structures.py 605 B

12345678910111213141516171819
  1. # mypy: allow-untyped-defs
  2. from torch.utils.data.datapipes.datapipe import DataChunk
  3. from torch.utils.data.datapipes.dataframe import dataframe_wrapper as df_wrapper
  4. __all__ = ["DataChunkDF", ]
  5. class DataChunkDF(DataChunk):
  6. """DataChunkDF iterating over individual items inside of DataFrame containers, to access DataFrames user `raw_iterator`."""
  7. def __iter__(self):
  8. for df in self.items:
  9. yield from df_wrapper.iterate(df)
  10. def __len__(self):
  11. total_len = 0
  12. for df in self.items:
  13. total_len += df_wrapper.get_len(df)
  14. return total_len