Proper Release alongside paper (#59)

* Also run pre-commit on dev * Update docs requirements * Remove old requirements * Remove no longer needed setup.py * No longer target incompatible Python versions * Add credits and improve settings * Wrap FFT for specific array structures in Exponax (#4) * Add custom wrapped FFT calls * Rewrite derivative routines to use wrapped FFT calls * Make number of spatial dims optional * Run black * Correctly declare as optional * Use wrapped routines * Use wrapped routines * Use wrapped routines * Use wrapped routines * Use wrapped routines * Use wrapped routines * Export the spectral submodule * Add more details on usage * Start documentation of spectral module * Stepper restructure (#21) * Move reaction into stepper * Adapt imports in tests * Move linear timesteppers into separete files * Move generic physical stepper into submodule * Rename to nonlinear * Correctly export in submodule * Move utilities * Export generic utilities * Fully translate linear generic stepper * Fully translate convection stepper * Fully translate generic nonlinear stepper * Fully translate generic Gradient norm stepper * Correctly forward dealiasing fraction * No need to set dealiasing fraction * Fully translate generic polynomial stepper * Don't translate vorticity convection * Remove normalized submodule * Remove no longer relevant exports * Export generic stepper * Fix issues with dealiasing_fraction arguments * Adapt tests * Remove commented sections * Adapt cross-references * New docs layout to better reflect the API * Add a boilerplate for the overview * Consistent docs (#26) * Add documentation to base * Add documentation to etdrk core * Adapt to Equinox style * Change to Equinox style * Change to Equinox style * Change to Equinox style * Fix typo * Consistent documentation * Improve nonlin fun base documentation * Change to Equinox style * Improve docs for convection * Improve docs for gradient norm * Improve docs for combined general nonlinearity * Fix docstring * Improve and fix documentation * Fix broken link * Adapt to Equinox style * Improve documentation * Add clarification * better wording * Fix convection difficulty computation * Improve doc of generic gradient norm stepper * Add docstring * Add documentation * Add documentation * Add documentation * Use correct tuple typing (#27) * Add type hints (#28) * Fourier repeated stepper (#29) * Implement logic of fourier substepping * Forward .step() to call signature with shape check * Add hints from Base Stepper * Add hints on cost-saving * Also add hint to constructor * Add a test for the new repeated stepper * Add CI for running a test on PRs * Fix repeated Python versions * Fourier Interpolation and Up/Downsampling (#33) * Add simple Fourier interpolator * Add 1D tests * Add tests in 2d * Loosen tests * Fix the generalization to 2d * Add tests in 3d * Forward indexing convention * Correctly use ellipsis * Add documentation * Add documentation * Add utility to get modes slices * Add test for mode slice generation * Return tuples * Utility to move between resolutions * Add tests for mapping between resolutions * Fix issue with oddball sanity * Add new utilities to documentation * Add documentation * Add return type hint * Simple showcase of new functions * Properly rerun notebook * Add PyTest dependency * Fix tests to use tuples * Loosen requirements on (decaying) Burgers problem * Add requirement on 3d volume renderer * Use quotes to correctly get 3.10 * Ignore test requiring GPU * Remove vape dependency because we no longer test for it * Spectral updates (#35) * Preliminary function to compute the spectrum * Potential improvement to spectrum computation * Add hint on enstrophy * Fix on wavenumber norm computation * Rename to better reflect its effect on oddball mode * Add tests on filter mask * Improve first half of spectral documentation * Merge common subexpressions * Extend and fix documentation * Add experimental convinience function to extract the Fourier coefficients * Unify the creation of scaling arrays * Test scaling array for norm_compensation * Add tests for coefficient extraction in 1D * Test for coefficient extration in 2D * Enhance docstring * Remove experimental function * Fix spectrum function * Export get_spectrum instead of make_incompressible * Adapt docs * Use proper links * Enhance docstring * Enhance docstring * Test the spectrum creation * Restructure docstring * Tests helper utilities for fft setup * Hyperviscous kdv [BREAKING] (#36) * Have the dispersion operator being created seperately * Make KdV hyper-viscous by default * Remove unnecessary copy * Requires new coefficients to work with changed defaults in KdV * Fourier Coefficient Extractor (#38) * Add function * Enhance docstring * Change rounding default * Add coefficient extractor to docs * Quicker spectrum in 1D * More validation (#39) * Add note * Add 2d and 3d advection * Add diffusion analytical solution in 2D * Add diffusion test in 3D * Ensure all presets are working * Employ a fix for single facets * Use Exponax viz routines * Also employ fix for faceted animations * Fix defaults for Kolmogorov * Change to exponax viz routines * Add 3D dynamics * Add hint on where to find the reference qualitative rollouts * Re-Execute Notebook * Re-Execute Notebook * Re-Execute notebook * Start with a validation notebook for KdV * Test if KdV without convection is purely linear * Finished KdV soliton comparison * Faster way to compute the spectrum binning in higher dimensions * Improved version of kolmogorov comparison * Improved Metrics System (#40) * Start implementing metrics tests * Starting a notebook on validating the metrics * [WIP] Start generalizing the fourier nRMSE to allow for derivative based losses * Set up a new structure for metrics * New export structure * Overwrite the existing metrics submodule * Adapt tests because domain_extent is now considered correctly * Test against analytical solution * Fix parameterization of test * Start set up with a new fourier metrics implementation * Ensure correct shape * Add first test on Fourier aggregation * Add basic abolute metrics, more TODO * Port the correlation metric * New API in docs * Port new spatial metrics to API docs * Rework correlation documentation * Add drafts on an advanced background notebook on metrics and their relation to functional norms * Fix the scaling in the Fourier aggregation * Use pytest approx * Fix scaling * Add equivalence test of Fourier and spatial aggregation * Rework implementations * Remove mean correlation * Add soboleb based metrics * Add helper function to apply metric functions on batched states * Test with non-unit domain * Start with a low-level introduction to metrics * Add basics notebook * Adapt API docs * Add to API docs * Add documentation to spatial metrics * Reference does not be zero in normalized metrics * Add documentation * Fix documentation * Test Sobolev against manual adding of derivatives with fourier-base metrics * Fix typo * Add documentation * Add tests on Filtering * Add notes on preliminary notebook * Fix broken links * Also add warning to advanced metrics notebook * Continue with the metric convergence notebook, still WIP * Add example on Sobolev usage * Finish validation notebook * Add temporal convergence notebook (#41) * Add warning for KS conservative in higher dimensions * Symmetric Metrics & Metrics improvements (#44) * Extend docstring for symmetric metrics * Improve aggregator docstring * Fix that identity is only guaranteed for L^p norm * Fix docstring * Implement symmetric mode * Fix missing sqrt * Add symmetric versions for MAE, MSE, and RMSE * Add new symmetric metrics to documentation * Add test for symmetric metrics * Ensure mean is only taken over axis zero in case there are additional axes * Add details * Add non-conservative convection term (#46) * add non-conservative convection term * fix errors in docstring * update docstring * use the dealiased value for non-conservative convection * Fix non-forwarded conservative flag * Small Improvements to Gaussian Random Field (#48) * Retain mean mode from white noise * Add comment on why there is a 2.0 in denominator * Polynomial Stepper Fix (#49) * Have general diffusion coefficient consisten with norm/diff * Use correct sign on quadratic nonlinearity * Conservative KS should use conservative convection (#50) * Conservative KS should use conservative convection * Adapt tests * Consistent Generic Steppers (#52) * Use consistent attribute names * Forward API changes to tests * Forward changes to qualitative rollout * Ensure single-channel convection nonlin is conservative (#53) * Rework KS Attributes (#54) * Remove name of diffusivities from combustion KS * Remove name of diffusivities from conservative KS * Foward changes to tests * Fix missing factor of 0.5 * Rework 1D example notebook * Docs tweak (#55) * Change to proper markdown * Add KS conservative docstring * Re-Execute IC notebook * Rerun general and normalized notebookt * Rerun subclassing example * Rerun showcase 2D * Add solver overview in 3D * Add 3D Solver overview to docs * Automatically deduce version number --------- Co-authored-by: Qiang <137192063+qiauil@users.noreply.github.com>
Ceyron · Oct 23, 2024 · 56becef · 56becef
1 parent 8959079
commit 56becef
Show file tree

Hide file tree

Showing 160 changed files with 78,534 additions and 49,822 deletions.
diff --git a/.github/workflows/pre_commit.yml b/.github/workflows/pre_commit.yml
@@ -6,6 +6,7 @@ on:
   push:
     branches:
       - main
+      - dev
 
 jobs:
   pre-commit:

diff --git a/.github/workflows/run_tests.yml b/.github/workflows/run_tests.yml
@@ -0,0 +1,34 @@
+name: Run tests
+
+on:
+  pull_request:
+
+jobs:
+  run-test:
+    strategy:
+      matrix:
+        python-version: [ "3.10", "3.11", "3.12" ]
+        os: [ ubuntu-latest ]
+      fail-fast: false
+    runs-on: ${{ matrix.os }}
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v2
+
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v2
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install -r ./tests/requirements.txt
+
+      - name: Checks with pre-commit
+        uses: pre-commit/action@v2.0.3
+
+      - name: Test with pytest
+        run: |
+          python -m pip install .
+          python -m pytest tests/
diff --git a/.gitignore copy b/.gitignore copy
diff --git a/docs/api/stepper/physical/poisson.md → docs/api/stepper/additional/poisson.md b/docs/api/stepper/physical/poisson.md → docs/api/stepper/additional/poisson.md
diff --git a/docs/api/stepper/difficulty/convection.md → ...neric/difficulty/difficulty_convection.md b/docs/api/stepper/difficulty/convection.md → ...neric/difficulty/difficulty_convection.md
@@ -1,6 +1,6 @@
 # Convection
 
-::: exponax.normalized.DifficultyConvectionStepper
+::: exponax.stepper.generic.DifficultyConvectionStepper
     options:
         members:
             - __init__

diff --git a/docs/api/stepper/difficulty/gradient_norm.md → ...ic/difficulty/difficulty_gradient_norm.md b/docs/api/stepper/difficulty/gradient_norm.md → ...ic/difficulty/difficulty_gradient_norm.md
@@ -1,6 +1,6 @@
 # Gradient Norm
 
-::: exponax.normalized.DifficultyGradientNormStepper  
+::: exponax.stepper.generic.DifficultyGradientNormStepper  
     options:
         members:
             - __init__

diff --git a/docs/api/stepper/difficulty/linear.md → ...r/generic/difficulty/difficulty_linear.md b/docs/api/stepper/difficulty/linear.md → ...r/generic/difficulty/difficulty_linear.md
@@ -1,6 +1,6 @@
 # Linear
 
-::: exponax.normalized.DifficultyLinearStepper  
+::: exponax.stepper.generic.DifficultyLinearStepper  
     options:
         members:
             - __init__

diff --git a/docs/api/stepper/difficulty/nonlinear.md → ...eneric/difficulty/difficulty_nonlinear.md b/docs/api/stepper/difficulty/nonlinear.md → ...eneric/difficulty/difficulty_nonlinear.md
@@ -1,6 +1,6 @@
 # Nonlinear
 
-::: exponax.normalized.DifficultyGeneralNonlinearStepper
+::: exponax.stepper.generic.DifficultyNonlinearStepper
     options:
         members:
             - __init__

diff --git a/docs/api/stepper/difficulty/polynomial.md → ...neric/difficulty/difficulty_polynomial.md b/docs/api/stepper/difficulty/polynomial.md → ...neric/difficulty/difficulty_polynomial.md
@@ -1,6 +1,6 @@
 # Polynomial
 
-::: exponax.normalized.DifficultyPolynomialStepper  
+::: exponax.stepper.generic.DifficultyPolynomialStepper  
     options:
         members:
             - __init__

diff --git a/docs/api/stepper/normalized/convection.md → ...neric/normalized/normalized_convection.md b/docs/api/stepper/normalized/convection.md → ...neric/normalized/normalized_convection.md
@@ -1,6 +1,6 @@
 # Convection
 
-::: exponax.normalized.NormalizedConvectionStepper  
+::: exponax.stepper.generic.NormalizedConvectionStepper  
     options:
         members:
             - __init__

diff --git a/docs/api/stepper/normalized/gradient_norm.md → ...ic/normalized/normalized_gradient_norm.md b/docs/api/stepper/normalized/gradient_norm.md → ...ic/normalized/normalized_gradient_norm.md
@@ -1,6 +1,6 @@
 # Gradient NormA
 
-::: exponax.normalized.NormalizedGradientNormStepper  
+::: exponax.stepper.generic.NormalizedGradientNormStepper  
     options:
         members:
             - __init__

diff --git a/docs/api/stepper/normalized/linear.md → ...r/generic/normalized/normalized_linear.md b/docs/api/stepper/normalized/linear.md → ...r/generic/normalized/normalized_linear.md
@@ -1,5 +1,5 @@
 
-::: exponax.normalized.NormalizedLinearStepper
+::: exponax.stepper.generic.NormalizedLinearStepper
     options:
         members:
             - __init__

diff --git a/docs/api/stepper/normalized/nonlinear.md → ...eneric/normalized/normalized_nonlinear.md b/docs/api/stepper/normalized/nonlinear.md → ...eneric/normalized/normalized_nonlinear.md
@@ -1,6 +1,6 @@
 # Nonlinear
 
-::: exponax.normalized.NormalizedGeneralNonlinearStepper  
+::: exponax.stepper.generic.NormalizedNonlinearStepper  
     options:
         members:
             - __init__

diff --git a/docs/api/stepper/normalized/polynomial.md → ...neric/normalized/normalized_polynomial.md b/docs/api/stepper/normalized/polynomial.md → ...neric/normalized/normalized_polynomial.md
@@ -1,6 +1,6 @@
 # Polynomial
 
-::: exponax.normalized.NormalizedPolynomialStepper  
+::: exponax.stepper.generic.NormalizedPolynomialStepper  
     options:
         members:
             - __init__

diff --git a/...er/physical/general/general_convection.md → ...er/generic/physical/general_convection.md b/...er/physical/general/general_convection.md → ...er/generic/physical/general_convection.md
@@ -1,6 +1,6 @@
 # General Convection Stepper
 
-::: exponax.stepper.GeneralConvectionStepper
+::: exponax.stepper.generic.GeneralConvectionStepper
     options:
         members:
             - __init__

diff --git a/...physical/general/general_gradient_norm.md → ...generic/physical/general_gradient_norm.md b/...physical/general/general_gradient_norm.md → ...generic/physical/general_gradient_norm.md
@@ -1,6 +1,6 @@
 # General Gradient Norm Stepper
 
-::: exponax.stepper.GeneralGradientNormStepper
+::: exponax.stepper.generic.GeneralGradientNormStepper
     options:
         members:
             - __init__

diff --git a/...tepper/physical/general/general_linear.md → ...tepper/generic/physical/general_linear.md b/...tepper/physical/general/general_linear.md → ...tepper/generic/physical/general_linear.md
@@ -1,6 +1,6 @@
 # General Linear Stepper
 
-::: exponax.stepper.GeneralLinearStepper
+::: exponax.stepper.generic.GeneralLinearStepper
     options:
         members:
             - __init__

diff --git a/...per/physical/general/general_nonlinear.md → ...per/generic/physical/general_nonlinear.md b/...per/physical/general/general_nonlinear.md → ...per/generic/physical/general_nonlinear.md
@@ -1,6 +1,6 @@
 # General Nonlinear Stepper
 
-::: exponax.stepper.GeneralNonlinearStepper
+::: exponax.stepper.generic.GeneralNonlinearStepper
     options:
         members:
             - __init__

diff --git a/...er/physical/general/general_polynomial.md → ...er/generic/physical/general_polynomial.md b/...er/physical/general/general_polynomial.md → ...er/generic/physical/general_polynomial.md
@@ -1,6 +1,6 @@
 # General Polynomial Stepper
 
-::: exponax.stepper.GeneralPolynomialStepper
+::: exponax.stepper.generic.GeneralPolynomialStepper
     options:
         members:
             - __init__

diff --git a/...l/general/general_vorticity_convection.md → .../physical/general_vorticity_convection.md b/...l/general/general_vorticity_convection.md → .../physical/general_vorticity_convection.md
@@ -1,6 +1,6 @@
 # General Vorticity Convection Stepper
 
-::: exponax.stepper.GeneralVorticityConvectionStepper
+::: exponax.stepper.generic.GeneralVorticityConvectionStepper
     options:
         members:
             - __init__

diff --git a/.../api/stepper/physical/linear/advection.md → docs/api/stepper/linear/advection.md b/.../api/stepper/physical/linear/advection.md → docs/api/stepper/linear/advection.md
diff --git a/...er/physical/linear/advection_diffusion.md → ...api/stepper/linear/advection_diffusion.md b/...er/physical/linear/advection_diffusion.md → ...api/stepper/linear/advection_diffusion.md
diff --git a/.../api/stepper/physical/linear/diffusion.md → docs/api/stepper/linear/diffusion.md b/.../api/stepper/physical/linear/diffusion.md → docs/api/stepper/linear/diffusion.md
diff --git a/...api/stepper/physical/linear/dispersion.md → docs/api/stepper/linear/dispersion.md b/...api/stepper/physical/linear/dispersion.md → docs/api/stepper/linear/dispersion.md
diff --git a/...tepper/physical/linear/hyper_diffusion.md → docs/api/stepper/linear/hyper_diffusion.md b/...tepper/physical/linear/hyper_diffusion.md → docs/api/stepper/linear/hyper_diffusion.md
diff --git a/docs/api/stepper/physical/burgers.md → docs/api/stepper/nonlinear/burgers.md b/docs/api/stepper/physical/burgers.md → docs/api/stepper/nonlinear/burgers.md
diff --git a/docs/api/stepper/physical/kdv.md → docs/api/stepper/nonlinear/kdv.md b/docs/api/stepper/physical/kdv.md → docs/api/stepper/nonlinear/kdv.md
diff --git a/docs/api/stepper/physical/ks.md → docs/api/stepper/nonlinear/ks.md b/docs/api/stepper/physical/ks.md → docs/api/stepper/nonlinear/ks.md
diff --git a/docs/api/stepper/physical/ks_cons.md → docs/api/stepper/nonlinear/ks_cons.md b/docs/api/stepper/physical/ks_cons.md → docs/api/stepper/nonlinear/ks_cons.md
diff --git a/docs/api/stepper/physical/navier_stokes.md → docs/api/stepper/nonlinear/navier_stokes.md b/docs/api/stepper/physical/navier_stokes.md → docs/api/stepper/nonlinear/navier_stokes.md
diff --git a/docs/api/stepper/normalized/vorticity_convection.md b/docs/api/stepper/normalized/vorticity_convection.md
diff --git a/docs/api/stepper/overview.md b/docs/api/stepper/overview.md
@@ -0,0 +1,23 @@
+# Overview
+
+`Exponax` comes with many pre-built dynamics. They are all based on the idea of
+solving semi-linear PDEs
+
+$$
+\partial_t u = \mathcal{L} u + \mathcal{N}(u)
+$$
+
+where $\mathcal{L}$ is a linear operator and $\mathcal{N}$ is a non-linear
+differential operator.
+
+TODO: add classification of dynamics based on:
+
+- linear or nonlinear or reaction-diffusion
+- specific (i.e. concrete equations like advection, Burgers etc.) or generic
+  based on providing a collection of linear and nonlinear coefficients
+- if generic: whether it uses a physical, normalized or difficulty-based
+  interface
+- in higher dimensions: whether it behaves isotropic or anisotropic (with
+  different coefficients per direction)
+
+Then the resulting behavior can also be classified into ...
diff --git a/...physical/reaction_diffusion/allen_cahn.md → docs/api/stepper/reaction/allen_cahn.md b/...physical/reaction_diffusion/allen_cahn.md → docs/api/stepper/reaction/allen_cahn.md
@@ -1,6 +1,6 @@
 # Allen-Cahn
 
-::: exponax.reaction.AllenCahn
+::: exponax.stepper.reaction.AllenCahn
     options:
         members:
             - __init__

diff --git a/...sical/reaction_diffusion/cahn_hilliard.md → docs/api/stepper/reaction/cahn_hilliard.md b/...sical/reaction_diffusion/cahn_hilliard.md → docs/api/stepper/reaction/cahn_hilliard.md
@@ -1,6 +1,6 @@
 # Cahn-Hilliad
 
-::: exponax.reaction.CahnHilliard
+::: exponax.stepper.reaction.CahnHilliard
     options:
         members:
             - __init__

diff --git a/...physical/reaction_diffusion/fisher_kpp.md → docs/api/stepper/reaction/fisher_kpp.md b/...physical/reaction_diffusion/fisher_kpp.md → docs/api/stepper/reaction/fisher_kpp.md
@@ -1,6 +1,6 @@
 # Fisher-KPP
 
-::: exponax.reaction.FisherKPP
+::: exponax.stepper.reaction.FisherKPP
     options:
         members:
             - __init__

diff --git a/...physical/reaction_diffusion/gray_scott.md → docs/api/stepper/reaction/gray_scott.md b/...physical/reaction_diffusion/gray_scott.md → docs/api/stepper/reaction/gray_scott.md
@@ -1,6 +1,6 @@
 # Gray-Scott
 
-::: exponax.reaction.GrayScott
+::: exponax.stepper.reaction.GrayScott
     options:
         members:
             - __init__
-Original file line number
+Diff line change
@@ Expand Up / @@ -6,6 +6,7 @@ on: @@
       push:
         branches:
           - main
+          - dev
     jobs:
       pre-commit:
@@ Expand Down @@