tomodrgn.analysis.recursive_load_dataframe#

recursive_load_dataframe(volumeseries_star_path: str, tomo_id_column: str) DataFrame[source]#

Create merged dataframe containing:

  1. imageseries star file used to train model (referenced in train_vae config.pkl)

  2. volumeseries star file specified here

  3. any *.pkl file found recursively within this notebook’s directory which contains a numpy array with first axis shape matching the number of particles in the imageseries star file.

Data are added assuming all indexing matches imageseries star file particle order.

Parameters:
  • volumeseries_star_path – absolute path to volume series star file, must reference the same set of particles referenced by the starfile used for tomodrgn train_vae

  • tomo_id_column – full string name of column containing unique values per tomogram in volseries star file

Returns:

pandas dataframe containing all described data