graph2mat.tools.lightning.data

Data loading for pytorch_lightning workflows.

Classes

MatrixDataModule([out_matrix, basis_files, ...])

class graph2mat.tools.lightning.data.MatrixDataModule(out_matrix: Literal['density_matrix', 'hamiltonian', 'energy_density_matrix', 'dynamical_matrix'] | None = None, basis_files: str | None = None, no_basis: dict | None = None, basis_table: BasisTableWithEdges | None = None, root_dir: str = '.', train_runs: str | None = None, val_runs: str | None = None, test_runs: str | None = None, predict_structs: str | None = None, runs_json: str | None = None, symmetric_matrix: bool = False, sub_point_matrix: bool = True, batch_size: int = 5, loader_threads: int = 1, copy_root_to_tmp: bool = False, store_in_memory: bool = False, rotating_pool_size: int | None = None, initial_node_feats: str = 'OneHotZ')

Bases: LightningDataModule

__init__(out_matrix: Literal['density_matrix', 'hamiltonian', 'energy_density_matrix', 'dynamical_matrix'] | None = None, basis_files: str | None = None, no_basis: dict | None = None, basis_table: BasisTableWithEdges | None = None, root_dir: str = '.', train_runs: str | None = None, val_runs: str | None = None, test_runs: str | None = None, predict_structs: str | None = None, runs_json: str | None = None, symmetric_matrix: bool = False, sub_point_matrix: bool = True, batch_size: int = 5, loader_threads: int = 1, copy_root_to_tmp: bool = False, store_in_memory: bool = False, rotating_pool_size: int | None = None, initial_node_feats: str = 'OneHotZ')

Parameters:

• out_matrix ('density_matrix', 'hamiltonian', 'energy_density_matrix', 'dynamical_matrix')

• basis_files (Union[str, None])

• basis_table (Union[BasisTableWithEdges, None])

• root_dir (str)

• train_runs (Optional[str])

• val_runs (Optional[str])

• test_runs (Optional[str])

• predict_structs (Optional[str])

• runs_json (Optional[str]) – Path to json-file with a dictionary where the keys are train/val/test/predict. and the dictionary values are list of paths to the run files relative to root_dir The paths will be overwritten by train_runs/val_runs/test_runs/predict_structs if given.

• symmetric_matrix (bool)

• sub_point_matrix (bool)

• batch_size (int)

• loader_threads (int)

• copy_root_to_tmp (bool)

• store_in_memory (bool) – If true, will load the dataset into host memory, otherwise it will be read from disk

• rotating_pool_size (int) – If given, the training data will be continously loaded into a smaller poool of this size. The data in the active pool can be used several times before it is swapped out with new data. This is useful if the data loading is slow. Note that the notion of epochs will not be meaningful when this kind of loading is used. This will not affect test/val/predict data.

predict_dataloader()

An iterable or collection of iterables specifying prediction samples.

For more information about multiple dataloaders, see this section.

It’s recommended that all data downloads and preparation happen in prepare_data.

• predict

• prepare_data

• setup

Note

Lightning tries to add the correct sampler for distributed and arbitrary hardware There is no need to set it yourself.

Returns:

A torch.utils.data.DataLoader or a sequence of them specifying prediction samples.

prepare_data()

Use this to download and prepare data. Downloading and saving data with multiple processes (distributed settings) will result in corrupted data. Lightning ensures this method is called only within a single process, so you can safely add your downloading logic within.

Warning

DO NOT set state to the model (use setup instead) since this is NOT called on every device

Example:

def prepare_data(self):
    # good
    download_data()
    tokenize()
    etc()

    # bad
    self.split = data_split
    self.some_state = some_other_state()

In a distributed environment, prepare_data can be called in two ways (using prepare_data_per_node)

1. Once per node. This is the default and is only called on LOCAL_RANK=0.

2. Once in total. Only called on GLOBAL_RANK=0.

Example:

# DEFAULT
# called once per node on LOCAL_RANK=0 of that node
class LitDataModule(LightningDataModule):
    def __init__(self):
        super().__init__()
        self.prepare_data_per_node = True


# call on GLOBAL_RANK=0 (great for shared file systems)
class LitDataModule(LightningDataModule):
    def __init__(self):
        super().__init__()
        self.prepare_data_per_node = False

This is called before requesting the dataloaders:

model.prepare_data()
initialize_distributed()
model.setup(stage)
model.train_dataloader()
model.val_dataloader()
model.test_dataloader()
model.predict_dataloader()

setup(stage: str)

Called at the beginning of fit (train + validate), validate, test, or predict. This is a good hook when you need to build models dynamically or adjust something about them. This hook is called on every process when using DDP.

Parameters:

stage – either 'fit', 'validate', 'test', or 'predict'

Example:

class LitModel(...):
    def __init__(self):
        self.l1 = None

    def prepare_data(self):
        download_data()
        tokenize()

        # don't do this
        self.something = else

    def setup(self, stage):
        data = load_data(...)
        self.l1 = nn.Linear(28, data.num_classes)

teardown(stage: str)

Called at the end of fit (train + validate), validate, test, or predict.

Parameters:

stage – either 'fit', 'validate', 'test', or 'predict'

test_dataloader()

An iterable or collection of iterables specifying test samples.

For more information about multiple dataloaders, see this section.

For data processing use the following pattern:

• download in prepare_data

• process and split in setup

However, the above are only necessary for distributed processing.

Warning

do not assign state in prepare_data

• test

• prepare_data

• setup

Note

Lightning tries to add the correct sampler for distributed and arbitrary hardware. There is no need to set it yourself.

Note

If you don’t need a test dataset and a test_step, you don’t need to implement this method.

train_dataloader()

An iterable or collection of iterables specifying training samples.

For more information about multiple dataloaders, see this section.

The dataloader you return will not be reloaded unless you set :paramref:`~pytorch_lightning.trainer.trainer.Trainer.reload_dataloaders_every_n_epochs` to a positive integer.

For data processing use the following pattern:

• download in prepare_data

• process and split in setup

However, the above are only necessary for distributed processing.

Warning

do not assign state in prepare_data

• fit

• prepare_data

• setup

Note

Lightning tries to add the correct sampler for distributed and arbitrary hardware. There is no need to set it yourself.

val_dataloader()

An iterable or collection of iterables specifying validation samples.

For more information about multiple dataloaders, see this section.

The dataloader you return will not be reloaded unless you set :paramref:`~pytorch_lightning.trainer.trainer.Trainer.reload_dataloaders_every_n_epochs` to a positive integer.

It’s recommended that all data downloads and preparation happen in prepare_data.

• fit

• validate

• prepare_data

• setup

Note

Lightning tries to add the correct sampler for distributed and arbitrary hardware There is no need to set it yourself.

Note

If you don’t need a validation dataset and a validation_step, you don’t need to implement this method.