Header

class torchdata.datapipes.iter.Header(source_datapipe: IterDataPipe[T_co], limit: int = 10)

Yields elements from the source DataPipe from the start, up to the specfied limit (functional name: header). This DataPipe can also be used to manually set the length of a DataPipe to a certain value; you can do so by calling dp.header(desired_len).

Parameters:

• source_datapipe – the DataPipe from which elements will be yielded

• limit – the number of elements to yield before stopping

Example

>>> from torchdata.datapipes.iter import IterableWrapper
>>> dp = IterableWrapper(range(10))
>>> header_dp = dp.header(3)
>>> list(header_dp)
[0, 1, 2]