diff --git a/Neural Networks/RNN-Traffic-Prediction/.gitattributes b/Neural Networks/RNN-Traffic-Prediction/.gitattributes new file mode 100644 index 00000000..dfe07704 --- /dev/null +++ b/Neural Networks/RNN-Traffic-Prediction/.gitattributes @@ -0,0 +1,2 @@ +# Auto detect text files and perform LF normalization +* text=auto diff --git a/Neural Networks/RNN-Traffic-Prediction/1-RNN.ipynb b/Neural Networks/RNN-Traffic-Prediction/1-RNN.ipynb new file mode 100644 index 00000000..4a45c355 --- /dev/null +++ b/Neural Networks/RNN-Traffic-Prediction/1-RNN.ipynb @@ -0,0 +1,547 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 1, + "id": "initial_id", + "metadata": { + "collapsed": true, + "ExecuteTime": { + "end_time": "2023-12-19T11:03:06.646836Z", + "start_time": "2023-12-19T11:03:05.207381Z" + } + }, + "outputs": [], + "source": [ + "import numpy as np\n", + "import pandas as pd\n", + "import matplotlib.pyplot as plt\n", + "import torch\n", + "from torch import nn\n", + "from utils import *\n", + "from torch.utils.data import Dataset, DataLoader" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "outputs": [ + { + "data": { + "text/plain": "
", + "image/png": "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\n" + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "# 数据可视化,以前500次测量为例,省略了部分代码\n", + "# plt.plot(data[0:500])\n", + "# plt.title(\"Traffic Flow By Time\")\n", + "# plt.xlabel(\"Time\")\n", + "# plt.ylabel(\"Traffic Flow\")\n", + "# plt.show()" + ], + "metadata": { + "collapsed": false, + "ExecuteTime": { + "end_time": "2023-12-19T10:51:02.733621Z", + "start_time": "2023-12-19T10:51:02.646652Z" + } + }, + "id": "5c0711251dada4c1" + }, + { + "cell_type": "code", + "execution_count": 3, + "outputs": [], + "source": [ + "# 自定义数据集，用于创建Dataloader\n", + "class my_Dataset(Dataset):\n", + " def __init__(self, features, labels):\n", + " self.X = features\n", + " self.y = labels\n", + "\n", + " def __getitem__(self, index):\n", + " return self.X[index], self.y[index]\n", + "\n", + " def __len__(self):\n", + " return self.X.shape[0]" + ], + "metadata": { + "collapsed": false, + "ExecuteTime": { + "end_time": "2023-12-19T11:03:29.330372Z", + "start_time": "2023-12-19T11:03:29.325339Z" + } + }, + "id": "bd9fc1fa4d90b61c" + }, + { + "cell_type": "code", + "execution_count": 4, + "outputs": [], + "source": [ + "# Traffic数据集\n", + "class TrafficDataset:\n", + " def __init__(self):\n", + " #只取第一个传感器每5分钟采集到的车流量数据这一个维度来进行预测，其余传感器同理\n", + " self.raw_data = np.load('./dataset/traffic-flow/traffic.npz')['data'][:,0,0]\n", + " # 数据标准化\n", + " self.min = self.raw_data.min()\n", + " self.max = self.raw_data.max()\n", + " self.raw_data = (self.raw_data - self.min) / (self.max - self.min)\n", + " # 数据反标准化\n", + " def denormalize(self, x):\n", + " return x * (self.max - self.min) + self.min\n", + "\n", + " def construct_set(self, train_por,test_por, window_size): \n", + " # 补全构造过程\n", + " train_data_length=round(len(self.raw_data)*train_por)\n", + " train_set_raw_data=self.raw_data[0:train_data_length]\n", + " train_x=[]\n", + " train_y=[]\n", + " #窗口化处理\n", + " for i in range(train_set_raw_data.shape[0] - window_size + 1):\n", + " train_x.append(train_set_raw_data[i:i+window_size].tolist())\n", + " #每次新进来的数据作为上一个输入数据的输出结果，即“标签”\n", + " for i in range(window_size,len(train_set_raw_data)):\n", + " train_y.append(train_set_raw_data[i].tolist())\n", + " train_x=train_x[:-1]\n", + " \n", + " val_data_length=round(len(self.raw_data)*(train_por+test_por))\n", + " val_set_raw_data=self.raw_data[train_data_length:val_data_length]\n", + " val_x=[]\n", + " val_y=[]\n", + " for i in range(val_set_raw_data.shape[0] - window_size + 1):\n", + " val_x.append(val_set_raw_data[i:i+window_size].tolist())\n", + " for i in range(window_size,len(val_set_raw_data)):\n", + " val_y.append(val_set_raw_data[i].tolist())\n", + " val_x=val_x[:-1]\n", + " \n", + " test_set_raw_data=self.raw_data[val_data_length:]\n", + " test_x=[]\n", + " test_y=[]\n", + " for i in range(test_set_raw_data.shape[0] - window_size + 1):\n", + " test_x.append(test_set_raw_data[i:i+window_size].tolist())\n", + " for i in range(window_size,len(test_set_raw_data)):\n", + " test_y.append(test_set_raw_data[i].tolist())\n", + " test_x=test_x[:-1]\n", + "\n", + " train_set = my_Dataset(torch.Tensor(train_x), torch.Tensor(train_y))\n", + " val_set = my_Dataset(torch.Tensor(val_x), torch.Tensor(val_y))\n", + " test_set = my_Dataset(torch.Tensor(test_x), torch.Tensor(test_y))\n", + " return train_set, val_set, test_set" + ], + "metadata": { + "collapsed": false, + "ExecuteTime": { + "end_time": "2023-12-19T11:03:31.143471Z", + "start_time": "2023-12-19T11:03:31.141375Z" + } + }, + "id": "9925005738f38bb" + }, + { + "cell_type": "code", + "execution_count": 5, + "outputs": [], + "source": [ + "#创建数据集\n", + "TrafficData=TrafficDataset()" + ], + "metadata": { + "collapsed": false, + "ExecuteTime": { + "end_time": "2023-12-19T11:03:32.654702Z", + "start_time": "2023-12-19T11:03:32.496069Z" + } + }, + "id": "ba543889d8ed6ecb" + }, + { + "cell_type": "code", + "execution_count": 6, + "outputs": [], + "source": [ + "#构造数据\n", + "train_set,val_set,test_set=TrafficData.construct_set(0.6,0.2,12)" + ], + "metadata": { + "collapsed": false, + "ExecuteTime": { + "end_time": "2023-12-19T11:03:34.389043Z", + "start_time": "2023-12-19T11:03:34.382077Z" + } + }, + "id": "8df7e5675d0d6018" + }, + { + "cell_type": "code", + "execution_count": 7, + "outputs": [], + "source": [ + "#升维，便于网络处理\n", + "train_set.X=torch.unsqueeze(train_set.X,2)\n", + "val_set.X=torch.unsqueeze(val_set.X,2)\n", + "test_set.X=torch.unsqueeze(test_set.X,2)" + ], + "metadata": { + "collapsed": false, + "ExecuteTime": { + "end_time": "2023-12-19T11:03:36.565335Z", + "start_time": "2023-12-19T11:03:36.560315Z" + } + }, + "id": "f7ea14373370133a" + }, + { + "cell_type": "code", + "execution_count": 8, + "outputs": [], + "source": [ + "#手动创建RNN网络\n", + "class MyRNN(nn.Module):\n", + " def __init__(self, input_size, hidden_size, output_size):\n", + " \"\"\"\n", + " :param input_size: 指定输入数据的维度。例如，对于简单的时间序列预测问题，每一步的输入均为一个采样值，因此input_size=1.\n", + " :param hidden_size: 指定隐藏状态的维度。这个值并不受输入和输出控制，但会影响模型的容量。\n", + " :param output_size: 指定输出数据的维度。此值取决于具体的预测要求。例如，对简单的时间序列预测问题，output_size=1.\n", + " \"\"\"\n", + " super().__init__()\n", + " self.hidden_size = hidden_size\n", + " \n", + " # 可学习参数的维度设置，可以类比一下全连接网络的实现。其维度取决于输入数据的维度，以及指定的隐藏状态维度。\n", + " self.w_h = nn.Parameter(torch.rand(input_size, hidden_size))\n", + " self.u_h = nn.Parameter(torch.rand(hidden_size, hidden_size))\n", + " self.b_h = nn.Parameter(torch.zeros(hidden_size))\n", + " \n", + " self.w_y = nn.Parameter(torch.rand(hidden_size, output_size))\n", + " self.b_y = nn.Parameter(torch.zeros(output_size))\n", + " \n", + " # 准备激活函数。Dropout函数可选。\n", + " self.tanh = nn.Tanh()\n", + " self.leaky_relu = nn.LeakyReLU()\n", + " \n", + " # 可选：使用性能更好的参数初始化函数\n", + " for param in self.parameters():\n", + " if param.dim() > 1:\n", + " nn.init.xavier_uniform_(param)\n", + " \n", + " def forward(self, x):\n", + " \"\"\"\n", + " :param x: 输入序列。一般来说，此输入包含三个维度：batch，序列长度，以及每条数据的特征。\n", + " \"\"\"\n", + " batch_size = x.size(0)\n", + " seq_len = x.size(1)\n", + " \n", + " # 初始化隐藏状态，一般设为全0。由于是内部新建的变量，需要同步设备位置。\n", + " h = torch.zeros(batch_size, self.hidden_size).to(x.device)\n", + " # RNN实际上只能一步一步处理序列。因此需要用循环迭代。\n", + " y_list = []\n", + " for i in range(seq_len):\n", + " h = self.tanh(torch.matmul(x[:, i, :], self.w_h) + \n", + " torch.matmul(h, self.u_h) + self.b_h) # (batch_size, hidden_size)\n", + " y = self.leaky_relu(torch.matmul(h, self.w_y) + self.b_y) # (batch_size, output_size)\n", + " y_list.append(y)\n", + " # 一般来说，RNN的返回值为最后一步的隐藏状态，以及每一步的输出状态。\n", + " return torch.stack(y_list, dim=1), h" + ], + "metadata": { + "collapsed": false, + "ExecuteTime": { + "end_time": "2023-12-19T11:03:38.959095Z", + "start_time": "2023-12-19T11:03:38.955551Z" + } + }, + "id": "b6e32c1be19a4bd3" + }, + { + "cell_type": "code", + "execution_count": 9, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "torch.Size([32, 12, 1]) torch.Size([32, 32])\n" + ] + } + ], + "source": [ + "device = 'mps' # MacOS\n", + "batch_size = 32\n", + "seq_len = 12\n", + "input_size = 2\n", + "hidden_size = 32\n", + "output_size = 1\n", + "\n", + "x = torch.rand(batch_size, seq_len, input_size).to(device)\n", + "rnn = MyRNN(input_size, hidden_size, output_size).to(device)\n", + "hidden, y = rnn(x)\n", + "print(hidden.shape, y.shape)" + ], + "metadata": { + "collapsed": false, + "ExecuteTime": { + "end_time": "2023-12-19T11:03:47.261489Z", + "start_time": "2023-12-19T11:03:47.160379Z" + } + }, + "id": "8a4644957f3a307b" + }, + { + "cell_type": "code", + "execution_count": 10, + "outputs": [], + "source": [ + "#封装DataLoader\n", + "train_loader = DataLoader(train_set, batch_size=batch_size, shuffle=True, num_workers=0, drop_last=True)\n", + "val_loader = DataLoader(val_set, batch_size=batch_size, shuffle=False, num_workers=0, drop_last=True)\n", + "test_loader = DataLoader(test_set, batch_size=batch_size, shuffle=False, num_workers=0, drop_last=True)" + ], + "metadata": { + "collapsed": false, + "ExecuteTime": { + "end_time": "2023-12-19T11:03:49.535049Z", + "start_time": "2023-12-19T11:03:49.530787Z" + } + }, + "id": "986e2c5d359864d5" + }, + { + "cell_type": "code", + "execution_count": 11, + "outputs": [], + "source": [ + "#设置超参数\n", + "device = torch.device(\"mps\")\n", + "input_size = train_set.X.shape[-1]\n", + "hidden_size = 64\n", + "output_size = 1\n", + "seq_len = 12\n", + "lr = 0.0001\n", + "epochs = 80\n", + "loss_func = nn.MSELoss()\n", + "\n", + "my_rnn = MyRNN(input_size, hidden_size, output_size).to(device)\n", + "\n", + "optimizer = torch.optim.Adam(my_rnn.parameters(), lr)" + ], + "metadata": { + "collapsed": false, + "ExecuteTime": { + "end_time": "2023-12-19T11:03:53.526506Z", + "start_time": "2023-12-19T11:03:53.015525Z" + } + }, + "id": "3a794ac0a3f1555" + }, + { + "cell_type": "code", + "execution_count": 12, + "outputs": [], + "source": [ + "#评估函数\n", + "def mape(y_true, y_pred):\n", + " y_true, y_pred = np.array(y_true), np.array(y_pred)\n", + " non_zero_index = (y_true > 0)\n", + " y_true = y_true[non_zero_index]\n", + " y_pred = y_pred[non_zero_index]\n", + "\n", + " mape = np.abs((y_true - y_pred) / y_true)\n", + " mape[np.isinf(mape)] = 0\n", + " return np.mean(mape) * 100" + ], + "metadata": { + "collapsed": false, + "ExecuteTime": { + "end_time": "2023-12-19T11:03:55.093778Z", + "start_time": "2023-12-19T11:03:55.087792Z" + } + }, + "id": "ca88dd4138c940b0" + }, + { + "cell_type": "code", + "execution_count": 13, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "*** epoch1, train loss 0.0147, train rmse 55.9098, val loss 0.0028, val rmse 32.253209, time use 4.055s\n", + "*** epoch2, train loss 0.0022, train rmse 29.8158, val loss 0.0024, val rmse 29.998086, time use 3.913s\n", + "*** epoch3, train loss 0.0020, train rmse 28.3517, val loss 0.0023, val rmse 29.328176, time use 3.915s\n", + "*** epoch4, train loss 0.0019, train rmse 27.6689, val loss 0.0022, val rmse 28.818703, time use 3.913s\n", + "*** epoch5, train loss 0.0018, train rmse 27.1487, val loss 0.0023, val rmse 29.412237, time use 3.882s\n", + "*** epoch6, train loss 0.0018, train rmse 27.0165, val loss 0.0021, val rmse 28.247436, time use 3.825s\n", + "*** epoch7, train loss 0.0018, train rmse 26.9224, val loss 0.0021, val rmse 28.274240, time use 3.826s\n", + "*** epoch8, train loss 0.0018, train rmse 26.7306, val loss 0.0024, val rmse 30.115352, time use 3.850s\n", + "*** epoch9, train loss 0.0018, train rmse 26.5642, val loss 0.0021, val rmse 28.103900, time use 3.847s\n", + "*** epoch10, train loss 0.0017, train rmse 26.3491, val loss 0.0021, val rmse 28.168528, time use 3.851s\n", + "*** epoch11, train loss 0.0017, train rmse 26.3540, val loss 0.0021, val rmse 27.697132, time use 3.911s\n", + "*** epoch12, train loss 0.0017, train rmse 26.2239, val loss 0.0021, val rmse 27.923170, time use 3.890s\n", + "*** epoch13, train loss 0.0017, train rmse 26.3193, val loss 0.0020, val rmse 27.729645, time use 3.820s\n", + "*** epoch14, train loss 0.0017, train rmse 26.1079, val loss 0.0021, val rmse 27.904116, time use 3.842s\n", + "*** epoch15, train loss 0.0017, train rmse 26.0709, val loss 0.0021, val rmse 27.744417, time use 3.840s\n", + "*** epoch16, train loss 0.0017, train rmse 26.1314, val loss 0.0020, val rmse 27.676445, time use 3.854s\n", + "*** epoch17, train loss 0.0017, train rmse 26.1999, val loss 0.0022, val rmse 28.739625, time use 3.859s\n", + "*** epoch18, train loss 0.0017, train rmse 26.1475, val loss 0.0020, val rmse 27.503687, time use 3.871s\n", + "*** epoch19, train loss 0.0017, train rmse 26.0901, val loss 0.0020, val rmse 27.466388, time use 3.864s\n", + "*** epoch20, train loss 0.0017, train rmse 26.0980, val loss 0.0020, val rmse 27.473209, time use 3.832s\n", + "*** epoch21, train loss 0.0017, train rmse 26.0270, val loss 0.0020, val rmse 27.474716, time use 3.804s\n", + "*** epoch22, train loss 0.0017, train rmse 25.9111, val loss 0.0020, val rmse 27.389196, time use 3.820s\n", + "*** epoch23, train loss 0.0017, train rmse 26.0672, val loss 0.0021, val rmse 27.787729, time use 3.836s\n", + "*** epoch24, train loss 0.0017, train rmse 26.0017, val loss 0.0020, val rmse 27.646956, time use 3.848s\n", + "*** epoch25, train loss 0.0017, train rmse 26.0015, val loss 0.0020, val rmse 27.343402, time use 3.851s\n", + "*** epoch26, train loss 0.0017, train rmse 26.0259, val loss 0.0020, val rmse 27.346988, time use 3.857s\n", + "*** epoch27, train loss 0.0017, train rmse 25.9359, val loss 0.0020, val rmse 27.648784, time use 3.873s\n", + "*** epoch28, train loss 0.0017, train rmse 26.1408, val loss 0.0020, val rmse 27.332677, time use 3.816s\n", + "*** epoch29, train loss 0.0017, train rmse 25.9372, val loss 0.0020, val rmse 27.319761, time use 3.814s\n", + "*** epoch30, train loss 0.0017, train rmse 25.9813, val loss 0.0020, val rmse 27.334266, time use 3.812s\n", + "*** epoch31, train loss 0.0017, train rmse 25.8559, val loss 0.0020, val rmse 27.436516, time use 3.836s\n", + "*** epoch32, train loss 0.0017, train rmse 26.0346, val loss 0.0021, val rmse 28.052707, time use 3.862s\n", + "*** epoch33, train loss 0.0017, train rmse 25.8875, val loss 0.0020, val rmse 27.359147, time use 3.859s\n", + "*** epoch34, train loss 0.0017, train rmse 25.9423, val loss 0.0022, val rmse 28.462561, time use 3.866s\n", + "*** epoch35, train loss 0.0017, train rmse 25.8537, val loss 0.0020, val rmse 27.264209, time use 3.847s\n", + "*** epoch36, train loss 0.0017, train rmse 25.9300, val loss 0.0020, val rmse 27.509944, time use 3.810s\n", + "*** epoch37, train loss 0.0017, train rmse 25.8902, val loss 0.0020, val rmse 27.307579, time use 3.825s\n", + "*** epoch38, train loss 0.0017, train rmse 25.9006, val loss 0.0020, val rmse 27.345387, time use 3.848s\n", + "*** epoch39, train loss 0.0017, train rmse 25.8137, val loss 0.0020, val rmse 27.219189, time use 3.840s\n", + "*** epoch40, train loss 0.0017, train rmse 25.8570, val loss 0.0020, val rmse 27.599184, time use 3.848s\n", + "*** epoch41, train loss 0.0017, train rmse 25.7942, val loss 0.0020, val rmse 27.683255, time use 3.865s\n", + "*** epoch42, train loss 0.0017, train rmse 25.7896, val loss 0.0020, val rmse 27.335772, time use 3.875s\n", + "*** epoch43, train loss 0.0017, train rmse 25.8175, val loss 0.0020, val rmse 27.142766, time use 3.816s\n", + "*** epoch44, train loss 0.0017, train rmse 25.9386, val loss 0.0020, val rmse 27.256869, time use 3.822s\n", + "*** epoch45, train loss 0.0017, train rmse 25.8039, val loss 0.0020, val rmse 27.491751, time use 3.832s\n", + "*** epoch46, train loss 0.0017, train rmse 25.7810, val loss 0.0020, val rmse 27.116621, time use 3.846s\n", + "*** epoch47, train loss 0.0017, train rmse 25.7910, val loss 0.0020, val rmse 27.120560, time use 3.852s\n", + "*** epoch48, train loss 0.0017, train rmse 25.8839, val loss 0.0020, val rmse 27.372101, time use 3.876s\n", + "*** epoch49, train loss 0.0017, train rmse 25.9491, val loss 0.0020, val rmse 27.180260, time use 3.875s\n", + "*** epoch50, train loss 0.0017, train rmse 25.8008, val loss 0.0020, val rmse 27.202318, time use 3.861s\n", + "*** epoch51, train loss 0.0017, train rmse 25.9265, val loss 0.0021, val rmse 27.749886, time use 3.845s\n", + "*** epoch52, train loss 0.0017, train rmse 25.7136, val loss 0.0020, val rmse 27.099871, time use 3.856s\n", + "*** epoch53, train loss 0.0017, train rmse 25.7649, val loss 0.0020, val rmse 27.271203, time use 3.848s\n", + "*** epoch54, train loss 0.0017, train rmse 25.6730, val loss 0.0020, val rmse 27.131278, time use 3.862s\n", + "*** epoch55, train loss 0.0017, train rmse 25.8009, val loss 0.0020, val rmse 27.071622, time use 3.861s\n", + "*** epoch56, train loss 0.0017, train rmse 25.7266, val loss 0.0020, val rmse 27.130786, time use 3.866s\n", + "*** epoch57, train loss 0.0017, train rmse 25.7280, val loss 0.0020, val rmse 27.559120, time use 3.909s\n", + "*** epoch58, train loss 0.0017, train rmse 25.8104, val loss 0.0020, val rmse 27.075410, time use 3.813s\n", + "*** epoch59, train loss 0.0017, train rmse 25.6838, val loss 0.0020, val rmse 27.271316, time use 3.820s\n", + "*** epoch60, train loss 0.0017, train rmse 25.7400, val loss 0.0020, val rmse 27.379531, time use 3.805s\n", + "*** epoch61, train loss 0.0017, train rmse 25.6797, val loss 0.0020, val rmse 27.260481, time use 3.836s\n", + "*** epoch62, train loss 0.0017, train rmse 25.6446, val loss 0.0020, val rmse 27.082587, time use 3.820s\n", + "*** epoch63, train loss 0.0017, train rmse 25.7089, val loss 0.0020, val rmse 27.181398, time use 3.858s\n", + "*** epoch64, train loss 0.0017, train rmse 25.7784, val loss 0.0020, val rmse 27.056911, time use 3.882s\n", + "*** epoch65, train loss 0.0017, train rmse 25.7111, val loss 0.0020, val rmse 27.075189, time use 3.834s\n", + "*** epoch66, train loss 0.0017, train rmse 25.6829, val loss 0.0020, val rmse 27.022135, time use 3.827s\n", + "*** epoch67, train loss 0.0017, train rmse 25.7524, val loss 0.0020, val rmse 27.277432, time use 3.799s\n", + "*** epoch68, train loss 0.0017, train rmse 25.6617, val loss 0.0020, val rmse 27.019056, time use 3.838s\n", + "*** epoch69, train loss 0.0016, train rmse 25.6185, val loss 0.0020, val rmse 27.039056, time use 3.840s\n", + "*** epoch70, train loss 0.0017, train rmse 25.6978, val loss 0.0020, val rmse 27.053015, time use 3.842s\n", + "*** epoch71, train loss 0.0017, train rmse 25.7275, val loss 0.0020, val rmse 27.075493, time use 3.869s\n", + "*** epoch72, train loss 0.0017, train rmse 25.8320, val loss 0.0020, val rmse 27.148115, time use 3.860s\n", + "*** epoch73, train loss 0.0017, train rmse 25.6917, val loss 0.0020, val rmse 27.128534, time use 3.805s\n", + "*** epoch74, train loss 0.0017, train rmse 25.7821, val loss 0.0020, val rmse 27.051987, time use 3.781s\n", + "*** epoch75, train loss 0.0017, train rmse 25.6910, val loss 0.0020, val rmse 27.446969, time use 3.780s\n", + "*** epoch76, train loss 0.0017, train rmse 25.7446, val loss 0.0020, val rmse 27.338370, time use 3.785s\n", + "*** epoch77, train loss 0.0017, train rmse 25.7886, val loss 0.0020, val rmse 27.047652, time use 3.797s\n", + "*** epoch78, train loss 0.0017, train rmse 25.7128, val loss 0.0019, val rmse 27.031457, time use 3.816s\n", + "*** epoch79, train loss 0.0017, train rmse 25.6337, val loss 0.0020, val rmse 27.176096, time use 3.826s\n", + "*** epoch80, train loss 0.0017, train rmse 25.6600, val loss 0.0020, val rmse 27.569620, time use 3.852s\n", + "\n", + "****************************************\n", + "Final result:\n", + "Get best validation rmse 27.0191 at epoch 67\n", + "Total time 307.78s\n", + "Test result:\n", + "Test RMSE: 27.72809789202586 Test MAE: 21.731021717861967 Test MAPE: 8.06397219096218\n" + ] + } + ], + "source": [ + "#训练\n", + "train_loss_lst, val_loss_lst, \\\n", + " train_score_lst, val_score_lst, stop_epoch = train(my_rnn, train_loader, val_loader, test_loader,\n", + " loss_func, TrafficData.denormalize, optimizer, epochs,\n", + " early_stop=20, device=device, output_model=None)" + ], + "metadata": { + "collapsed": false, + "ExecuteTime": { + "end_time": "2023-12-19T11:09:54.475044Z", + "start_time": "2023-12-19T11:03:58.269558Z" + } + }, + "id": "4384d9947043ba99" + }, + { + "cell_type": "code", + "execution_count": 15, + "outputs": [ + { + "data": { + "text/plain": "
", + "image/png": "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\n" + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "text/plain": "
", + "image/png": "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\n" + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "#模型结果可视化\n", + "visualize(stop_epoch, train_loss_lst, val_loss_lst, y_label='Loss')\n", + "plot_metric(train_score_lst)" + ], + "metadata": { + "collapsed": false, + "ExecuteTime": { + "end_time": "2023-12-19T11:10:21.250538Z", + "start_time": "2023-12-19T11:10:20.941235Z" + } + }, + "id": "77c76adef4ec54ff" + }, + { + "cell_type": "code", + "execution_count": null, + "outputs": [], + "source": [], + "metadata": { + "collapsed": false + }, + "id": "452966495837ceaa" + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 2 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython2", + "version": "2.7.6" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/Neural Networks/RNN-Traffic-Prediction/utils.py b/Neural Networks/RNN-Traffic-Prediction/utils.py new file mode 100644 index 00000000..6c542ba9 --- /dev/null +++ b/Neural Networks/RNN-Traffic-Prediction/utils.py @@ -0,0 +1,174 @@ +import math +import torch +from torch.utils import data +import torch.nn as nn +from matplotlib import pyplot as plt +from sklearn.metrics import mean_squared_error as mse_fn, mean_absolute_error as mae_fn +import numpy as np +import time + + +def mape_fn(y, pred): + mask = y != 0 + y = y[mask] + pred = pred[mask] + mape = np.abs((y - pred) / y) + mape = np.mean(mape) * 100 + return mape + + +def eval(y, pred): + y = y.cpu().numpy() + pred = pred.cpu().numpy() + mse = mse_fn(y, pred) + rmse = math.sqrt(mse) + mae = mae_fn(y, pred) + mape = mape_fn(y, pred) + return [rmse, mae, mape] + + +# 测试函数（用于分类） +def test(net, output_model, data_iter, loss_fn, denormalize_fn, device='cpu'): + rmse, mae, mape = 0, 0, 0 + batch_count = 0 + total_loss = 0.0 + net.eval() + if output_model is not None: + output_model.eval() + for X, Y in data_iter: + X = X.to(device).float() + Y = Y.to(device).float() + output, hidden = net(X) + if output_model is not None: + y_hat = output_model(output[:, -1, :].squeeze(-1)).squeeze(-1) + else: + y_hat = output[:, -1, :].squeeze(-1) + loss = loss_fn(y_hat, Y) + + Y = denormalize_fn(Y) + y_hat = denormalize_fn(y_hat) + a, b, c = eval(Y.detach(), y_hat.detach()) + rmse += a + mae += b + mape += c + total_loss += loss.detach().cpu().numpy().tolist() + batch_count += 1 + return [rmse / batch_count, mae / batch_count, mape / batch_count], total_loss / batch_count + + +def train(net, train_iter, val_iter, test_iter, loss_fn, denormalize_fn, optimizer, num_epoch, + early_stop=10, device='cpu', output_model=None, is_print=True, is_print_batch=False): + train_loss_lst = [] + val_loss_lst = [] + train_score_lst = [] + val_score_lst = [] + epoch_time = [] + + best_epoch = 0 + best_val_rmse = 9999 + early_stop_flag = 0 + for epoch in range(num_epoch): + net.train() + if output_model is not None: + output_model.train() + epoch_loss = 0 + batch_count = 0 + batch_time = [] + rmse, mae, mape = 0, 0, 0 + for X, Y in train_iter: + batch_s = time.time() + X = X.to(device).float() + Y = Y.to(device).float() + optimizer.zero_grad() + output, hidden = net(X) + if output_model is not None: + y_hat = output_model(output[:, -1, :].squeeze(-1)).squeeze() + else: + y_hat = output[:, -1, :].squeeze(-1) + loss = loss_fn(y_hat, Y) + loss.backward() + optimizer.step() + + Y = denormalize_fn(Y) + y_hat = denormalize_fn(y_hat) + a, b, c = eval(Y.detach(), y_hat.detach()) + rmse += a + mae += b + mape += c + epoch_loss += loss.detach().cpu().numpy().tolist() + batch_count += 1 + # sample_num += X.shape[0] + + batch_time.append(time.time() - batch_s) + if is_print and is_print_batch: + print('epoch-batch: %d-%d, train loss %.4f, time use %.3fs' % + (epoch + 1, batch_count, epoch_loss, batch_time[-1])) + + train_loss = epoch_loss / batch_count + train_loss_lst.append(train_loss) + train_score_lst.append([rmse/batch_count, mae/batch_count, mape/batch_count]) + + # 验证集 + val_score, val_loss = test(net, output_model, val_iter, loss_fn, denormalize_fn, device) + val_score_lst.append(val_score) + val_loss_lst.append(val_loss) + + epoch_time.append(np.array(batch_time).sum()) + + # 打印本轮训练结果 + if is_print: + print('*** epoch%d, train loss %.4f, train rmse %.4f, val loss %.4f, val rmse %.6f, time use %.3fs' % + (epoch + 1, train_loss, train_score_lst[-1][0], val_loss, val_score[0], epoch_time[-1])) + + # 早停 + if val_score[0] < best_val_rmse: + best_val_rmse = val_score[0] + best_epoch = epoch + early_stop_flag = 0 + else: + early_stop_flag += 1 + if early_stop_flag == early_stop: + print(f'\nThe model has not been improved for {early_stop} rounds. Stop early!') + break + + # 输出最终训练结果 + print(f'\n{"*" * 40}\nFinal result:') + print(f'Get best validation rmse {np.array(val_score_lst)[:, 0].min() :.4f} ' + f'at epoch {best_epoch}') + print(f'Total time {np.array(epoch_time).sum():.2f}s') + print() + + # 计算测试集效果 + test_score, test_loss = test(net, output_model, test_iter, loss_fn, denormalize_fn, device) + print('Test result:') + print(f'Test RMSE: {test_score[0]} Test MAE: {test_score[1]} Test MAPE: {test_score[2]}') + return train_loss_lst, val_loss_lst, train_score_lst, val_score_lst, epoch + + +def visualize(num_epochs, train_data, test_data, x_label='epoch', y_label='loss'): + x = np.arange(0, num_epochs + 1).astype(dtype=np.int) + plt.plot(x, train_data, label=f"train_{y_label}", linewidth=1.5) + plt.plot(x, test_data, label=f"val_{y_label}", linewidth=1.5) + plt.xlabel(x_label) + plt.ylabel(y_label) + plt.legend() + plt.show() + + +def plot_metric(score_log): + score_log = np.array(score_log) + + plt.figure(figsize=(10, 6), dpi=300) + plt.subplot(2, 2, 1) + plt.plot(score_log[:, 0], c='#d28ad4') + plt.ylabel('RMSE') + + plt.subplot(2, 2, 2) + plt.plot(score_log[:, 1], c='#e765eb') + plt.ylabel('MAE') + + plt.subplot(2, 2, 3) + plt.plot(score_log[:, 2], c='#6b016d') + plt.ylabel('MAPE') + + plt.show()