nos.data package¶

Submodules¶

nos.data.transmission_loss module¶

class nos.data.transmission_loss.TLDataset(path: Path, n_samples: int = -1, v_transform: Literal['quantile', 'min_max', 'normalize'] = 'quantile')¶: Bases: OperatorDataset

class nos.data.transmission_loss.TLDatasetCompact(path: Path, n_samples: int = -1, v_transform: Literal['quantile', 'min_max', 'normalize'] = 'quantile')¶

Bases: OperatorDataset

Transmission loss dataset, with bigger evaluation space.

nos.data.transmission_loss.get_min_max_transform(src: Tensor) → Transform¶

nos.data.transmission_loss.get_n_unique(df: DataFrame, n_samples: int = -1)¶

nos.data.transmission_loss.get_normalize_transform(src: Tensor) → Transform¶

nos.data.transmission_loss.get_tl_compact(path: Path, n_samples: int = -1)¶

nos.data.transmission_loss.get_tl_frame(path: Path, n_samples: int = -1)¶

nos.data.transmission_loss.get_tl_from_path(path: Path)¶

nos.data.transmission_loss.get_unique_crystals(df: DataFrame) → DataFrame¶

nos.data.xdmf_to_torch module¶

nos.data.xdmf_to_torch.get_array(data_dir: Path, element: Element) → tensor¶

Gets the first DataItem from a given element.

Parameters:

data_dir – directory in which both the xdmf and h5 files are located
element – parent

Returns:

array of all elements stored in the first DataItem for the given element.

nos.data.xdmf_to_torch.xdmf_to_torch(file: Path) → dict¶

Converts a xdmf file to a dictionary with all values stored within it.

Parameters:: file – path to the xdmf file (h5 file needs to be located in the same dir).
Returns:: Dictionary containing topology, geometry, frequencies, and complex values.

Module contents¶

class nos.data.TLDataset(path: Path, n_samples: int = -1, v_transform: Literal['quantile', 'min_max', 'normalize'] = 'quantile')¶: Bases: OperatorDataset

class nos.data.TLDatasetCompact(path: Path, n_samples: int = -1, v_transform: Literal['quantile', 'min_max', 'normalize'] = 'quantile')¶