Merge pull request #29 from paulxshen/dev

Dev
paulxshen · Mar 31, 2024 · 040477a · 040477a
2 parents 134c82c + 09ab154
commit 040477a
Show file tree

Hide file tree

Showing 7 changed files with 23 additions and 17 deletions.
diff --git a/README.md b/README.md
@@ -1,7 +1,5 @@
 Working draft of a differentiable FDTD engine for inverse design in photonics, acoustics & RF.
-
-Documentation at https://github.io/differentiable-fdtd-beta-prerelease
-
+Documentation at https://paulxshen.github.io/Luminescent.jl/
 Discussion & updates at https://discourse.julialang.org/t/pre-ann-differentiable-fdtd-for-inverse-design-in-photonics-acoustics-and-rf/105405/12
 ![a](bend-pre.gif)
 ![b](bend-post.gif)
diff --git a/examples/inverse_design_waveguide_bend/inverse_design_waveguide_bend.jl b/examples/inverse_design_waveguide_bend/inverse_design_waveguide_bend.jl
@@ -100,8 +100,9 @@ power_profile = F.(real.(Jy .* conj.(Mz)))
 power_profile /= norm(power_profile)
 
 if ongpu
-    using CUDA, Flux
-    @assert CUDA.functional()
+    using Flux
+    # using CUDA
+    # @assert CUDA.functional()
     u0, model, static_mask, μ, σ, σm, field_padding, source_instances =
         gpu.((u0, model, static_mask, μ, σ, σm, field_padding, source_instances))
     merge!(configs, (; u0, field_padding, source_instances))

diff --git a/examples/periodic_scattering/periodic_scattering.jl b/examples/periodic_scattering/periodic_scattering.jl
@@ -61,8 +61,9 @@ p = apply(geometry_staggering; p...)
 
 # move to gpu
 if dogpu
-    using CUDA, Flux
-    @assert CUDA.functional()
+    using Flux
+    # using CUDA
+    # @assert CUDA.functional()
     u0, p, field_padding, source_instances = gpu.((u0, p, field_padding, source_instances))
 end
 

diff --git a/examples/periodic_scattering/periodic_scattering.mp4 b/examples/periodic_scattering/periodic_scattering.mp4
diff --git a/examples/quarter_wavelength_antenna/quarter_wavelength_antenna.jl b/examples/quarter_wavelength_antenna/quarter_wavelength_antenna.jl
@@ -2,7 +2,7 @@
 We simulate a quarter wavelength antenna above conductor ground plane and compute its nearfield radiation pattern
 =#
 
-using UnPack, LinearAlgebra, GLMakie
+using UnPack, LinearAlgebra, GLMakie, CoordinateTransformations
 using GLMakie: volume
 using Luminescent, LuminescentVisualization
 
@@ -45,8 +45,9 @@ p = apply(geometry_staggering, p)
 
 # move to gpu
 if dogpu
-    using CUDA, Flux
-    @assert CUDA.functional()
+    using Flux
+    # using CUDA
+    # @assert CUDA.functional()
     u0, p, field_padding, source_instances = gpu.((u0, p, field_padding, source_instances))
 end
 

diff --git a/makedocs/src/index.md b/makedocs/src/index.md
@@ -2,7 +2,7 @@
 
  Prerelease. First stable release planned for April . Until then, accuracy not validated. Report bugs on [Github](https://github.com/paulxshen/Luminescent.jl) - we usually respond within a day
 ## Overview
-Generative design meets Maxwell's Equations. Differentiable FDTD package for inverse design & topology optimization in semiconductor photonics, acoustics and RF. GPU and automatic differentiation (AD) compatible. Uses AD by `Zygote.jl` for adjoint optimization. Integrates with [`Jello.jl`](https://github.com/paulxshen/Jello.jl) to generate length scale controlled paramaterized geometry . Staggered Yee grid update with fully featured boundary conditions & sources in 1d/2d/3d. Customizable physics to potentially incorporate dynamics like heat transfer, charge transport.
+Generative design meets Maxwell's Equations. GPU and automatic differentiation (AD) compatible FDTD package in Julia for inverse design & topology optimization in semiconductor photonics, acoustics and RF. Uses AD by `Zygote.jl` for adjoint optimization. Integrates with [`Jello.jl`](https://github.com/paulxshen/Jello.jl) to generate length scale controlled paramaterized geometry . Staggered Yee grid update with fully featured boundary conditions & sources in 1d/2d/3d. Customizable physics to potentially incorporate dynamics like heat transfer, charge transport.
 ## Gallery
 
 ### Generative Inverse design of compact silicon photonics splitter 
@@ -13,12 +13,17 @@ Generative design meets Maxwell's Equations. Differentiable FDTD package for inv
 ![](assets/periodic_scattering.mp4)
 Please star us on [Github](https://github.com/paulxshen/Luminescent.jl) if you like our work :)
 ## Installation
-Install via 
+Install Julia preferably with VSCODE. Then add our package  via 
 ```
+using Pkg
 Pkg.add(url="https://github.com/paulxshen/Luminescent.jl")
 Pkg.add(url="https://github.com/paulxshen/LuminescentVisualization.jl")
 ```
-`LuminescentVisualization.jl` contains visualization utilities
+`LuminescentVisualization.jl` contains visualization utilities. If you're running tutorials also add
+```
+]add UnPack, BSON,DataStructures, StatsBase, Zygote, Jello, GLMakie, CoordinateTransformations,AbbreviatedStackTraces,CUDA,Flux
+```
+Omit CUDA if you don't have NVidia GPU
 ## Quickstart
 Please refer to the first tutorial 
 ## People

diff --git a/scripts/setup.jl b/scripts/setup.jl
@@ -1,8 +1,8 @@
 using Pkg
-# ]add UnPack, BSON,DataStructures, StatsBase, Documenter,ImageTransformations,Interpolations, Zygote, Optim,ArrayPadding, Porcupine, Jello, GLMakie, Functors, Lazy,AbbreviatedStackTraces,ReverseStackTraces,CUDA,Flux
+# ]add UnPack, BSON,DataStructures, StatsBase, Documenter,ImageTransformations,Interpolations, Zygote, Optim,ArrayPadding, Porcupine, Jello, GLMakie, Functors, CoordinateTransformations,Lazy,AbbreviatedStackTraces,ReverseStackTraces,CUDA,Flux
 # Pkg.add(url="https://github.com/paulxshen/ArrayPadding.jl")
-Pkg.add(url="https://github.com/paulxshen/Jello.jl")
+# Pkg.add(url="https://github.com/paulxshen/Jello.jl")
 
 # Pkg.add(url="https://github.com/paulxshen/Porcupine.jl")
-# Pkg.add(url="https://github.com/paulxshen/Luminescent.jl")
-# Pkg.add(url="https://github.com/paulxshen/LuminescentVisualization.jl")
+Pkg.add(url="https://github.com/paulxshen/Luminescent.jl")
+Pkg.add(url="https://github.com/paulxshen/LuminescentVisualization.jl")