You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Does ReverseDiff.jl have to be a dependency of this package? Alternatively, does this package have to be in the dependency graph of OrdinaryDiffEq? This is causing RD.jl to be loaded when DiffEq is loaded, and RD.jl contributes a full 3 seconds to the load time of OrdinaryDiffEq.jl, just for this package to provide the method Base.getindex(b::LazyBufferCache, u::ReverseDiff.TrackedArray)
julia> @time_imports using OrdinaryDiffEq
3.1 ms DocStringExtensions 67.85% compilation time
0.1 ms Reexport
0.2 ms Requires
0.0 ms SuiteSparse
2.1 ms ArrayInterfaceCore
0.5 ms Statistics
1.5 ms StaticArraysCore
371.5 ms StaticArrays
8.6 ms FunctionWrappers
0.2 ms MuladdMacro
4.5 ms OrderedCollections
0.1 ms UnPack
0.2 ms Parameters
11.8 ms FiniteDiff 60.53% compilation time (18% recompilation)
4.4 ms IrrationalConstants
0.5 ms DiffRules
1.4 ms DiffResults
20.2 ms Preferences
0.1 ms OpenLibm_jll
0.2 ms NaNMath
0.4 ms DelimitedFiles
0.2 ms Compat
45.6 ms ChainRulesCore
0.4 ms ChangesOfVariables
0.4 ms InverseFunctions
0.5 ms LogExpFunctions
0.2 ms JLLWrappers
3.1 ms OpenSpecFun_jll 70.08% compilation time
12.2 ms SpecialFunctions
8.3 ms MacroTools
0.2 ms CommonSubexpressions
62.0 ms ForwardDiff
0.7 ms ConstructionBase
57.3 ms Setfield 80.68% compilation time
19.5 ms RecipesBase
0.1 ms ZygoteRules
0.2 ms Adapt
0.2 ms ArrayInterfaceStaticArraysCore
117.0 ms FillArrays
0.1 ms DataValueInterfaces
0.5 ms DataAPI
0.1 ms IteratorInterfaceExtensions
0.0 ms TableTraits
9.6 ms Tables
2.4 ms GPUArraysCore
20.8 ms RecursiveArrayTools
10.4 ms IterativeSolvers
0.0 ms IfElse
13.9 ms Static
9.6 ms ArrayInterface
37.1 ms OffsetArrays
0.4 ms ArrayInterfaceOffsetArrays
0.7 ms ArrayInterfaceStaticArrays
0.1 ms SIMDTypes
0.8 ms ManualMemory
2.7 ms LayoutPointers
0.7 ms CpuId
144.2 ms CPUSummary 91.24% compilation time
0.1 ms BitTwiddlingConvenienceFunctions
64.7 ms HostCPUFeatures 75.37% compilation time
324.7 ms VectorizationBase
2.6 ms SLEEFPirates
44.4 ms ThreadingUtilities 79.12% compilation time
13.4 ms PolyesterWeave 72.62% compilation time
2.5 ms CloseOpenIntervals
0.1 ms SnoopPrecompile
2.1 ms SIMDDualNumbers
175.1 ms LoopVectorization
141.3 ms StrideArraysCore 2.60% compilation time
0.7 ms Polyester
134.2 ms TriangularSolve 5.41% compilation time
179.8 ms RecursiveFactorization
0.1 ms CommonSolve
0.5 ms FunctionWrappersWrappers
110.5 ms SciMLBase
4.2 ms NonlinearSolve
0.3 ms FastBroadcast
59.3 ms DataStructures
0.3 ms SortingAlgorithms
6.3 ms Missings
0.2 ms StatsAPI
23.7 ms StatsBase
16.0 ms PDMats
1.1 ms Rmath_jll
59.5 ms Rmath 90.40% compilation time
1.9 ms Calculus
11.5 ms DualNumbers
0.6 ms HypergeometricFunctions
5.2 ms StatsFuns
1.7 ms QuadGK
0.9 ms DensityInterface
174.8 ms Distributions
0.2 ms Tricks
22.6 ms DiffEqBase 54.41% compilation time
0.1 ms FastClosures
12.3 ms KLU
7.6 ms FastLapackInterface
5.1 ms Krylov
61.7 ms KrylovKit 15.50% compilation time
25.1 ms LinearSolve 16.88% compilation time
2974.4 ms ReverseDiff
0.6 ms PreallocationTools
21.3 ms GenericSchur
2.5 ms ExponentialUtilities
4.0 ms Distances
4.2 ms NLSolversBase
3.7 ms LineSearches
1.9 ms NLsolve
2.6 ms SimpleTraits
3.7 ms ArnoldiMethod
0.4 ms Inflate
33.9 ms Graphs
0.2 ms VertexSafeGraphs
19.4 ms SparseDiffTools 30.17% compilation time
0.1 ms ArrayInterfaceGPUArrays
1632.2 ms OrdinaryDiffEq
This was measured on Julia Version 1.9.0-DEV.1307 (2022-09-07) Commit 83658d265d8 (3 days old master)
The text was updated successfully, but these errors were encountered:
Does ReverseDiff.jl have to be a dependency of this package? Alternatively, does this package have to be in the dependency graph of OrdinaryDiffEq? This is causing RD.jl to be loaded when DiffEq is loaded, and RD.jl contributes a full 3 seconds to the load time of OrdinaryDiffEq.jl, just for this package to provide the method
Base.getindex(b::LazyBufferCache, u::ReverseDiff.TrackedArray)
This was measured on Julia Version 1.9.0-DEV.1307 (2022-09-07) Commit 83658d265d8 (3 days old master)
The text was updated successfully, but these errors were encountered: