vindy package

Subpackages

• vindy.callbacks package
  • Submodules
  • vindy.callbacks.pdf_threshold_callback module
    • PDFThresholdCallback
      • PDFThresholdCallback.__init__()
      • PDFThresholdCallback.cancel_coefficients()
      • PDFThresholdCallback.on_epoch_end()
      • PDFThresholdCallback.on_train_end()
  • vindy.callbacks.save_coefficients_callback module
    • SaveCoefficientsCallback
      • SaveCoefficientsCallback.__init__()
      • SaveCoefficientsCallback.on_epoch_end()
  • vindy.callbacks.spotify_callback module
    • SpotifyCallback
      • SpotifyCallback.on_epoch_end()
  • vindy.callbacks.threshold_prune_callback module
    • ThresholdPruneCallback
      • ThresholdPruneCallback.__init__()
      • ThresholdPruneCallback.on_epoch_end()
      • ThresholdPruneCallback.on_train_end()
      • ThresholdPruneCallback.prune_weights()
  • Module contents
• vindy.distributions package
  • Submodules
  • vindy.distributions.base_distribution module
    • BaseDistribution
      • BaseDistribution.call()
      • BaseDistribution.KL_divergence()
      • BaseDistribution.call()
      • BaseDistribution.plot()
      • BaseDistribution.prob_density_fcn()
      • BaseDistribution.reverse_log()
      • BaseDistribution.variance()
  • vindy.distributions.gaussian module
    • Gaussian
      • Gaussian.KL_divergence()
      • Gaussian.__init__()
      • Gaussian.call()
      • Gaussian.log_var_to_deviation()
      • Gaussian.prob_density_fcn()
      • Gaussian.variance()
      • Gaussian.variance_to_log_scale()
  • vindy.distributions.laplace module
    • Laplace
      • Laplace.KL_divergence()
      • Laplace.__init__()
      • Laplace.call()
      • Laplace.prob_density_fcn()
      • Laplace.variance()
      • Laplace.variance_to_log_scale()
  • Module contents
• vindy.layers package
  • Submodules
  • vindy.layers.sindy_layer module
    • SindyLayer
      • SindyLayer.__init__()
      • SindyLayer.assert_arguments()
      • SindyLayer.call()
      • SindyLayer.coefficient_matrix_shape
      • SindyLayer.concat_features()
      • SindyLayer.features()
      • SindyLayer.fill_coefficient_matrix()
      • SindyLayer.get_feature_names()
      • SindyLayer.get_prunable_weights()
      • SindyLayer.get_sindy_coeffs()
      • SindyLayer.init_weigths()
      • SindyLayer.integrate()
      • SindyLayer.kernel_shape
      • SindyLayer.loss_trackers
      • SindyLayer.model_equation_to_str()
      • SindyLayer.print()
      • SindyLayer.prune_weights()
      • SindyLayer.rhs_()
      • SindyLayer.set_mask()
  • vindy.layers.vindy_layer module
    • LogVarL1L2Regularizer
    • VindyLayer
      • VindyLayer.__init__()
      • VindyLayer.assert_additional_args()
      • VindyLayer.call()
      • VindyLayer.call_uq()
      • VindyLayer.get_sindy_coeffs()
      • VindyLayer.init_weigths()
      • VindyLayer.integrate_uq()
      • VindyLayer.kl_loss()
      • VindyLayer.loss_trackers
      • VindyLayer.pdf_thresholding()
      • VindyLayer.visualize_coefficients()
  • Module contents
• vindy.libraries package
  • Submodules
  • vindy.libraries.base_library module
    • BaseLibrary
      • BaseLibrary.get_names()
  • vindy.libraries.exponential_library module
    • ExponentialLibrary
      • ExponentialLibrary.get_names()
  • vindy.libraries.force_library module
    • ForceLibrary
      • ForceLibrary.get_names()
  • vindy.libraries.fourier_library module
    • FourierLibrary
      • FourierLibrary.get_names()
  • vindy.libraries.polynomial_library module
    • PolynomialLibrary
      • PolynomialLibrary.__init__()
      • PolynomialLibrary.get_names()
      • PolynomialLibrary.loop_rec()
      • PolynomialLibrary.loop_rec_names()
      • PolynomialLibrary.poly_higher_order()
  • Module contents
• vindy.networks package
  • Submodules
  • vindy.networks.autoencoder_sindy module
    • AutoencoderSindy
      • AutoencoderSindy.__init__()
      • AutoencoderSindy.assert_arguments()
      • AutoencoderSindy.build_decoder()
      • AutoencoderSindy.build_encoder()
      • AutoencoderSindy.build_loss()
      • AutoencoderSindy.build_model()
      • AutoencoderSindy.calc_latent_time_derivatives()
      • AutoencoderSindy.compile()
      • AutoencoderSindy.create_loss_trackers()
      • AutoencoderSindy.decode()
      • AutoencoderSindy.encode()
      • AutoencoderSindy.get_loss()
      • AutoencoderSindy.get_loss_2nd()
      • AutoencoderSindy.get_loss_rec()
      • AutoencoderSindy.get_trainable_weights()
      • AutoencoderSindy.reconstruct()
      • AutoencoderSindy.reconstruction_loss()
  • vindy.networks.base_model module
    • BaseModel
      • BaseModel.assert_arguments()
      • BaseModel.build_sindy()
      • BaseModel.concatenate_sindy_input()
      • BaseModel.define_scaling()
      • BaseModel.evaluate_sindy_layer()
      • BaseModel.fit()
      • BaseModel.flatten3d()
      • BaseModel.flatten_dummy()
      • BaseModel.get_int_loss()
      • BaseModel.integrate()
      • BaseModel.load()
      • BaseModel.print()
      • BaseModel.rescale()
      • BaseModel.save()
      • BaseModel.scale()
      • BaseModel.sindy_coeffs()
      • BaseModel.split_inputs()
      • BaseModel.test_step()
      • BaseModel.train_step()
      • BaseModel.unflatten3d()
      • BaseModel.vis_modes()
  • vindy.networks.identification_network module
    • IdentificationNetwork
      • IdentificationNetwork.__init__()
      • IdentificationNetwork.build_loss()
      • IdentificationNetwork.build_model()
      • IdentificationNetwork.create_loss_trackers()
      • IdentificationNetwork.get_loss()
      • IdentificationNetwork.get_loss_2nd()
      • IdentificationNetwork.get_trainable_weights()
  • vindy.networks.sindy_network module
  • vindy.networks.variational_autoencoder_sindy module
  • vindy.networks.veni module
    • VENI
      • VENI.build_encoder()
      • VENI.call()
      • VENI.create_loss_trackers()
      • VENI.encode()
      • VENI.kl_loss()
      • VENI.reconstruction_loss()
  • Module contents
• vindy.utils package
  • Submodules
  • vindy.utils.utils module
    • add_lognormal_noise()
    • coefficient_distribution_gif()
    • coefficient_distributions_to_csv()
    • create_result_directory()
    • get_config()
    • get_latent_initial_conditions()
    • log_model_summary()
    • perform_forward_uq()
    • perform_inference()
    • plot_coefficients_train_history()
    • plot_inference_results()
    • plot_train_history()
    • set_seed()
    • switch_data_format()
    • uq_plots()
    • validate_data_path()
  • Module contents

Module contents

VENI VINDy VICI (vindy) package.

High-level package description.

This package provides tools for data-driven surrogate modeling, callbacks, distributions, layers and network architectures used for system identification and discovery of governing equations.

Modules

networks

Neural network architectures for system identification and variational-autoencoder based models.

callbacks

Training callbacks for monitoring and logging.

distributions

Probability distributions used in the variational encoder/decoder.

layers

Custom Keras layers for SINDy and related operations.

Notes

Docstrings in the project are written in NumPy style and parsed by Sphinx using the napoleon extension.