OCDocker.OCScore.Optimization.future.DNN module

Module with a helper to perform the optimization of the future DNN pipeline.

It is imported as:

import OCDocker.OCScore.Optimization.future.DNN as ocdnn_future

OCDocker.OCScore.Optimization.future.DNN.optimize_NN_future(df_path, storage_id, base_models_folder, data={}, storage='sqlite:///NN_optimization.db', use_pdb_train=True, no_scores=False, only_scores=False, use_PCA=False, best_ao_params=None, pca_type=80, pca_model='', encoder_dims=(16, 256), autoencoder=True, multiencoder=False, run_autoencoder_optimization=True, num_processes_autoencoder=8, total_trials_autoencoder=2000, run_NN_optimization=True, num_processes_NN=4, total_trials_NN=50, explained_variance=0.95, random_seed=42, load_if_exists=True, use_gpu=True, parallel_backend='joblib', verbose=False, use_future=True, future_config=None)[source]

Optimize the future DNN pipeline.

Parameters:

  • df_path (str) – Path to the dataset.

  • storage_id (int) – Storage ID.

  • base_models_folder (str) – Base folder for models.

  • data (dict, optional) – Preloaded data dict (ignored in future pipeline if empty).

  • use_future (bool, optional) – If False, fallback to the current pipeline. Default True.

  • storage (str) –

  • use_pdb_train (bool) –

  • no_scores (bool) –

  • only_scores (bool) –

  • use_PCA (bool) –

  • best_ao_params (dict | None) –

  • pca_type (int) –

  • pca_model (str | PCA) –

  • encoder_dims (tuple[int, int]) –

  • autoencoder (bool) –

  • multiencoder (bool) –

  • run_autoencoder_optimization (bool) –

  • num_processes_autoencoder (int) –

  • total_trials_autoencoder (int) –

  • run_NN_optimization (bool) –

  • num_processes_NN (int) –

  • total_trials_NN (int) –

  • explained_variance (float) –

  • random_seed (int) –

  • load_if_exists (bool) –

  • use_gpu (bool) –

  • parallel_backend (str) –

  • verbose (bool) –

  • future_config (dict | None) –

Return type:

None

Notes

Example usage:

optimize_NN_future(df_path, 1, “./models”, total_trials_NN=20)

OCDocker.OCScore.Optimization.future.DNN.optimize(df_path, storage_id, base_models_folder, data={}, storage='sqlite:///NN_optimization.db', use_pdb_train=True, no_scores=False, only_scores=False, use_PCA=False, best_ao_params=None, pca_type=80, pca_model='', encoder_dims=(16, 256), autoencoder=True, multiencoder=False, run_autoencoder_optimization=True, num_processes_autoencoder=8, total_trials_autoencoder=2000, run_NN_optimization=True, num_processes_NN=4, total_trials_NN=50, explained_variance=0.95, random_seed=42, load_if_exists=True, use_gpu=True, parallel_backend='joblib', verbose=False, use_future=True, future_config=None)

Optimize the future DNN pipeline.

Parameters:

  • df_path (str) – Path to the dataset.

  • storage_id (int) – Storage ID.

  • base_models_folder (str) – Base folder for models.

  • data (dict, optional) – Preloaded data dict (ignored in future pipeline if empty).

  • use_future (bool, optional) – If False, fallback to the current pipeline. Default True.

  • storage (str) –

  • use_pdb_train (bool) –

  • no_scores (bool) –

  • only_scores (bool) –

  • use_PCA (bool) –

  • best_ao_params (dict | None) –

  • pca_type (int) –

  • pca_model (str | PCA) –

  • encoder_dims (tuple[int, int]) –

  • autoencoder (bool) –

  • multiencoder (bool) –

  • run_autoencoder_optimization (bool) –

  • num_processes_autoencoder (int) –

  • total_trials_autoencoder (int) –

  • run_NN_optimization (bool) –

  • num_processes_NN (int) –

  • total_trials_NN (int) –

  • explained_variance (float) –

  • random_seed (int) –

  • load_if_exists (bool) –

  • use_gpu (bool) –

  • parallel_backend (str) –

  • verbose (bool) –

  • future_config (dict | None) –

Return type:

None

Notes

Example usage:

optimize_NN_future(df_path, 1, “./models”, total_trials_NN=20)