arange and linspace in PyTorch

Buy Me a Coffee☕

*My post explains logspace().

arange() can create the 1D tensor of zero or integers or floating-point numbers between start and end-1(start<=x<=end-1)as shown below:

*Memos:

• arange() can be used with torch but not with a tensor.
• The 1st argument with torch is start(Optional-Default:0-Type:int, float, complex or bool): *Memos
  • It must be lower than or equal to end.
  • The 0D tensor of int, float, complex or bool also works.
• The 2nd argument with torch is end(Required-Type:int, float, complex or bool): *Memos:
  • It must be greater than or equal to start.
  • The 0D tensor of int, float, complex or bool also works.
• The 3rd argument with torch is step(Optional-Default:1-Type:int, float, complex or bool): *Memos:
  • It must be greater than 0.
  • The 0D tensor of int, float, complex or bool also works.
• There is dtype argument with torch(Optional-Default:None-Type:dtype): *Memos:
  • If it's None, it's inferred from start, end or step, then for floating-point numbers, get_default_dtype() is used. *My post explains get_default_dtype() and set_default_dtype().
  • dtype= must be used.
  • My post explains dtype argument.
• There is device argument with torch(Optional-Default:None-Type:str, int or device()): *Memos:
  • If it's None, get_default_device() is used. *My post explains get_default_device() and set_default_device().
  • device= must be used.
  • My post explains device argument.
• There is requires_grad argument with torch(Optional-Default:False-Type:bool): *Memos:
  • requires_grad= must be used.
  • My post explains requires_grad argument.
• There is out argument with torch(Optional-Default:None-Type:tensor): *Memos:
  • out= must be used.
  • My post explains out argument.
• There is range() which is similar to arange() but range() is deprecated.

import torch

torch.arange(end=5)
# tensor([0, 1, 2, 3, 4])

torch.arange(start=5, end=15)
# tensor([5, 6, 7, 8, 9, 10, 11, 12, 13, 14])

torch.arange(start=5, end=15, step=3)
# tensor([5, 8, 11, 14])

torch.arange(start=-5, end=5)
# tensor([-5, -4, -3, -2, -1, 0, 1, 2, 3, 4])

torch.arange(start=-5, end=5, step=3)
torch.arange(start=torch.tensor(-5),
             end=torch.tensor(5),
             step=torch.tensor(3))
# tensor([-5, -2, 1, 4])

torch.arange(start=-5., end=5., step=3.)
torch.arange(start=torch.tensor(-5.),
             end=torch.tensor(5.),
             step=torch.tensor(3.))
# tensor([-5., -2., 1., 4.])

torch.arange(start=-5.+0.j, end=5.+0.j, step=3.+0.j)
torch.arange(start=torch.tensor(-5.+0.j),
             end=torch.tensor(5.+0.j),
             step=torch.tensor(3.+0.j))
# tensor([-5., -2., 1., 4.])

torch.arange(start=False, end=True, step=True)
torch.arange(start=torch.tensor(False),
             end=torch.tensor(True),
             step=torch.tensor(True))
# tensor([0])

linspace() can create the 1D tensor of the zero or more integers, floating-point numbers or complex numbers evenly spaced between start and end(start<=x<=end) as shown below:

*Memos:

• linspace() can be used with torch but not with a tensor.
• The 1st argument with torch is start(Required-Type:int, float, complex or bool). *The 0D tensor of int, float, complex or bool also works.
• The 2nd argument with torch is end(Required-Type:int, float, complex or bool). *The 0D tensor of int, float, complex or bool also works.
• The 3rd argument with torch is steps(Required-Type:int): *Memos:
  • It must be greater than or equal to 0.
  • The 0D tensor of int also works.
• There is dtype argument with torch(Optional-Default:None-Type:dtype): *Memos:
  • If it's None, it's inferred from start, end or step, then for floating-point numbers, get_default_dtype() is used. *My post explains get_default_dtype() and set_default_dtype().
  • Setting start and end of integer type is not enough to create the 1D tensor of integer type so integer type with dtype must be set.
  • dtype= must be used.
  • My post explains dtype argument.
• There is device argument with torch(Optional-Default:None-Type:str, int or device()): *Memos:
  • If it's None, get_default_device() is used. *My post explains get_default_device() and set_default_device().
  • device= must be used.
  • My post explains device argument.
• There is requires_grad argument with torch(Optional-Default:False-Type:bool): *Memos:
  • requires_grad= must be used.
  • My post explains requires_grad argument.
• There is out argument with torch(Optional-Default:None-Type:tensor): *Memos:
  • out= must be used.
  • My post explains out argument.

import torch

torch.linspace(start=10, end=20, steps=0)
torch.linspace(start=20, end=10, steps=0)
# tensor([])

torch.linspace(start=10., end=20., steps=1)
tensor([10.])

torch.linspace(start=20, end=10, steps=1)
# tensor([20.])

torch.linspace(start=10., end=20., steps=2)
# tensor([10., 20.])

torch.linspace(start=20, end=10, steps=2)
# tensor([20., 10.])

torch.linspace(start=10., end=20., steps=3)
# tensor([10., 15., 20.])

torch.linspace(start=20, end=10, steps=3)
# tensor([20., 15., 10.])

torch.linspace(start=10., end=20., steps=4)
# tensor([10.0000, 13.3333, 16.6667, 20.0000])

torch.linspace(start=20., end=10., steps=4)
# tensor([20.0000, 16.6667, 13.3333, 10.0000])

torch.linspace(start=10, end=20, steps=4, dtype=torch.int64)
torch.linspace(start=torch.tensor(10),
               end=torch.tensor(20),
               steps=torch.tensor(4),
               dtype=torch.int64)
# tensor([10.0000, 13.3333, 16.6667, 20.0000])

torch.linspace(start=10.+6.j, end=20.+3.j, steps=4)
torch.linspace(start=torch.tensor(10.+6.j),
               end=torch.tensor(20.+3.j),
               steps=torch.tensor(4))
# tensor([10.0000+6.j, 13.3333+5.j, 16.6667+4.j, 20.0000+3.j])

torch.linspace(start=False, end=True, steps=4)
torch.linspace(start=torch.tensor(True),
               end=torch.tensor(False),
               steps=torch.tensor(4))
# tensor([0.0000, 0.3333, 0.6667, 1.0000])

torch.linspace(start=10, end=20, steps=4, dtype=torch.int64)
torch.linspace(start=torch.tensor(10),
               end=torch.tensor(20),
               steps=torch.tensor(4), dtype=torch.int64)
# tensor([10.0000, 13.3333, 16.6667, 20.0000])