Skip to content

Conversation

@benedict-96
Copy link
Collaborator

Implemented PGHNNs.

Things that should probably be done before this pr is merged:

  • get rid of type piracy (see applychain, the method for functor of Chain and Jacobian) in file generalized_hamiltonian_neural_network.jl,
  • define what is QPT2 in AbstractNeuralNetworks and ultimately get rid of what QPT is now,
  • extend the DataLoader to the parameter-dependent scenario.

…ral_network.jl (ii) created generalized_hamiltonian_neural_network where I defined energy functions (iii) moved basic functionality form standard_hamiltonian_neural_network.jl to new file hamiltonian_neural_network.jl.
@codecov
Copy link

codecov bot commented Jul 15, 2025

Codecov Report

❌ Patch coverage is 41.04167% with 283 lines in your changes missing coverage. Please review.
✅ Project coverage is 59.42%. Comparing base (7475500) to head (7b10ae2).

Files with missing lines Patch % Lines
src/layers/forcing_dissipation_layers.jl 0.00% 52 Missing ⚠️
src/pullbacks/symbolic_hnn_pullback.jl 9.80% 46 Missing ⚠️
...tectures/generalized_hamiltonian_neural_network.jl 64.84% 45 Missing ⚠️
src/layers/parametric_resnet_layer.jl 0.00% 35 Missing ⚠️
src/layers/wide_resnet.jl 0.00% 21 Missing ⚠️
src/data_loader/parametric_data_loader.jl 70.00% 18 Missing ⚠️
src/architectures/parametric_resnet.jl 0.00% 16 Missing ⚠️
...s/forced_generalized_hamiltonian_neural_network.jl 0.00% 15 Missing ⚠️
src/architectures/forced_sympnet.jl 0.00% 13 Missing ⚠️
src/optimizers/optimizer.jl 11.11% 8 Missing ⚠️
... and 4 more
Additional details and impacted files
@@            Coverage Diff             @@
##             main     #207      +/-   ##
==========================================
- Coverage   61.82%   59.42%   -2.40%     
==========================================
  Files         124      132       +8     
  Lines        3790     4222     +432     
==========================================
+ Hits         2343     2509     +166     
- Misses       1447     1713     +266     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

…s as default when network was agnostic to Q/P.
… sublayers in a forcing layers, which should be independent of the total number of layers.
@benedict-96 benedict-96 removed the request for review from michakraus November 29, 2025 10:43
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

Development

Successfully merging this pull request may close these issues.

2 participants