请给我python代码要求如下：1对PTB心电数据集进行滤波与降噪的预处理，画出波形，PTB心电数据集地址D:\下载\ptb-xl-a-large-publicly-available-electrocardiography-dataset-1.0.3\records100为；2.使用CNN+RNN进行特征提取；3.使用RNN+CNN进行复原，画出波形；4使用pytorch

好的，以下是符合您要求的 Python 代码：

import wfdb
import numpy as np
import matplotlib.pyplot as plt
import torch
import torch.nn as nn
import torch.optim as optim

# 读取心电数据
record = wfdb.rdrecord('D:/下载/ptb-xl-a-large-publicly-available-electrocardiography-dataset-1.0.3/records100/00000')
signal = record.p_signal.flatten()

# 设置滤波器参数
fs = record.fs
nyq = 0.5 * fs
low = 5 / nyq
high = 15 / nyq
order = 2

# Butterworth 滤波器
b, a = signal.butter(order, [low, high], btype='band')
filtered_signal = signal.filtfilt(b, a)

# 绘制滤波后的心电图波形
plt.plot(filtered_signal)
plt.title('Filtered ECG Signal')
plt.xlabel('Time (ms)')
plt.ylabel('Amplitude (mV)')
plt.show()

# 噪声降噪
noisy_signal = filtered_signal + np.random.normal(0, 0.5, len(filtered_signal))
denoised_signal = signal.wiener(noisy_signal)

# 绘制降噪后的心电图波形
plt.plot(denoised_signal)
plt.title('Denoised ECG Signal')
plt.xlabel('Time (ms)')
plt.ylabel('Amplitude (mV)')
plt.show()

# 特征提取CNN + RNN
class ECGNet(nn.Module):
    def __init__(self):
        super(ECGNet, self).__init__()
        self.conv1 = nn.Conv1d(1, 64, kernel_size=5, stride=1, padding=2)
        self.bn1 = nn.BatchNorm1d(64)
        self.pool1 = nn.MaxPool1d(kernel_size=2, stride=2)
        self.conv2 = nn.Conv1d(64, 128, kernel_size=5, stride=1, padding=2)
        self.bn2 = nn.BatchNorm1d(128)
        self.pool2 = nn.MaxPool1d(kernel_size=2, stride=2)
        self.lstm = nn.LSTM(input_size=128, hidden_size=64, batch_first=True)
        self.fc1 = nn.Linear(64, 1)
        
    def forward(self, x):
        x = self.conv1(x)
        x = self.bn1(x)
        x = nn.functional.relu(x)
        x = self.pool1(x)
        x = self.conv2(x)
        x = self.bn2(x)
        x = nn.functional.relu(x)
        x = self.pool2(x)
        x = x.permute(0, 2, 1)
        _, (h_n, _) = self.lstm(x)
        x = self.fc1(h_n.squeeze(0))
        return x

# 复原 RNN + CNN
class ECGRecover(nn.Module):
    def __init__(self):
        super(ECGRecover, self).__init__()
        self.fc1 = nn.Linear(1, 64)
        self.lstm = nn.LSTM(input_size=64, hidden_size=128, batch_first=True)
        self.upsample1 = nn.ConvTranspose1d(128, 64, kernel_size=4, stride=2, padding=1)
        self.bn1 = nn.BatchNorm1d(64)
        self.upsample2 = nn.ConvTranspose1d(64, 1, kernel_size=4, stride=2, padding=1)
        self.bn2 = nn.BatchNorm1d(1)
        
    def forward(self, x):
        x = self.fc1(x)
        x = x.unsqueeze(1)
        x = x.repeat(1, 128, 1)
        x, _ = self.lstm(x)
        x = self.upsample1(x)
        x = self.bn1(x)
        x = nn.functional.relu(x)
        x = self.upsample2(x)
        x = self.bn2(x)
        x = nn.functional.relu(x)
        return x.squeeze(1)

# 训练模型
ecg_net = ECGNet()
ecg_recover = ECGRecover()
criterion = nn.MSELoss()
optimizer = optim.Adam(list(ecg_net.parameters()) + list(ecg_recover.parameters()), lr=0.001)

train_data = np.array_split(denoised_signal, len(denoised_signal) // 2000)
train_data = [torch.FloatTensor(x.reshape(1, -1)) for x in train_data]

for epoch in range(100):
    running_loss = 0.0
    for i, data in enumerate(train_data):
        optimizer.zero_grad()
        ecg_features = ecg_net(data)
        ecg_recovered = ecg_recover(ecg_features)
        loss = criterion(ecg_recovered, data)
        loss.backward()
        optimizer.step()
        running_loss += loss.item()
    print('Epoch %d: Loss = %.4f' % (epoch+1, running_loss / len(train_data)))

# 绘制复原后的心电图波形
with torch.no_grad():
    ecg_features = ecg_net(torch.FloatTensor(denoised_signal.reshape(1, -1)))
    ecg_recovered = ecg_recover(ecg_features)
    plt.plot(ecg_recovered.numpy().flatten())
    plt.title('Recovered ECG Signal')
    plt.xlabel('Time (ms)')
    plt.ylabel('Amplitude (mV)')
    plt.show()

这段代码使用了 PyTorch 框架，实现了对 PTB 心电数据集进行滤波与降噪的预处理，使用 CNN + RNN 进行特征提取，使用 RNN + CNN 进行复原，并绘制了波形图。如果您需要对代码中的参数或模型结构进行修改以得到更好的结果，可以自行进行尝试。