wave_classify/STEP2/wave_dataset-Copy1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 构造数据集\n",
    "## 相差处理\n",
    "处理各个初始文件，3900为一组，后一组减去前一组。\n",
    "首先明确各个文件名"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['hand-data-2019-12-08-19-44-33.txt',\n",
       " 'hand-data-2019-12-08-19-58-45.txt',\n",
       " 'foot-data-2019-12-08-19-36-03.txt',\n",
       " 'stick-data-2019-12-08-19-47-49.txt',\n",
       " 'stick-data-2019-12-08-19-45-44.txt',\n",
       " 'foot-data-2019-12-08-20-02-28.txt',\n",
       " 'stick-data-2019-12-08-20-06-46.txt',\n",
       " 'hand-data-2019-12-08-19-57-31.txt',\n",
       " 'foot-data-2019-12-08-20-05-39.txt',\n",
       " 'foot-data-2019-12-08-19-37-11.txt',\n",
       " 'foot-data-2019-12-08-20-04-36.txt',\n",
       " 'stick-data-2019-12-08-20-08-47.txt',\n",
       " 'hand-data-2019-12-08-20-01-10.txt',\n",
       " 'stick-data-2019-12-08-20-07-45.txt',\n",
       " 'hand-data-2019-12-08-19-42-17.txt',\n",
       " 'hand-data-2019-12-08-19-43-25.txt',\n",
       " 'foot-data-2019-12-08-19-23-49.txt',\n",
       " 'stick-data-2019-12-08-19-48-56.txt',\n",
       " 'foot-data-2019-12-08-19-38-42.txt',\n",
       " 'foot-data-2019-12-08-20-03-31.txt',\n",
       " 'stick-data-2019-12-08-19-46-41.txt',\n",
       " 'hand-data-2019-12-08-19-41-17.txt',\n",
       " 'stick-data-2019-12-08-20-09-52.txt',\n",
       " 'hand-data-2019-12-08-20-00-07.txt']"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "files = os.listdir('data2000')\n",
    "files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def loadData(data):\n",
    "    '''\n",
    "    相差法处理原始数据,得到有效数据\n",
    "    '''\n",
    "    data = np.array(data)\n",
    "    valid_data = []\n",
    "    interval = 3750\n",
    "    for i in range(0,len(data)-interval,interval):\n",
    "        temp = [x for x in map(abs, data[i+interval:i+interval*2]-data[i:i+interval])]\n",
    "        valid_data.append(temp)\n",
    "    print('共{}组数据，每组数据有{}个点'.format(len(valid_data),len(valid_data[0])))\n",
    "    return valid_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "foot-data-2019-12-08-19-36-03.txt is finished.\n",
      "foot-data-2019-12-08-20-02-28.txt is finished.\n",
      "foot-data-2019-12-08-20-05-39.txt is finished.\n",
      "foot-data-2019-12-08-19-37-11.txt is finished.\n",
      "foot-data-2019-12-08-20-04-36.txt is finished.\n",
      "foot-data-2019-12-08-19-23-49.txt is finished.\n",
      "foot-data-2019-12-08-19-38-42.txt is finished.\n",
      "foot-data-2019-12-08-20-03-31.txt is finished.\n"
     ]
    }
   ],
   "source": [
    "# 处理脚踏数据\n",
    "data = []\n",
    "for i in files:\n",
    "    if 'foot' in i:\n",
    "        with open(os.path.join('data2000',i),'r') as f:\n",
    "            for line in f:\n",
    "                data.append(int(line.strip()))\n",
    "            print('{} is finished.'.format(i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "共1999组数据，每组数据有3750个点\n"
     ]
    }
   ],
   "source": [
    "foot_data = loadData(data)    \n",
    "# 保存脚踏数据\n",
    "with open('foot_data.txt','w') as f:\n",
    "    for line in foot_data:\n",
    "        f.write(str(line)+'\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "hand-data-2019-12-08-19-44-33.txt is finished.\n",
      "hand-data-2019-12-08-19-58-45.txt is finished.\n",
      "hand-data-2019-12-08-19-57-31.txt is finished.\n",
      "hand-data-2019-12-08-20-01-10.txt is finished.\n",
      "hand-data-2019-12-08-19-42-17.txt is finished.\n",
      "hand-data-2019-12-08-19-43-25.txt is finished.\n",
      "hand-data-2019-12-08-19-41-17.txt is finished.\n",
      "hand-data-2019-12-08-20-00-07.txt is finished.\n",
      "共1999组数据，每组数据有3750个点\n"
     ]
    }
   ],
   "source": [
    "# 处理手触数据\n",
    "data = []\n",
    "for i in files:\n",
    "    if 'hand' in i:\n",
    "        with open(os.path.join('data2000',i),'r') as f:\n",
    "            for line in f:\n",
    "                data.append(int(line.strip()))\n",
    "            print('{} is finished.'.format(i))\n",
    "\n",
    "hand_data = loadData(data)\n",
    "# 保存手触数据\n",
    "with open('hand_data.txt','w') as f:\n",
    "    for line in hand_data:\n",
    "        f.write(str(line)+'\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "stick-data-2019-12-08-19-47-49.txt is finished.\n",
      "stick-data-2019-12-08-19-45-44.txt is finished.\n",
      "stick-data-2019-12-08-20-06-46.txt is finished.\n",
      "stick-data-2019-12-08-20-08-47.txt is finished.\n",
      "stick-data-2019-12-08-20-07-45.txt is finished.\n",
      "stick-data-2019-12-08-19-48-56.txt is finished.\n",
      "stick-data-2019-12-08-19-46-41.txt is finished.\n",
      "stick-data-2019-12-08-20-09-52.txt is finished.\n",
      "共1999组数据，每组数据有3750个点\n"
     ]
    }
   ],
   "source": [
    "# 处理棒击数据\n",
    "data = []\n",
    "for i in files:\n",
    "    if 'stick' in i:\n",
    "        with open(os.path.join('data2000',i),'r') as f:\n",
    "            for line in f:\n",
    "                data.append(int(line.strip()))\n",
    "            print('{} is finished.'.format(i))\n",
    "\n",
    "stick_data = loadData(data)\n",
    "# 保存手触数据\n",
    "with open('stick_data.txt','w') as f:\n",
    "    for line in stick_data:\n",
    "        f.write(str(line)+'\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,8))\n",
    "plt.subplot(221)\n",
    "plt.title('foot_data')\n",
    "plt.plot(foot_data[-1])\n",
    "plt.subplot(222)\n",
    "plt.title('hand_data')\n",
    "plt.plot(hand_data[-1])\n",
    "plt.subplot(223)\n",
    "plt.title('stick_data')\n",
    "plt.plot(stick_data[-1])\n",
    "# plt.subplot(224)\n",
    "# plt.title('normal_data')\n",
    "# plt.plot(normal_data[11])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 划分数据集\n",
    "每类数据共499个，则训练数据417个，验证数据41个，测试数据41个。\\\n",
    "建立`train.txt`, `train_label.txt`, `val.txt`, `val_label.txt`, `test.txt`, `test_label.txt`。\n",
    "1. 依次取出四种数据，打乱内部顺序\n",
    "2. 取前417个放入`train_data`，接着41个放入`val_data`，最后41个放入`test_data`。\n",
    "3. 构造依次417个foot，hand，stick，normal的`train_label`，同样的还有`val_label`,`test_label`\n",
    "4. 利用`sklearn.utils.shuffle()`函数同步打乱data与其label并放入各自文件夹"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.utils import shuffle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def handle_data(name,data,train_data,val_data,test_data,train_label,val_label,test_label):\n",
    "    '''\n",
    "    拿出数据，打乱，存放\n",
    "    '''\n",
    "    data = shuffle(data)\n",
    "    for i in range(1665):\n",
    "        train_data.append(data[i])\n",
    "    for i in range(1665,1665+167):\n",
    "        val_data.append(data[i])\n",
    "    for i in range(1665+167,1999):\n",
    "        test_data.append(data[i])\n",
    "        \n",
    "    # 将label编码为整数，foot-1，hand-2,stick-3\n",
    "    if name == 'foot': name = 0\n",
    "    elif name == 'hand': name = 1\n",
    "    else: name = 2\n",
    "    for i in range(1665):\n",
    "        train_label.append(name)\n",
    "    for i in range(167):\n",
    "        val_label.append(name)\n",
    "        test_label.append(name)\n",
    "    return train_data,val_data,test_data,train_label,val_label,test_label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "当前有1665条训练数据，167条验证数据，167条测试数据\n",
      "当前有1665个训练标签，167个验证标签，167个测试标签\n",
      "当前有3330条训练数据，334条验证数据，334条测试数据\n",
      "当前有3330个训练标签，334个验证标签，334个测试标签\n",
      "当前有4995条训练数据，501条验证数据，501条测试数据\n",
      "当前有4995个训练标签，501个验证标签，501个测试标签\n"
     ]
    }
   ],
   "source": [
    "file_name = ['foot','hand','stick']\n",
    "train_data,val_data,test_data,train_label,val_label,test_label = [],[],[],[],[],[]\n",
    "for name in file_name:\n",
    "    data = []\n",
    "    with open(name+'_data.txt','r') as f:\n",
    "        for line in f:\n",
    "            data.append(line.strip('[]\\n'))\n",
    "    handle_data(name,data,train_data,val_data,test_data,train_label,val_label,test_label)\n",
    "    print('当前有{}条训练数据，{}条验证数据，{}条测试数据'.format(len(train_data),len(val_data),len(test_data)))\n",
    "    print('当前有{}个训练标签，{}个验证标签，{}个测试标签'.format(len(train_label),len(val_label),len(test_label)))#"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 成对打乱数据\n",
    "train_data, train_label = shuffle(train_data, train_label)\n",
    "val_data, val_label = shuffle(val_data, val_label)\n",
    "test_data, test_label = shuffle(test_data, test_label)\n",
    "# 写入文件\n",
    "with open('train_data.txt','w') as f:\n",
    "    for line in train_data:\n",
    "        f.write(line+'\\n')\n",
    "with open('val_data.txt','w') as f:\n",
    "    for line in val_data:\n",
    "        f.write(line+'\\n')\n",
    "with open('test_data.txt','w') as f:\n",
    "    for line in test_data:\n",
    "        f.write(line+'\\n')\n",
    "with open('train_label.txt','w') as f:\n",
    "    for line in train_label:\n",
    "        f.write(str(line)+'\\n')\n",
    "with open('val_label.txt','w') as f:\n",
    "    for line in val_label:\n",
    "        f.write(str(line)+'\\n')\n",
    "with open('test_label.txt','w') as f:\n",
    "    for line in test_label:\n",
    "        f.write(str(line)+'\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}