深度学习(李沐)—————线性回归

%matplotlib inline
import random
import torch
from d2l import torch as d2l

"""训练样本"""

def synthetic_data(w,b,num_examples):
    """生成y=Xw+b+噪声"""
    X = torch.normal(0,1,(num_examples,len(w)))  
    """torch.normal（0，1）表示生成均值为0 方差为1的随机数  列度是w 行数是样本 """
    y = torch.matmul(X,w)+b  
    """matmul表示矩阵相乘 """
    y += torch.normal(0,0.01,y.shape) 
    """表示生成均值为0 方差为0.01的随机数  列度是w 行数是样本  形状与y的形状一样"""
    return X,y.reshape((-1,1)) 
    """reshape -1表示自动计算 后面1表示固定1列 即列向量为1"""

true_w = torch.tensor([2,-3.4])
"""reshape -1表示自动计算 后面1表示固定1列 即列向量为1"""
true_b = 4.2
features,labels = synthetic_data(true_w ,true_b,1000)
"""生成特征以及标注"""

"""对训练样本进行展示"""
print('features:',features[0],'\nlabel:',labels[0])

   
d2l.set_figsize()
d2l.plt.scatter(features[:,1].detach().numpy(),labels.detach().numpy(),1);

 训练样本显示结果

"""data_iter函数接收批量大小、特征矩阵哈标签向量作为输入，生成大小为batch_size的小批量""" 
def data_iter(batch_size,features,labels):
    num_examples = len(features)
    indices = list(range(num_examples))
    random.shuffle(indices)
    """打乱下标，以便后面随即访问样本"""
    for i in range(0,num_examples,batch_size):
        batch_indices = torch.tensor(
            indices[i:min(i+batch_size,num_examples)])
        yield features[batch_indices],labels[batch_indices]
        """yield就是return返回一个值，并记住返回值的位置，下次迭代从此位置后开始-------return返回函数会终止，yield不会"""

batch_size = 10

for X,y in data_iter(batch_size,features,labels):
    print(X,'\n',y)
    break

此处运行结果如下

    
"""定义初始化参数模型"""
w = torch.normal(0,0.01,size=(2,1),requires_grad=True)
"""requires_grad=True 需要梯度计算"""
b = torch.zeros(1,requires_grad=True)
"""偏差"""



"""定义模型"""
def linreg(X,w,b):
    """线性回归模型"""
    return torch.matmul(X,w)+b


"""定义损失函数"""
def squared_loss(y_hat,y):
    """均方损失"""
    return (y_hat-y.reshape(y_hat.shape))**2/2


"""定义优化算法"""
def sgd(params,lr,batch_size):
    """小批量随机梯度下降"""
    with torch.no_grad():
        for param in params:
            param -= lr * param.grad / batch_size
            param.grad.zero_()

 训练过程

"""训练过程"""
lr = 0.03
num_epochs = 3
net = linreg
loss = squared_loss

for epoch in range(num_epochs):
    for X,y in data_iter(batch_size,features,labels):
        l = loss(net(X,w,b),y)
        l.sum().backward()
        sgd([w,b],lr,batch_size) 
        """使用参数的梯度更新参数"""
    with torch.no_grad():
        train_l = loss(net(features,w,b),labels)
        print(f'epoch{epoch+1},loss{float(train_l.mean()):f}')

比较真实参数和通过训练学到的参数来评估训练的成功程度

"""比较真实参数和通过训练学到的参数来评估训练的成功程度""" 
print(f'w的估计误差:{true_w - w.reshape(true_w.shape)}')
print(f'b的估计误差:{true_b - b}')

完整代码如下

%matplotlib inline
import random
import torch
from d2l import torch as d2l

"""训练样本"""

def synthetic_data(w,b,num_examples):
    """生成y=Xw+b+噪声"""
    X = torch.normal(0,1,(num_examples,len(w)))  
    """torch.normal（0，1）表示生成均值为0 方差为1的随机数  列度是w 行数是样本 """
    y = torch.matmul(X,w)+b  
    """matmul表示矩阵相乘 """
    y += torch.normal(0,0.01,y.shape) 
    """表示生成均值为0 方差为0.01的随机数  列度是w 行数是样本  形状与y的形状一样"""
    return X,y.reshape((-1,1)) 
    """reshape -1表示自动计算 后面1表示固定1列 即列向量为1"""

true_w = torch.tensor([2,-3.4])
"""reshape -1表示自动计算 后面1表示固定1列 即列向量为1"""
true_b = 4.2
features,labels = synthetic_data(true_w ,true_b,1000)
"""生成特征以及标注"""
print('features:',features[0],'\nlabel:',labels[0])

   
d2l.set_figsize()
d2l.plt.scatter(features[:,1].detach().numpy(),labels.detach().numpy(),1);




def data_iter(batch_size,features,labels):
    num_examples = len(features)
    indices = list(range(num_examples))
    random.shuffle(indices)
    """打乱下标，以便后面随即访问样本"""
    for i in range(0,num_examples,batch_size):
        batch_indices = torch.tensor(
            indices[i:min(i+batch_size,num_examples)])
        yield features[batch_indices],labels[batch_indices]
        """yield就是return返回一个值，并记住返回值的位置，下次迭代从此位置后开始-------return返回函数会终止，yield不会"""

batch_size = 10

for X,y in data_iter(batch_size,features,labels):
    print(X,'\n',y)
    break
    
    
    
    
"""定义初始化参数模型"""
w = torch.normal(0,0.01,size=(2,1),requires_grad=True)
"""requires_grad=True 需要梯度计算"""
b = torch.zeros(1,requires_grad=True)
"""偏差"""



"""定义模型"""
def linreg(X,w,b):
    """线性回归模型"""
    return torch.matmul(X,w)+b


"""定义损失函数"""
def squared_loss(y_hat,y):
    """均方损失"""
    return (y_hat-y.reshape(y_hat.shape))**2/2


"""定义优化算法"""
def sgd(params,lr,batch_size):
    """小批量随机梯度下降"""
    with torch.no_grad():
        for param in params:
            param -= lr * param.grad / batch_size
            param.grad.zero_()
            
            
"""训练过程"""
lr = 0.03
num_epochs = 3
net = linreg
loss = squared_loss

for epoch in range(num_epochs):
    for X,y in data_iter(batch_size,features,labels):
        l = loss(net(X,w,b),y)
        l.sum().backward()
        sgd([w,b],lr,batch_size) 
        """使用参数的梯度更新参数"""
    with torch.no_grad():
        train_l = loss(net(features,w,b),labels)
        print(f'epoch{epoch+1},loss{float(train_l.mean()):f}')
        
"""比较真实参数和通过训练学到的参数来评估训练的成功程度""" 
print(f'w的估计误差:{true_w - w.reshape(true_w.shape)}')
print(f'b的估计误差:{true_b - b}')