# Sciml\_train not defined

**URL:** <https://discourse.julialang.org/t/sciml-train-not-defined/102884>\
**Category:** New to Julia\
**Created:** [August 16, 2023, 9:12pm UTC](https://discourse.julialang.org/t/sciml-train-not-defined/102884 "2023-08-16T21:12:24Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![mirandaholmes](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/mirandaholmes/32/52290_2.png) [@mirandaholmes](https://discourse.julialang.org/u/mirandaholmes)\
**Post date:** [August 16, 2023, 9:12pm UTC](https://discourse.julialang.org/t/sciml-train-not-defined/102884/1 "2023-08-16T21:12:24Z")

</div>

Hi, I just started using Julia, on a Macbook Air with M2 chip. I am trying to run the Lotka-Volterra optimization example ([https://docs.juliahub.com/DiffEqFlux/BdO4p/1.10.3/examples/LV-ODE/](https://docs.juliahub.com/DiffEqFlux/BdO4p/1.10.3/examples/LV-ODE/)) verbatim, after installing all the required packages. I continue getting this error:

ERROR: UndefVarError: `sciml_train` not defined  
Stacktrace:  
[1] getproperty(x::Module, f::Symbol)  
@ Base ./Base.jl:31  
[2] top-level scope  
@ ~/Library/CloudStorage/Dropbox/Work/JuliaOptTests/test\_odeopt\_ml.jl:67

I have tried updating the packages, restarting, etc, but I continue to get this error. What am I doing wrong? Is there a bug in the package?

---

<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:** [August 16, 2023, 10:06pm UTC](https://discourse.julialang.org/t/sciml-train-not-defined/102884/2 "2023-08-16T22:06:06Z")

</div>

The problem is that is not the right documentation. If you use the package’s documentation things will work. I think you’re looking for:

[https://docs.sciml.ai/SciMLSensitivity/stable/tutorials/parameter\_estimation\_ode/](https://docs.sciml.ai/SciMLSensitivity/stable/tutorials/parameter_estimation_ode/)

---

<div class="post-metadata">

**Author:** ![mirandaholmes](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/mirandaholmes/32/52290_2.png) [@mirandaholmes](https://discourse.julialang.org/u/mirandaholmes)\
**Post date:** [August 17, 2023, 7:06pm UTC](https://discourse.julialang.org/t/sciml-train-not-defined/102884/3 "2023-08-17T19:06:29Z")

</div>

Thank you! I didn’t know to watch out for different documentation for different versions. (The one above is what pops up in google when I search for “Julia ODE parameter estimation”).

I have another question – the example in the documentation above uses a package OptimizationPolyalgorithms. Where can I find documentation about this package? It doesn’t seem to be listed in the list of packages in the Optimization.jl documentation – see eg the left-hand side bar in [https://docs.sciml.ai/Optimization/stable/optimization\_packages/optimisers/#Installation:-OptimizationFlux.jl](https://docs.sciml.ai/Optimization/stable/optimization_packages/optimisers/#Installation:-OptimizationFlux.jl), and the only documentation I can find online through juliahub appears to be empty: [https://docs.juliahub.com/OptimizationPolyalgorithms/2YgTO/0.1.0/](https://docs.juliahub.com/OptimizationPolyalgorithms/2YgTO/0.1.0/) .

Is there documentation on this package somewhere? If not, would a similar package be OptimizationOptimisers?

---

<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:** [August 17, 2023, 10:19pm UTC](https://discourse.julialang.org/t/sciml-train-not-defined/102884/4 "2023-08-17T22:19:02Z")

</div>

> [@mirandaholmes](#):
>
> Is there documentation on this package somewhere? If not, would a similar package be OptimizationOptimisers?

There just isn’t yet. I need to write something on the polyalgorithms, that one is my fault 😅 . You can just not use the polyalgorithm for now if you don’t want, using Adam or BFGS is fine, this is just more robust.