄⠈⠂⠀⠀⠐⠒⢄⠠⠄⠁⠁⠀⢀⣀⠀⠀⠃⠂⠀⡐⠄⠄⠀⠀⢀⠀⠍⠠⠄⡠⠌⎥
⎢⠀⠐⠐⠁⠀⠰⢨⢀⠂⠀⠀⡠⡂⠐⡒⠄⠀⠆⠘⢘⢂⠀⠀⠈⠈⡀⠆⡘⠌⠃⠀⠂⡄⠀⠈⠀⢂⠐⠁⠀⎥
⎢⠀⠁⠑⠀⢀⡄⠂⠀⠂⢀⡁⡀⠲⢈⡠⠀⢀⠀⠀⠆⠀⢂⢔⠀⠀⠡⠄⠠⠠⢠⠄⠈⢈⠐⠋⡨⡀⡀⠌⠀⎥
⎢⠀⠠⠀⡈⠐⡀⠂⠀⠠⠀⠈⢅⠁⠀⠀⠀⠀⠀⡄⢉⠀⠅⠁⠠⠁⢐⠀⠀⡀⠈⠀⠀⠀⠠⠠⠄⠁⠀⢀⢠⎥
⎢⠐⣠⢂⠊⡠⡲⠘⠀⠀⠁⠐⠄⠍⠁⢄⠀⠠⠀⠀⠄⠐⠀⡀⢋⠀⠠⠄⠡⠀⠁⠐⡀⠀⠀⠈⡩⠀⡀⠠⠐⎥
⎢⠀⠉⠁⠀⠈⢐⠰⠂⠀⢄⠀⠠⠆⡌⢅⠀⣑⠅⠓⠐⡄⢀⠅⡀⠠⠁⡀⠡⠀⠠⠠⡐⠈⡄⠔⠴⠤⡡⡀⠁⎥
⎢⠁⡁⠀⠀⠝⠃⠉⠀⠰⡀⠀⠈⠐⡌⠂⠀⠠⢡⠈⡀⠱⠐⠀⢀⢀⠵⠀⠒⡂⠐⠈⣤⣀⡀⡌⠀⠠⠈⠀⢀⎥
⎢⠂⢀⠰⠀⡀⠁⠈⢉⢑⡂⢀⡈⣀⠂⢁⠀⠊⢀⡀⢀⡀⠄⡂⠈⠀⠁⡀⠁⠂⠀⡄⠀⠀⠀⠐⠄⡜⠈⠀⠀⎥
⎢⠠⡀⠂⠀⢄⠄⠀⠠⡀⢁⠂⢈⢰⠉⠂⡀⠂⠠⠀⠐⡀⠠⠂⡀⡀⢀⢑⢀⠐⠐⠠⡄⠀⠀⠐⠀⢒⠄⢁⠌⎥
⎢⠀⢀⠌⠐⠂⢀⠉⠌⠄⢁⠰⠁⡀⠀⡀⠄⠀⠈⡀⠀⢄⢀⠀⢢⠆⠊⡀⠤⡀⠀⡈⠐⠀⠅⠀⠌⢎⠐⠁⠅⎥
⎢⠊⠀⠀⠀⠰⡰⡨⡀⠀⢀⡄⠀⡀⡠⠰⠀⠀⠀⠀⣒⢆⣁⠂⠈⠀⠐⠐⠂⡈⠀⠃⠂⠠⠠⠌⠠⠨⠂⠢⡀⎥
⎢⠁⢀⡒⠂⠈⢩⢀⠀⢨⡄⡀⢄⠈⡂⠐⠀⡀⠠⠤⡰⢀⠀⠀⠀⠀⡄⠀⠪⠀⠁⠈⠀⡄⠠⠃⠀⠈⢠⡡⠄⎥
⎢⠀⡂⠕⠡⠀⠆⠠⠅⠁⠀⡑⠀⢌⠀⡂⠂⠅⠩⠁⢀⠑⠀⢠⠄⠠⡆⠂⠣⠀⠄⠂⠢⠂⢰⠄⠖⠥⠐⠂⠈⎥
⎢⠐⠀⠀⡄⠀⠌⠠⢄⢀⠌⡀⠎⠐⡁⠐⠁⡀⠣⠁⠂⠃⠁⢴⠃⠀⡁⠉⠊⠀⠒⠁⠀⠀⠨⠈⠀⠠⠒⡄⠀⎥
⎢⡂⠀⠠⠍⠀⠀⠄⡀⠔⢀⠁⠈⢄⣅⢄⡂⢈⠰⠠⠃⡁⠱⠒⢀⡀⠀⠀⠙⠂⠈⠁⠀⠘⠔⠄⠀⠒⡀⠂⣀⎥
⎣⠈⠀⠈⢀⠉⠀⠀⠑⠀⠀⡐⠀⠆⠀⢈⠀⠀⡄⠂⠉⠈⠀⡘⠀⠀⠈⠒⠐⢒⢉⠈⠁⠁⠀⠂⠀⠠⠂⢀⠄⎦
julia> @benchmark getindex($x, i, j) setup=(i=rand(1:1000); j=rand(1:1000)) gctrial=false gcsample=false evals=1_000 samples=10_000
BenchmarkTools.Trial: 10000 samples with 1000 evaluations.
Range (min … max): 2.083 ns … 26.459 ns ┊ GC (min … max): 0.00% … 0.00%
Time (median): 2.875 ns ┊ GC (median): 0.00%
Time (mean ± σ): 2.870 ns ± 0.813 ns ┊ GC (mean ± σ): 0.00% ± 0.00%
█ ▁ ▁
▂▂▄██▄▁▂▂▂▂▁▁▁▂▂▂▂▂▁▃█▅▄▃▁▂▂▂▅▆▄▁▄▃▂▂▃▁█▄▄▄▃▃▁▂▂▂▁▂▁▂▂▁▂▂▂ ▃
2.08 ns Histogram: frequency by time 4.12 ns <
Memory estimate: 0 bytes, allocs estimate: 0.
julia> nonzeros = findall(x .!= 0);
julia> @benchmark getindex($x, i, j) setup=((i, j)=Tuple(rand(nonzeros))) gctrial=false gcsample=false evals=1_000 samples=10_000
BenchmarkTools.Trial: 10000 samples with 1000 evaluations.
Range (min … max): 16.000 ns … 22.415 μs ┊ GC (min … max): 0.00% … 99.71%
Time (median): 19.959 ns ┊ GC (median): 0.00%
Time (mean ± σ): 24.752 ns ± 252.301 ns ┊ GC (mean ± σ): 16.53% ± 2.21%
▂█▇▃ ▁▁
▂▃████▇▅▅▅▅▅▇████▇▆▅▅▅▅▅▄▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▂▂▂▁▁▂▂▂▂▂▂▂ ▃
16 ns Histogram: frequency by time 36.5 ns <
Memory estimate: 64 bytes, allocs estimate: 2.
julia> @benchmark getindex($x, i, j) setup=((i, j)=rand((rand(1:1000), rand(1:1000)), Tuple(rand(nonzeros)))) gctrial=false gcsample=false evals=1_000 samples=10_000 seconds=10
BenchmarkTools.Trial: 10000 samples with 1000 evaluations.
Range (min … max): 14.125 ns … 99.250 ns ┊ GC (min … max): 0.00% … 0.00%
Time (median): 16.875 ns ┊ GC (median): 0.00%
Time (mean ± σ): 17.289 ns ± 3.290 ns ┊ GC (mean ± σ): 0.00% ± 0.00%
▁▂ ▁▁▃▂▆█▇▃▄▃
▁▁▂▄███████████████▆▆▆▅▃▃▃▂▂▂▂▂▂▂▂▂▁▂▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁ ▃
14.1 ns Histogram: frequency by time 25.8 ns <
Memory estimate: 64 bytes, allocs estimate: 2.
julia> @less getindex(x, 1, 1)
@RCI @propagate_inbounds function getindex(A::AbstractSparseMatrixCSC{T}, i0::In
teger, i1::Integer) where T
@boundscheck checkbounds(A, i0, i1)
r1 = Int(@inbounds getcolptr(A)[i1])
r2 = Int(@inbounds getcolptr(A)[i1+1]-1)
(r1 > r2) && return zero(T)
r1 = searchsortedfirst(view(rowvals(A), r1:r2), i0) + r1 - 1
((r1 > r2) || (rowvals(A)[r1] != i0)) ? zero(T) : nonzeros(A)[r1]
end |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There should be little performance impact:
See JuliaLang/julia#51177 and JuliaLang/julia#51176 for motivation