自己动手写数据输入并计算pytorch

import torch
from IPython import display
from matplotlib import pyplot as plt
import numpy as np
import random

num_inputs = 2  # （特征数）为2。
num_examples = 1000  # 训练数据集样本数为1000
true_w = [2, -3.4]  # 线性回归模型真实权᯿
true_b = 4.2  # 偏差   torch.from_numpy是将 numpy 转成 tensor
# loc(float)：此概率分布的均值（对应着整个分布的中心centre
# scale(float)：此概率分布的标准差（对应于分布的宽度，scale越大，图形越矮胖；scale越小，图形越瘦高）
# size(int or tuple of ints)：输出的shape，默认为None，只输出一个值
features = torch.from_numpy(np.random.normal(0, 1, (num_examples, num_inputs))).to(torch.float32) #生产1000个数据
labels = true_w[0] * features[:, 0] + true_w[1] * features[:, 1] + true_b  #生成1000个labels
labels += torch.from_numpy(np.random.normal(0, 0.01, size=labels.size()))  #增加噪声
print(features)


def use_svg_display():
    # ⽤⽮量图显示
    display.set_matplotlib_formats('svg')


def set_figsize(figsize=(3.5, 2.5)):
    use_svg_display()
    # 设置图的尺⼨
    plt.rcParams['figure.figsize'] = figsize


print(features[0], labels[0])
# # 在../d2lzh_pytorch⾥⾯添加上⾯两个函数后就可以这样导⼊
# import sys
# sys.path.append("..")
# from d2lzh_pytorch import *
set_figsize()  #.numpy（）数据类型转换
plt.scatter(features[:, 1].numpy(), labels.numpy(), 1)

plt.show()


# 本函数已保存在d2lzh包中⽅便以后使⽤
def data_iter(batch_size, features, labels):
    num_examples = len(features)
    indices = list(range(num_examples))
    random.shuffle(indices)  # 样本的读取顺序是随机的   以batch_size为步长 i是batch_size的整数
    for i in range(0, num_examples, batch_size):
        j = torch.LongTensor(indices[i: min(i + batch_size,
                                            num_examples)])  # 最后⼀次可能不⾜⼀个batch
    yield features.index_select(0, j), labels.index_select(0, j)


batch_size = 10
for X, y in data_iter(batch_size, features, labels):
    print(X, y)
    break

w = torch.tensor(np.random.normal(0, 0.01, (num_inputs, 1)),
                 dtype=torch.float32)
b = torch.zeros(1, dtype=torch.float32)
w.requires_grad_(requires_grad=True)
b.requires_grad_(requires_grad=True)


def linreg(X, w, b):  # 本函数已保存在d2lzh_pytorch包中⽅便以后使⽤
    return torch.mm(X, w) + b


def squared_loss(y_hat, y):  # 本函数已保存在d2lzh_pytorch包中⽅便以后使
    # 注意这⾥返回的是向量, 另外, pytorch⾥的MSELoss并没有除以 2
    return (y_hat - y.view(y_hat.size())) ** 2 / 2


def sgd(params, lr, batch_size):  # 本函数已保存在d2lzh_pytorch包中⽅便
    for param in params:
        param.data -= lr * param.grad / batch_size  #


lr = 0.03
num_epochs = 3
net = linreg
loss = squared_loss
for epoch in range(num_epochs):  # 训练模型⼀共需要num_epochs个迭代周期
    for X, y in data_iter(batch_size, features, labels):
        l = loss(net(X, w, b), y).sum() #求和操作
        l.backward()  # ⼩批量的损失对模型参数求梯度
        sgd([w, b], lr, batch_size)  # 使⽤⼩批量随机梯度下降迭代模型参数

        # 不要忘了梯度清零
        w.grad.data.zero_()
        b.grad.data.zero_()
    train_l = loss(net(features, w, b), labels)
    print('epoch %d, loss %f' % (epoch + 1, train_l.mean().item()))

print(true_w, '\n', w)
print(true_b, '\n', b)