campa.tl.load_full_data_dict

load_full_data_dict(exp, keys=('x', 'y', 'obj_ids', 'latent'), data_dirs=None, save_dir='aggregated/full_data')[source]

Load mpp_datas used in experiment in a dict.

NOTE: this might take a few minutes, as all data needs to be loaded.

Parameters

  • exp (Experiment) – Experiment from which to load the mpp_datas.

  • keys (Iterable[str]) – Controls which numpy data matrices are being loaded. Passed to MPPData.from_data_dir(), with optional_keys set to empty list. Excluding mpp here speeds up loading.

  • data_dirs (Iterable[str] | None) – Directories that should be loaded, if None, all data_dirs are loaded.

  • save_dir (str) – Directory in which the data to be loaded is stored, relative to {exp.dir}/{exp.name}.

Return type

Dictionary with data_dirs as keys and MPPData as values.