Cetaceo documentation

Add your content using reStructuredText syntax. See the reStructuredText documentation for details.

Contents

• data package
  • Submodules
  • data.HDF5Reader module
    • HDF5Dataset
    • HDF5Reader
      • HDF5Reader.count_datasets()
      • HDF5Reader.load_data_to_tensor()
  • data.TorchReader module
    • TorchDataset
      • TorchDataset.rescale_data()
      • TorchDataset.scale_data()
  • data.VTUReader module
    • VTUDataset
      • VTUDataset.load_mesh()
    • VTUReader
      • VTUReader.get_cell_data()
      • VTUReader.get_centers_coordinates()
      • VTUReader.get_data_names()
      • VTUReader.get_groupID()
  • data.dataset module
    • BaseDataset
      • BaseDataset.isscaled
      • BaseDataset.process_data()
      • BaseDataset.rescale_data()
      • BaseDataset.rescale_x()
      • BaseDataset.rescale_y()
      • BaseDataset.scale_data()
  • Module contents
• evaluators package
  • Submodules
  • evaluators.evaluator module
    • Evaluator
      • Evaluator.print_metrics()
  • evaluators.regression_evaluator module
    • RegressionEvaluator
      • RegressionEvaluator.R2()
      • RegressionEvaluator.ae_q()
      • RegressionEvaluator.l2_error()
      • RegressionEvaluator.mean_absolute_error()
      • RegressionEvaluator.mean_relative_error()
      • RegressionEvaluator.mean_squared_error()
      • RegressionEvaluator.print_metrics()
      • RegressionEvaluator.tolerance
  • evaluators.regression_evaluator_plotter module
    • RegressionEvaluatorPlotter
  • Module contents
• models package
  • Subpackages
    • models.pinn package
      • Submodules
      • models.pinn.boundary_condition module
      • models.pinn.pinn module
      • models.pinn.piratenet module
      • models.pinn.simple_mlp module
      • Module contents
  • Submodules
  • models.KANSINupm_v1 module
    • KANSIN
      • KANSIN.define_checkpoint()
    • SineLayer
      • SineLayer.forward()
      • SineLayer.init_weights()
  • models.KAN_v4_1 module
    • ChebyshevLayer
      • ChebyshevLayer.forward()
    • JacobiLayer
      • JacobiLayer.forward()
    • KAN
      • KAN.create_optimized_model()
      • KAN.fit()
      • KAN.forward()
      • KAN.load()
      • KAN.predict()
      • KAN.save()
    • TaylorLayer
      • TaylorLayer.forward()
  • models.KANupm_v4 module
    • ChebyLayer
      • ChebyLayer.forward()
    • ChebyLayer_ant
      • ChebyLayer_ant.forward()
    • ChebyLayer_v2
      • ChebyLayer_v2.forward()
    • JacobiLayer
      • JacobiLayer.forward()
    • KANupm
      • KANupm.exam()
      • KANupm.fit()
      • KANupm.forward()
      • KANupm.load()
      • KANupm.predict()
      • KANupm.save()
    • MakeDataset_kan
    • TaylorLayer
      • TaylorLayer.forward()
  • models.KANupm_v4_1 module
    • ChebyLayer_ant
      • ChebyLayer_ant.forward()
    • ChebyLayer_v2
      • ChebyLayer_v2.forward()
    • ChebyLayer_v3
      • ChebyLayer_v3.forward()
    • JacobiLayer
      • JacobiLayer.forward()
    • KANupm
      • KANupm.exam()
      • KANupm.fit()
      • KANupm.forward()
      • KANupm.load()
      • KANupm.predict()
      • KANupm.print_hours()
      • KANupm.save()
    • MakeDataset_kan
    • TaylorLayer
      • TaylorLayer.forward()
  • models.KANupm_v4_2 module
    • ChebyLayer_ant
      • ChebyLayer_ant.forward()
    • ChebyLayer_v2
      • ChebyLayer_v2.forward()
    • ChebyLayer_v3
      • ChebyLayer_v3.forward()
    • JacobiLayer
      • JacobiLayer.forward()
    • KANupm
      • KANupm.exam()
      • KANupm.fit()
      • KANupm.forward()
      • KANupm.load()
      • KANupm.predict()
      • KANupm.print_hours()
      • KANupm.save()
    • MakeDataset_kan
    • TaylorLayer
      • TaylorLayer.forward()
  • models.KANupm_v6_1 module
    • ChebyLayer_ant
      • ChebyLayer_ant.forward()
    • ChebyLayer_v2
      • ChebyLayer_v2.forward()
    • ChebyLayer_v3
      • ChebyLayer_v3.forward()
    • JacobiLayer
      • JacobiLayer.forward()
    • KANupm
      • KANupm.count_trainable_params()
      • KANupm.exam()
      • KANupm.fit()
      • KANupm.fit_scored()
      • KANupm.forward()
      • KANupm.get_weights()
      • KANupm.load()
      • KANupm.plot_structure()
      • KANupm.predict()
      • KANupm.print_hours()
      • KANupm.save()
    • MakeDatasetScored_kan
    • MakeDataset_kan
    • TaylorLayer
      • TaylorLayer.forward()
    • WeightedMSELoss
      • WeightedMSELoss.forward()
  • models.KANupm_v8 module
    • ChebyLayer_ant
      • ChebyLayer_ant.forward()
    • ChebyLayer_v2
      • ChebyLayer_v2.forward()
    • ChebyLayer_v3
      • ChebyLayer_v3.forward()
    • JacobiLayer
      • JacobiLayer.forward()
    • KANupm
      • KANupm.count_trainable_params()
      • KANupm.define_checkpoint()
      • KANupm.exam()
      • KANupm.fit()
      • KANupm.fit_scored()
      • KANupm.forward()
      • KANupm.get_weights()
      • KANupm.load()
      • KANupm.plot_structure()
      • KANupm.predict()
      • KANupm.print_hours()
      • KANupm.save()
    • MakeDatasetScored_kan
    • MakeDatasetScored_kan_ant
    • MakeDataset_kan
    • NormPLoss
      • NormPLoss.forward()
    • SineLayer
      • SineLayer.forward()
      • SineLayer.init_weights()
    • StatisticalLoss
      • StatisticalLoss.forward()
    • TaylorLayer
      • TaylorLayer.forward()
  • models.VAE module
    • VAE
      • VAE.decode()
      • VAE.encode()
      • VAE.forward()
      • VAE.loss_function()
      • VAE.reparameterize()
      • VAE.sample()
  • models.elm module
    • ELM
      • ELM.fit()
    • LocELM
      • LocELM.fit()
  • models.mlp module
    • MLP
      • MLP.create_optimized_model()
      • MLP.fit()
      • MLP.forward()
      • MLP.load()
      • MLP.predict()
      • MLP.save()
  • models.model_interface module
    • Model
      • Model.create_optimized_model()
      • Model.fit()
      • Model.load()
      • Model.predict()
      • Model.save()
  • Module contents
• optimization package
  • Submodules
  • optimization.optimizer_interface module
    • OptunaOptimizer
      • OptunaOptimizer.optimize()
  • optimization.optuna_base_optimizer module
    • OptunaBaseOptimizer
      • OptunaBaseOptimizer.optimization_params
      • OptunaBaseOptimizer.optimize()
  • Module contents
• pipeline module
  • Pipeline
    • Pipeline.evaluate()
    • Pipeline.model
    • Pipeline.run()
• plotting package
  • Submodules
  • plotting.metrics_study module
  • plotting.plot_2d_mesh module
    • plot_mesh_2d()
    • plot_mesh_3d_interactive()
  • plotting.plotter module
    • Plotter
      • Plotter.plot()
  • plotting.true_vs_pred_plot module
    • TrueVsPredPlotter
      • TrueVsPredPlotter.plot()
  • Module contents
• utils package
  • Submodules
  • utils.Metadata module
    • Metadata
      • Metadata.add_nested_field()
      • Metadata.add_repeating_block()
      • Metadata.generate_next_model_name()
      • Metadata.save_to_csv()
      • Metadata.set_field()
  • utils.aerodynamics module
    • lift_drag_coeff()
  • utils.lift-drag module
  • utils.nastran_reader module
  • utils.path_manager module
    • PathManager
      • PathManager.create_directory()
  • utils.setseed module
    • set_seed()
  • utils.splitting module
  • utils.zone_metrics module
    • zone_metrics_calculation()
    • zone_metrics_plotter()
  • Module contents

---

Examples

• Examples
  • How to train an MLP using the Pipeline
    • Import classes and define paths
    • Define sklearn scalers if needed
    • Create datasets
    • Evaluators and Plotters
    • Model creation
    • Run the pipeline
  • How to optimize with optuna an MLP using the Pipeline
    • Import classes and define paths
    • Define sklearn scalers if needed
    • Create datasets
    • Evaluator
    • Optimization
    • Run the pipeline
  • Train a PINN (Physics Informed Neural Network) with a custom PDE
    • Define the collocation points
    • Define the PDE
    • Define the boundary conditions
      • Dataset creation
    • Train the pinn
      • The model is retrained with an LBFGS optimizer to improve its acuraccy
    • Make the predictions and plot the results
      • Evaluation using RegressionEvaluator
    • Save and load the model
  • Train a PINN (Physics Informed Neural Network) Shrodinger equation
    • Define BC points and Collocation points
    • Define PDE loss function
    • Define Boundary conditions
    • Create Dataset
    • PINN and NN definiton
    • PINN training with Adam optimizer
    • PINN training with LBFGS optimizer
    • Save model

---

Comparison RL vs PSO

• Airfoil Optimization with PSO
• Airfoil Optimization with RL
• Airfoil Optimization with PSO preserving maximum thickness
• Airfoil Optimization with RL preserving maximum thickness