量化分析/分析方法/Regression/modelling-sp500-index-price-based-on-US-economic-indicators.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "0aa8c164-afb4-48c2-95d6-7e819f9e0ea2",
   "metadata": {},
   "source": [
    "# 预测SP500指数走势\n",
    "\n",
    "## 一、前言\n",
    "- 作者：Griddb\n",
    "- 标题：Modelling S&P 500 Index Price Based on U.S. Economic Indicators using Python and GridDB\n",
    "- 目的：**利用机器学习中的Regression Model 基于美国经济指标预测SP500走势**\n",
    "- 原文地址：https://griddb.net/en/blog/modelling-sp-500-index-price-based-on-u-s-economic-indicators-using-python-and-griddb/\n",
    "- 源码地址：无\n",
    "- 数据源： `finance.yahoo.com` and `fred.stlouisfed.org`\n",
    "\n",
    "**声明**\n",
    "- 本文只是出于学习的目的将其翻译成中文并附上一定的注释，并没有打算将其用作盈利的目的。\n",
    "- 原文并没有提供原始数据集，我通过作者提供的数据出处，收集数据并清洗后放在文件夹中\n",
    "- 由于yahoo finance上获取的SP500 index数据只有1985年打后的，所以跟原文的数据有一点出入\n",
    "- 这里没有使用GridDB数据库，而是直接通过csv格式读取数据\n",
    "- 数据集的获取方式放到文章的最后\n",
    "\n",
    "## 二、研究目的\n",
    "经济指标已被用于许多的研究当中，而且很多的统计方法都结合了这些数据用于预测股票价格。本文的目的是根据美国经济指标对标准普尔500指数的月度价格进行建模， 然后执行统计测试，最后构建机器学习模型。\n",
    "\n",
    "## 三、数据指标处理\n",
    "对于SP500指数来说，对其影响最大的指标可归纳为以下类别：一般宏观经济指标，劳动力市场指标（失业率和就业报告），房地产指标，信贷市场指标，货币供应指标，消费者（家庭）金融行为指标和商品市场指标。\n",
    "- 数据周期为 1985-01-01 至 2022-06-01 之间\n",
    "- SP500指数收盘价和美国经济指标数据频率均为月度\n",
    "\n",
    "数据集各变量（指标）的描述：\n",
    "\n",
    "- 标准普尔500指数各月价格（单位：美元）`SP500`\n",
    "- 利率，美国贴现率（单位：每年百分比）`INTDSRUSM193N`\n",
    "- 商业和工业贷款，所有商业银行（单位：数十亿美元）`BUSLOANS`\n",
    "- 银行优惠贷款利率（单位：百分比）`MPRIME`\n",
    "- 联邦基金有效利率（联邦基金利率是存款机构隔夜交易联邦资金的利率。（单位：百分比）`FEDFUNDS`\n",
    "- 流通中的货币（单位：数十亿美元）`CURRCIR`\n",
    "- 个人储蓄率（个人储蓄占个人可支配收入（DPI）的百分比，通常称为“个人储蓄率”，计算为个人储蓄与DPI的比率。（单位：百分比）`PSAVERT`\n",
    "- 在许可证发放地点授权的新私人住房单位：总单位（单位：数十亿美元）`PERMIT`\n",
    "- 工业生产：总指数（工业生产指数（INDPRO）是衡量位于美国制造业，采矿业，电力和天然气公用事业的所有设施的实际产出的经济指标）（单位：指数2017 = 100）`INDPRO`\n",
    "- 个人储蓄（单位：数十亿美元）`PMSAVE`\n",
    "- 机动车零售：国内汽车 （单位：百万辆）`DAUTOSAAR`\n",
    "- 失业水平 （单位： 数千人）`UNEMPLOY`\n",
    "- 所有城市居民消费价格指数：美国城市平均值的所有项目（单位：指数1982-1984=100）`CPIAUCSL`\n",
    "\n",
    "上述数据集除了SP500之外，都可以单独从`fred.stlouisfed.org`上免费下载得到；而SP500可以从需要从`finance.yahoo.com`上下载，由于yahoo已经不再提供免费下载，所以要通过`yfinance`库获取，具体方法在文章结尾。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "6a53f4fe-29f4-4c57-a0fe-24ddf5728dac",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import datetime\n",
    "\n",
    "import statsmodels.api as sm\n",
    "from statsmodels.stats.outliers_influence import variance_inflation_factor\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8fb5b1e1-635e-4cc0-b7db-0af6ab7b9a93",
   "metadata": {},
   "source": [
    "## 四、数据探索"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "0fe64232-8826-4720-9194-7fd5261eec24",
   "metadata": {},
   "outputs": [],
   "source": [
    "us_economic_data = pd.read_csv('us_economic_data.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "6b40f634-63d5-4413-9420-fb8b98750daf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><div id=91d2a087-e8e6-4468-bdfa-699208b995c5 style=\"display:none; background-color:#9D6CFF; color:white; width:200px; height:30px; padding-left:5px; border-radius:4px; flex-direction:row; justify-content:space-around; align-items:center;\" onmouseover=\"this.style.backgroundColor='#BA9BF8'\" onmouseout=\"this.style.backgroundColor='#9D6CFF'\" onclick=\"window.commands?.execute('create-mitosheet-from-dataframe-output');\">See Full Dataframe in Mito</div> <script> if (window.commands.hasCommand('create-mitosheet-from-dataframe-output')) document.getElementById('91d2a087-e8e6-4468-bdfa-699208b995c5').style.display = 'flex' </script> <table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>DATE</th>\n",
       "      <th>SP500</th>\n",
       "      <th>PERMIT</th>\n",
       "      <th>INDPRO</th>\n",
       "      <th>FEDFUNDS</th>\n",
       "      <th>PSAVERT</th>\n",
       "      <th>CURRCIR</th>\n",
       "      <th>MPRIME</th>\n",
       "      <th>BUSLOANS</th>\n",
       "      <th>INTDSRUSM193N</th>\n",
       "      <th>DAUTOSAAR</th>\n",
       "      <th>CPIAUCSL</th>\n",
       "      <th>UNEMPLOY</th>\n",
       "      <th>PMSAVE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1985-01-01</td>\n",
       "      <td>179.630005</td>\n",
       "      <td>1660.0</td>\n",
       "      <td>54.7203</td>\n",
       "      <td>8.35</td>\n",
       "      <td>10.3</td>\n",
       "      <td>180.656</td>\n",
       "      <td>10.61</td>\n",
       "      <td>472.9593</td>\n",
       "      <td>8.00</td>\n",
       "      <td>8.244</td>\n",
       "      <td>105.7</td>\n",
       "      <td>8423</td>\n",
       "      <td>312.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1985-02-01</td>\n",
       "      <td>181.179993</td>\n",
       "      <td>1662.0</td>\n",
       "      <td>54.9100</td>\n",
       "      <td>8.50</td>\n",
       "      <td>9.1</td>\n",
       "      <td>178.242</td>\n",
       "      <td>10.50</td>\n",
       "      <td>475.6222</td>\n",
       "      <td>8.00</td>\n",
       "      <td>8.308</td>\n",
       "      <td>106.3</td>\n",
       "      <td>8321</td>\n",
       "      <td>273.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1985-03-01</td>\n",
       "      <td>180.660004</td>\n",
       "      <td>1727.0</td>\n",
       "      <td>54.9841</td>\n",
       "      <td>8.58</td>\n",
       "      <td>8.7</td>\n",
       "      <td>179.037</td>\n",
       "      <td>10.50</td>\n",
       "      <td>480.2038</td>\n",
       "      <td>8.00</td>\n",
       "      <td>8.380</td>\n",
       "      <td>106.8</td>\n",
       "      <td>8339</td>\n",
       "      <td>262.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1985-04-01</td>\n",
       "      <td>179.830002</td>\n",
       "      <td>1664.0</td>\n",
       "      <td>54.8853</td>\n",
       "      <td>8.27</td>\n",
       "      <td>9.9</td>\n",
       "      <td>180.854</td>\n",
       "      <td>10.50</td>\n",
       "      <td>478.9746</td>\n",
       "      <td>8.00</td>\n",
       "      <td>8.474</td>\n",
       "      <td>107.0</td>\n",
       "      <td>8395</td>\n",
       "      <td>304.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1985-05-01</td>\n",
       "      <td>189.550003</td>\n",
       "      <td>1709.0</td>\n",
       "      <td>54.9520</td>\n",
       "      <td>7.97</td>\n",
       "      <td>11.1</td>\n",
       "      <td>182.460</td>\n",
       "      <td>10.31</td>\n",
       "      <td>480.7533</td>\n",
       "      <td>7.81</td>\n",
       "      <td>8.227</td>\n",
       "      <td>107.2</td>\n",
       "      <td>8302</td>\n",
       "      <td>349.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table></div>"
      ],
      "text/plain": [
       "         DATE       SP500  PERMIT   INDPRO  FEDFUNDS  PSAVERT  CURRCIR  \\\n",
       "0  1985-01-01  179.630005  1660.0  54.7203      8.35     10.3  180.656   \n",
       "1  1985-02-01  181.179993  1662.0  54.9100      8.50      9.1  178.242   \n",
       "2  1985-03-01  180.660004  1727.0  54.9841      8.58      8.7  179.037   \n",
       "3  1985-04-01  179.830002  1664.0  54.8853      8.27      9.9  180.854   \n",
       "4  1985-05-01  189.550003  1709.0  54.9520      7.97     11.1  182.460   \n",
       "\n",
       "   MPRIME  BUSLOANS  INTDSRUSM193N  DAUTOSAAR  CPIAUCSL  UNEMPLOY  PMSAVE  \n",
       "0   10.61  472.9593           8.00      8.244     105.7      8423   312.2  \n",
       "1   10.50  475.6222           8.00      8.308     106.3      8321   273.6  \n",
       "2   10.50  480.2038           8.00      8.380     106.8      8339   262.2  \n",
       "3   10.50  478.9746           8.00      8.474     107.0      8395   304.7  \n",
       "4   10.31  480.7533           7.81      8.227     107.2      8302   349.5  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "us_economic_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "db53bba7-41f6-474b-872e-ba75a82875bc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><div id=968e1279-657d-48eb-bb3e-7704c6388cf5 style=\"display:none; background-color:#9D6CFF; color:white; width:200px; height:30px; padding-left:5px; border-radius:4px; flex-direction:row; justify-content:space-around; align-items:center;\" onmouseover=\"this.style.backgroundColor='#BA9BF8'\" onmouseout=\"this.style.backgroundColor='#9D6CFF'\" onclick=\"window.commands?.execute('create-mitosheet-from-dataframe-output');\">See Full Dataframe in Mito</div> <script> if (window.commands.hasCommand('create-mitosheet-from-dataframe-output')) document.getElementById('968e1279-657d-48eb-bb3e-7704c6388cf5').style.display = 'flex' </script> <table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SP500</th>\n",
       "      <th>PERMIT</th>\n",
       "      <th>INDPRO</th>\n",
       "      <th>FEDFUNDS</th>\n",
       "      <th>PSAVERT</th>\n",
       "      <th>CURRCIR</th>\n",
       "      <th>MPRIME</th>\n",
       "      <th>BUSLOANS</th>\n",
       "      <th>INTDSRUSM193N</th>\n",
       "      <th>DAUTOSAAR</th>\n",
       "      <th>CPIAUCSL</th>\n",
       "      <th>UNEMPLOY</th>\n",
       "      <th>PMSAVE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DATE</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1985-01-01</th>\n",
       "      <td>179.630005</td>\n",
       "      <td>1660.0</td>\n",
       "      <td>54.7203</td>\n",
       "      <td>8.35</td>\n",
       "      <td>10.3</td>\n",
       "      <td>180.656</td>\n",
       "      <td>10.61</td>\n",
       "      <td>472.9593</td>\n",
       "      <td>8.00</td>\n",
       "      <td>8.244</td>\n",
       "      <td>105.7</td>\n",
       "      <td>8423</td>\n",
       "      <td>312.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1985-02-01</th>\n",
       "      <td>181.179993</td>\n",
       "      <td>1662.0</td>\n",
       "      <td>54.9100</td>\n",
       "      <td>8.50</td>\n",
       "      <td>9.1</td>\n",
       "      <td>178.242</td>\n",
       "      <td>10.50</td>\n",
       "      <td>475.6222</td>\n",
       "      <td>8.00</td>\n",
       "      <td>8.308</td>\n",
       "      <td>106.3</td>\n",
       "      <td>8321</td>\n",
       "      <td>273.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1985-03-01</th>\n",
       "      <td>180.660004</td>\n",
       "      <td>1727.0</td>\n",
       "      <td>54.9841</td>\n",
       "      <td>8.58</td>\n",
       "      <td>8.7</td>\n",
       "      <td>179.037</td>\n",
       "      <td>10.50</td>\n",
       "      <td>480.2038</td>\n",
       "      <td>8.00</td>\n",
       "      <td>8.380</td>\n",
       "      <td>106.8</td>\n",
       "      <td>8339</td>\n",
       "      <td>262.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1985-04-01</th>\n",
       "      <td>179.830002</td>\n",
       "      <td>1664.0</td>\n",
       "      <td>54.8853</td>\n",
       "      <td>8.27</td>\n",
       "      <td>9.9</td>\n",
       "      <td>180.854</td>\n",
       "      <td>10.50</td>\n",
       "      <td>478.9746</td>\n",
       "      <td>8.00</td>\n",
       "      <td>8.474</td>\n",
       "      <td>107.0</td>\n",
       "      <td>8395</td>\n",
       "      <td>304.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1985-05-01</th>\n",
       "      <td>189.550003</td>\n",
       "      <td>1709.0</td>\n",
       "      <td>54.9520</td>\n",
       "      <td>7.97</td>\n",
       "      <td>11.1</td>\n",
       "      <td>182.460</td>\n",
       "      <td>10.31</td>\n",
       "      <td>480.7533</td>\n",
       "      <td>7.81</td>\n",
       "      <td>8.227</td>\n",
       "      <td>107.2</td>\n",
       "      <td>8302</td>\n",
       "      <td>349.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table></div>"
      ],
      "text/plain": [
       "                 SP500  PERMIT   INDPRO  FEDFUNDS  PSAVERT  CURRCIR  MPRIME  \\\n",
       "DATE                                                                          \n",
       "1985-01-01  179.630005  1660.0  54.7203      8.35     10.3  180.656   10.61   \n",
       "1985-02-01  181.179993  1662.0  54.9100      8.50      9.1  178.242   10.50   \n",
       "1985-03-01  180.660004  1727.0  54.9841      8.58      8.7  179.037   10.50   \n",
       "1985-04-01  179.830002  1664.0  54.8853      8.27      9.9  180.854   10.50   \n",
       "1985-05-01  189.550003  1709.0  54.9520      7.97     11.1  182.460   10.31   \n",
       "\n",
       "            BUSLOANS  INTDSRUSM193N  DAUTOSAAR  CPIAUCSL  UNEMPLOY  PMSAVE  \n",
       "DATE                                                                        \n",
       "1985-01-01  472.9593           8.00      8.244     105.7      8423   312.2  \n",
       "1985-02-01  475.6222           8.00      8.308     106.3      8321   273.6  \n",
       "1985-03-01  480.2038           8.00      8.380     106.8      8339   262.2  \n",
       "1985-04-01  478.9746           8.00      8.474     107.0      8395   304.7  \n",
       "1985-05-01  480.7533           7.81      8.227     107.2      8302   349.5  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Changing the column name and then setting the date as an index od the dataframe\n",
    "df = us_economic_data\n",
    "df['DATE'] = pd.to_datetime(df['DATE'])\n",
    "df = df.set_index('DATE')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0137d003-7dc7-4dcd-84b3-46f1a989fd97",
   "metadata": {},
   "source": [
    "````{margin}\n",
    "Augmented Dickey-Fuller Test 是一种统计学的检测方法，用于检测一个有某种趋势的时间序列的平稳性。是一种重要的单根检测方法。\n",
    "\n",
    "其初始假设null hypothesis是该序列不稳定，检测可以有两种途径，一是统计量小于特定拒绝域值；二是p-value大于相应域值。如果是，则拒绝假设，认为序列是稳定的。\n",
    "````\n",
    "\n",
    "我们执行了Augmented Dickey-Fuller来确定SP500指数收盘价是否具有平稳性。之后我们对收盘价和每个自变量（经济指标）之间的数据进行可视化，以确定变量（经济指标）之间是否存在线性关系，从而仅选择其具有那些具有线性关系的变量用作建模。\n",
    "\n",
    "其次，我们需要使用VIF测试检查自变量（经济指标）之间是否相互关联。最后，由于指标是在不同的统计范围上测量的，因此使用z score标准对数据进行了归一化，以产生准确可靠的模式。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "c64fa377-2d1d-44fd-8732-01cdaeeca29b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def tsplot(y, figsize=(12, 7), syle='bmh'):\n",
    "    \n",
    "    if not isinstance(y, pd.Series):\n",
    "        y = pd.Series(y)\n",
    "        \n",
    "    with plt.style.context(style='bmh'):\n",
    "        fig = plt.figure(figsize=figsize)\n",
    "        layout = (2,1)\n",
    "        ts_ax = plt.subplot2grid(layout, (0,0), colspan=2)\n",
    "        y.plot(ax=ts_ax)\n",
    "        p_value = sm.tsa.stattools.adfuller(y)[1]\n",
    "        ts_ax.set_title('Time Series Analysis Plots\\n Dickey-Fuller: p={0:.5f}'.format(p_value))\n",
    "\n",
    "        plt.tight_layout()\n",
    "        \n",
    "tsplot(df['SP500'])\n",
    "\n",
    "data_diff = df['SP500'] - df['SP500'].shift(1)\n",
    "\n",
    "tsplot(data_diff[1:])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "688059b9-0123-4ee7-8ad7-4b95f65746b4",
   "metadata": {},
   "source": [
    "````{margin}\n",
    "方差扩大因子(variance inflation factor)简称VIF，是表征自变数观察值之间复共线性程度的数值。\n",
    "````\n",
    "\n",
    "如果我们在Dickey Fuller Test的假设中取5%作为置信水平，我们可以看到值远大于0.05，因此可以判断SP500是不具备平稳性的。\n",
    "\n",
    "为了使数据保持平稳，有必要创建一个新的因变量，该因变量是根据当前每月价格减去上个月的价格计算得出的。在这种情况下，重新计算前面提到的测试，并且发现数据是平稳的，因为p-value显示为0。\n",
    "\n",
    "下一步是找出自变量是否彼此不相关。为此选择了 VIF（Variance Inflating factor, 方差膨胀因子）测试。方差膨胀因子 （VIF） 用于检验回归模型中是否存在多重共线性。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "1b91bf6e-edca-4ff1-a845-9892ef68e7e0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['SP500', 'PERMIT', 'INDPRO', 'FEDFUNDS', 'PSAVERT', 'CURRCIR', 'MPRIME',\n",
       "       'BUSLOANS', 'INTDSRUSM193N', 'DAUTOSAAR', 'CPIAUCSL', 'UNEMPLOY',\n",
       "       'PMSAVE'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "5a78c8a3-4f72-4251-912a-1542c8c4e3be",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "          feature          VIF\n",
      "0   INTDSRUSM193N    97.796219\n",
      "1        BUSLOANS   182.967472\n",
      "2          MPRIME   694.827208\n",
      "3        FEDFUNDS   327.213964\n",
      "4         CURRCIR   365.219235\n",
      "5         PSAVERT    91.198794\n",
      "6          PMSAVE    40.048740\n",
      "7       DAUTOSAAR    63.977266\n",
      "8        UNEMPLOY    62.133967\n",
      "9          INDPRO  1337.341806\n",
      "10         PERMIT    37.411240\n",
      "11       CPIAUCSL  3058.034391\n"
     ]
    }
   ],
   "source": [
    "# the independent variables set\n",
    "X = df[['INTDSRUSM193N', 'BUSLOANS', 'MPRIME', 'FEDFUNDS', 'CURRCIR',\n",
    "       'PSAVERT', 'PMSAVE', 'DAUTOSAAR', 'UNEMPLOY', 'INDPRO', 'PERMIT',\n",
    "       'CPIAUCSL']]\n",
    "  \n",
    "# VIF dataframe\n",
    "vif_data = pd.DataFrame()\n",
    "vif_data[\"feature\"] = X.columns\n",
    "  \n",
    "# calculating VIF for each feature\n",
    "vif_data[\"VIF\"] = [variance_inflation_factor(X.values, i)\n",
    "                          for i in range(len(X.columns))]\n",
    "  \n",
    "print(vif_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ddaef966-13ad-4806-8420-ffc8fb640e1e",
   "metadata": {},
   "source": [
    "从上述表中可以看出，每个变量的 VIF 测试值远高于 10。在这种情况下，具有最高 VIF 值的指标将被删除，并重新计算整个集合。只要值小于 10 的所有变量都保留下来。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "107bf568-29e0-4669-93f0-80493e7c6631",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    feature        VIF\n",
      "0  FEDFUNDS   6.149062\n",
      "1   CURRCIR   7.253224\n",
      "2   PSAVERT   7.123685\n",
      "3    PERMIT  10.382944\n"
     ]
    }
   ],
   "source": [
    "# the independent variables set\n",
    "X = df[[  'FEDFUNDS', 'CURRCIR','PSAVERT', 'PERMIT',]]\n",
    "  \n",
    "# VIF dataframe\n",
    "vif_data = pd.DataFrame()\n",
    "vif_data[\"feature\"] = X.columns\n",
    "  \n",
    "# calculating VIF for each feature\n",
    "vif_data[\"VIF\"] = [variance_inflation_factor(X.values, i)\n",
    "                          for i in range(len(X.columns))]\n",
    "  \n",
    "print(vif_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f49fa7ed-7a53-4231-a9f6-14d9392938d1",
   "metadata": {},
   "source": [
    "我们发现其中 11 个经济指标不符合线性回归规则。\n",
    "\n",
    "我们这阶段最后一步就是将这 3 个指标与SP500价格指数整合在一起，并根据z score标准对其进行标准化。\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "fe5e3966-bc85-4821-81cc-2b861f561168",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_new = df[['SP500','FEDFUNDS','CURRCIR','PSAVERT','PERMIT']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "65f3ea4b-6e8b-4aeb-8ab8-a5088ff254ae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><div id=cc9d960a-33df-42cd-9d02-95be41349c11 style=\"display:none; background-color:#9D6CFF; color:white; width:200px; height:30px; padding-left:5px; border-radius:4px; flex-direction:row; justify-content:space-around; align-items:center;\" onmouseover=\"this.style.backgroundColor='#BA9BF8'\" onmouseout=\"this.style.backgroundColor='#9D6CFF'\" onclick=\"window.commands?.execute('create-mitosheet-from-dataframe-output');\">See Full Dataframe in Mito</div> <script> if (window.commands.hasCommand('create-mitosheet-from-dataframe-output')) document.getElementById('cc9d960a-33df-42cd-9d02-95be41349c11').style.display = 'flex' </script> <table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SP500</th>\n",
       "      <th>FEDFUNDS</th>\n",
       "      <th>CURRCIR</th>\n",
       "      <th>PSAVERT</th>\n",
       "      <th>PERMIT</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DATE</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1985-01-01</th>\n",
       "      <td>-1.207185</td>\n",
       "      <td>1.749872</td>\n",
       "      <td>-1.190784</td>\n",
       "      <td>1.118462</td>\n",
       "      <td>0.751727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1985-02-01</th>\n",
       "      <td>-1.205417</td>\n",
       "      <td>1.803265</td>\n",
       "      <td>-1.195481</td>\n",
       "      <td>0.702642</td>\n",
       "      <td>0.756793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1985-03-01</th>\n",
       "      <td>-1.206010</td>\n",
       "      <td>1.831741</td>\n",
       "      <td>-1.193934</td>\n",
       "      <td>0.564035</td>\n",
       "      <td>0.921442</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1985-04-01</th>\n",
       "      <td>-1.206957</td>\n",
       "      <td>1.721396</td>\n",
       "      <td>-1.190398</td>\n",
       "      <td>0.979855</td>\n",
       "      <td>0.761859</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1985-05-01</th>\n",
       "      <td>-1.195873</td>\n",
       "      <td>1.614611</td>\n",
       "      <td>-1.187274</td>\n",
       "      <td>1.395675</td>\n",
       "      <td>0.875847</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table></div>"
      ],
      "text/plain": [
       "               SP500  FEDFUNDS   CURRCIR   PSAVERT    PERMIT\n",
       "DATE                                                        \n",
       "1985-01-01 -1.207185  1.749872 -1.190784  1.118462  0.751727\n",
       "1985-02-01 -1.205417  1.803265 -1.195481  0.702642  0.756793\n",
       "1985-03-01 -1.206010  1.831741 -1.193934  0.564035  0.921442\n",
       "1985-04-01 -1.206957  1.721396 -1.190398  0.979855  0.761859\n",
       "1985-05-01 -1.195873  1.614611 -1.187274  1.395675  0.875847"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_z_scaled = df_new.copy()\n",
    "  \n",
    "# apply normalization techniques\n",
    "for column in df_z_scaled.columns:\n",
    "    df_z_scaled[column] = (df_z_scaled[column] -\n",
    "                           df_z_scaled[column].mean()) / df_z_scaled[column].std()    \n",
    "  \n",
    "# view normalized data   \n",
    "df_z_scaled.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "89e6b321-870e-41be-9fcb-0a36300278f0",
   "metadata": {},
   "source": [
    "## 五、建模\n",
    "现在，让我们继续在数据集上构建和评估机器学习模型。我们将首先为模型创建 `features` 和 `labels`，并将它们拆分为训练和测试示例。测试大小已保留为总数据集大小的 20%。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "544cf844-ab59-4b03-8a5d-03a0bb81fbe7",
   "metadata": {},
   "outputs": [],
   "source": [
    "features = df_z_scaled.drop(columns = ['SP500'], axis = 1)\n",
    "labels = df_z_scaled[['SP500']]\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(features, labels, test_size = 0.2, random_state = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "51172605-93fa-40d4-8867-34d971306000",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of training data: (352, 4)\n",
      "Shape of the training target data: (352, 1)\n",
      "Shape of test data: (88, 4)\n",
      "Shape of the test target data: (88, 1)\n"
     ]
    }
   ],
   "source": [
    "print(f\"Shape of training data: {X_train.shape}\")\n",
    "print(f\"Shape of the training target data: {y_train.shape}\")\n",
    "\n",
    "print(f\"Shape of test data: {X_test.shape}\")\n",
    "print(f\"Shape of the test target data: {y_test.shape}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ac0d4404-1003-4fdb-8182-3e2969efb014",
   "metadata": {},
   "source": [
    "### （1）Linear Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "c95adcad-f558-4c44-a9ef-d3fe08dc7a3f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9543343215374887\n"
     ]
    }
   ],
   "source": [
    "regr = LinearRegression()\n",
    "  \n",
    "regr.fit(X_train, y_train)\n",
    "print(regr.score(X_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "a8020cd4-9704-4b96-b7c5-557e5402d7c0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><div id=196d884c-638e-4703-a6f5-620f9d4a641c style=\"display:none; background-color:#9D6CFF; color:white; width:200px; height:30px; padding-left:5px; border-radius:4px; flex-direction:row; justify-content:space-around; align-items:center;\" onmouseover=\"this.style.backgroundColor='#BA9BF8'\" onmouseout=\"this.style.backgroundColor='#9D6CFF'\" onclick=\"window.commands?.execute('create-mitosheet-from-dataframe-output');\">See Full Dataframe in Mito</div> <script> if (window.commands.hasCommand('create-mitosheet-from-dataframe-output')) document.getElementById('196d884c-638e-4703-a6f5-620f9d4a641c').style.display = 'flex' </script> <table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>FEDFUNDS</td>\n",
       "      <td>0.182386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>CURRCIR</td>\n",
       "      <td>1.119148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PSAVERT</td>\n",
       "      <td>-0.027447</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>PERMIT</td>\n",
       "      <td>0.118363</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table></div>"
      ],
      "text/plain": [
       "          0         0\n",
       "0  FEDFUNDS  0.182386\n",
       "1   CURRCIR  1.119148\n",
       "2   PSAVERT -0.027447\n",
       "3    PERMIT  0.118363"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "coefficients = pd.concat([pd.DataFrame(X.columns),pd.DataFrame(np.transpose(regr.coef_))], axis = 1)\n",
    "coefficients"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "92d8febc-6f17-45a4-a519-8dc2fb751cad",
   "metadata": {},
   "source": [
    "`CURRCIR (Currency in Circulation)`具有最高的系数，这意味着对模型的影响最大。\n",
    "\n",
    "在模型适合我们的训练数据后，我们可以继续预测我们的测试集，以评估模型性能。让我们将预测存储在`predicted`中。\n",
    "\n",
    "评估指标：\n",
    "1. 决定系数 （R^2）。这是数据中模型置信度的最重要指标，在回归模型的所有描述中都是必需的。\n",
    "2. 均方误差 （MSE）。此值测量实际值和预测值之间的均方根误差。\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "685906c8-9b35-4138-ac7a-ccc92df3a4d9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE:  0.048140955259271395\n",
      "RMSE:  0.024070477629635698\n",
      "r2 score for Random Forest model is 0.9543343215374887\n"
     ]
    }
   ],
   "source": [
    "y_pred = regr.predict(X_test)\n",
    "mse = mean_squared_error(y_test, y_pred)\n",
    "print(\"MSE: \", mse)\n",
    "print(\"RMSE: \", mse*(1/2.0)) \n",
    "\n",
    "r2 = r2_score(y_test, y_pred)\n",
    "print('r2 score for Random Forest model is', r2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "98589844-05b6-45df-b8e1-17073042e8c2",
   "metadata": {},
   "source": [
    "### （2）RandomForest Regression"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ec166a62-0f90-49bf-912e-59da1b5b5225",
   "metadata": {},
   "source": [
    "下面使用了随机森林回归的方法，因为`Random forest regression` 考虑了多种预测结果，所以基于随机森林的预测更准确、算法更稳定。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "53b37750-9695-4247-bb43-29e7f41bbb30",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R-squared: 0.9988945870353322\n"
     ]
    }
   ],
   "source": [
    "rfr = RandomForestRegressor()\n",
    "\n",
    "rfr.fit(X_train, y_train)\n",
    "\n",
    "score = rfr.score(X_train, y_train)\n",
    "print(\"R-squared:\", score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "9c236495-eb17-40e8-ae2b-d48c97b03aa3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE:  0.00568486950353933\n",
      "RMSE:  0.002842434751769665\n",
      "r2 score for Random Forest model is 0.9946074310023176\n"
     ]
    }
   ],
   "source": [
    "ypred = rfr.predict(X_test)\n",
    "\n",
    "mse = mean_squared_error(y_test, ypred)\n",
    "print(\"MSE: \", mse)\n",
    "print(\"RMSE: \", mse*(1/2.0)) \n",
    "\n",
    "r2 = r2_score(y_test, ypred)\n",
    "print('r2 score for Random Forest model is', r2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "f5142dd5-c373-480f-af50-9b597d7dd76a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(5,5))\n",
    "plt.scatter(y_test['SP500'].values, ypred, c='crimson')\n",
    "\n",
    "\n",
    "p1 = max(max(ypred), max(y_test['SP500'].values))\n",
    "p2 = min(min(ypred), min(y_test['SP500'].values))\n",
    "plt.plot([p1, p2], [p1, p2], 'b-')\n",
    "plt.xlabel('True Values', fontsize=15)\n",
    "plt.ylabel('Predictions', fontsize=15)\n",
    "plt.axis('equal')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b010845f-fe56-43b4-9d43-34c9e3e9e5d4",
   "metadata": {},
   "source": [
    "根据两种回归方法得出的 R2 和 MSE/RMSE，随机森林模型在这例子中显然较优， 其统计线性回归模型中提高了 6%，错误率降低了 20 倍。\n",
    "\n",
    "## 【数据获取过程】"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "01096ab1-a1a6-4595-8e3c-1108df8de9d1",
   "metadata": {
    "tags": [
     "remove-output"
    ]
   },
   "outputs": [],
   "source": [
    "from pandas_datareader import data as pdr\n",
    "\n",
    "import yfinance as yf\n",
    "yf.pdr_override() \n",
    "\n",
    "# download dataframe\n",
    "data = pdr.get_data_yahoo(\"^GSPC\", start=\"1985-01-01\", end=\"2022-06-01\", interval = \"1mo\")\n",
    "data = data.rename(columns={'Close':'SP500'})\n",
    "data.index.names=['DATE']\n",
    "data = data[['SP500']]\n",
    "\n",
    "data.to_csv('../data/sp500.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b4206c80-67dc-419b-a5df-29ccbf54a82c",
   "metadata": {
    "tags": [
     "remove-output"
    ]
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import os\n",
    "import glob\n",
    "\n",
    "PATH_DIR = '../data'\n",
    "file_paths = sorted(glob.iglob(PATH_DIR + '/*'), key=os.path.getmtime)\n",
    "\n",
    "df_merge=pd.read_csv(PATH_DIR+'/sp500.csv')\n",
    "df_merge = df_merge.set_index('DATE')\n",
    "\n",
    "for fp in file_paths:\n",
    "    if 'sp500' not in fp:       \n",
    "        df = pd.read_csv(fp)\n",
    "        df = df.set_index('DATE')\n",
    "        df_merge = df_merge.join(df, how='inner')\n",
    "df_merge.to_csv('../us_economic_data.csv')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}