gluonts.dataset.arrow.file module#

class gluonts.dataset.arrow.file.ArrowFile(path: pathlib.Path)[source]#

Bases: gluonts.dataset.arrow.file.File

property batch_offsets#

decoder: gluonts.dataset.arrow.dec.ArrowDecoder#

iter_batches()[source]#

location_for(idx)[source]#

metadata() → Dict[str, str][source]#

path: pathlib.Path#

reader: pyarrow.ipc.RecordBatchFileReader#

property schema#

class gluonts.dataset.arrow.file.ArrowStreamFile(path: pathlib.Path)[source]#

Bases: gluonts.dataset.arrow.file.File

metadata() → Dict[str, str][source]#

path: pathlib.Path#

class gluonts.dataset.arrow.file.File[source]#

Bases: object

SUFFIXES = {'.arrow', '.parquet'}#

static infer(path: pathlib.Path) → Union[gluonts.dataset.arrow.file.ArrowFile, gluonts.dataset.arrow.file.ParquetFile, gluonts.dataset.arrow.file.ArrowStreamFile][source]#

Return either ArrowFile or ArrowStreamFile by inspecting provided path.

Arrow’s random-access format starts with ARROW1, so we peek the provided file for it.

abstract metadata() → Dict[str, str][source]#

class gluonts.dataset.arrow.file.ParquetFile(path: pathlib.Path)[source]#

Bases: gluonts.dataset.arrow.file.File

metadata() → Dict[str, str][source]#

path: pathlib.Path#

reader: pyarrow.parquet.ParquetFile#