[WIP] Setup DiffEqGPU integration #122
Merged
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
MWE:
```py
from diffeqpy import de
def f(u,p,t):
x, y, z = u
sigma, rho, beta = p
return [sigma * (y - x), x * (rho - z) - y, x * y - beta * z]
u0 = [1.0,0.0,0.0]
tspan = (0., 100.)
p = [10.0,28.0,8/3]
prob = de.ODEProblem(f, u0, tspan, p)
fast_prob = de.jit(prob)
sol = de.solve(fast_prob,saveat=0.01)
import random
def prob_func(prob,i,rep):
de.remake(prob,u0=[random.uniform(0, 1)*u0[i] for i in range(0,3)],
p=[random.uniform(0, 1)*p[i] for i in range(0,3)])
ensembleprob = de.EnsembleProblem(fast_prob, prob_func = prob_func, safetycopy=False)
sol = de.solve(ensembleprob,de.Tsit5(),de.EnsembleSerial(),trajectories=10000,saveat=0.01)
from diffeqpy import cuda
sol = de.solve(ensembleprob,cuda.GPUTsit5(),cuda.EnsembleGPUKernel(cuda.CUDABackend()),trajectories=10000,saveat=0.01)
```
Currently fails at the first ensemble solve:
```py
sol = de.solve(ensembleprob,de.Tsit5(),de.EnsembleSerial(),trajectories=10000,saveat=0.01)
```
```
>>> sol = de.solve(ensembleprob,de.Tsit5(),de.EnsembleSerial(),trajectories=10000,saveat=0.01)
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "C:\Users\accou\.julia\packages\PythonCall\qTEA1\src\jlwrap\any.jl", line 208, in __call__
return self._jl_callmethod($(pyjl_methodnum(pyjlany_call)), args, kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
juliacall.JuliaError: MethodError: no method matching init(::Py, ::OrdinaryDiffEq.Tsit5{typeof(OrdinaryDiffEq.trivial_limiter!), typeof(OrdinaryDiffEq.trivial_limiter!), Static.False}; saveat::Float64)
Closest candidates are:
init(!Matched::SciMLBase.OptimizationProblem, ::Any, !Matched::Any...; kwargs...)
@ SciMLBase C:\Users\accou\.julia\packages\SciMLBase\McEqc\src\solve.jl:146
init(!Matched::SciMLBase.PDEProblem, ::SciMLBase.AbstractDEAlgorithm, !Matched::Any...; kwargs...)
@ DiffEqBase C:\Users\accou\.julia\packages\DiffEqBase\xSmHR\src\solve.jl:1116
init(!Matched::SciMLBase.AbstractJumpProblem, ::Any...; kwargs...)
@ DiffEqBase C:\Users\accou\.julia\packages\DiffEqBase\xSmHR\src\solve.jl:499
...
Stacktrace:
[1] solve(::Py, ::Vararg{Any}; kwargs::Base.Pairs{Symbol, Float64, Tuple{Symbol}, NamedTuple{(:saveat,), Tuple{Float64}}})
@ CommonSolve C:\Users\accou\.julia\packages\CommonSolve\JfpfI\src\CommonSolve.jl:23
[2] batch_func(i::Int64, prob::SciMLBase.EnsembleProblem{SciMLBase.ODEProblem{Vector{Float64}, Tuple{Float64, Float64}, true, Vector{Float64}, SciMLBase.ODEFunction{true, SciMLBase.AutoSpecialize, ModelingToolkit.var"#k#545"{RuntimeGeneratedFunctions.RuntimeGeneratedFunction{(:ˍ₋arg1, :ˍ₋arg2, :t), ModelingToolkit.var"#_RGF_ModTag", ModelingToolkit.var"#_RGF_ModTag", (0x946926fe, 0xab7b8dc2, 0x707dbba2, 0x8b060826, 0xe63fdf5b), Nothing}, RuntimeGeneratedFunctions.RuntimeGeneratedFunction{(:ˍ₋out, :ˍ₋arg1, :ˍ₋arg2, :t), ModelingToolkit.var"#_RGF_ModTag", ModelingToolkit.var"#_RGF_ModTag", (0xaf9b47a5, 0x8d86fb18, 0x65c64d40, 0xf8480c5a, 0x8614cffa), Nothing}}, LinearAlgebra.UniformScaling{Bool}, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Vector{Symbol}, Symbol, Vector{Symbol}, ModelingToolkit.var"#637#generated_observed#555"{Bool, ModelingToolkit.ODESystem, Dict{Any, Any}, Vector{SymbolicUtils.BasicSymbolic{Real}}}, Nothing, ModelingToolkit.ODESystem}, Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}}, SciMLBase.StandardODEProblem}, Py, typeof(SciMLBase.DEFAULT_OUTPUT_FUNC), typeof(SciMLBase.DEFAULT_REDUCTION), Nothing}, alg::OrdinaryDiffEq.Tsit5{typeof(OrdinaryDiffEq.trivial_limiter!), typeof(OrdinaryDiffEq.trivial_limiter!), Static.False}; kwargs::Base.Pairs{Symbol, Float64, Tuple{Symbol}, NamedTuple{(:saveat,), Tuple{Float64}}})
@ SciMLBase C:\Users\accou\.julia\packages\SciMLBase\McEqc\src\ensemble\basic_ensemble_solve.jl:100
[3] (::SciMLBase.var"#604#605"{Base.Pairs{Symbol, Float64, Tuple{Symbol}, NamedTuple{(:saveat,), Tuple{Float64}}}, SciMLBase.EnsembleProblem{SciMLBase.ODEProblem{Vector{Float64}, Tuple{Float64, Float64}, true, Vector{Float64}, SciMLBase.ODEFunction{true, SciMLBase.AutoSpecialize, ModelingToolkit.var"#k#545"{RuntimeGeneratedFunctions.RuntimeGeneratedFunction{(:ˍ₋arg1, :ˍ₋arg2, :t), ModelingToolkit.var"#_RGF_ModTag", ModelingToolkit.var"#_RGF_ModTag", (0x946926fe, 0xab7b8dc2, 0x707dbba2, 0x8b060826, 0xe63fdf5b), Nothing}, RuntimeGeneratedFunctions.RuntimeGeneratedFunction{(:ˍ₋out, :ˍ₋arg1, :ˍ₋arg2, :t), ModelingToolkit.var"#_RGF_ModTag", ModelingToolkit.var"#_RGF_ModTag", (0xaf9b47a5, 0x8d86fb18, 0x65c64d40, 0xf8480c5a, 0x8614cffa), Nothing}}, LinearAlgebra.UniformScaling{Bool}, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Vector{Symbol}, Symbol, Vector{Symbol}, ModelingToolkit.var"#637#generated_observed#555"{Bool, ModelingToolkit.ODESystem, Dict{Any, Any}, Vector{SymbolicUtils.BasicSymbolic{Real}}}, Nothing, ModelingToolkit.ODESystem}, Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}}, SciMLBase.StandardODEProblem}, Py, typeof(SciMLBase.DEFAULT_OUTPUT_FUNC), typeof(SciMLBase.DEFAULT_REDUCTION), Nothing}, OrdinaryDiffEq.Tsit5{typeof(OrdinaryDiffEq.trivial_limiter!), typeof(OrdinaryDiffEq.trivial_limiter!), Static.False}})(i::Int64)
@ SciMLBase C:\Users\accou\.julia\packages\SciMLBase\McEqc\src\ensemble\basic_ensemble_solve.jl:154
[4] responsible_map
@ C:\Users\accou\.julia\packages\SciMLBase\McEqc\src\ensemble\basic_ensemble_solve.jl:147 [inlined]
[5] #solve_batch#603
@ C:\Users\accou\.julia\packages\SciMLBase\McEqc\src\ensemble\basic_ensemble_solve.jl:153 [inlined]
[6] solve_batch
@ C:\Users\accou\.julia\packages\SciMLBase\McEqc\src\ensemble\basic_ensemble_solve.jl:152 [inlined]
[7] macro expansion
@ .\timing.jl:393 [inlined]
[8] __solve(prob::SciMLBase.EnsembleProblem{SciMLBase.ODEProblem{Vector{Float64}, Tuple{Float64, Float64}, true, Vector{Float64}, SciMLBase.ODEFunction{true, SciMLBase.AutoSpecialize, ModelingToolkit.var"#k#545"{RuntimeGeneratedFunctions.RuntimeGeneratedFunction{(:ˍ₋arg1, :ˍ₋arg2, :t), ModelingToolkit.var"#_RGF_ModTag", ModelingToolkit.var"#_RGF_ModTag", (0x946926fe, 0xab7b8dc2, 0x707dbba2, 0x8b060826, 0xe63fdf5b), Nothing}, RuntimeGeneratedFunctions.RuntimeGeneratedFunction{(:ˍ₋out, :ˍ₋arg1, :ˍ₋arg2, :t), ModelingToolkit.var"#_RGF_ModTag", ModelingToolkit.var"#_RGF_ModTag", (0xaf9b47a5, 0x8d86fb18, 0x65c64d40, 0xf8480c5a, 0x8614cffa), Nothing}}, LinearAlgebra.UniformScaling{Bool}, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Vector{Symbol}, Symbol, Vector{Symbol}, ModelingToolkit.var"#637#generated_observed#555"{Bool, ModelingToolkit.ODESystem, Dict{Any, Any}, Vector{SymbolicUtils.BasicSymbolic{Real}}}, Nothing, ModelingToolkit.ODESystem}, Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}}, SciMLBase.StandardODEProblem}, Py, typeof(SciMLBase.DEFAULT_OUTPUT_FUNC), typeof(SciMLBase.DEFAULT_REDUCTION), Nothing}, alg::OrdinaryDiffEq.Tsit5{typeof(OrdinaryDiffEq.trivial_limiter!), typeof(OrdinaryDiffEq.trivial_limiter!), Static.False}, ensemblealg::SciMLBase.EnsembleSerial; trajectories::Int64, batch_size::Int64, pmap_batch_size::Int64, kwargs::Base.Pairs{Symbol, Float64, Tuple{Symbol}, NamedTuple{(:saveat,), Tuple{Float64}}})
@ SciMLBase C:\Users\accou\.julia\packages\SciMLBase\McEqc\src\ensemble\basic_ensemble_solve.jl:64
[9] solve(::SciMLBase.EnsembleProblem{SciMLBase.ODEProblem{Vector{Float64}, Tuple{Float64, Float64}, true, Vector{Float64}, SciMLBase.ODEFunction{true, SciMLBase.AutoSpecialize, ModelingToolkit.var"#k#545"{RuntimeGeneratedFunctions.RuntimeGeneratedFunction{(:ˍ₋arg1, :ˍ₋arg2, :t), ModelingToolkit.var"#_RGF_ModTag", ModelingToolkit.var"#_RGF_ModTag", (0x946926fe, 0xab7b8dc2, 0x707dbba2, 0x8b060826, 0xe63fdf5b), Nothing}, RuntimeGeneratedFunctions.RuntimeGeneratedFunction{(:ˍ₋out, :ˍ₋arg1, :ˍ₋arg2, :t), ModelingToolkit.var"#_RGF_ModTag", ModelingToolkit.var"#_RGF_ModTag", (0xaf9b47a5, 0x8d86fb18, 0x65c64d40, 0xf8480c5a, 0x8614cffa), Nothing}}, LinearAlgebra.UniformScaling{Bool}, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Vector{Symbol}, Symbol, Vector{Symbol}, ModelingToolkit.var"#637#generated_observed#555"{Bool, ModelingToolkit.ODESystem, Dict{Any, Any}, Vector{SymbolicUtils.BasicSymbolic{Real}}}, Nothing, ModelingToolkit.ODESystem}, Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}}, SciMLBase.StandardODEProblem}, Py, typeof(SciMLBase.DEFAULT_OUTPUT_FUNC), typeof(SciMLBase.DEFAULT_REDUCTION), Nothing}, ::OrdinaryDiffEq.Tsit5{typeof(OrdinaryDiffEq.trivial_limiter!), typeof(OrdinaryDiffEq.trivial_limiter!), Static.False}, ::Vararg{Any}; kwargs::Base.Pairs{Symbol, Real, Tuple{Symbol, Symbol}, NamedTuple{(:saveat, :trajectories), Tuple{Float64, Int64}}})
@ DiffEqBase C:\Users\accou\.julia\packages\DiffEqBase\xSmHR\src\solve.jl:1053
[10] pyjlany_call(self::typeof(CommonSolve.solve), args_::Py, kwargs_::Py)
@ PythonCall C:\Users\accou\.julia\packages\PythonCall\qTEA1\src\jlwrap\any.jl:34
[11] _pyjl_callmethod(f::Any, self_::Ptr{PythonCall.C.PyObject}, args_::Ptr{PythonCall.C.PyObject}, nargs::Int64)
@ PythonCall C:\Users\accou\.julia\packages\PythonCall\qTEA1\src\jlwrap\base.jl:69
[12] _pyjl_callmethod(o::Ptr{PythonCall.C.PyObject}, args::Ptr{PythonCall.C.PyObject})
@ PythonCall.C C:\Users\accou\.julia\packages\PythonCall\qTEA1\src\cpython\jlwrap.jl:47
```
|
@LilithHafner do you know why that conversion of Py -> EnsembleProblem isn't done? |
|
I think the issue is that the keyword argument ensembleprob = de.EnsembleProblem(fast_prob, prob_func = prob_func, safetycopy=False)`is not converted. If I change that line to ensembleprob = de.EnsembleProblem(fast_prob, safetycopy=False)`Then the non-cuda solve works (or maybe it hangs, idk. It works when I lower trajectories to 1000). |
from diffeqpy import de
def f(u,p,t):
x, y, z = u
sigma, rho, beta = p
return [sigma * (y - x), x * (rho - z) - y, x * y - beta * z]
u0 = [1.0,0.0,0.0]
tspan = (0., 100.)
p = [10.0,28.0,8/3]
prob = de.ODEProblem(f, u0, tspan, p)
fast_prob = de.jit(prob)
sol = de.solve(fast_prob,saveat=0.01)
import random
def prob_func(prob,i,rep):
de.remake(prob,u0=[random.uniform(0, 1)*u0[i] for i in range(0,3)],
p=[random.uniform(0, 1)*p[i] for i in range(0,3)])
ensembleprob = de.EnsembleProblem(fast_prob, safetycopy=False)
sol = de.solve(ensembleprob,cuda.GPUTsit5(),cuda.EnsembleGPUKernel(CUDABackend()),trajectories=10000,saveat=0.01)@utkarsh530 can DiffEqGPU not auto-convert to SA? |
from diffeqpy import de
def f(u,p,t):
x, y, z = u
sigma, rho, beta = p
return [sigma * (y - x), x * (rho - z) - y, x * y - beta * z]
u0 = [1.0,0.0,0.0]
tspan = (0., 100.)
p = [10.0,28.0,8/3]
prob = de.ODEProblem(f, u0, tspan, p)
fast_prob = de.jit(prob)
sol = de.solve(fast_prob,saveat=0.01)
import random
def prob_func(prob,i,rep):
de.remake(prob,u0=[random.uniform(0, 1)*u0[i] for i in range(0,3)],
p=[random.uniform(0, 1)*p[i] for i in range(0,3)])
ensembleprob = de.EnsembleProblem(fast_prob, safetycopy=False)
sol = de.solve(ensembleprob,de.Tsit5(),de.EnsembleSerial(),trajectories=10000,saveat=0.01)
from diffeqpy import cuda
sol = de.solve(ensembleprob,cuda.GPUTsit5(),cuda.EnsembleGPUKernel(CUDABackend()),trajectories=10000,saveat=0.01)This works with SciML/DiffEqGPU.jl#308 |
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.
MWE:
Currently fails at the first ensemble solve: