1_Implied_Volatility_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline\n",
    "\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.linear_model import Lasso\n",
    "import statsmodels.api as sm\n",
    "import statsmodels.formula.api as smf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def func_mae(x1,x2):\n",
    "    return np.mean(np.abs(x1-x2))\n",
    "\n",
    "def lasso_cv(X_train,y_train,alp_list,split_n,intercept_boolen,MAX_ITER=10000):\n",
    "    kf = KFold(n_splits = split_n)\n",
    "    \n",
    "    mse_ave = []\n",
    "    mae_ave = []\n",
    "    \n",
    "    for t in alp_list:\n",
    "        mse = []\n",
    "        mae = []\n",
    "        ls = Lasso(alpha = t,fit_intercept = intercept_boolen,max_iter = MAX_ITER)\n",
    "        \n",
    "        for k, (train, test) in enumerate(kf.split(X_train)):\n",
    "            ls.fit(X_train.loc[train],y_train.loc[train])\n",
    "            predictions = ls.predict(X_train.loc[test])\n",
    "            mse.append(sm.tools.eval_measures.mse(predictions,y_train.loc[test]))\n",
    "            mae.append(func_mae(predictions,y_train.loc[test]))\n",
    "            \n",
    "        mse_ave.append(np.mean(mse))\n",
    "        mae_ave.append(np.mean(mae))\n",
    "    \n",
    "    df = pd.DataFrame({'alpha':alp_list,'MSE':mse_ave,'MAE':mae_ave})\n",
    "    a_star = df.sort_values(by='MSE').reset_index().loc[0,'alpha']\n",
    "    mse_star = df.sort_values(by='MSE').reset_index().loc[0,'MSE']\n",
    "    \n",
    "    return [a_star,mse_star]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>cchv10</th>\n",
       "      <th>cchv22</th>\n",
       "      <th>cchv250</th>\n",
       "      <th>cchv44</th>\n",
       "      <th>cchv5</th>\n",
       "      <th>cchv66</th>\n",
       "      <th>gkhv10</th>\n",
       "      <th>gkhv22</th>\n",
       "      <th>gkhv250</th>\n",
       "      <th>...</th>\n",
       "      <th>rshv44</th>\n",
       "      <th>rshv5</th>\n",
       "      <th>rshv66</th>\n",
       "      <th>yzhv10</th>\n",
       "      <th>yzhv22</th>\n",
       "      <th>yzhv250</th>\n",
       "      <th>yzhv44</th>\n",
       "      <th>yzhv5</th>\n",
       "      <th>yzhv66</th>\n",
       "      <th>vix</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>2015-12-01</td>\n",
       "      <td>0.259215</td>\n",
       "      <td>0.279535</td>\n",
       "      <td>0.421753</td>\n",
       "      <td>0.233623</td>\n",
       "      <td>0.358053</td>\n",
       "      <td>0.390261</td>\n",
       "      <td>0.204379</td>\n",
       "      <td>0.240595</td>\n",
       "      <td>0.373034</td>\n",
       "      <td>...</td>\n",
       "      <td>0.229824</td>\n",
       "      <td>0.244163</td>\n",
       "      <td>0.332196</td>\n",
       "      <td>0.214024</td>\n",
       "      <td>0.267942</td>\n",
       "      <td>0.433284</td>\n",
       "      <td>0.270279</td>\n",
       "      <td>0.257701</td>\n",
       "      <td>0.389017</td>\n",
       "      <td>32.826614</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>2015-12-02</td>\n",
       "      <td>0.366014</td>\n",
       "      <td>0.315227</td>\n",
       "      <td>0.423600</td>\n",
       "      <td>0.257679</td>\n",
       "      <td>0.538290</td>\n",
       "      <td>0.394065</td>\n",
       "      <td>0.225499</td>\n",
       "      <td>0.249751</td>\n",
       "      <td>0.373265</td>\n",
       "      <td>...</td>\n",
       "      <td>0.231295</td>\n",
       "      <td>0.259192</td>\n",
       "      <td>0.330025</td>\n",
       "      <td>0.234991</td>\n",
       "      <td>0.273084</td>\n",
       "      <td>0.433438</td>\n",
       "      <td>0.272863</td>\n",
       "      <td>0.296591</td>\n",
       "      <td>0.388105</td>\n",
       "      <td>31.359858</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>2015-12-03</td>\n",
       "      <td>0.360918</td>\n",
       "      <td>0.314335</td>\n",
       "      <td>0.423487</td>\n",
       "      <td>0.257439</td>\n",
       "      <td>0.537456</td>\n",
       "      <td>0.341263</td>\n",
       "      <td>0.237105</td>\n",
       "      <td>0.254178</td>\n",
       "      <td>0.373515</td>\n",
       "      <td>...</td>\n",
       "      <td>0.231586</td>\n",
       "      <td>0.276978</td>\n",
       "      <td>0.327804</td>\n",
       "      <td>0.244212</td>\n",
       "      <td>0.276789</td>\n",
       "      <td>0.433650</td>\n",
       "      <td>0.273368</td>\n",
       "      <td>0.310326</td>\n",
       "      <td>0.379251</td>\n",
       "      <td>31.862687</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>2015-12-04</td>\n",
       "      <td>0.383709</td>\n",
       "      <td>0.298086</td>\n",
       "      <td>0.423778</td>\n",
       "      <td>0.265814</td>\n",
       "      <td>0.418484</td>\n",
       "      <td>0.312630</td>\n",
       "      <td>0.239585</td>\n",
       "      <td>0.250902</td>\n",
       "      <td>0.373581</td>\n",
       "      <td>...</td>\n",
       "      <td>0.230796</td>\n",
       "      <td>0.251610</td>\n",
       "      <td>0.306779</td>\n",
       "      <td>0.248477</td>\n",
       "      <td>0.274844</td>\n",
       "      <td>0.433757</td>\n",
       "      <td>0.274230</td>\n",
       "      <td>0.288362</td>\n",
       "      <td>0.350372</td>\n",
       "      <td>32.419264</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>2015-12-07</td>\n",
       "      <td>0.383769</td>\n",
       "      <td>0.280394</td>\n",
       "      <td>0.423455</td>\n",
       "      <td>0.260403</td>\n",
       "      <td>0.419736</td>\n",
       "      <td>0.311930</td>\n",
       "      <td>0.237847</td>\n",
       "      <td>0.222077</td>\n",
       "      <td>0.373553</td>\n",
       "      <td>...</td>\n",
       "      <td>0.230363</td>\n",
       "      <td>0.202105</td>\n",
       "      <td>0.286753</td>\n",
       "      <td>0.247135</td>\n",
       "      <td>0.246933</td>\n",
       "      <td>0.433665</td>\n",
       "      <td>0.273243</td>\n",
       "      <td>0.252466</td>\n",
       "      <td>0.334571</td>\n",
       "      <td>32.171463</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 38 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        date    cchv10    cchv22   cchv250    cchv44     cchv5    cchv66  \\\n",
       "0 2015-12-01  0.259215  0.279535  0.421753  0.233623  0.358053  0.390261   \n",
       "1 2015-12-02  0.366014  0.315227  0.423600  0.257679  0.538290  0.394065   \n",
       "2 2015-12-03  0.360918  0.314335  0.423487  0.257439  0.537456  0.341263   \n",
       "3 2015-12-04  0.383709  0.298086  0.423778  0.265814  0.418484  0.312630   \n",
       "4 2015-12-07  0.383769  0.280394  0.423455  0.260403  0.419736  0.311930   \n",
       "\n",
       "     gkhv10    gkhv22   gkhv250  ...    rshv44     rshv5    rshv66    yzhv10  \\\n",
       "0  0.204379  0.240595  0.373034  ...  0.229824  0.244163  0.332196  0.214024   \n",
       "1  0.225499  0.249751  0.373265  ...  0.231295  0.259192  0.330025  0.234991   \n",
       "2  0.237105  0.254178  0.373515  ...  0.231586  0.276978  0.327804  0.244212   \n",
       "3  0.239585  0.250902  0.373581  ...  0.230796  0.251610  0.306779  0.248477   \n",
       "4  0.237847  0.222077  0.373553  ...  0.230363  0.202105  0.286753  0.247135   \n",
       "\n",
       "     yzhv22   yzhv250    yzhv44     yzhv5    yzhv66        vix  \n",
       "0  0.267942  0.433284  0.270279  0.257701  0.389017  32.826614  \n",
       "1  0.273084  0.433438  0.272863  0.296591  0.388105  31.359858  \n",
       "2  0.276789  0.433650  0.273368  0.310326  0.379251  31.862687  \n",
       "3  0.274844  0.433757  0.274230  0.288362  0.350372  32.419264  \n",
       "4  0.246933  0.433665  0.273243  0.252466  0.334571  32.171463  \n",
       "\n",
       "[5 rows x 38 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_pickle('regr_df_new.pkl')\n",
    "del df['iv']\n",
    "df2 = pd.read_csv('SSE510050VIX(1).csv',index_col=0)\n",
    "df2['date'] = df2['tradedate'].apply(pd.to_datetime)\n",
    "df2 = df2[['date','vix']]\n",
    "df3 = pd.merge(left=df,right=df2,how='inner',on='date')\n",
    "df3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "X=df3.iloc[:,1:37]\n",
    "y=df3.iloc[:,37]\n",
    "length = int(len(y)*0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cchv10</th>\n",
       "      <th>cchv22</th>\n",
       "      <th>cchv250</th>\n",
       "      <th>cchv44</th>\n",
       "      <th>cchv5</th>\n",
       "      <th>cchv66</th>\n",
       "      <th>gkhv10</th>\n",
       "      <th>gkhv22</th>\n",
       "      <th>gkhv250</th>\n",
       "      <th>gkhv44</th>\n",
       "      <th>...</th>\n",
       "      <th>rshv250</th>\n",
       "      <th>rshv44</th>\n",
       "      <th>rshv5</th>\n",
       "      <th>rshv66</th>\n",
       "      <th>yzhv10</th>\n",
       "      <th>yzhv22</th>\n",
       "      <th>yzhv250</th>\n",
       "      <th>yzhv44</th>\n",
       "      <th>yzhv5</th>\n",
       "      <th>yzhv66</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.259215</td>\n",
       "      <td>0.279535</td>\n",
       "      <td>0.421753</td>\n",
       "      <td>0.233623</td>\n",
       "      <td>0.358053</td>\n",
       "      <td>0.390261</td>\n",
       "      <td>0.204379</td>\n",
       "      <td>0.240595</td>\n",
       "      <td>0.373034</td>\n",
       "      <td>0.219368</td>\n",
       "      <td>...</td>\n",
       "      <td>0.377166</td>\n",
       "      <td>0.229824</td>\n",
       "      <td>0.244163</td>\n",
       "      <td>0.332196</td>\n",
       "      <td>0.214024</td>\n",
       "      <td>0.267942</td>\n",
       "      <td>0.433284</td>\n",
       "      <td>0.270279</td>\n",
       "      <td>0.257701</td>\n",
       "      <td>0.389017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.366014</td>\n",
       "      <td>0.315227</td>\n",
       "      <td>0.423600</td>\n",
       "      <td>0.257679</td>\n",
       "      <td>0.538290</td>\n",
       "      <td>0.394065</td>\n",
       "      <td>0.225499</td>\n",
       "      <td>0.249751</td>\n",
       "      <td>0.373265</td>\n",
       "      <td>0.224471</td>\n",
       "      <td>...</td>\n",
       "      <td>0.377064</td>\n",
       "      <td>0.231295</td>\n",
       "      <td>0.259192</td>\n",
       "      <td>0.330025</td>\n",
       "      <td>0.234991</td>\n",
       "      <td>0.273084</td>\n",
       "      <td>0.433438</td>\n",
       "      <td>0.272863</td>\n",
       "      <td>0.296591</td>\n",
       "      <td>0.388105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0.360918</td>\n",
       "      <td>0.314335</td>\n",
       "      <td>0.423487</td>\n",
       "      <td>0.257439</td>\n",
       "      <td>0.537456</td>\n",
       "      <td>0.341263</td>\n",
       "      <td>0.237105</td>\n",
       "      <td>0.254178</td>\n",
       "      <td>0.373515</td>\n",
       "      <td>0.225533</td>\n",
       "      <td>...</td>\n",
       "      <td>0.377320</td>\n",
       "      <td>0.231586</td>\n",
       "      <td>0.276978</td>\n",
       "      <td>0.327804</td>\n",
       "      <td>0.244212</td>\n",
       "      <td>0.276789</td>\n",
       "      <td>0.433650</td>\n",
       "      <td>0.273368</td>\n",
       "      <td>0.310326</td>\n",
       "      <td>0.379251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0.383709</td>\n",
       "      <td>0.298086</td>\n",
       "      <td>0.423778</td>\n",
       "      <td>0.265814</td>\n",
       "      <td>0.418484</td>\n",
       "      <td>0.312630</td>\n",
       "      <td>0.239585</td>\n",
       "      <td>0.250902</td>\n",
       "      <td>0.373581</td>\n",
       "      <td>0.225541</td>\n",
       "      <td>...</td>\n",
       "      <td>0.377394</td>\n",
       "      <td>0.230796</td>\n",
       "      <td>0.251610</td>\n",
       "      <td>0.306779</td>\n",
       "      <td>0.248477</td>\n",
       "      <td>0.274844</td>\n",
       "      <td>0.433757</td>\n",
       "      <td>0.274230</td>\n",
       "      <td>0.288362</td>\n",
       "      <td>0.350372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>0.383769</td>\n",
       "      <td>0.280394</td>\n",
       "      <td>0.423455</td>\n",
       "      <td>0.260403</td>\n",
       "      <td>0.419736</td>\n",
       "      <td>0.311930</td>\n",
       "      <td>0.237847</td>\n",
       "      <td>0.222077</td>\n",
       "      <td>0.373553</td>\n",
       "      <td>0.225019</td>\n",
       "      <td>...</td>\n",
       "      <td>0.377377</td>\n",
       "      <td>0.230363</td>\n",
       "      <td>0.202105</td>\n",
       "      <td>0.286753</td>\n",
       "      <td>0.247135</td>\n",
       "      <td>0.246933</td>\n",
       "      <td>0.433665</td>\n",
       "      <td>0.273243</td>\n",
       "      <td>0.252466</td>\n",
       "      <td>0.334571</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 36 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     cchv10    cchv22   cchv250    cchv44     cchv5    cchv66    gkhv10  \\\n",
       "0  0.259215  0.279535  0.421753  0.233623  0.358053  0.390261  0.204379   \n",
       "1  0.366014  0.315227  0.423600  0.257679  0.538290  0.394065  0.225499   \n",
       "2  0.360918  0.314335  0.423487  0.257439  0.537456  0.341263  0.237105   \n",
       "3  0.383709  0.298086  0.423778  0.265814  0.418484  0.312630  0.239585   \n",
       "4  0.383769  0.280394  0.423455  0.260403  0.419736  0.311930  0.237847   \n",
       "\n",
       "     gkhv22   gkhv250    gkhv44  ...   rshv250    rshv44     rshv5    rshv66  \\\n",
       "0  0.240595  0.373034  0.219368  ...  0.377166  0.229824  0.244163  0.332196   \n",
       "1  0.249751  0.373265  0.224471  ...  0.377064  0.231295  0.259192  0.330025   \n",
       "2  0.254178  0.373515  0.225533  ...  0.377320  0.231586  0.276978  0.327804   \n",
       "3  0.250902  0.373581  0.225541  ...  0.377394  0.230796  0.251610  0.306779   \n",
       "4  0.222077  0.373553  0.225019  ...  0.377377  0.230363  0.202105  0.286753   \n",
       "\n",
       "     yzhv10    yzhv22   yzhv250    yzhv44     yzhv5    yzhv66  \n",
       "0  0.214024  0.267942  0.433284  0.270279  0.257701  0.389017  \n",
       "1  0.234991  0.273084  0.433438  0.272863  0.296591  0.388105  \n",
       "2  0.244212  0.276789  0.433650  0.273368  0.310326  0.379251  \n",
       "3  0.248477  0.274844  0.433757  0.274230  0.288362  0.350372  \n",
       "4  0.247135  0.246933  0.433665  0.273243  0.252466  0.334571  \n",
       "\n",
       "[5 rows x 36 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    32.826614\n",
       "1    31.359858\n",
       "2    31.862687\n",
       "3    32.419264\n",
       "4    32.171463\n",
       "Name: vix, dtype: float64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Rolling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "predict_values = []\n",
    "true_values = []\n",
    "alp_star = []\n",
    "\n",
    "for i in range(0,length):\n",
    "    alp_start = 0.006\n",
    "    alp_stop =  0.011\n",
    "    alp = np.linspace(start=alp_start,stop=alp_stop,num=11)\n",
    "    alp = [round(i,4) for i in alp]\n",
    "    \n",
    "    X_train = X.iloc[i:len(y)-length+i,:].reset_index(drop=True)\n",
    "    y_train = y.iloc[i:len(y)-length+i].reset_index(drop=True)\n",
    "    a = lasso_cv(X_train,y_train,alp,10,False,10000)[0]\n",
    "    \n",
    "    while(a==alp_stop):\n",
    "        alp_start += 0.005\n",
    "        alp_stop += 0.005\n",
    "        alp = np.linspace(start=alp_start,stop=alp_stop,num=11)\n",
    "        alp = [round(i,4) for i in alp]\n",
    "        a = lasso_cv(X_train,y_train,alp,10,False,10000)[0]\n",
    "    \n",
    "    ls = Lasso(alpha = a,fit_intercept = False,max_iter = 10000)\n",
    "    ls.fit(X_train,y_train)\n",
    "    prediction = ls.predict(X.iloc[len(y)-length+i:len(y)-length+i+1])[0]\n",
    "    true_values.append(y[len(y)-length+i])\n",
    "    predict_values.append(prediction)\n",
    "    alp_star.append(a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x264a3334788>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.lineplot(np.arange(len(predict_values)),true_values,label='true')\n",
    "sns.lineplot(np.arange(len(predict_values)),predict_values,label='predict')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_date=df3.loc[len(X)-length:,'date']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Static"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train=X.iloc[0:len(X)-length,:]\n",
    "y_train=y[0:len(X)-length]\n",
    "X_test=X.iloc[len(X)-length:,:].reset_index(drop=True)\n",
    "y_test=y[len(X)-length:].reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def func_mae(x1,x2):\n",
    "    return np.mean(np.abs(x1-x2))\n",
    "\n",
    "\n",
    "kf = KFold(n_splits=10)\n",
    "alp = np.linspace(start=0.006,stop=0.02,num=40)\n",
    "\n",
    "mse_ave = []\n",
    "mae_ave = []\n",
    "\n",
    "for t in alp:\n",
    "    mse = []\n",
    "    mae = []\n",
    "    ls = Lasso(alpha=t,fit_intercept=False,max_iter=10000)\n",
    "    for k, (train, test) in enumerate(kf.split(X_train)):\n",
    "        ls.fit(X_train.loc[train],y_train.loc[train])\n",
    "        predictions = ls.predict(X_train.loc[test])\n",
    "        mse.append(sm.tools.eval_measures.mse(predictions,y_train.loc[test]))\n",
    "        mae.append(func_mae(predictions,y_train.loc[test]))\n",
    "    mse_ave.append(np.mean(mse))\n",
    "    mae_ave.append(np.mean(mae))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x264a96f0688>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.scatterplot(alp,mse_ave)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.008153846153846154"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a_star=pd.DataFrame({'alpha':alp,'mse':mse_ave}).sort_values(by='mse').reset_index().iloc[0,1]\n",
    "a_star"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>feature</th>\n",
       "      <th>coefficient</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>cchv22</td>\n",
       "      <td>3.296907</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>cchv250</td>\n",
       "      <td>3.913722</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>cchv44</td>\n",
       "      <td>18.953319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>cchv5</td>\n",
       "      <td>0.874950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>gkhv10</td>\n",
       "      <td>19.655932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>gkhv22</td>\n",
       "      <td>11.144272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>gkhv44</td>\n",
       "      <td>26.261632</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>gkhv5</td>\n",
       "      <td>7.160772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>gkyzhv5</td>\n",
       "      <td>-2.587251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>gkyzhv66</td>\n",
       "      <td>21.707756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>10</td>\n",
       "      <td>rshv66</td>\n",
       "      <td>0.462886</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>11</td>\n",
       "      <td>yzhv22</td>\n",
       "      <td>17.573455</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>12</td>\n",
       "      <td>yzhv5</td>\n",
       "      <td>6.717206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>13</td>\n",
       "      <td>yzhv66</td>\n",
       "      <td>5.087382</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     feature  coefficient\n",
       "0     cchv22     3.296907\n",
       "1    cchv250     3.913722\n",
       "2     cchv44    18.953319\n",
       "3      cchv5     0.874950\n",
       "4     gkhv10    19.655932\n",
       "5     gkhv22    11.144272\n",
       "6     gkhv44    26.261632\n",
       "7      gkhv5     7.160772\n",
       "8    gkyzhv5    -2.587251\n",
       "9   gkyzhv66    21.707756\n",
       "10    rshv66     0.462886\n",
       "11    yzhv22    17.573455\n",
       "12     yzhv5     6.717206\n",
       "13    yzhv66     5.087382"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ls = Lasso(alpha=a_star,fit_intercept=False,max_iter=10000)\n",
    "ls.fit(X_train,y_train)\n",
    "coefficient_matrix=pd.DataFrame({'feature':X_train.columns,'coefficient':ls.coef_})\n",
    "coefficient_matrix[coefficient_matrix.iloc[:,1]!=0].reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test_predict=ls.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x264a97c2ac8>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.distplot(y_test-y_test_predict,bins=40)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "15.910265396474948"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sm.tools.eval_measures.mse(y_test,y_test_predict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6.482071731137561"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sm.tools.eval_measures.mse(y_train,ls.predict(X_train))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>rolling predict</th>\n",
       "      <th>static predict</th>\n",
       "      <th>real vix</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>2019-03-01</td>\n",
       "      <td>27.962047</td>\n",
       "      <td>27.919863</td>\n",
       "      <td>29.965378</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>2019-03-04</td>\n",
       "      <td>29.655264</td>\n",
       "      <td>29.635086</td>\n",
       "      <td>32.584331</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>2019-03-05</td>\n",
       "      <td>29.292779</td>\n",
       "      <td>29.366355</td>\n",
       "      <td>33.718041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>2019-03-06</td>\n",
       "      <td>29.156175</td>\n",
       "      <td>29.193196</td>\n",
       "      <td>35.120822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>2019-03-07</td>\n",
       "      <td>29.857601</td>\n",
       "      <td>29.556089</td>\n",
       "      <td>36.295277</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>323</td>\n",
       "      <td>2020-07-03</td>\n",
       "      <td>21.937558</td>\n",
       "      <td>18.710070</td>\n",
       "      <td>27.789083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>324</td>\n",
       "      <td>2020-07-06</td>\n",
       "      <td>28.810273</td>\n",
       "      <td>22.970015</td>\n",
       "      <td>35.705403</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>325</td>\n",
       "      <td>2020-07-07</td>\n",
       "      <td>30.921841</td>\n",
       "      <td>24.366067</td>\n",
       "      <td>35.787732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>326</td>\n",
       "      <td>2020-07-08</td>\n",
       "      <td>32.201464</td>\n",
       "      <td>25.508971</td>\n",
       "      <td>37.600200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>327</td>\n",
       "      <td>2020-07-09</td>\n",
       "      <td>32.499841</td>\n",
       "      <td>25.803094</td>\n",
       "      <td>38.152322</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>328 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          date  rolling predict  static predict   real vix\n",
       "0   2019-03-01        27.962047       27.919863  29.965378\n",
       "1   2019-03-04        29.655264       29.635086  32.584331\n",
       "2   2019-03-05        29.292779       29.366355  33.718041\n",
       "3   2019-03-06        29.156175       29.193196  35.120822\n",
       "4   2019-03-07        29.857601       29.556089  36.295277\n",
       "..         ...              ...             ...        ...\n",
       "323 2020-07-03        21.937558       18.710070  27.789083\n",
       "324 2020-07-06        28.810273       22.970015  35.705403\n",
       "325 2020-07-07        30.921841       24.366067  35.787732\n",
       "326 2020-07-08        32.201464       25.508971  37.600200\n",
       "327 2020-07-09        32.499841       25.803094  38.152322\n",
       "\n",
       "[328 rows x 4 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results = pd.DataFrame({'date':test_date.reset_index(drop=True),'rolling predict':predict_values,\n",
    "                        'static predict':y_test_predict,'real vix':true_values})\n",
    "results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [],
   "source": [
    "results.to_csv('vix.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\yangl\\Anaconda3\\lib\\site-packages\\pandas\\plotting\\_matplotlib\\converter.py:103: FutureWarning: Using an implicitly registered datetime converter for a matplotlib plotting method. The converter was registered by pandas on import. Future versions of pandas will require you to explicitly register matplotlib converters.\n",
      "\n",
      "To register the converters:\n",
      "\t>>> from pandas.plotting import register_matplotlib_converters\n",
      "\t>>> register_matplotlib_converters()\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x264a9629808>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,6))\n",
    "sns.lineplot(test_date,true_values,label='true')\n",
    "sns.lineplot(test_date,predict_values,label='predict by rolling')\n",
    "sns.lineplot(test_date,y_test_predict,label='predict by static model')\n",
    "plt.title('Rolling V.S. Static')\n",
    "plt.legend(loc=2)"
   ]
  }
 ],
 "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": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}