import torch
import numpy as np
from torch import nn
from torch.autograd import Variable
import torch.nn.functional as F
import matplotlib.pyplot as plt
def plot_decision_boundary(model, x, y):
# Set min and max values and give it some padding
x_min, x_max = x[:, 0].min() - 1, x[:, 0].max() + 1
y_min, y_max = x[:, 1].min() - 1, x[:, 1].max() + 1
h = 0.01
# Generate a grid of points with distance h between them
xx, yy = np.meshgrid(np.arange(x_min, x_max, h), np.arange(y_min, y_max, h))
# Predict the function value for the whole grid
Z = model(np.c_[xx.ravel(), yy.ravel()])
Z = Z.reshape(xx.shape)
# Plot the contour and training examples
plt.contourf(xx, yy, Z, cmap=plt.cm.Spectral)
plt.ylabel('x2')
plt.xlabel('x1')
plt.scatter(x[:, 0], x[:, 1], c=y.reshape(-1), s=40, cmap=plt.cm.Spectral)
def draw_picture():
np.random.seed(1)
m = 400 # 樣本數量
N = int(m/2) # 每一類的點的個數
D = 2 # 維度
x = np.zeros((m, D))
y = np.zeros((m, 1), dtype='uint8') # label 向量，0 表示紅色，1 表示藍色
a = 4
for j in range(2):
ix = range(N*j,N*(j+1))
t = np.linspace(j*3.12,(j+1)*3.12,N) + np.random.randn(N)*0.2 # theta
r = a*np.sin(4*t) + np.random.randn(N)*0.2 # radius
x[ix] = np.c_[r*np.sin(t), r*np.cos(t)]
y[ix] = j
plt.scatter(x[:, 0], x[:, 1], c=y.reshape(-1), s=40, cmap=plt.cm.Spectral)
plt.show()
plt.close()
return x,y
def logistic_regression(x,w,b):
return torch.mm(x, w) + b
def plot_logistic(x,w,b):
x = Variable(torch.from_numpy(x).float())
out = F.sigmoid(logistic_regression(x,w,b))
out = (out > 0.5) * 1
return out.data.numpy()
def logistic_train(x,y):
x = torch.from_numpy(x).float()
y = torch.from_numpy(y).float()
w = nn.Parameter(torch.randn(2, 1))
b = nn.Parameter(torch.zeros(1))
optimizer = torch.optim.SGD([w, b], 1e-1)
criterion = nn.BCEWithLogitsLoss()
for e in range(100):
out = logistic_regression(Variable(x),w,b)
loss = criterion(out, Variable(y))
optimizer.zero_grad()
loss.backward()
optimizer.step()
if (e + 1) % 20 == 0:
print('epoch: {}, loss: {}'.format(e+1, loss.item()))
plot_decision_boundary(lambda x: plot_logistic(x,w,b), x.numpy(), y.numpy())
plt.title('logistic regression')
plt.show()
plt.close()
def two_network_train(x,y):
# 定義兩層神經網絡的參數
w1 = nn.Parameter(torch.randn(2, 4) * 0.01) # 隱藏層神經元個數 2
b1 = nn.Parameter(torch.zeros(4))
w2 = nn.Parameter(torch.randn(4, 1) * 0.01)
b2 = nn.Parameter(torch.zeros(1))
# 定義模型
def two_network(x):
x1 = torch.mm(x, w1) + b1
x1 = F.tanh(x1) # 使用 PyTorch 自帶的 tanh 激活函數
x2 = torch.mm(x1, w2) + b2
return x2
x = torch.from_numpy(x).float()
y = torch.from_numpy(y).float()
optimizer = torch.optim.SGD([w1, w2, b1, b2], 1.)
criterion = nn.BCEWithLogitsLoss()
# 我們訓練 10000 次
for e in range(10000):
out = two_network(Variable(x))
loss = criterion(out, Variable(y))
optimizer.zero_grad()
loss.backward()
optimizer.step()
if (e + 1) % 1000 == 0:
print('epoch: {}, loss: {}'.format(e+1, loss.item()))
def plot_network(x):
x = Variable(torch.from_numpy(x).float())
x1 = torch.mm(x, w1) + b1
x1 = F.tanh(x1)
x2 = torch.mm(x1, w2) + b2
out = F.sigmoid(x2)
out = (out > 0.5) * 1
return out.data.numpy()
plot_decision_boundary(lambda x: plot_network(x), x.numpy(), y.numpy())
plt.title('2 layer network')
plt.show()
plt.close()
def sequential_train(x,y):
x = torch.from_numpy(x).float()
y = torch.from_numpy(y).float()
# Sequential
seq_net = nn.Sequential(
nn.Linear(2, 4), # PyTorch 中的線性層，wx + b
nn.Tanh(),
nn.Linear(4, 1)
)
# 通過 parameters 可以取得模型的參數
param = seq_net.parameters()
criterion = nn.BCEWithLogitsLoss()
# 定義優化器
optim = torch.optim.SGD(param, 1.)
# 我們訓練 10000 次
for e in range(10000):
out = seq_net(Variable(x))
loss = criterion(out, Variable(y))
optim.zero_grad()
loss.backward()
optim.step()
if (e + 1) % 1000 == 0:
print('epoch: {}, loss: {}'.format(e+1, loss.item()))
def plot_seq(x):
out = F.sigmoid(seq_net(Variable(torch.from_numpy(x).float()))).data.numpy()
out = (out > 0.5) * 1
return out
plot_decision_boundary(lambda x: plot_seq(x), x.numpy(), y.numpy())
plt.title('sequential')
plt.show()
plt.close()
#兩種保存模型的辦法
# 1.將參數和模型保存在一起
torch.save(seq_net, 'save_seq_net.pth')
# 2.保存模型參數
torch.save(seq_net.state_dict(), 'save_seq_net_params.pth')
def load_moudle():
# 讀取保存的模型
seq_net1 = torch.load('save_seq_net.pth')
print(seq_net1)
print(seq_net1[0].weight)
seq_net2 = nn.Sequential(
nn.Linear(2, 4),
nn.Tanh(),
nn.Linear(4, 1)
)
seq_net2.load_state_dict(torch.load('save_seq_net_params.pth'))
print(seq_net2)
print(seq_net2[0].weight)
def module_train(x,y):
class module_net(nn.Module):
def __init__(self, num_input, num_hidden, num_output):
super(module_net, self).__init__()
self.layer1 = nn.Linear(num_input, num_hidden)
self.layer2 = nn.Tanh()
self.layer3 = nn.Linear(num_hidden, num_output)
def forward(self, x):
x = self.layer1(x)
x = self.layer2(x)
x = self.layer3(x)
return x
x = torch.from_numpy(x).float()
y = torch.from_numpy(y).float()
mo_net = module_net(2, 4, 1)
optim = torch.optim.SGD(mo_net.parameters(), 1.)
criterion = nn.BCEWithLogitsLoss()
# 我們訓練 10000 次
for e in range(10000):
out = mo_net(Variable(x))
loss = criterion(out, Variable(y))
optim.zero_grad()
loss.backward()
optim.step()
if (e + 1) % 1000 == 0:
print('epoch: {}, loss: {}'.format(e+1, loss.item()))
if __name__ == "__main__":
x,y=draw_picture()
logistic_train(x,y)# 自定義兩層神經網絡
sequential_train(x,y)# 定義sequential神經網絡
load_moudle()
module_train(x,y)# 定義module神經網絡
