第三章-利用autograd/Variable实现线性回归
import torch as t
from torch.autograd import Variable as V
from matplotlib import pyplot as plt
from IPython import display
#为了在不同的计算机上运行时下面的输出一致,设置随机种子,每次得到的随机数是固定的
t.manual_seed(1000)
def get_fake_data(batch_size = 8):
# 产生随机数据: y = x*2+3,加上了一些噪声
x = t.randn(batch_size,1)*20
y = x*2+(1+t.randn(batch_size,1))*3
return x,y
# 随机初始化参数
w = V(t.randn(1,1),requires_grad = True)
b = V(t.zeros(1,1),requires_grad = True)
lr = 0.0001 #学习率
for ii in range(5000):
x,y = get_fake_data()
x,y = V(x),V(y)
#forward:计算loss
y_pred = x.mm(w)+b.expand_as(y)
loss = 0.5*(y_pred-y)**2
loss = loss.sum()
#backwrad:计算计算梯度
loss.backward()
#更新参数
w.data.sub_(lr * w.grad.data)
b.data.sub_(lr * w.grad.data)
#只使用本次数据的梯度,所以要梯度清零
w.grad.data.zero_()
b.grad.data.zero_()
if ii%1000 == 0:
#画图
display.clear_output(wait=True)
x = t.arange(0,20).float().view(-1,1) # -1代表不确定的数,注意 t.arange的输出结果为 Long
y = x.mm(w.data) + b.data.expand_as(x)
plt.plot(x.numpy(),y.numpy()) # predicted
x2,y2 = get_fake_data(batch_size=20)
plt.scatter(x2.numpy(),y2.numpy()) # true data
plt.xlim(0,20)
plt.ylim(0,41)
plt.show()
plt.pause(0.5)
print(w.data.item(),b.data.item())
结果:
1.98751699924469 3.1595630645751953