TensorDictBase.pop#

TensorDictBase.pop(key: NestedKey, default: Any = NO_DEFAULT) → Tensor | TensorCollection#

Removes an entry from the tensordict and returns its value.

A nested key removes one entry of a nested tensordict and keeps the nested tensordict. A locked tensordict raises RuntimeError when the key is present (see lock_()).

Parameters:

key (NestedKey) – the key of the entry to remove.
default (Any, optional) – the value returned when the key is missing, including a nested key that runs through a tensor. Without it, a missing key raises KeyError and a nested key that runs through a tensor raises ValueError.

Returns:

The value of the removed entry, or default.

Examples

>>> import torch
>>> from tensordict import TensorDict
>>> td = TensorDict(a=torch.zeros(3), b=TensorDict(c=torch.ones(3)))
>>> td.pop(("b", "c"))
tensor([1., 1., 1.])
>>> sorted(td.keys(include_nested=True))
['a', 'b']
>>> td.pop("missing", None) is None
True

Every tensordict class and every tensorclass has this method. For the other methods, see Methods by task.

TensorDictBase.pop#

Docs

Tutorials

Resources