reference/generated/tensordict.TensorDictBase.__eq__

Run in Google Colab

Colab

Download Notebook

Notebook

View on GitHub

GitHub

TensorDictBase.__eq__

TensorDictBase.__eq__(other: Any) → Self | bool

Compares the tensordict with other leaf by leaf with ==.

other can be a tensordict or a dict with the same keys, paired with the tensordict key by key (a KeyError is raised if the keys differ); a tensor, which is aligned with the batch dimensions of each leaf; or a number. A batch size that differs from the tensordict’s is broadcast with it. An other of another type, such as None, gives False. The other comparisons (!=, <, <=, >, >=) work the same way, except for such an other: != gives True, and <, <=, > and >= give False and a FutureWarning. Starting with TensorDict 0.17, the default will change to returning NotImplemented, so that Python raises a TypeError, as it does for tensors and for &.

The result is a new tensordict that is not locked and, except with a tensor other of one or more dims, has no dim names. &, ~ and the arithmetic operators keep the dim names and the lock.

Parameters:

other (TensorDictBase, dict, Tensor or Number) – the value to compare with.

Returns:

A new tensordict of boolean leaves, or False.

Examples

>>> import torch
>>> from tensordict import TensorDict
>>> td = TensorDict(a=torch.tensor([1.0, 2.0]), b=torch.zeros(2), batch_size=[2])
>>> (td == 1)["a"]
tensor([ True, False])
>>> (td == td.clone()).all()
True
>>> (td >= torch.tensor([2.0, 0.0]))["b"]
tensor([False,  True])
>>> td == None
False

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