*Memos:
- My post explains diag() and diagflat().
- My post explains diagonal().
- My post explains diag_embed().
- My post explains device().
eye() can create the 2D tensor with zero or more 1., 1, 1.+0.j or True on the diagonal and zero or more 0., 0, 0.+0.j or False elsewhere as shown below:
*Memos:
-
eye()can be used with torch but not with a tensor. - The returned tensor has zero or more floating-point numbers(Default), integers, complex numbers or boolean values.
- The 1st argument(
int) withtorchisn(Required) which is the number of rows. - The 2nd argument(
int) withtorchism(Optional-Default:n) which is the number of columns. - There is
dtypeargument(torch.dtype) (Optional-Default:None) withtorch. *Memos:- If
None, set_default_dtype() is used. -
dtypecan also accept int(), float() and bool() but not complex() which are python built-in functions. -
dtype=must be used.
- If
- There is
deviceargument(str,intor torch.device) (Optional-Default:cpu) withtorch. *Memos:-
cpu,cuda,ipu,xpu,mkldnn,opengl,opencl,ideep,hip,ve,fpga,ort,xla,lazy,vulkan,mps,meta,hpu,mtiaorprivateuseonecan be set todevice. - Setting
0uses GPU(CUDA). -
device=must be used.
-
import torch
torch.eye(n=0)
torch.eye(n=0, device='cpu')
torch.eye(n=0, device=torch.device(device='cpu'))
torch.eye(n=0, device=torch.device(type='cpu'))
# tensor([], size=(0, 0))
torch.eye(n=1)
# tensor([[1.]])
torch.eye(n=2)
# tensor([[1., 0.],
# [0., 1.]])
torch.eye(n=3)
# tensor([[1., 0., 0.],
# [0., 1., 0.],
# [0., 0., 1.]])
torch.eye(n=4)
# tensor([[1., 0., 0., 0.],
# [0., 1., 0., 0.],
# [0., 0., 1., 0.],
# [0., 0., 0., 1.]])
torch.eye(n=4, m=0)
# tensor([], size=(4, 0))
torch.eye(n=4, m=1)
# tensor([[1.],
# [0.],
# [0.],
# [0.]])
torch.eye(n=4, m=2)
# tensor([[1., 0.],
# [0., 1.],
# [0., 0.],
# [0., 0.]])
torch.eye(n=4, m=3)
# tensor([[1., 0., 0.],
# [0., 1., 0.],
# [0., 0., 1.],
# [0., 0., 0.]])
torch.eye(n=4, m=4)
# tensor([[1., 0., 0., 0.],
# [0., 1., 0., 0.],
# [0., 0., 1., 0.],
# [0., 0., 0., 1.]])
torch.eye(n=4, m=5)
# tensor([[1., 0., 0., 0., 0.],
# [0., 1., 0., 0., 0.],
# [0., 0., 1., 0., 0.],
# [0., 0., 0., 1., 0.]])
torch.eye(n=4, m=6)
# tensor([[1., 0., 0., 0., 0., 0.],
# [0., 1., 0., 0., 0., 0.],
# [0., 0., 1., 0., 0., 0.],
# [0., 0., 0., 1., 0., 0.]])
torch.eye(n=4, m=6, dtype=torch.float64, device='cuda:0')
torch.eye(n=4, m=6, dtype=torch.float64, device='cuda')
torch.eye(n=4, m=6, dtype=torch.float64, device=0)
torch.eye(n=4, m=6, dtype=float, device=torch.device('cuda:0'))
torch.eye(n=4, m=6, dtype=float, device=torch.device(type='cuda'))
torch.eye(n=4, m=6, dtype=float, device=torch.device(type='cuda', index=0))
# tensor([[1., 0., 0., 0., 0., 0.],
# [0., 1., 0., 0., 0., 0.],
# [0., 0., 1., 0., 0., 0.],
# [0., 0., 0., 1., 0., 0.]], device='cuda:0', dtype=torch.float64)
torch.eye(n=4, m=6, dtype=torch.int64)
torch.eye(n=4, m=6, dtype=int)
# tensor([[1, 0, 0, 0, 0, 0],
# [0, 1, 0, 0, 0, 0],
# [0, 0, 1, 0, 0, 0],
# [0, 0, 0, 1, 0, 0]], dtype=torch.int32)
torch.eye(n=4, m=6, dtype=torch.complex64)
# tensor([[1.+0.j, 0.+0.j, 0.+0.j, 0.+0.j, 0.+0.j, 0.+0.j],
# [0.+0.j, 1.+0.j, 0.+0.j, 0.+0.j, 0.+0.j, 0.+0.j],
# [0.+0.j, 0.+0.j, 1.+0.j, 0.+0.j, 0.+0.j, 0.+0.j],
# [0.+0.j, 0.+0.j, 0.+0.j, 1.+0.j, 0.+0.j, 0.+0.j]])
torch.eye(n=4, m=6, dtype=torch.bool)
torch.eye(n=4, m=6, dtype=bool)
# tensor([[True, False, False, False, False, False],
# [False, True, False, False, False, False],
# [False, False, True, False, False, False],
# [False, False, False, True, False, False]])
Top comments (0)