# NeuralPDE tutorial not running

**URL:** <https://discourse.julialang.org/t/neuralpde-tutorial-not-running/110188>\
**Category:** General Usage\
**Tags:** question, sciml\
**Created:** [February 14, 2024, 12:59am UTC](https://discourse.julialang.org/t/neuralpde-tutorial-not-running/110188 "2024-02-14T00:59:44Z")\
**Posts on this page:** 8\
**Page:** 1

<div class="post-metadata">

**Author:** ![Y\_Sungtaek\_JU](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/y_sungtaek_ju/32/207011_2.png) [@Y\_Sungtaek\_JU](https://discourse.julialang.org/u/Y_Sungtaek_JU)\
**Post date:** [February 14, 2024, 12:59am UTC](https://discourse.julialang.org/t/neuralpde-tutorial-not-running/110188/1 "2024-02-14T00:59:44Z")

</div>

I am new to NeuralPDE.jl. I tried the second tutorial for PDE PINN (defining systems of PDEs for PINNs)

[https://docs.sciml.ai/NeuralPDE/stable/tutorials/systems/#systems](https://docs.sciml.ai/NeuralPDE/stable/tutorials/systems/#systems)

But I got the error message below for the line:

res = Optimization.solve(prob, BFGS(); callback = callback, maxiters = 5000)

ERROR: type OptimizationState has no field depvar

I am using VSCode on Mac M2 and Julia 1.10.0. I created a new environment and added the latest of all the required packages.

Any help would be greatly appreciated!

* * *

Stacktrace:  
[1] getproperty  
@ Base ./Base.jl:37 [inlined]  
[2] macro expansion  
@ NeuralPDE ~/.julia/packages/NeuralPDE/5b5Gb/src/discretize.jl:140 [inlined]  
[3] macro expansion  
@ NeuralPDE ~/.julia/packages/RuntimeGeneratedFunctions/Yo8zx/src/RuntimeGeneratedFunctions.jl:163 [inlined]  
[4] macro expansion  
@ NeuralPDE ./none:0 [inlined]  
[5] generated\_callfunc(::RuntimeGeneratedFunctions.RuntimeGeneratedFunction{…}, ::Matrix{…}, ::Optimization.OptimizationState{…}, ::Vector{…}, ::typeof(NeuralPDE.numeric\_derivative), ::NeuralPDE.var"#300#307"{…}, ::NeuralPDE.var"#32#33", ::Nothing)  
@ NeuralPDE ./none:0  
[6] (::RuntimeGeneratedFunctions.RuntimeGeneratedFunction{…})(::Matrix{…}, ::Optimization.OptimizationState{…}, ::Vector{…}, ::Function, ::Function, ::Function, ::Nothing)  
@ RuntimeGeneratedFunctions ~/.julia/packages/RuntimeGeneratedFunctions/Yo8zx/src/RuntimeGeneratedFunctions.jl:150  
[7] (::NeuralPDE.var"#240#241"{…})(cord::Matrix{…}, θ::Optimization.OptimizationState{…})  
@ NeuralPDE ~/.julia/packages/NeuralPDE/5b5Gb/src/discretize.jl:161  
[8] (::NeuralPDE.var"#integrand#138"{…})(x::Matrix{…}, θ::Optimization.OptimizationState{…})  
@ NeuralPDE ~/.julia/packages/NeuralPDE/5b5Gb/src/training\_strategies.jl:316  
[9] (::BatchIntegralFunction{…})(::Matrix{…}, ::Vararg{…})  
@ SciMLBase ~/.julia/packages/SciMLBase/QSc1r/src/scimlfunctions.jl:2200  
[10] \_\_solvebp\_call(prob::IntegralProblem{…}, alg::Integrals.CubatureJLh, sensealg::Integrals.ReCallVJP{…}, domain::Tuple{…}, p::Optimization.OptimizationState{…}; reltol::Float64, abstol::Float64, maxiters::Int64)  
@ IntegralsCubatureExt ~/.julia/packages/Integrals/CazE7/ext/IntegralsCubatureExt.jl:23  
[11] \_\_solvebp\_call(::Integrals.IntegralCache{…}, ::Integrals.CubatureJLh, ::Vararg{…}; kwargs::@Kwargs{…})  
@ Integrals ~/.julia/packages/Integrals/CazE7/src/common.jl:115  
[12] _solvebp(::Integrals.IntegralCache{…}, ::Vararg{…}; kwargs::@Kwargs{…})  
@ Integrals ~/.julia/packages/Integrals/CazE7/src/Integrals.jl:65  
[13] solve!(cache::Integrals.IntegralCache{…})  
@ Integrals ~/.julia/packages/Integrals/CazE7/src/common.jl:105  
[14] solve(prob::IntegralProblem{…}, alg::Integrals.CubatureJLh; kwargs::@Kwargs{…})  
@ Integrals ~/.julia/packages/Integrals/CazE7/src/common.jl:101  
[15] solve  
@ ~/.julia/packages/Integrals/CazE7/src/common.jl:98 [inlined]  
[16] (::NeuralPDE.var"#134#137"{…})(lb::Vector{…}, ub::Vector{…}, loss::NeuralPDE.var"#240#241"{…}, θ::Optimization.OptimizationState{…})  
@ NeuralPDE ~/.julia/packages/NeuralPDE/5b5Gb/src/training\_strategies.jl:320  
[17] (::NeuralPDE.var"#135#139"{…})(θ::Optimization.OptimizationState{…})  
@ NeuralPDE ~/.julia/packages/NeuralPDE/5b5Gb/src/training\_strategies.jl:326  
[18] (::var"#8#11"{…})(l_::NeuralPDE.var"#135#139"{…})  
@ Main ~/Library/CloudStorage/GoogleDrive-stju@g.ucla.edu/My Drive/STJulia/ST\_PINN/NeuralPDE\_tutorial\_1.jl:45  
[19] iterate(g::Base.Generator, s::Vararg{Any})  
@ Base ./generator.jl:47 [inlined]  
[20] \_collect(c::Vector{…}, itr::Base.Generator{…}, ::Base.EltypeUnknown, isz::Base.HasShape{…})  
@ Base ./array.jl:854  
[21] collect\_similar(cont::Vector{…}, itr::Base.Generator{…})  
@ Base ./array.jl:763  
[22] map(f::Function, A::Vector{NeuralPDE.var"#135#139"{…} where loss\_function})  
@ Base ./abstractarray.jl:3282  
[23] (::var"#7#10")(p::Optimization.OptimizationState{…}, l::Float64)  
@ Main ~/Library/CloudStorage/GoogleDrive-stju@g.ucla.edu/My Drive/STJulia/ST\_PINN/NeuralPDE\_tutorial\_1.jl:45  
[24] (::OptimizationOptimJL.var"#\_cb#12"{OptimizationCache{…}})(trace::OptimizationState{Float64, BFGS{…}})  
@ OptimizationOptimJL ~/.julia/packages/OptimizationOptimJL/yMF3E/src/OptimizationOptimJL.jl:142  
[25] update!(tr::Vector{…}, iteration::Int64, f\_x::Float64, grnorm::Float64, dt::Dict{…}, store\_trace::Bool, show\_trace::Bool, show\_every::Int64, callback::OptimizationOptimJL.var"#\_cb#12"{…}, trace\_simplex::Bool)  
@ Optim ~/.julia/packages/Optim/EJwLF/src/utilities/update.jl:25  
[26] update!(tr::Vector{…}, iteration::Int64, f\_x::Float64, grnorm::Float64, dt::Dict{…}, store\_trace::Bool, show\_trace::Bool, show\_every::Int64, callback::OptimizationOptimJL.var"#\_cb#12"{…}, trace\_simplex::Bool)  
@ Optim ~/.julia/packages/Optim/EJwLF/src/utilities/update.jl:11 [inlined]  
[27] trace!(tr::Vector{…}, d::TwiceDifferentiable{…}, state::Optim.BFGSState{…}, iteration::Int64, method::BFGS{…}, options::Optim.Options{…}, curr\_time::Float64)  
@ Optim ~/.julia/packages/Optim/EJwLF/src/multivariate/solvers/first\_order/bfgs.jl:191  
[28] optimize(d::TwiceDifferentiable{…}, initial\_x::ComponentArrays.ComponentVector{…}, method::BFGS{…}, options::Optim.Options{…}, state::Optim.BFGSState{…})  
@ Optim ~/.julia/packages/Optim/EJwLF/src/multivariate/optimize/optimize.jl:50  
[29] optimize(d::TwiceDifferentiable{…}, initial\_x::ComponentArrays.ComponentVector{…}, method::BFGS{…}, options::Optim.Options{…})  
@ Optim ~/.julia/packages/Optim/EJwLF/src/multivariate/optimize/optimize.jl:36  
[30] \_\_solve(cache::OptimizationCache{…})  
@ OptimizationOptimJL ~/.julia/packages/OptimizationOptimJL/yMF3E/src/OptimizationOptimJL.jl:212  
[31] solve!(cache::OptimizationCache{…})  
@ SciMLBase ~/.julia/packages/SciMLBase/QSc1r/src/solve.jl:179  
[32] solve(::OptimizationProblem{…}, ::BFGS{…}; kwargs::@Kwargs{…})  
@ SciMLBase ~/.julia/packages/SciMLBase/QSc1r/src/solve.jl:96  
[33] top-level scope  
@ ~/Library/CloudStorage/GoogleDrive-stju@g.ucla.edu/My Drive/STJulia/ST\_PINN/NeuralPDE\_tutorial\_1.jl:50  
Some type information was truncated. Use `show(err)` to see complete types.

---

<div class="post-metadata">

**Author:** ![ChrisRackauckas](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/chrisrackauckas/32/77_2.png) [@ChrisRackauckas](https://discourse.julialang.org/u/ChrisRackauckas)\
**Post date:** [February 14, 2024, 2:12am UTC](https://discourse.julialang.org/t/neuralpde-tutorial-not-running/110188/2 "2024-02-14T02:12:05Z")

</div>

What version? `]st`?

---

<div class="post-metadata">

**Author:** ![Y\_Sungtaek\_JU](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/y_sungtaek_ju/32/207011_2.png) [@Y\_Sungtaek\_JU](https://discourse.julialang.org/u/Y_Sungtaek_JU)\
**Post date:** [February 14, 2024, 4:35pm UTC](https://discourse.julialang.org/t/neuralpde-tutorial-not-running/110188/3 "2024-02-14T16:35:49Z")

</div>

Oh, sorry for not including that info. Here it is.

[b2108857] Lux v0.5.17  
[961ee093] ModelingToolkit v8.75.0  
[315f7962] NeuralPDE v5.11.0  
[7f7a1694] Optimization v3.22.0  
[36348300] OptimizationOptimJL v0.2.2  
[91a5bcdd] Plots v1.40.1

---

<div class="post-metadata">

**Author:** ![ChrisRackauckas](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/chrisrackauckas/32/77_2.png) [@ChrisRackauckas](https://discourse.julialang.org/u/ChrisRackauckas)\
**Post date:** [February 17, 2024, 11:37am UTC](https://discourse.julialang.org/t/neuralpde-tutorial-not-running/110188/4 "2024-02-17T11:37:52Z")

</div>

Was that in a fresh REPL?

---

<div class="post-metadata">

**Author:** ![Y\_Sungtaek\_JU](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/y_sungtaek_ju/32/207011_2.png) [@Y\_Sungtaek\_JU](https://discourse.julialang.org/u/Y_Sungtaek_JU)\
**Post date:** [February 17, 2024, 8:04pm UTC](https://discourse.julialang.org/t/neuralpde-tutorial-not-running/110188/5 "2024-02-17T20:04:53Z")

</div>

Yes. I made a dedicated environment to run these tutorials for NeuralPDE.

Thanks for looking into this!

---

<div class="post-metadata">

**Author:** ![SathvikBhagavan](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/sathvikbhagavan/32/205417_2.png) [@SathvikBhagavan](https://discourse.julialang.org/u/SathvikBhagavan)\
**Post date:** [February 18, 2024, 4:01am UTC](https://discourse.julialang.org/t/neuralpde-tutorial-not-running/110188/6 "2024-02-18T04:01:26Z")

</div>

The tutorial is slightly outdated. The issue is in the callback functions.

```julia
callback = function (p, l)
           println("loss: ", l)
           println("pde_losses: ", map(l_ -> l_(p.u), pde_inner_loss_functions))
           println("bcs_losses: ", map(l_ -> l_(p.u), bcs_inner_loss_functions))
           return false
       end

```

should fix it. Notice doing `p.u` instead of `p`. This is because of OptimizationOptimJL@0.2 introduced `OptimizationState` ([[WIP]Stats and State by Vaibhavdixit02 · Pull Request #652 · SciML/Optimization.jl · GitHub](https://github.com/SciML/Optimization.jl/pull/652)).

---

<div class="post-metadata">

**Author:** ![Y\_Sungtaek\_JU](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/y_sungtaek_ju/32/207011_2.png) [@Y\_Sungtaek\_JU](https://discourse.julialang.org/u/Y_Sungtaek_JU)\
**Post date:** [February 19, 2024, 8:10am UTC](https://discourse.julialang.org/t/neuralpde-tutorial-not-running/110188/7 "2024-02-19T08:10:51Z")

</div>

Thanks!

---

<div class="post-metadata">

**Author:** ![Maysam\_Gholampour](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/maysam_gholampour/32/218184_2.png) [@Maysam\_Gholampour](https://discourse.julialang.org/u/Maysam_Gholampour)\
**Post date:** [February 27, 2024, 6:43am UTC](https://discourse.julialang.org/t/neuralpde-tutorial-not-running/110188/8 "2024-02-27T06:43:16Z")

</div>

thanks.