Volume preserving transformer

Project.toml

@@ -41,7 +41,6 @@ GeometricEquations = "0.16"
 GeometricIntegrators = "0.13"
 HDF5 = "0.16, 0.17"
 KernelAbstractions = "0.9"
-Lux = "0.4, 0.5"
 NNlib = "0.8, 0.9"
 ProgressMeter = "1"
 SafeTestsets = "0.1"
@@ -55,11 +54,10 @@ Aqua = "4c88cf16-eb10-579e-8560-4a9242c79595"
 ChainRulesTestUtils = "cdddcdb0-9152-4a09-a978-84456f9df70a"
 GeometricEquations = "c85262ba-a08a-430a-b926-d29770767bf2"
 GeometricProblems = "18cb22b4-ad41-5c80-9c5f-710df63fbdc9"
-Lux = "b2108857-7c20-44ae-9111-449ecde12c47"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 RungeKutta = "fb486d5c-30a0-4a8a-8415-a8b4ace5a6f7"
 SafeTestsets = "1bc83da4-3b8d-516f-aca4-4fe02f6d838f"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
 
 [targets]
-test = ["ChainRulesTestUtils", "Lux", "Random", "SafeTestsets", "Test"]
+test = ["ChainRulesTestUtils", "Random", "SafeTestsets", "Test"]

code_generation/matrixinverse.jl

@@ -0,0 +1,21 @@
+using Symbolics
+
+function matrix_inverse(n)
+    @variables A[1:n,1:n]
+
+    B = inv(collect(A))
+
+    if n ≤ 6
+        B = simplify.(B)
+    end
+
+    build_function(B, A)[2]
+end
+
+for n in 2:8
+    println("$(n)x$(n)...")
+    expr = matrix_inverse(n)
+    str = replace(replace(replace(string(expr), r"#= [^*\s]* =#" => ""), r"\n[\s]*\n" => "\n"), "Num" => "")
+    str = replace(str, "function (ˍ₋out, A)" => "function matinv$(n)x$(n)(ˍ₋out, A)")
+    write("matrixinverse_$(n)x$(n).jl", str)
+end

docs/src/tikz/Makefile

@@ -22,8 +22,6 @@ convert_with_pdftocairo: $(MYDIR)/*.pdf
 	done
 
 png:
-	pdftocairo 	-png -r 500		-transp -singlefile	 logo_with_name.pdf								logo_with_name
-	pdftocairo 	-png -r 500		-transp -singlefile	 logo_with_name_dark.pdf						logo_with_name_dark
 	pdftocairo  -png -r $(res1) -transp -singlefile  adam_optimizer.pdf               	 			adam_optimizer  
 	pdftocairo  -png -r $(res1) -transp -singlefile  adam_optimizer_dark.pdf               	 		adam_optimizer_dark                     
 	pdftocairo  -png -r $(res1) -transp -singlefile  general_optimization.pdf		 	 			general_optimization   
@@ -36,9 +34,9 @@ png:
 	pdftocairo	-png -r $(res2) -transp -singlefile	 sympnet_architecture_dark.pdf					sympnet_architecture_dark
 	pdftocairo 	-png -r $(res3) -transp -singlefile  structs_visualization.pdf						structs_visualization
 	pdftocairo 	-png -r $(res3) -transp -singlefile  structs_visualization_dark.pdf					structs_visualization_dark
-	pdftocairo 	-png -r $(res1) -transp -singlefile	 logo.pdf										logo
 	pdftocairo  -png -r 500     -transp -singlefile  logo_with_name.pdf                             logo_with_name
 	pdftocairo  -png -r 500     -transp -singlefile  logo_with_name_dark.pdf                        logo_with_name_dark
+	pdftocairo 	-png -r $(res1) -transp -singlefile	 logo.pdf										logo 
 	pdftocairo  -png -r $(res1) -transp -singlefile  symplectic_autoencoder.pdf						symplectic_autoencoder
 	pdftocairo  -png -r $(res1) -transp -singlefile  symplectic_autoencoder_dark.pdf				symplectic_autoencoder_dark
 	pdftocairo  -png -r $(res1) -transp -singlefile  solution_manifold_2.pdf						solution_manifold_2

docs/src/tutorials/sympnet_tutorial.md

@@ -78,7 +78,10 @@ H:(q,p)\in\mathbb{R}^2 \mapsto \frac{1}{2}p^2-cos(q) \in \mathbb{R}.
 Here we generate pendulum data with the script `GeometricMachineLearning/scripts/pendulum.jl`:
 
 ```@example sympnet
-using GeometricMachineLearning
+using GeometricMachineLearning # hide
+import Random # hide
+
+Random.seed!(1234)
 
 # load script
 include("../../../scripts/pendulum.jl")
@@ -165,7 +168,7 @@ The train function will change the parameters of the neural networks and gives a
 
 The trainings data `data_q` and `data_p` must be matrices of $\mathbb{R}^{n\times d}$ where $n$ is the length of data and $d$ is the half of the dimension of the system, i.e `data_q[i,j]` is $q_j(t_i)$ where $(t_1,...,t_n)$ are the corresponding time of the training data.
 
-Then we can make prediction. Let's compare the initial data with a prediction starting from the same phase space point using the provided function Iterate_Sympnet:
+Then we can make prediction. Let's compare the initial data with a prediction starting from the same phase space point using the provided function `iterate`:
 
 ```@example sympnet
 ics = (q=qp_data.q[:,1], p=qp_data.p[:,1])

scripts/Project.toml

@@ -1,10 +1,13 @@
 [deps]
-Flux = "587475ba-b771-5e3f-ad9e-33799f191a9c"
+CUDA = "052768ef-5323-5732-b1bb-66c8b64840ba"
+GeometricEquations = "c85262ba-a08a-430a-b926-d29770767bf2"
 GeometricIntegrators = "dcce2d33-59f6-5b8d-9047-0defad88ae06"
 GeometricMachineLearning = "194d25b2-d3f5-49f0-af24-c124f4aa80cc"
+GeometricProblems = "18cb22b4-ad41-5c80-9c5f-710df63fbdc9"
+GeometricSolutions = "7843afe4-64f4-4df4-9231-049495c56661"
 LaTeXStrings = "b964fa9f-0449-5b57-a5c2-d3ea65f4040f"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
-Lux = "b2108857-7c20-44ae-9111-449ecde12c47"
 Plots = "91a5bcdd-55d7-5caf-9e0b-520d859cae80"
+ProfileView = "c46f51b8-102a-5cf2-8d2c-8597cb0e0da7"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 Zygote = "e88e6eb3-aa80-5325-afca-941959d7151f"