pytorch （四）F.softmax cross_entropy confusion_matrix stack

import numpy as np

import torch.nn.functional as F

import torch

>>> F.softmax(x, dim=0)
tensor([[0.3333, 0.3333, 0.3333, 0.3333, 0.3333],
        [0.3333, 0.3333, 0.3333, 0.3333, 0.3333],
        [0.3333, 0.3333, 0.3333, 0.3333, 0.3333]])
>>> x
tensor([[1., 2., 3., 4., 5.],
        [1., 2., 3., 4., 5.],
        [1., 2., 3., 4., 5.]])
>>> F.softmax(x, dim=1)
tensor([[0.0117, 0.0317, 0.0861, 0.2341, 0.6364],
        [0.0117, 0.0317, 0.0861, 0.2341, 0.6364],
        [0.0117, 0.0317, 0.0861, 0.2341, 0.6364]])
>>> sum([0.0117, 0.0317, 0.0861, 0.2341, 0.6364])
1.0
>>>

softmax功能是把一组数归一为和为1. 

如v = [a , b , c ,d ],, s = sum(v) softmax变成[a/ s, b/s, c/s,d/s]

 

>>> from sklearn.metrics import confusion_matrix
>>> confusion_matrix([1, 2, 1, 3,3,3,4],[1,3,1,4,3,3,4])
array([[2, 0, 0, 0],
       [0, 0, 1, 0],
       [0, 0, 2, 1],
       [0, 0, 0, 1]], dtype=int64)

import numpy as np 
import torch.nn.functional as F
import torch
x = np.array([[1, 2,3,4,5],
             [1, 2,3,4,5],
             [1, 2,3,4,5]]).astype(np.float32)
y = np.array([1, 1, 0])
x = torch.from_numpy(x)
y = torch.from_numpy(y).long()

soft_out = F.softmax(x,dim=1)
log_soft_out = torch.log(soft_out)
loss = F.nll_loss(log_soft_out, y)
print(soft_out)
print(log_soft_out)
print(loss)

loss = F.cross_entropy(x, y)
print(loss)

    Example::

        >>> # input is of size N x C = 3 x 5
        >>> input = torch.randn(3, 5, requires_grad=True)
        >>> # each element in target has to have 0 <= value < C
        >>> target = torch.tensor([1, 0, 4])
        >>> output = F.nll_loss(F.log_softmax(input), target)
        >>> output.backward()