ncem.estimators.EstimatorCVAE.get_data

EstimatorCVAE.get_data(data_origin: str, data_path: str, radius: Optional[int], n_rings: int = 1, graph_covar_selection: Optional[Union[List[str], Tuple[str]]] = None, node_label_space_id: str = 'type', node_feature_space_id: str = 'standard', use_covar_node_position: bool = False, use_covar_node_label: bool = False, use_covar_graph_covar: bool = False, domain_type: str = 'image', robustness: Optional[float] = None, robustness_seed: int = 1, n_top_genes: Optional[int] = None, segmentation_robustness: Optional[List[float]] = None, resimulate_nodes: bool = False, resimulate_nodes_w_depdency: bool = False, resimulate_nodes_sparsity_rate: float = 0.5)

Get data used in estimator classes.

Parameters

• data_origin (str) – Data origin.

• data_path (str) – Data path.

• radius (int, optional) – Radius.

• n_rings (int) – Number of rings of neighbors for grid data.

• graph_covar_selection (list, tuple, optional) – Selected graph covariates.

• node_label_space_id (str) – Node label space id.

• node_feature_space_id (str) – Node feature space id.

• use_covar_node_position (bool) – Whether to use node position as covariate.

• use_covar_node_label (bool) – Whether to use node label as covariate.

• use_covar_graph_covar (bool) – Whether to use graph covariates.

• domain_type (str) – Covariate that is used as domain.

• robustness (float, optional) – Optional fraction of images for robustness test.

• robustness_seed (int) – Seed for robustness analysis

• n_top_genes (int, optional) – N top genes for highly variable gene selection.

• segmentation_robustness (list, optional) – Parameters for segmentation robustness fit, float for fraction of nodes and float for signal overflow.

Raises

ValueError – If sub-selected covar_selection could not be found, node_label_space_id or node_feature_space_id not recognized