:py:mod:`deeprvat.deeprvat.train` ================================= .. py:module:: deeprvat.deeprvat.train .. autodoc2-docstring:: deeprvat.deeprvat.train :allowtitles: Module Contents --------------- Classes ~~~~~~~ .. list-table:: :class: autosummary longtable :align: left * - :py:obj:`MultiphenoDataset ` - .. autodoc2-docstring:: deeprvat.deeprvat.train.MultiphenoDataset :summary: * - :py:obj:`MultiphenoBaggingData ` - .. autodoc2-docstring:: deeprvat.deeprvat.train.MultiphenoBaggingData :summary: Functions ~~~~~~~~~ .. list-table:: :class: autosummary longtable :align: left * - :py:obj:`cli ` - .. autodoc2-docstring:: deeprvat.deeprvat.train.cli :summary: * - :py:obj:`subset_samples ` - .. autodoc2-docstring:: deeprvat.deeprvat.train.subset_samples :summary: * - :py:obj:`make_dataset_ ` - .. autodoc2-docstring:: deeprvat.deeprvat.train.make_dataset_ :summary: * - :py:obj:`make_dataset ` - .. autodoc2-docstring:: deeprvat.deeprvat.train.make_dataset :summary: * - :py:obj:`run_bagging ` - .. autodoc2-docstring:: deeprvat.deeprvat.train.run_bagging :summary: * - :py:obj:`train ` - .. autodoc2-docstring:: deeprvat.deeprvat.train.train :summary: * - :py:obj:`best_training_run ` - .. autodoc2-docstring:: deeprvat.deeprvat.train.best_training_run :summary: Data ~~~~ .. list-table:: :class: autosummary longtable :align: left * - :py:obj:`logger ` - .. autodoc2-docstring:: deeprvat.deeprvat.train.logger :summary: * - :py:obj:`METRICS ` - .. autodoc2-docstring:: deeprvat.deeprvat.train.METRICS :summary: * - :py:obj:`OPTIMIZERS ` - .. autodoc2-docstring:: deeprvat.deeprvat.train.OPTIMIZERS :summary: * - :py:obj:`ACTIVATIONS ` - .. autodoc2-docstring:: deeprvat.deeprvat.train.ACTIVATIONS :summary: * - :py:obj:`DEFAULT_OPTIMIZER ` - .. autodoc2-docstring:: deeprvat.deeprvat.train.DEFAULT_OPTIMIZER :summary: API ~~~ .. py:data:: logger :canonical: deeprvat.deeprvat.train.logger :value: 'getLogger(...)' .. autodoc2-docstring:: deeprvat.deeprvat.train.logger .. py:data:: METRICS :canonical: deeprvat.deeprvat.train.METRICS :value: None .. autodoc2-docstring:: deeprvat.deeprvat.train.METRICS .. py:data:: OPTIMIZERS :canonical: deeprvat.deeprvat.train.OPTIMIZERS :value: None .. autodoc2-docstring:: deeprvat.deeprvat.train.OPTIMIZERS .. py:data:: ACTIVATIONS :canonical: deeprvat.deeprvat.train.ACTIVATIONS :value: None .. autodoc2-docstring:: deeprvat.deeprvat.train.ACTIVATIONS .. py:data:: DEFAULT_OPTIMIZER :canonical: deeprvat.deeprvat.train.DEFAULT_OPTIMIZER :value: None .. autodoc2-docstring:: deeprvat.deeprvat.train.DEFAULT_OPTIMIZER .. py:function:: cli() :canonical: deeprvat.deeprvat.train.cli .. autodoc2-docstring:: deeprvat.deeprvat.train.cli .. py:function:: subset_samples(input_tensor: torch.Tensor, covariates: torch.Tensor, y: torch.Tensor, min_variant_count: int) -> typing.Tuple[torch.Tensor, torch.Tensor, torch.Tensor] :canonical: deeprvat.deeprvat.train.subset_samples .. autodoc2-docstring:: deeprvat.deeprvat.train.subset_samples .. py:function:: make_dataset_(debug: bool, pickle_only: bool, compression_level: int, training_dataset_file: typing.Optional[str], config_file: typing.Union[str, pathlib.Path], input_tensor_out_file: str, covariates_out_file: str, y_out_file: str) :canonical: deeprvat.deeprvat.train.make_dataset_ .. autodoc2-docstring:: deeprvat.deeprvat.train.make_dataset_ .. py:function:: make_dataset(debug: bool, pickle_only: bool, compression_level: int, training_dataset_file: typing.Optional[str], config_file: str, input_tensor_out_file: str, covariates_out_file: str, y_out_file: str) :canonical: deeprvat.deeprvat.train.make_dataset .. autodoc2-docstring:: deeprvat.deeprvat.train.make_dataset .. py:class:: MultiphenoDataset(data: typing.Dict[str, typing.Dict], batch_size: int, split: str = 'train', cache_tensors: bool = False, temp_dir: typing.Optional[str] = None, chunksize: int = 1000) :canonical: deeprvat.deeprvat.train.MultiphenoDataset Bases: :py:obj:`torch.utils.data.Dataset` .. autodoc2-docstring:: deeprvat.deeprvat.train.MultiphenoDataset .. rubric:: Initialization .. autodoc2-docstring:: deeprvat.deeprvat.train.MultiphenoDataset.__init__ .. py:method:: __len__() :canonical: deeprvat.deeprvat.train.MultiphenoDataset.__len__ .. autodoc2-docstring:: deeprvat.deeprvat.train.MultiphenoDataset.__len__ .. py:method:: __getitem__(index) :canonical: deeprvat.deeprvat.train.MultiphenoDataset.__getitem__ .. autodoc2-docstring:: deeprvat.deeprvat.train.MultiphenoDataset.__getitem__ .. py:method:: index_input_tensor_zarr(pheno: str, indices: numpy.ndarray) :canonical: deeprvat.deeprvat.train.MultiphenoDataset.index_input_tensor_zarr .. autodoc2-docstring:: deeprvat.deeprvat.train.MultiphenoDataset.index_input_tensor_zarr .. py:class:: MultiphenoBaggingData(data: typing.Dict[str, typing.Dict], train_proportion: float, sample_with_replacement: bool = True, upsampling_factor: int = 1, batch_size: typing.Optional[int] = None, num_workers: typing.Optional[int] = 0, pin_memory: bool = False, cache_tensors: bool = False, temp_dir: typing.Optional[str] = None, chunksize: int = 1000, deterministic: bool = False) :canonical: deeprvat.deeprvat.train.MultiphenoBaggingData Bases: :py:obj:`pytorch_lightning.LightningDataModule` .. autodoc2-docstring:: deeprvat.deeprvat.train.MultiphenoBaggingData .. rubric:: Initialization .. autodoc2-docstring:: deeprvat.deeprvat.train.MultiphenoBaggingData.__init__ .. py:method:: upsample() -> numpy.ndarray :canonical: deeprvat.deeprvat.train.MultiphenoBaggingData.upsample .. autodoc2-docstring:: deeprvat.deeprvat.train.MultiphenoBaggingData.upsample .. py:method:: train_dataloader() :canonical: deeprvat.deeprvat.train.MultiphenoBaggingData.train_dataloader .. autodoc2-docstring:: deeprvat.deeprvat.train.MultiphenoBaggingData.train_dataloader .. py:method:: val_dataloader() :canonical: deeprvat.deeprvat.train.MultiphenoBaggingData.val_dataloader .. autodoc2-docstring:: deeprvat.deeprvat.train.MultiphenoBaggingData.val_dataloader .. py:function:: run_bagging(config: typing.Dict, data: typing.Dict[str, typing.Dict], log_dir: str, checkpoint_file: typing.Optional[str] = None, trial: typing.Optional[optuna.trial.Trial] = None, trial_id: typing.Optional[int] = None, debug: bool = False, deterministic: bool = False) -> typing.Optional[float] :canonical: deeprvat.deeprvat.train.run_bagging .. autodoc2-docstring:: deeprvat.deeprvat.train.run_bagging .. py:function:: train(debug: bool, deterministic: bool, training_gene_file: typing.Optional[str], n_trials: int, trial_id: typing.Optional[int], sample_file: typing.Optional[str], phenotype: typing.Tuple[typing.Tuple[str, str, str, str]], config_file: str, log_dir: str, hpopt_file: str) :canonical: deeprvat.deeprvat.train.train .. autodoc2-docstring:: deeprvat.deeprvat.train.train .. py:function:: best_training_run(debug: bool, log_dir: str, checkpoint_dir: str, hpopt_db: str, config_file_out: str) :canonical: deeprvat.deeprvat.train.best_training_run .. autodoc2-docstring:: deeprvat.deeprvat.train.best_training_run