Super Kai (Kazuya Ito)
Posted on • Edited on
positive, neg and copysign in PyTorch
positive() can just get the same tensor as the input tensor which is the 0D or more D tensor of zero or more elements as shown below:
*Memos:
positive()can be used withtorch or a tensor.- The 1st argument(
input) withtorchor using a tensor(Required-Type:tensorofint,floatorcomplex).
importtorchmy_tensor=torch.tensor(7)torch.positive(input=my_tensor)my_tensor.positive()# tensor(7)my_tensor=torch.tensor([7,-1,5,-7,-9,-3,0,6])torch.positive(input=my_tensor)# tensor([7, -1, 5, -7, -9, -3, 0, 6])my_tensor=torch.tensor([[7,-1,5,-7],[-9,-3,0,6]])torch.positive(input=my_tensor)# tensor([[7, -1, 5, -7],# [-9, -3, 0, 6]])my_tensor=torch.tensor([[[7,-1],[5,-7]],[[-9,-3],[0,6]]])torch.positive(input=my_tensor)# tensor([[[7, -1], [5, -7]],# [[-9, -3], [0, 6]]])my_tensor=torch.tensor([[[7.,-1.],[5.,-7.]],[[-9.,-3.],[0.,6.]]])torch.positive(input=my_tensor)# tensor([[[7., -1.], [5., -7.]],# [[-9., -3.], [0., 6.]]])my_tensor=torch.tensor([[[7.+0.j,-1.+0.j],[5.+0.j,-7.+0.j]],[[-9.+0.j,-3.+0.j],[0.+0.j,6.+0.j]]])torch.positive(input=my_tensor)# tensor([[[7.+0.j, -1.+0.j],# [5.+0.j, -7.+0.j]],# [[-9.+0.j, -3.+0.j],# [0.+0.j, 6.+0.j]]])neg() can get the 0D or more D tensor of the zero or more elements changed from+ to- and from- to+ from the 0D or more D tensor of zero or more elements as shown below:
*Memos:
neg()can be used withtorchor a tensor.- The 1st argument(
input) withtorchor using a tensor(Required-Type:tensorofint,floatorcomplex). - There is
outargument withtorch(Optional-Default:None-Type:tensor):*Memos:out=must be used.- My post explains
outargument.
- negative() is the alias of
neg().
importtorchmy_tensor=torch.tensor(7)torch.neg(input=my_tensor)my_tensor.neg()# tensor(-7)my_tensor=torch.tensor([7,-1,5,-7,-9,-3,0,6])torch.neg(input=my_tensor)# tensor([-7, 1, -5, 7, 9, 3, 0, -6])my_tensor=torch.tensor([[7,-1,5,-7],[-9,-3,0,6]])torch.neg(input=my_tensor)# tensor([[-7, 1, -5, 7],# [9, 3, 0, -6]])my_tensor=torch.tensor([[[7,-1],[5,-7]],[[-9,-3],[0,6]]])torch.neg(input=my_tensor)# tensor([[[-7, 1], [-5, 7]],# [[9, 3], [0, -6]]])my_tensor=torch.tensor([[[7.,-1.],[5.,-7.]],[[-9.,-3.],[0.,6.]]])torch.neg(input=my_tensor)# tensor([[[-7., 1.], [-5., 7.]],# [[9., 3.], [-0., -6.]]])my_tensor=torch.tensor([[[7.+0.j,-1.+0.j],[5.+0.j,-7.+0.j]],[[-9.+0.j,-3.+0.j],[0.+0.j,6.+0.j]]])torch.neg(input=my_tensor)# tensor([[[-7.+0.j, 1.+0.j], [-5.+0.j, 7.+0.j]],# [[9.+0.j, 3.+0.j], [0.+0.j, -6.+0.j]]])copysign() can get the 0D or more D tensor of the zero or more floating point numbers changed+ and- byother tensor from two of 0D or more D tensors as shown below:
*Memos:
copysign()can be used withtorchor a tensor.- The 1st argument(
input) withtorchor using a tensor(Required-Type:tensorofint,float,bool). - The 2st argument with
torchor the 1st argument isother(Required-Type:tensororscalarofint,floatorbool). *The sign(+or-) is applied to the returned tensor. - There is
outargument withtorch(Optional-Default:None-Type:tensor):*Memos:out=must be used.- My post explains
outargument.
importtorchtensor1=torch.tensor([[7.,-1.,5.,-7.],[-9.,-3.,0.,6.]])tensor2=torch.tensor([-1.,0.,1.,2.])torch.copysign(input=tensor1,other=tensor2)tensor1.copysign(other=tensor2)# tensor([[-7., 1., 5., 7.],# [-9., 3., 0., 6.]])torch.copysign(input=tensor1,other=2.)# tensor([[7., 1., 5., 7.],# [9., 3., 0., 6.]])tensor1=torch.tensor([[[7.,-1.],[5.,-7.]],[[-9.,-3.],[0.,6.]]])tensor2=torch.tensor(-1.)torch.copysign(input=tensor1,other=tensor2)# tensor([[[-7., -1.], [-5., -7.]],# [[-9., -3.], [-0., -6.]]])torch.copysign(input=tensor1,other=2.)# tensor([[[7., 1.], [5., 7.]],# [[9., 3.], [0., 6.]]])tensor1=torch.tensor([[[7,-1],[5,-7]],[[-9,-3],[0,6]]])tensor2=torch.tensor(-1)torch.copysign(input=tensor1,other=tensor2)torch.copysign(input=tensor1,other=-1)# tensor([[[-7., -1.], [-5., -7.]],# [[-9., -3.], [-0., -6.]]])torch.copysign(input=tensor1,other=2)# tensor([[[7., 1.], [5., 7.]],# [[9., 3.], [0., 6.]]])tensor1=torch.tensor([[[True,False],[True,False]],[[False,True],[False,True]]])tensor2=torch.tensor(True)torch.copysign(input=tensor1,other=tensor2)torch.copysign(input=tensor1,other=False)# tensor([[[1., 0.], [1., 0.]],# [[0., 1.], [0., 1.]]])Top comments(0)
Subscribe
For further actions, you may consider blocking this person and/orreporting abuse
