PYTHON

pytorch view -1 meaning

If there is any situation that you don't know how many rows you want but are sure of the number of columns, then you can specify this with a -1. (Note that you can extend this to tensors with more dimensions. Only one of the axis value can be -1). This is a way of telling the library: "give me a tensor that has these many columns and you compute the appropriate number of rows that is necessary to make this happen".

view(-1) in pytorch

import torch

x = torch.arange(6)

print(x.view(3, -1))      # inferred size will be 2 as 6 / 3 = 2
# tensor([[ 0.,  1.],
#         [ 2.,  3.],
#         [ 4.,  5.]])

print(x.view(-1, 6))      # inferred size will be 1 as 6 / 6 = 1
# tensor([[ 0.,  1.,  2.,  3.,  4.,  5.]])

print(x.view(1, -1, 2))   # inferred size will be 3 as 6 / (1 * 2) = 3
# tensor([[[ 0.,  1.],
#          [ 2.,  3.],
#          [ 4.,  5.]]])

# print(x.view(-1, 5))    # throw error as there's no int N so that 5 * N = 6
# RuntimeError: invalid argument 2: size '[-1 x 5]' is invalid for input with 6 elements

print(x.view(-1, -1, 3))  # throw error as only one dimension can be inferred
# RuntimeError: invalid argument 1: only one dimension can be inferred

view(-1 1) pytorch

view is similar to numpy's reshape
"view" shares the underlying data with the original tensor, so it is really
a view into the old tensor instead of creating a brand new one