Add symbol based indexing for interpolated solutions #66

sharanry · 2021-06-18T12:21:11Z

This PR requires SciML/RecursiveArrayTools.jl#143 and does the following:

Fills extra fields of DiffEqArray offered by Enable symbol based indexing of interpolated solutions by adding extra fields to DiffEqArray type RecursiveArrayTools.jl#143 after interpolating to enable symbol based indexing of eliminated and reduced states.
Add tests for the same.

devmotion · 2021-06-18T12:31:13Z

I always thought at some point one could unify the standard interpolations in SciMLBase and OrdinaryDiffEq etc. Maybe the additional arguments might make this more difficult? I guess it doesn't matter though since there is no push in this direction currently AFAIK.

ChrisRackauckas · 2021-06-18T12:32:50Z

The lazy interpolations of the Verner methods make this a bit hard. Keeping the SciMLBase ones light makes sense IMO since the OrdinaryDiffEq.jl ones do a lot of very special stuff that nothing else needs.

devmotion · 2021-06-18T12:35:13Z

Yes, I had only the Hermite and linear interpolations in mind.

ChrisRackauckas · 2021-06-18T12:36:42Z

That would make dense=false a type unstable change because it would change the underlying type for the interpolation, instead of just checking for the bool and doing the linear behavior. Constant prop may handle it, but 🤷

test/downstream/observables.jl

src/solutions/solution_interface.jl

codecov · 2021-06-21T11:14:52Z

Codecov Report

Merging #66 (b0341e0) into master (2c8d34d) will increase coverage by 2.35%.
The diff coverage is 80.00%.

@@            Coverage Diff             @@
##           master      #66      +/-   ##
==========================================
+ Coverage   11.00%   13.36%   +2.35%     
==========================================
  Files          39       39              
  Lines        2852     2888      +36     
==========================================
+ Hits          314      386      +72     
+ Misses       2538     2502      -36

Impacted Files	Coverage Δ
src/interpolation.jl	`0.00% <0.00%> (ø)`
src/solutions/ode_solutions.jl	`40.00% <82.14%> (+23.01%)`	⬆️
src/solutions/solution_interface.jl	`10.63% <100.00%> (+10.63%)`	⬆️
src/scimlfunctions.jl	`7.50% <0.00%> (+0.73%)`	⬆️
src/problems/ode_problems.jl	`12.30% <0.00%> (+1.37%)`	⬆️
src/operators/operators.jl	`5.26% <0.00%> (+5.26%)`	⬆️
src/function_wrappers.jl	`25.00% <0.00%> (+25.00%)`	⬆️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 2c8d34d...b0341e0. Read the comment docs.

src/utils.jl

ChrisRackauckas · 2021-06-28T22:45:09Z

Seems like most tests are failing?

…ry/SciMLBase.jl into sy/interp_elim_var_access

test/downstream/observables.jl

src/solutions/ode_solutions.jl

devmotion · 2021-06-29T06:08:07Z

src/solutions/ode_solutions.jl

+      sol.interp(t,idxs,deriv,sol.prob.p,continuity)
+  elseif issymbollike(idxs)
+    if t isa Real
+      interp_sol = augment(sol.interp([t],nothing,deriv,sol.prob.p,continuity), sol)


Can't we use

Suggested change

interp_sol = augment(sol.interp([t],nothing,deriv,sol.prob.p,continuity), sol)

interp_sol = augment(sol.interp(t,nothing,deriv,sol.prob.p,continuity), sol)

We will have to add an extra constructor to DiffEqArray in RecursiveArrayTools.jl to support this. https://github.com/SciML/RecursiveArrayTools.jl/blob/master/src/vector_of_array.jl#L42-L46 doesn't cover the case where A.u is Vector{Real}.

~~Will be added in SciML/RecursiveArrayTools.jl#148~~

The current API of interpolation outputs Vector{<:Real} when the input t isa Real. DiffEqArray expects a vector of times t (Vector{<:Real}) and vector of corresponding states u (Vector{Vector{<:Real}}).

These two factors together make it difficult to accommodate what you are suggesting without making substantial changes to interpolation and DiffEqArray indexing.

devmotion · 2021-06-29T06:09:48Z

src/solutions/ode_solutions.jl

+    end
+  else
+    if t isa Real
+      interp_sol = augment(sol.interp([t],nothing,deriv,sol.prob.p,continuity), sol)


Same here, can't we use

Suggested change

interp_sol = augment(sol.interp([t],nothing,deriv,sol.prob.p,continuity), sol)

interp_sol = augment(sol.interp(t,nothing,deriv,sol.prob.p,continuity), sol)

sharanry · 2021-06-29T10:17:39Z

src/solutions/ode_solutions.jl

+function (sol::ODESolution)(t::AbstractVector{<:Real},::Type{Val{0}},idxs::AbstractVector,continuity)
+  if any(.!issymbollike.(idxs))
+    error("Incorrect specification of `idxs`")
+  end
+  interp_sol = augment(sol.interp(t,nothing,Val{0},sol.prob.p,continuity), sol)
+  u = [[interp_sol[idx][i] for idx in idxs] for i in 1:length(t)]
+  DiffEqArray(u, t)
+end
+
+for T in 1:3
+  function (sol::ODESolution)(t,::Type{Val{T}},idxs,continuity)
+    error("Higher-order interpolation is not implemented.")
+  end
+end


Is this approach okay to handle higher-order interpolation? @devmotion

Just don't define these functions?

I guess the idea was to have a user-readable error. @YingboMa Any thoughts?

How about

function (sol::ODESolution)(t::AbstractVector{<:Real},::Type{Val{N}},idxs::AbstractVector,continuity) where N N == 0 || error("Higher-order interpolation is not implemented.")

@YingboMa I think erroring out at sol(t, idxs=[...]) stage for higher-order interpolation is wrong (as @devmotion already mentioned). I just realized that it has been already been implemented in many scenarios.

https://github.com/SciML/SciMLBase.jl/blob/master/src/interpolation.jl#L274-L314
https://github.com/SciML/SciMLBase.jl/blob/master/src/interpolation.jl#L335-L369
https://github.com/SciML/SciMLBase.jl/blob/master/src/interpolation.jl#L387
https://github.com/SciML/SciMLBase.jl/blob/master/src/interpolation.jl#L414
https://github.com/SciML/SciMLBase.jl/blob/master/src/interpolation.jl#L439
https://github.com/SciML/SciMLBase.jl/blob/master/src/interpolation.jl#L465

I have set a default at interpolant stage b97346c .

…ts pass

sharanry · 2021-07-01T11:53:49Z

Weird. The downstream tests seem to pass locally.

ChrisRackauckas · 2021-07-01T12:03:15Z

https://github.com/SciML/SciMLBase.jl/pull/66/checks?check_run_id=2961581350#step:6:881 that one is slightly stochastic (we need to make it take more samples), so that's fine and would almost certainly pass if re-tested.

https://github.com/SciML/SciMLBase.jl/pull/66/checks?check_run_id=2961581076 is worrisome: do your symbolic versions match?

sharanry · 2021-07-01T12:29:42Z

@ChrisRackauckas I updated it to v1.2.1 and reconfirmed SciMLBase's downstream tests. Could you restart just that particular workflow?

sharanry · 2021-07-01T12:38:41Z

Okay, it errored again. I will take a closer look.

sharanry · 2021-07-01T15:12:49Z

@YingboMa @ChrisRackauckas It looks like issymbollike() is erroneously returning false on Symbolics.Num type object only in a sandboxed downstream test environment. It correctly returns true when I normally import ModellingToolkit/RecursiveArrayTools. Any idea what could possibly lead to such behavior?

ChrisRackauckas · 2021-07-01T15:20:24Z

We need to just get rid of that hack. @YingboMa @shashi can we just add istree to https://github.com/JuliaSymbolics/TermInterface.jl/blob/master/src/TermInterface.jl and depend on it here for this?

shashi · 2021-07-01T18:39:37Z

We are renaming istree i today isterm so yes it’s already in TermInterface. We just need to use it in SU :p

ChrisRackauckas · 2021-07-01T18:41:25Z

Okay yes, so try and prioritize that and we'll change this from the hacky thing to some traits and proper dispatches.

sharanry · 2021-07-05T17:43:03Z

@ChrisRackauckas The tests for this PR will now probably pass once SciML/RecursiveArrayTools.jl#152 is merged and released.

ChrisRackauckas · 2021-07-05T20:35:11Z

Shit that was a mistake. Reverting.

This should be a nice improvement overall to the health of the debugging experience. For example, the code from this post (https://discourse.julialang.org/t/optimizationmoi-ipopt-violating-inequality-constraint/92608) led to a question that took a bit to understand. But now when you run ```julia import Optimization import OptimizationMOI, Ipopt const AV{T} = AbstractVector{T} function model_constraints!(out::AV{<:Real}, u::AV{<:Real}, data) # Model parameters dt, a, b = u out[1] = a - 1/dt # Must be < 0 @info "Must be NEGATIVE: $(out[1])" end function model_variance(u::AV{T}, data::AV{<:Real}) where T<:Real # Model parameters dt, a, b = u # Compute variance variance = zeros(T, length(data)) variance[1] = one(T) for t in 1:(length(data) - 1) variance[t+1] = (1 - dt * a) * variance[t] + dt * data[t]^2 + dt * b end variance end function model_loss(u::AV{T}, data::AV{<:Real})::T where T<:Real variance = model_variance(u, data) loglik::T = zero(T) for (r, var) in zip(data, variance) loglik += -(log(2π) + log(var) + r^2 / var) / 2 end -loglik / length(data) end function model_fit(u0::AV{T}, data::AV{<:Real}) where T<:Real func = Optimization.OptimizationFunction( model_loss, Optimization.AutoForwardDiff(), cons=model_constraints! ) prob = Optimization.OptimizationProblem( func, u0, data, # 0 < dt < 1 && 1 < a < Inf && 0 < b < Inf lb=T[0.0, 1.0, 0.0], ub=T[1.0, Inf, Inf], # ^dt ^a ^b ^dt ^a ^b <= model parameters lcons=T[-Inf], ucons=T[0.0] # a - 1/dt < 0 ) sol = Optimization.solve(prob, Ipopt.Optimizer()) sol.u end let data = [ 2.1217711584057386, -0.28350145551002465, 2.3593492969513004, 0.192856733601849, 0.4566485836385113, 1.332717934013979, -1.286716619379847, 0.9868669960185211, 2.2358674776395224, -2.7933975791568098, 1.2555871497124622, 1.276879759908467, -0.8392016987911409, -1.1580875182201849, 0.33201646080578456, -0.17212553408696898, 1.1275285626369556, 0.23041139849229036, 1.648423577528424, 2.384823597473343, -0.4005518932539747, -1.117737311211693, -0.9490152960583265, -1.1454539355078672, 1.4158585811404159, -0.18926972177257692, -0.2867541528181491, -1.2077459688543788, -0.6397173049620141, 0.66147783407023, 0.049805188778543466, 0.902540117368457, -0.7018417933284938, 0.47342354473843684, 1.2620345361591596, -1.1483844812087018, -0.06487285080802752, 0.39020117013487715, -0.38454491504165356, 1.5125786171885645, -0.6751768274451174, 0.490916740658628, 0.012872300530924086, 0.46532447715746716, 0.34734421531357157, 0.3830452463549559, -0.8730874028738718, 0.4333151627834603, -0.40396180775692375, 2.0794821773418497, -0.5392735774960918, 0.6519326323752113, -1.4844713145398716, 0.3688828625691108, 1.010912990717231, 0.5018274939956874, 0.36656889279915833, -0.11403975693239479, -0.6460314660359935, -0.41997005020823147, 0.9652752515820495, -0.37375868692702047, -0.5780729659197872, 2.642742798278919, 0.5076984117208074, -0.4906395089461916, -1.804352047187329, -0.8596663844837792, -0.7510485548262176, -0.07922589350581195, 1.7201304839487317, 0.9024493222130577, -1.8216089665357902, 1.3929269238775426, -0.08410752079538407, 0.6423068180438288, 0.6615201016351212, 0.18546977816594887, -0.717521690742993, -1.0224309324751113, 1.7748350222721971, 0.1929546575877559, -0.1581871639724676, 0.20198379311238596, -0.6919373947349301, -0.9253274269423383, 0.549366272989534, -1.9302106783541606, 0.7197247279281573, -1.220334158468621, -0.9187468058921053, -2.1452607604834184, -2.1558650694862687, -0.9387913392336701, -0.676637835687265, -0.16621998352492198, 0.5637177022958897, -0.5258315560278541, 0.8413359958184765, -0.9096866525337141 ] # u0 = [0 < dt < 1, 1 < a < 1/dt, 0 < b < Inf] u0 = [0.3, 2.3333333333333335, 0.33333333333333337] @Assert 0 < u0[1] < 1 @Assert 1 < u0[2] < 1 / u0[1] @Assert 0 < u0[3] < Inf @info "Optimizing..." u0 model_fit(u0, data) end ``` you get: ```julia DomainError detected in the user `f` function. This occurs when the domain of a function is violated. For example, `log(-1.0)` is undefined because `log` of a real number is defined to only output real numbers, but `log` of a negative number is complex valued and therefore Julia throws a DomainError by default. Cases to be aware of include: * `log(x)`, `sqrt(x)`, `cbrt(x)`, etc. where `x<0` * `x^y` for `x<0` floating point `y` (example: `(-1.0)^(1/2) == im`) Within the context of SciML, this error can occur within the solver process even if the domain constraint would not be violated in the solution due to adaptivity. For example, an ODE solver or optimization routine may check a step at `new_u` which violates the domain constraint, and if violated reject the step and use a smaller `dt`. However, the throwing of this error will have halted the solving process. Thus the recommended fix is to replace this function with the equivalent ones from NaNMath.jl (https://github.com/JuliaMath/NaNMath.jl) which returns a NaN instead of an error. The solver will then effectively use the NaN within the error control routines to reject the out of bounds step. Additionally, one could perform a domain transformation on the variables so that such an issue does not occur in the definition of `f`. For more information, check out the following FAQ page: https://docs.sciml.ai/Optimization/stable/API/FAQ/#The-Solver-Seems-to-Violate-Constraints-During-the-Optimization,-Causing-DomainErrors,-What-Can-I-Do-About-That? Note that detailed debugging information adds a small amount of overhead to SciML solves which can be disabled with the keyword argument `debug = NoDebug()`. The detailed original error message information from Julia reproduced below: ERROR: DomainError with -2.4941978436429695: log will only return a complex result if called with a complex argument. Try log(Complex(x)). Stacktrace: [1] (::SciMLBase.VerboseDebugFunction{typeof(SciMLBase.__solve)})(::SciMLBase.OptimizationProblem{true, SciMLBase.OptimizationFunction{true, Optimization.AutoForwardDiff{nothing}, typeof(model_loss), Nothing, Nothing, Nothing, typeof(model_constraints!), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}, Vector{Float64}, Vector{Float64}, Vector{Float64}, Vector{Float64}, Nothing, Vector{Float64}, Vector{Float64}, Nothing, Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}}}, ::Vararg{Any}) @ SciMLBase C:\Users\accou\.julia\dev\SciMLBase\src\debug.jl:66 [2] #solve#572 @ c:\Users\accou\.julia\dev\SciMLBase\src\solve.jl:87 [inlined] [3] solve @ c:\Users\accou\.julia\dev\SciMLBase\src\solve.jl:80 [inlined] [4] model_fit(u0::Vector{Float64}, data::Vector{Float64}) @ Main c:\Users\accou\OneDrive\Computer\Desktop\test.jl:77 [5] top-level scope @ c:\Users\accou\OneDrive\Computer\Desktop\test.jl:88 caused by: DomainError with -2.4941978436429695: log will only return a complex result if called with a complex argument. Try log(Complex(x)). Stacktrace: [1] throw_complex_domainerror(f::Symbol, x::Float64) @ Base.Math .\math.jl:33 [2] _log(x::Float64, base::Val{:ℯ}, func::Symbol) @ Base.Math .\special\log.jl:301 [3] log @ .\special\log.jl:267 [inlined] [4] model_loss(u::Vector{Float64}, data::Vector{Float64}) @ Main c:\Users\accou\OneDrive\Computer\Desktop\test.jl:60 [5] OptimizationFunction @ C:\Users\accou\.julia\dev\SciMLBase\src\scimlfunctions.jl:3580 [inlined] [6] eval_objective(moiproblem::OptimizationMOI.MOIOptimizationProblem{Float64, SciMLBase.OptimizationFunction{true, Optimization.AutoForwardDiff{nothing}, typeof(model_loss), Optimization.var"#57#74"{ForwardDiff.GradientConfig{ForwardDiff.Tag{Optimization.var"#56#73"{SciMLBase.OptimizationFunction{true, Optimization.AutoForwardDiff{nothing}, typeof(model_loss), Nothing, Nothing, Nothing, typeof(model_constraints!), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}, Vector{Float64}}, Float64}, Float64, 3, Vector{ForwardDiff.Dual{ForwardDiff.Tag{Optimization.var"#56#73"{SciMLBase.OptimizationFunction{true, Optimization.AutoForwardDiff{nothing}, typeof(model_loss), Nothing, Nothing, Nothing, typeof(model_constraints!), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}, Vector{Float64}}, Float64}, Float64, 3}}}, Optimization.var"#56#73"{SciMLBase.OptimizationFunction{true, Optimization.AutoForwardDiff{nothing}, typeof(model_loss), Nothing, Nothing, Nothing, typeof(model_constraints!), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}, Vector{Float64}}}, Optimization.var"#60#77"{ForwardDiff.HessianConfig{ForwardDiff.Tag{Optimization.var"#56#73"{SciMLBase.OptimizationFunction{true, Optimization.AutoForwardDiff{nothing}, typeof(model_loss), Nothing, Nothing, Nothing, typeof(model_constraints!), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}, Vector{Float64}}, Float64}, Float64, 3, Vector{ForwardDiff.Dual{ForwardDiff.Tag{Optimization.var"#56#73"{SciMLBase.OptimizationFunction{true, Optimization.AutoForwardDiff{nothing}, typeof(model_loss), Nothing, Nothing, Nothing, typeof(model_constraints!), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}, Vector{Float64}}, Float64}, ForwardDiff.Dual{ForwardDiff.Tag{Optimization.var"#56#73"{SciMLBase.OptimizationFunction{true, Optimization.AutoForwardDiff{nothing}, typeof(model_loss), Nothing, Nothing, Nothing, typeof(model_constraints!), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}, Vector{Float64}}, Float64}, Float64, 3}, 3}}, Vector{ForwardDiff.Dual{ForwardDiff.Tag{Optimization.var"#56#73"{SciMLBase.OptimizationFunction{true, Optimization.AutoForwardDiff{nothing}, typeof(model_loss), Nothing, Nothing, Nothing, typeof(model_constraints!), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}, Vector{Float64}}, Float64}, Float64, 3}}}, Optimization.var"#56#73"{SciMLBase.OptimizationFunction{true, Optimization.AutoForwardDiff{nothing}, typeof(model_loss), Nothing, Nothing, Nothing, typeof(model_constraints!), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}, Vector{Float64}}}, Optimization.var"#63#80", Optimization.var"#64#81"{SciMLBase.OptimizationFunction{true, Optimization.AutoForwardDiff{nothing}, typeof(model_loss), Nothing, Nothing, Nothing, typeof(model_constraints!), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}, Vector{Float64}}, Optimization.var"#66#83"{ForwardDiff.JacobianConfig{ForwardDiff.Tag{Optimization.var"#65#82"{Int64}, Float64}, Float64, 3, Vector{ForwardDiff.Dual{ForwardDiff.Tag{Optimization.var"#65#82"{Int64}, Float64}, Float64, 3}}}}, Optimization.var"#71#88"{Int64, Vector{ForwardDiff.HessianConfig{ForwardDiff.Tag{Optimization.var"#69#86"{Int64}, Float64}, Float64, 3, Vector{ForwardDiff.Dual{ForwardDiff.Tag{Optimization.var"#69#86"{Int64}, Float64}, ForwardDiff.Dual{ForwardDiff.Tag{Optimization.var"#69#86"{Int64}, Float64}, Float64, 3}, 3}}, Vector{ForwardDiff.Dual{ForwardDiff.Tag{Optimization.var"#69#86"{Int64}, Float64}, Float64, 3}}}}, Vector{Optimization.var"#69#86"{Int64}}}, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}, Vector{Float64}, Vector{Float64}, Matrix{Float64}, Matrix{Float64}, Matrix{Float64}}, x::Vector{Float64}) @ OptimizationMOI C:\Users\accou\.julia\packages\OptimizationMOI\cHl7S\src\OptimizationMOI.jl:82 [7] eval_objective(model::Ipopt.Optimizer, x::Vector{Float64}) @ Ipopt C:\Users\accou\.julia\packages\Ipopt\rQctM\src\MOI_wrapper.jl:514 [8] (::Ipopt.var"#eval_f_cb#1"{Ipopt.Optimizer})(x::Vector{Float64}) @ Ipopt C:\Users\accou\.julia\packages\Ipopt\rQctM\src\MOI_wrapper.jl:597 [9] _Eval_F_CB(n::Int32, x_ptr::Ptr{Float64}, x_new::Int32, obj_value::Ptr{Float64}, user_data::Ptr{Nothing}) @ Ipopt C:\Users\accou\.julia\packages\Ipopt\rQctM\src\C_wrapper.jl:38 [10] IpoptSolve(prob::Ipopt.IpoptProblem) @ Ipopt C:\Users\accou\.julia\packages\Ipopt\rQctM\src\C_wrapper.jl:442 [11] optimize!(model::Ipopt.Optimizer) @ Ipopt C:\Users\accou\.julia\packages\Ipopt\rQctM\src\MOI_wrapper.jl:727 [12] __solve(prob::SciMLBase.OptimizationProblem{true, SciMLBase.OptimizationFunction{true, Optimization.AutoForwardDiff{nothing}, typeof(model_loss), Nothing, Nothing, Nothing, typeof(model_constraints!), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}, Vector{Float64}, Vector{Float64}, Vector{Float64}, Vector{Float64}, Nothing, Vector{Float64}, Vector{Float64}, Nothing, Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}}}, opt::Ipopt.Optimizer; maxiters::Nothing, maxtime::Nothing, abstol::Nothing, reltol::Nothing, kwargs::Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}}) @ OptimizationMOI C:\Users\accou\.julia\packages\OptimizationMOI\cHl7S\src\OptimizationMOI.jl:381 [13] __solve(prob::SciMLBase.OptimizationProblem{true, SciMLBase.OptimizationFunction{true, Optimization.AutoForwardDiff{nothing}, typeof(model_loss), Nothing, Nothing, Nothing, typeof(model_constraints!), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}, Vector{Float64}, Vector{Float64}, Vector{Float64}, Vector{Float64}, Nothing, Vector{Float64}, Vector{Float64}, Nothing, Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}}}, opt::Ipopt.Optimizer) @ OptimizationMOI C:\Users\accou\.julia\packages\OptimizationMOI\cHl7S\src\OptimizationMOI.jl:327 [14] (::SciMLBase.VerboseDebugFunction{typeof(SciMLBase.__solve)})(::SciMLBase.OptimizationProblem{true, SciMLBase.OptimizationFunction{true, Optimization.AutoForwardDiff{nothing}, typeof(model_loss), Nothing, Nothing, Nothing, typeof(model_constraints!), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, typeof(SciMLBase.DEFAULT_OBSERVED_NO_TIME), Nothing, Nothing, Nothing, Nothing, Nothing, Nothing, Nothing}, Vector{Float64}, Vector{Float64}, Vector{Float64}, Vector{Float64}, Nothing, Vector{Float64}, Vector{Float64}, Nothing, Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}}}, ::Vararg{Any}) @ SciMLBase C:\Users\accou\.julia\dev\SciMLBase\src\debug.jl:59 [15] #solve#572 @ c:\Users\accou\.julia\dev\SciMLBase\src\solve.jl:87 [inlined] [16] solve @ c:\Users\accou\.julia\dev\SciMLBase\src\solve.jl:80 [inlined] [17] model_fit(u0::Vector{Float64}, data::Vector{Float64}) @ Main c:\Users\accou\OneDrive\Computer\Desktop\test.jl:77 [18] top-level scope @ c:\Users\accou\OneDrive\Computer\Desktop\test.jl:88 ```

Add symbol based indexing for interpolated solutions

31930e3

sharanry mentioned this pull request Jun 18, 2021

Enable symbol based indexing of interpolated solutions by adding extra fields to DiffEqArray type SciML/RecursiveArrayTools.jl#143

Merged

5 tasks

sharanry added 2 commits June 21, 2021 13:57

Fix tests and refactor code

7cb0c98

Fix typos

dab1017

sharanry commented Jun 21, 2021

View reviewed changes

test/downstream/observables.jl Outdated Show resolved Hide resolved

ChrisRackauckas reviewed Jun 21, 2021

View reviewed changes

src/solutions/solution_interface.jl Outdated Show resolved Hide resolved

Revert back to old example for downstream tests

a43dfb7

devmotion reviewed Jun 22, 2021

View reviewed changes

src/utils.jl Outdated Show resolved Hide resolved

sharanry added 2 commits June 25, 2021 15:59

Add back issymbollike and parameterless_type`

0a2e52e

Merge branch 'master' into sy/interp_elim_var_access

3984dd6

sharanry requested a review from ChrisRackauckas June 28, 2021 06:23

sharanry added 4 commits June 29, 2021 10:36

Enable passing symbols as idxs for interpolated solutions

c434789

Rollback dae and rode augmentation

6e7b7f4

Merge branch 'sy/interp_elim_var_access' of https://github.com/sharan…

263776f

…ry/SciMLBase.jl into sy/interp_elim_var_access

Remove unnecessary import

a62f085

sharanry commented Jun 29, 2021

View reviewed changes

test/downstream/observables.jl Outdated Show resolved Hide resolved

Fix test

c522415

devmotion reviewed Jun 29, 2021

View reviewed changes

Use multiple dispatch

9f44c49

sharanry commented Jun 29, 2021

View reviewed changes

sharanry added 2 commits June 29, 2021 19:39

Cover more corner cases and make OrdinaryDiffEq discrete problems tes…

92d6918

…ts pass

Clean code

0a9eb89

sharanry mentioned this pull request Jun 29, 2021

Add special cases for DiffEqArray constructor SciML/RecursiveArrayTools.jl#148

Merged

Fix bug

8f81510

sharanry added 3 commits July 1, 2021 11:14

Change higher-order interpolation error handling

958c9b4

Add indexing tests to runtests

807c453

Minor update

3e599bb

sharanry requested review from devmotion and YingboMa July 1, 2021 06:05

sharanry added 5 commits July 1, 2021 11:42

Make higher order interpolation error handling more general

142917e

stylistic changes

4ace146

add more error tests

3617e0e

Rollback erroring out at higher order derivatives

9b23e84

interpolant defaults to error

b97346c

sharanry force-pushed the sy/interp_elim_var_access branch from 5aff2b4 to b97346c Compare July 1, 2021 14:59

Fix issymbollike

b0341e0

sharanry mentioned this pull request Jul 5, 2021

Fix issymbollike SciML/RecursiveArrayTools.jl#152

Merged

ChrisRackauckas closed this Jul 5, 2021

ChrisRackauckas reopened this Jul 5, 2021

ChrisRackauckas merged commit 0ddde22 into SciML:master Jul 5, 2021

ChrisRackauckas mentioned this pull request Jul 5, 2021

Revert "Add symbol based indexing for interpolated solutions" #77

Merged

sharanry mentioned this pull request Jul 7, 2021

Fix issymbollike special case not covered by downstream tests #80

Merged

	interp_sol = augment(sol.interp([t],nothing,deriv,sol.prob.p,continuity), sol)
	interp_sol = augment(sol.interp(t,nothing,deriv,sol.prob.p,continuity), sol)

Uh oh!

Add symbol based indexing for interpolated solutions #66

Add symbol based indexing for interpolated solutions #66

Uh oh!

Conversation

sharanry commented Jun 18, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

devmotion commented Jun 18, 2021

Uh oh!

ChrisRackauckas commented Jun 18, 2021

Uh oh!

devmotion commented Jun 18, 2021

Uh oh!

ChrisRackauckas commented Jun 18, 2021

Uh oh!

Uh oh!

Uh oh!

codecov bot commented Jun 21, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Uh oh!

ChrisRackauckas commented Jun 28, 2021

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

sharanry Jun 29, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

sharanry Jul 1, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

sharanry commented Jul 1, 2021

Uh oh!

ChrisRackauckas commented Jul 1, 2021

Uh oh!

sharanry commented Jul 1, 2021

Uh oh!

sharanry commented Jul 1, 2021

Uh oh!

sharanry commented Jul 1, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ChrisRackauckas commented Jul 1, 2021

Uh oh!

shashi commented Jul 1, 2021

Uh oh!

ChrisRackauckas commented Jul 1, 2021

Uh oh!

sharanry commented Jul 5, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ChrisRackauckas commented Jul 5, 2021

Uh oh!

Uh oh!

sharanry commented Jun 18, 2021 •

edited

Loading

codecov bot commented Jun 21, 2021 •

edited

Loading

sharanry Jun 29, 2021 •

edited

Loading

sharanry Jul 1, 2021 •

edited

Loading

sharanry commented Jul 1, 2021 •

edited

Loading

sharanry commented Jul 5, 2021 •

edited

Loading