build based on c69efb2

dev/.documenter-siteinfo.json

@@ -1 +1 @@
-{"documenter":{"julia_version":"1.10.5","generation_timestamp":"2024-09-24T14:04:13","documenter_version":"1.7.0"}}
+{"documenter":{"julia_version":"1.10.5","generation_timestamp":"2024-09-24T15:39:56","documenter_version":"1.7.0"}}

dev/api/index.html

@@ -4,4 +4,4 @@
     rng = Xoshiro(0),
     init_params = Lux.initialparameters(rng, chain),
     eltype = Float64,
-    name)</code></pre><p>Create an <code>ODESystem</code> with a neural network inside.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/SciML/ModelingToolkitNeuralNets.jl/blob/c9e6d673b9bea1dd80ec6757e6beae38d336c92d/src/ModelingToolkitNeuralNets.jl#L15-L24">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../friction/">« Friction Model</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="auto">Automatic (OS)</option><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="catppuccin-latte">catppuccin-latte</option><option value="catppuccin-frappe">catppuccin-frappe</option><option value="catppuccin-macchiato">catppuccin-macchiato</option><option value="catppuccin-mocha">catppuccin-mocha</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.7.0 on <span class="colophon-date" title="Tuesday 24 September 2024 14:04">Tuesday 24 September 2024</span>. Using Julia version 1.10.5.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
+    name)</code></pre><p>Create an <code>ODESystem</code> with a neural network inside.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/SciML/ModelingToolkitNeuralNets.jl/blob/c69efb2d2344a9514244d427a86472a51fbea5a2/src/ModelingToolkitNeuralNets.jl#L15-L24">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../friction/">« Friction Model</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="auto">Automatic (OS)</option><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="catppuccin-latte">catppuccin-latte</option><option value="catppuccin-frappe">catppuccin-frappe</option><option value="catppuccin-macchiato">catppuccin-macchiato</option><option value="catppuccin-mocha">catppuccin-mocha</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.7.0 on <span class="colophon-date" title="Tuesday 24 September 2024 15:39">Tuesday 24 September 2024</span>. Using Julia version 1.10.5.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>

dev/assets/Project.toml

@@ -18,7 +18,7 @@ Lux = "1"
 ModelingToolkit = "9.9"
 ModelingToolkitNeuralNets = "1"
 ModelingToolkitStandardLibrary = "2.7"
-Optimization = "3.24"
+Optimization = "3.24, 4.0"
 OptimizationOptimisers = "0.2.1, 0.3"
 OrdinaryDiffEq = "6.74"
 Plots = "1"

dev/friction/index.html

@@ -72,7 +72,7 @@
  [8.965808955089017]
  [9.04079394249597]
  [9.115490076324514]
- [9.18990042447709]</code></pre><p>Let&#39;s plot it.</p><pre><code class="language-julia hljs">scatter(sol_ref, label = &quot;velocity&quot;)</code></pre><img src="0d481402.svg" alt="Example block output"/><p>That was the velocity. Let&#39;s also plot the friction force acting on the object throughout the simulation.</p><pre><code class="language-julia hljs">scatter(sol_ref.t, friction.(first.(sol_ref.u)), label = &quot;friction force&quot;)</code></pre><img src="4eabe0d9.svg" alt="Example block output"/><h2 id="Model-Setup"><a class="docs-heading-anchor" href="#Model-Setup">Model Setup</a><a id="Model-Setup-1"></a><a class="docs-heading-anchor-permalink" href="#Model-Setup" title="Permalink"></a></h2><p>Now, we will try to learn the same friction model using a neural network. We will use <a href="../api/#ModelingToolkitNeuralNets.NeuralNetworkBlock"><code>NeuralNetworkBlock</code></a> to define neural network as a component. The input of the neural network is the velocity and the output is the friction force. We connect the neural network with the model using <code>RealInputArray</code> and <code>RealOutputArray</code> blocks.</p><pre><code class="language-julia hljs">function friction_ude(Fu)
+ [9.18990042447709]</code></pre><p>Let&#39;s plot it.</p><pre><code class="language-julia hljs">scatter(sol_ref, label = &quot;velocity&quot;)</code></pre><img src="b41f059d.svg" alt="Example block output"/><p>That was the velocity. Let&#39;s also plot the friction force acting on the object throughout the simulation.</p><pre><code class="language-julia hljs">scatter(sol_ref.t, friction.(first.(sol_ref.u)), label = &quot;friction force&quot;)</code></pre><img src="cc41e81e.svg" alt="Example block output"/><h2 id="Model-Setup"><a class="docs-heading-anchor" href="#Model-Setup">Model Setup</a><a id="Model-Setup-1"></a><a class="docs-heading-anchor-permalink" href="#Model-Setup" title="Permalink"></a></h2><p>Now, we will try to learn the same friction model using a neural network. We will use <a href="../api/#ModelingToolkitNeuralNets.NeuralNetworkBlock"><code>NeuralNetworkBlock</code></a> to define neural network as a component. The input of the neural network is the velocity and the output is the friction force. We connect the neural network with the model using <code>RealInputArray</code> and <code>RealOutputArray</code> blocks.</p><pre><code class="language-julia hljs">function friction_ude(Fu)
     @variables y(t) = 0.0
     @constants Fu = Fu
     @named nn_in = RealInputArray(nin = 1)
@@ -131,26 +131,26 @@
 op = OptimizationProblem(of, x0, (prob, sol_ref, get_vars, get_refs))
 res = solve(op, Adam(5e-3); maxiters = 10000, callback = cb)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">retcode: Default
 u: 120-element Vector{Float64}:
- -2.3636356168896393
- -2.4192743320056964
-  2.0553216024376972
- -0.299925206928586
-  0.9657348367786275
- -2.4086786498364874
- -2.321253697551076
-  0.5512956241338778
- -2.436006830074837
- -2.4256518875480806
+ -2.3637128272169083
+ -2.419325259817553
+  2.055314750464916
+ -0.29997447699192004
+  0.9657138396582154
+ -2.408748160967478
+ -2.3213373506901007
+  0.5512794612474039
+ -2.436072303529551
+ -2.425698691324607
   ⋮
- -2.1175791469580205
-  1.231137921763269
- -1.0464754889905168
- -0.25115604090898497
- -1.5963390506010442
-  1.453488833212753
- -2.2825358674620055
- -1.411365516380704
- -1.5278017288524202</code></pre><h2 id="Visualization-of-results"><a class="docs-heading-anchor" href="#Visualization-of-results">Visualization of results</a><a id="Visualization-of-results-1"></a><a class="docs-heading-anchor-permalink" href="#Visualization-of-results" title="Permalink"></a></h2><p>We now have a trained neural network! We can check whether running the simulation of the model embedded with the neural network matches the data or not.</p><pre><code class="language-julia hljs">res_p = SciMLStructures.replace(Tunable(), prob.p, res.u)
+ -2.1176846790986064
+  1.2311814175719746
+ -1.0465605429246023
+ -0.2511701306776758
+ -1.5964125302790146
+  1.4535187677568038
+ -2.2826745226184655
+ -1.4114013904506582
+ -1.5278808184718486</code></pre><h2 id="Visualization-of-results"><a class="docs-heading-anchor" href="#Visualization-of-results">Visualization of results</a><a id="Visualization-of-results-1"></a><a class="docs-heading-anchor-permalink" href="#Visualization-of-results" title="Permalink"></a></h2><p>We now have a trained neural network! We can check whether running the simulation of the model embedded with the neural network matches the data or not.</p><pre><code class="language-julia hljs">res_p = SciMLStructures.replace(Tunable(), prob.p, res.u)
 res_prob = remake(prob, p = res_p)
 res_sol = solve(res_prob, Rodas4(), saveat = sol_ref.t)</code></pre><p>Also, it would be interesting to check the simulation before the training to get an idea of the starting point of the network.</p><pre><code class="language-julia hljs">initial_sol = solve(prob, Rodas4(), saveat = sol_ref.t)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">retcode: Success
 Interpolation: 1st order linear
@@ -197,6 +197,6 @@
  [12.001435208620208, 12.001435208620206]
  [12.123764312172343, 12.123764312172343]</code></pre><p>Now we plot it.</p><pre><code class="language-julia hljs">scatter(sol_ref, idxs = [model_true.y], label = &quot;ground truth velocity&quot;)
 plot!(res_sol, idxs = [sys.friction.y], label = &quot;velocity after training&quot;)
-plot!(initial_sol, idxs = [sys.friction.y], label = &quot;velocity before training&quot;)</code></pre><img src="bb177919.svg" alt="Example block output"/><p>It matches the data well! Let&#39;s also check the predictions for the friction force and whether the network learnt the friction model or not.</p><pre><code class="language-julia hljs">scatter(sol_ref.t, friction.(first.(sol_ref.u)), label = &quot;ground truth friction&quot;)
+plot!(initial_sol, idxs = [sys.friction.y], label = &quot;velocity before training&quot;)</code></pre><img src="02d32aba.svg" alt="Example block output"/><p>It matches the data well! Let&#39;s also check the predictions for the friction force and whether the network learnt the friction model or not.</p><pre><code class="language-julia hljs">scatter(sol_ref.t, friction.(first.(sol_ref.u)), label = &quot;ground truth friction&quot;)
 plot!(res_sol.t, getindex.(res_sol[sys.nn.output.u], 1),
-    label = &quot;friction from neural network&quot;)</code></pre><img src="d4f1cf24.svg" alt="Example block output"/><p>It learns the friction model well!</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../">« Home</a><a class="docs-footer-nextpage" href="../api/">API »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="auto">Automatic (OS)</option><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="catppuccin-latte">catppuccin-latte</option><option value="catppuccin-frappe">catppuccin-frappe</option><option value="catppuccin-macchiato">catppuccin-macchiato</option><option value="catppuccin-mocha">catppuccin-mocha</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.7.0 on <span class="colophon-date" title="Tuesday 24 September 2024 14:04">Tuesday 24 September 2024</span>. Using Julia version 1.10.5.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
+    label = &quot;friction from neural network&quot;)</code></pre><img src="b6d297a2.svg" alt="Example block output"/><p>It learns the friction model well!</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../">« Home</a><a class="docs-footer-nextpage" href="../api/">API »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="auto">Automatic (OS)</option><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="catppuccin-latte">catppuccin-latte</option><option value="catppuccin-frappe">catppuccin-frappe</option><option value="catppuccin-macchiato">catppuccin-macchiato</option><option value="catppuccin-mocha">catppuccin-mocha</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.7.0 on <span class="colophon-date" title="Tuesday 24 September 2024 15:39">Tuesday 24 September 2024</span>. Using Julia version 1.10.5.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>