class_project.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys,os\n",
    "import yfinance as yf\n",
    "import pandas as pd\n",
    "import requests\n",
    "from bs4 import BeautifulSoup\n",
    "import plotly.graph_objects as go\n",
    "from plotly.subplots import make_subplots\n",
    "import numpy as np \n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import warnings\n",
    "from datetime import datetime\n",
    "from datetime import timedelta\n",
    "from dateutil.relativedelta import relativedelta\n",
    "from IPython.display import display\n",
    "\n",
    "#os.chdir(\"/content\")\n",
    "#from colorsetup import colors, palette\n",
    "#sns.set_palette(palette)\n",
    "\n",
    "# ignore warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "%matplotlib inline\n",
    "plotsize = (13, 5)\n",
    "plt.rcParams['figure.figsize'] = plotsize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from matplotlib.pyplot import subplots\n",
    "from sklearn.linear_model import \\\n",
    "     (LinearRegression,\n",
    "      LogisticRegression,\n",
    "      Lasso)\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.pipeline import Pipeline\n",
    "from ISLP import load_data\n",
    "from ISLP.models import ModelSpec as MS\n",
    "from sklearn.model_selection import \\\n",
    "     (train_test_split,\n",
    "      GridSearchCV)\n",
    "\n",
    "import torch\n",
    "from torch import nn\n",
    "from torch.optim import RMSprop\n",
    "from torch.utils.data import TensorDataset\n",
    "\n",
    "from torchmetrics import (MeanAbsoluteError,\n",
    "                          R2Score)\n",
    "from torchinfo import summary\n",
    "\n",
    "from pytorch_lightning import Trainer\n",
    "from pytorch_lightning.loggers import CSVLogger"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from ISLP.torch import (SimpleDataModule,\n",
    "                        SimpleModule,\n",
    "                        ErrorTracker,\n",
    "                        rec_num_workers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import statsmodels.api as sm\n",
    "from ISLP import load_data\n",
    "from ISLP.models import (#ModelSpec as MS,\n",
    "                         summarize) # the enclosing round parentheses is to write codes in multiple lines. \n",
    "from ISLP import confusion_table\n",
    "from ISLP.models import contrast\n",
    "from sklearn.discriminant_analysis import \\\n",
    "     (LinearDiscriminantAnalysis as LDA,\n",
    "      QuadraticDiscriminantAnalysis as QDA)\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "# "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.cross_decomposition import PLSRegression\n",
    "from ISLP.models import \\\n",
    "     (Stepwise,\n",
    "      sklearn_selected,\n",
    "      sklearn_selection_path)\n",
    "#!pip install l0bnb #[yw] in case error\" \"Numba needs Numpy 1.20 or less\", upgrade Numba by: pip install numba --upgrade\n",
    "from l0bnb import fit_path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.api import OLS\n",
    "import sklearn.model_selection as skm\n",
    "import sklearn.linear_model as skl\n",
    "\n",
    "from ISLP import load_data\n",
    "\n",
    "from functools import partial"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.tree import (DecisionTreeClassifier as DTC,\n",
    "                          DecisionTreeRegressor as DTR,\n",
    "                          plot_tree,\n",
    "                          export_text)\n",
    "from sklearn.metrics import (accuracy_score,\n",
    "                             log_loss)\n",
    "from sklearn.ensemble import \\\n",
    "     (RandomForestRegressor as RF,\n",
    "      GradientBoostingRegressor as GBR)\n",
    "from ISLP.bart import BART\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.svm import SVC\n",
    "from ISLP.svm import plot as plot_svm\n",
    "from sklearn.metrics import RocCurveDisplay"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# User defined parameters. Change as needed. \n",
    "DATA_FILE_PATH = 'apple_20240301.csv' # csv file to save the downloaded data\n",
    "MV_AVG_WINDOW = 100 #moving avg window size to normalize the volume data\n",
    "STOCK_MV_DIRECTION_THRESHOLD = 0 # return treshold to determine the stock movement direction\n",
    "TEST_START_DATE = '2022-12-25' #testing start date \n",
    "#this date is chosen to make sure that the last batch, \n",
    "# when using batch size 64, has at least two examples. The last batch must have at least two examples \n",
    "# to computer R^2  when using Torch NN. \n",
    "MAX_LAG = 5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Obtain the data and write it to a csv file. Run it only once unless you want to update data\n",
    "# apple = yf.Ticker(\"AAPL\")\n",
    "# apple_share_price_data = apple.history(period=\"max\")\n",
    "# apple_share_price_data.to_csv(DATA_FILE_PATH)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# utility file\n",
    "def read_data(csv_file='apple_df.csv',dc='Date', verbose = False):\n",
    "  \"\"\"\n",
    "  Read a csv file into a dataframe and convert the date string into a date index\n",
    "  INPUT: \n",
    "  csv_file: complete csv_file path including file name\n",
    "  dc: Date column name\n",
    "  OUTPUT: \n",
    "  df: a dataFrame with date value as index\n",
    "  \"\"\"\n",
    "  df =pd.read_csv(csv_file)\n",
    "  \n",
    "  df[dc]=df[dc].apply(lambda x: x[:11])\n",
    "  df.set_index(dc, inplace=True)\n",
    "  df.index= pd.DatetimeIndex(df.index)\n",
    "  if verbose: \n",
    "    print(df.head())\n",
    "    print(\"\\n\")\n",
    "    print(\"*\"*20+\"DataFrame Info\"+\"*\"*20)\n",
    "    df.info()\n",
    "    print(\"\\n\")\n",
    "    print(df.describe().transpose())\n",
    "  return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 10895 entries, 1980-12-12 to 2024-03-01\n",
      "Data columns (total 7 columns):\n",
      " #   Column        Non-Null Count  Dtype  \n",
      "---  ------        --------------  -----  \n",
      " 0   Open          10895 non-null  float64\n",
      " 1   High          10895 non-null  float64\n",
      " 2   Low           10895 non-null  float64\n",
      " 3   Close         10895 non-null  float64\n",
      " 4   Volume        10895 non-null  int64  \n",
      " 5   Dividends     10895 non-null  float64\n",
      " 6   Stock Splits  10895 non-null  float64\n",
      "dtypes: float64(6), int64(1)\n",
      "memory usage: 680.9 KB\n"
     ]
    }
   ],
   "source": [
    "data_file_path = DATA_FILE_PATH\n",
    "apple = read_data(data_file_path)\n",
    "apple.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Data preprocessing utility\n",
    "### remove the rows with Volume =0, on which there was no transaction \n",
    "apple = apple[apple['Volume'] != 0]\n",
    "\n",
    "apple.drop(['Dividends','Stock Splits'], axis=1, inplace=True)\n",
    "apple['month'] = apple.index.month\n",
    "apple['month'] = apple['month'].astype('category')\n",
    "\n",
    "apple['day_of_week'] = apple.index.day_name()\n",
    "apple['day_of_week'] = apple['day_of_week'].astype('category')\n",
    "\n",
    "day_order = [ 'Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday']\n",
    "apple['day_of_week'] = apple['day_of_week'].cat.set_categories(day_order)\n",
    "\n",
    "apple['return']=np.log(apple.Close/(apple.Close.shift(1)))\n",
    "apple['volatility'] = np.fabs(apple['return'])\n",
    "\n",
    "window = MV_AVG_WINDOW\n",
    "apple['log_volume'] = np.log(apple['Volume']/apple['Volume'].rolling(window).mean().shift(1)) \n",
    "#rolling creates 99 Nan, shift creates 1 nan: togeter 100 nan\n",
    "\n",
    "direction_threshold = STOCK_MV_DIRECTION_THRESHOLD\n",
    "apple['direction'] = apple['return'] > direction_threshold\n",
    "\n",
    "apple.dropna(inplace=True)\n",
    "\n",
    "apple['train'] =  ~(apple.index >= TEST_START_DATE )  #this date is chosen to make sure that the last batch, \n",
    "# when using batch size 64, has at least two examples. The last batch must have at least two examples \n",
    "# to computer R^2  when using Torch NN. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Open', 'High', 'Low', 'Close', 'Volume', 'month', 'day_of_week',\n",
       "       'return', 'volatility', 'log_volume', 'direction', 'train'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "apple.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "#### Normalize the indicated columns with z-values\n",
    "cols = ['return', 'log_volume', 'volatility']\n",
    "X = pd.DataFrame(StandardScaler(\n",
    "                     with_mean=True,\n",
    "                     with_std=True).fit_transform(apple[cols]),\n",
    "                 columns=apple[cols].columns,\n",
    "                 index=apple.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculate the lag columns of the columns above \n",
    "max_lag = MAX_LAG\n",
    "for lag in range(1, max_lag+1):\n",
    "    for col in cols:\n",
    "        newcol = np.zeros(X.shape[0]) * np.nan\n",
    "        newcol[lag:] = X[col].values[:-lag]\n",
    "        X.insert(len(X.columns), \"{0}_{1}\".format(col, lag), newcol)#insert at the end\n",
    "X.insert(len(X.columns), 'train', apple['train']) #insert the col training identifier \n",
    "X = X.dropna() #drop max_lag=5 rows with nan\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['return', 'log_volume', 'volatility', 'return_1', 'log_volume_1',\n",
       "       'volatility_1', 'return_2', 'log_volume_2', 'volatility_2', 'return_3',\n",
       "       'log_volume_3', 'volatility_3', 'return_4', 'log_volume_4',\n",
       "       'volatility_4', 'return_5', 'log_volume_5', 'volatility_5', 'train'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train days:  10493 test days:  296\n"
     ]
    }
   ],
   "source": [
    "train_len, test_len = len(X[X['train']==True]), len(X[X['train']!=True])\n",
    "print(\"train days: \", train_len, \"test days: \", test_len)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note: \n",
    "\n",
    "- `X`: contains all the numerical columns (return, volativilty and log_volume) and the bool col `train` \n",
    "- `apple`: contains the categorial cols: `month` and `day_of_week`. \n",
    "\n",
    "These two dataframes should not be modified and should be treated as original dataframes. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Classification\n",
    "## Logistic Regression using statsmodels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test = X.loc[X['train']], X.loc[~X['train']] # X[X['train'] == True] equiv to  X.loc[X['train']]\n",
    "Y_train, Y_test = X_train['return']>0, X_test['return']>0 # generate the binary response: True is UP. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((10493, 19), (296, 19), (10493,), (296,))"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, X_test.shape, Y_train.shape, Y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>coef</th>\n",
       "      <th>std err</th>\n",
       "      <th>z</th>\n",
       "      <th>P&gt;|z|</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>intercept</th>\n",
       "      <td>-0.0495</td>\n",
       "      <td>0.020</td>\n",
       "      <td>-2.530</td>\n",
       "      <td>0.011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_5</th>\n",
       "      <td>-0.0083</td>\n",
       "      <td>0.019</td>\n",
       "      <td>-0.429</td>\n",
       "      <td>0.668</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_4</th>\n",
       "      <td>0.0508</td>\n",
       "      <td>0.020</td>\n",
       "      <td>2.593</td>\n",
       "      <td>0.010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_3</th>\n",
       "      <td>-0.0388</td>\n",
       "      <td>0.020</td>\n",
       "      <td>-1.979</td>\n",
       "      <td>0.048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_2</th>\n",
       "      <td>-0.0680</td>\n",
       "      <td>0.020</td>\n",
       "      <td>-3.444</td>\n",
       "      <td>0.001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_1</th>\n",
       "      <td>-0.0098</td>\n",
       "      <td>0.019</td>\n",
       "      <td>-0.507</td>\n",
       "      <td>0.612</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>log_volume_1</th>\n",
       "      <td>0.0203</td>\n",
       "      <td>0.019</td>\n",
       "      <td>1.046</td>\n",
       "      <td>0.296</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                coef  std err      z  P>|z|\n",
       "intercept    -0.0495    0.020 -2.530  0.011\n",
       "return_5     -0.0083    0.019 -0.429  0.668\n",
       "return_4      0.0508    0.020  2.593  0.010\n",
       "return_3     -0.0388    0.020 -1.979  0.048\n",
       "return_2     -0.0680    0.020 -3.444  0.001\n",
       "return_1     -0.0098    0.019 -0.507  0.612\n",
       "log_volume_1  0.0203    0.019  1.046  0.296"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictors = ['return_5', 'return_4','return_3', 'return_2', 'return_1', 'log_volume_1']\n",
    "design = MS(predictors)\n",
    "X_train_5 = design.fit_transform(X_train)\n",
    "y = Y_train\n",
    "glm = sm.GLM(y,\n",
    "             X_train_5,\n",
    "             family=sm.families.Binomial())\n",
    "results = glm.fit()\n",
    "summarize(results)"
   ]
  },
  {
   "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>Truth</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Predicted</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>False</th>\n",
       "      <td>133</td>\n",
       "      <td>131</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>True</th>\n",
       "      <td>14</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Truth      False  True \n",
       "Predicted              \n",
       "False        133    131\n",
       "True          14     18"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test_5 = design.transform(X_test)\n",
    "probs = results.predict(exog= X_test_5)\n",
    "\n",
    "labels = probs > 0.5\n",
    "confusion_table(labels, Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5101351351351351"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Test Accuracy\n",
    "np.mean(labels == Y_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             coef  std err      z  P>|z|\n",
      "intercept -0.0494    0.020 -2.529  0.011\n",
      "return_2  -0.0694    0.020 -3.515  0.000\n",
      "return_1  -0.0101    0.019 -0.522  0.601 \n",
      "\n",
      "Confusion Table: \n",
      " Truth      False  True \n",
      "Predicted              \n",
      "False        144    139\n",
      "True           3     10\n",
      "Test Accuracy:  0.5202702702702703\n"
     ]
    }
   ],
   "source": [
    "# Using two lags\n",
    "predictors = [ 'return_2', 'return_1']\n",
    "design = MS(predictors)\n",
    "X_train_2 = design.fit_transform(X_train)\n",
    "y = Y_train\n",
    "glm = sm.GLM(y,\n",
    "             X_train_2,\n",
    "             family=sm.families.Binomial())\n",
    "results = glm.fit()\n",
    "print(summarize(results), '\\n')\n",
    "X_test_2 = design.transform(X_test)\n",
    "probs = results.predict(exog= X_test_2)\n",
    "\n",
    "labels = probs > 0.5\n",
    "print(\"Confusion Table: \\n\", confusion_table(labels, Y_test) )\n",
    "### Test Accuracy\n",
    "print(\"Test Accuracy: \", np.mean(labels == Y_test) )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Logistic Regression using SK-learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_2, X_test_2 = [M.drop(columns=['intercept']) #sklearn automatically adds an intercept, drop it from the design matrix by statsmodels\n",
    "                   for M in [X_train_2, X_test_2]]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`sklearn`,  by default it fits\n",
    "something like the *ridge regression* version\n",
    "of logistic regression. This can\n",
    "be modified by appropriately setting the argument `C` below. Its default\n",
    "value is 1 but by setting it to a very large number, the algorithm converges to the same solution as the usual (unregularized)\n",
    "logistic regression estimator discussed above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>Truth</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Predicted</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>False</th>\n",
       "      <td>144</td>\n",
       "      <td>139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>True</th>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Truth      False  True \n",
       "Predicted              \n",
       "False        144    139\n",
       "True           3     10"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "logit = LogisticRegression(C=1e10, solver='liblinear')\n",
    "logit.fit(X_train_2, Y_train)\n",
    "logit_pred = logit.predict_proba(X_test_2)\n",
    "logit_labels = np.where(logit_pred[:,1] > .5, True, False)\n",
    "confusion_table(logit_labels, Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5202702702702703"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Logistic accuracy\n",
    "np.mean(logit_labels == Y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## LDA using Sk-learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-1 {color: black;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>LinearDiscriminantAnalysis(store_covariance=True)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">LinearDiscriminantAnalysis</label><div class=\"sk-toggleable__content\"><pre>LinearDiscriminantAnalysis(store_covariance=True)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "LinearDiscriminantAnalysis(store_covariance=True)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lda = LDA(store_covariance=True) #store the covariance of each class\n",
    "\n",
    "lda.fit(X_train_2, Y_train) # note sklearn use the order X,Y, opposite the order used by statsmodels: Y, X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.03383844,  0.0050342 ],\n",
       "       [-0.03609458, -0.0059984 ]])"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lda.means_ #array-like of shape (n_classes, n_features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([False,  True]), array([0.51234156, 0.48765844]))"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lda.classes_, lda.priors_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.97727746],\n",
       "       [-0.14330167]])"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lda.scalings_ #Scaling of the features in the space spanned by the class centroids. \n",
    "# Only available for ‘svd’ and ‘eigen’ solvers."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These values provide the linear combination of `Lag1`  and `Lag2`  that are used to form the LDA decision rule (boundary). In other words, these are the multipliers of the elements of $X=x$ in (4.24).\n",
    "  If $-0.977\\times `Lag1`  - 0.143 \\times `Lag2` $ is large, then the LDA classifier will predict a market increase, and if it is small, then the LDA classifier will predict a market decline. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>Truth</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Predicted</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>False</th>\n",
       "      <td>144</td>\n",
       "      <td>139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>True</th>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Truth      False  True \n",
       "Predicted              \n",
       "False        144    139\n",
       "True           3     10"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lda_pred = lda.predict(X_test_2)\n",
    "confusion_table(lda_pred, Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5202702702702703"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# LDA accuracy\n",
    "np.mean(lda_pred == Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lda_prob = lda.predict_proba(X_test_2) #ndarray of shape (n_samples, n_classes)\n",
    "np.all(\n",
    "       np.where(lda_prob[:,1] >= 0.5, True, False) == lda_pred\n",
    "       )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## QDA using Sk-Learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-2 {color: black;}#sk-container-id-2 pre{padding: 0;}#sk-container-id-2 div.sk-toggleable {background-color: white;}#sk-container-id-2 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-2 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-2 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-2 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-2 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-2 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-2 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-2 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-2 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-2 div.sk-item {position: relative;z-index: 1;}#sk-container-id-2 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-2 div.sk-item::before, #sk-container-id-2 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-2 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-2 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-2 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-2 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-2 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-2 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-2 div.sk-label-container {text-align: center;}#sk-container-id-2 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-2 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>QuadraticDiscriminantAnalysis(store_covariance=True)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" checked><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">QuadraticDiscriminantAnalysis</label><div class=\"sk-toggleable__content\"><pre>QuadraticDiscriminantAnalysis(store_covariance=True)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "QuadraticDiscriminantAnalysis(store_covariance=True)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "qda = QDA(store_covariance=True)\n",
    "qda.fit(X_train_2, Y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([[ 0.03383844,  0.0050342 ],\n",
       "        [-0.03609458, -0.0059984 ]]),\n",
       " array([0.51234156, 0.48765844]))"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "qda.means_, qda.priors_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([[ 1.10241163, -0.00317939],\n",
       "        [-0.00317939,  1.06730139]]),\n",
       " array([[0.93688266, 0.02681325],\n",
       "        [0.02681325, 0.97636331]]))"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "qda.covariance_[0], qda.covariance_[1] #Covariance of each class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "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>Truth</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Predicted</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>False</th>\n",
       "      <td>13</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>True</th>\n",
       "      <td>134</td>\n",
       "      <td>128</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Truth      False  True \n",
       "Predicted              \n",
       "False         13     21\n",
       "True         134    128"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "qda_pred = qda.predict(X_test_2)\n",
    "confusion_table(qda_pred, Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.47635135135135137"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# QDA Accuracy\n",
    "np.mean(qda_pred == Y_test)  # much worse than LDA"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Naive Bayes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-3 {color: black;}#sk-container-id-3 pre{padding: 0;}#sk-container-id-3 div.sk-toggleable {background-color: white;}#sk-container-id-3 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-3 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-3 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-3 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-3 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-3 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-3 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-3 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-3 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-3 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-3 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-3 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-3 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-3 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-3 div.sk-item {position: relative;z-index: 1;}#sk-container-id-3 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-3 div.sk-item::before, #sk-container-id-3 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-3 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-3 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-3 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-3 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-3 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-3 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-3 div.sk-label-container {text-align: center;}#sk-container-id-3 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-3 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-3\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GaussianNB()</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" checked><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GaussianNB</label><div class=\"sk-toggleable__content\"><pre>GaussianNB()</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "GaussianNB()"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "NB = GaussianNB()\n",
    "NB.fit(X_train_2, Y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([False,  True]), array([0.51234156, 0.48765844]))"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "NB.classes_, NB.class_prior_, "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.03383844,  0.0050342 ],\n",
       "       [-0.03609458, -0.0059984 ]])"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "NB.theta_  # class mean, "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.10220657, 1.06710286],\n",
       "       [0.93669957, 0.97617251]])"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "NB.var_ # Class variance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "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>Truth</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Predicted</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>False</th>\n",
       "      <td>12</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>True</th>\n",
       "      <td>135</td>\n",
       "      <td>131</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Truth      False  True \n",
       "Predicted              \n",
       "False         12     18\n",
       "True         135    131"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nb_labels = NB.predict(X_test_2)\n",
    "confusion_table(nb_labels, Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.4831081081081081"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# NB Accuracy\n",
    "np.mean(nb_labels == Y_test)  # much worse than LDA"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "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>Truth</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Predicted</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>False</th>\n",
       "      <td>79</td>\n",
       "      <td>79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>True</th>\n",
       "      <td>68</td>\n",
       "      <td>70</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Truth      False  True \n",
       "Predicted              \n",
       "False         79     79\n",
       "True          68     70"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn1 = KNeighborsClassifier(n_neighbors=1)\n",
    "X_train_2, X_test_2 = [np.asarray(X) for X in [X_train_2, X_test_2]]\n",
    "knn1.fit(X_train_2, Y_train)\n",
    "knn1_pred = knn1.predict(X_test_2)\n",
    "confusion_table(knn1_pred, Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5033783783783784"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Accuracy\n",
    "np.mean(knn1_pred == Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "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>Truth</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Predicted</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>False</th>\n",
       "      <td>79</td>\n",
       "      <td>79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>True</th>\n",
       "      <td>68</td>\n",
       "      <td>70</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Truth      False  True \n",
       "Predicted              \n",
       "False         79     79\n",
       "True          68     70"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn3 = KNeighborsClassifier(n_neighbors=3)\n",
    "X_train_2, X_test_2 = [np.asarray(X) for X in [X_train_2, X_test_2]]\n",
    "knn3.fit(X_train_2, Y_train)\n",
    "knn3_pred = knn3.predict(X_test_2)\n",
    "confusion_table(knn1_pred, Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.4594594594594595"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(knn3_pred == Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K=1: # predicted to Go up: 138,  # which did Go up 70, accuracy 50.7%\n",
      "K=2: # predicted to Go up: 61,  # which did Go up 32, accuracy 52.5%\n",
      "K=3: # predicted to Go up: 119,  # which did Go up 54, accuracy 45.4%\n",
      "K=4: # predicted to Go up: 71,  # which did Go up 29, accuracy 40.8%\n",
      "K=5: # predicted to Go up: 120,  # which did Go up 55, accuracy 45.8%\n"
     ]
    }
   ],
   "source": [
    "### Tuning the KNN parameters\n",
    "for K in range(1,6):\n",
    "    knn = KNeighborsClassifier(n_neighbors=K)\n",
    "    knn_pred = knn.fit(X_train_2, Y_train).predict(X_test_2)\n",
    "    C = confusion_table(knn_pred, Y_test)\n",
    "    templ = ('K={0:d}: # predicted to Go up: {1:>2},' +  # > for right alighment\n",
    "            '  # which did Go up {2:d}, accuracy {3:.1%}')\n",
    "    pred = C.loc[True].sum()\n",
    "    did_up = C.loc[True,True]\n",
    "    print(templ.format(\n",
    "          K,\n",
    "          pred,\n",
    "          did_up,\n",
    "          did_up / pred))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier as RFC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-4 {color: black;}#sk-container-id-4 pre{padding: 0;}#sk-container-id-4 div.sk-toggleable {background-color: white;}#sk-container-id-4 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-4 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-4 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-4 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-4 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-4 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-4 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-4 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-4 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-4 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-4 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-4 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-4 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-4 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-4 div.sk-item {position: relative;z-index: 1;}#sk-container-id-4 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-4 div.sk-item::before, #sk-container-id-4 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-4 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-4 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-4 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-4 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-4 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-4 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-4 div.sk-label-container {text-align: center;}#sk-container-id-4 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-4 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-4\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>RandomForestClassifier(n_estimators=500, random_state=0)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-4\" type=\"checkbox\" checked><label for=\"sk-estimator-id-4\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomForestClassifier</label><div class=\"sk-toggleable__content\"><pre>RandomForestClassifier(n_estimators=500, random_state=0)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "RandomForestClassifier(n_estimators=500, random_state=0)"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc = RFC(max_features='sqrt', random_state=0, n_estimators=500,)\n",
    "rfc.fit(X_train_2, Y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RFC Accuracy:  0.5033783783783784\n"
     ]
    }
   ],
   "source": [
    "rfc_pred = rfc.predict(X_test_2)\n",
    "confusion_table(rfc_pred, Y_test)\n",
    "print(\"RFC Accuracy: \", np.mean(rfc_pred == Y_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'C': 100, 'gamma': 1}"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# warning: takes a long time to run\n",
    "svm_rbf = SVC(kernel=\"rbf\", gamma=1, C=1)\n",
    "kfold = skm.KFold(5, \n",
    "                  random_state=0,\n",
    "                  shuffle=True)\n",
    "grid = skm.GridSearchCV(svm_rbf,\n",
    "                        {'C':[0.1,1,10,100,1000],\n",
    "                         'gamma':[0.5,1,2,3,4]},\n",
    "                        refit=True,\n",
    "                        cv=kfold,\n",
    "                        scoring='accuracy');\n",
    "grid.fit(X_train_5, Y_train)\n",
    "grid.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "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>Truth</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Predicted</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>False</th>\n",
       "      <td>57</td>\n",
       "      <td>80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>True</th>\n",
       "      <td>90</td>\n",
       "      <td>69</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Truth      False  True \n",
       "Predicted              \n",
       "False         57     80\n",
       "True          90     69"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_svm = grid.best_estimator_\n",
    "\n",
    "y_hat_test = best_svm.predict(X_test_5)\n",
    "confusion_table(y_hat_test, Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.42567567567567566, 0.4339622641509434)"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(y_hat_test == Y_test), 69/(69+90)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ROC curve\n",
    "roc_curve = RocCurveDisplay.from_estimator # shorthand\n",
    "fig, ax = subplots(figsize=(8,8))\n",
    "roc_curve(best_svm,\n",
    "          X_test_5,\n",
    "          Y_test,\n",
    "          name='Testing',\n",
    "          color='r',\n",
    "          ax=ax);\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Regression\n",
    "## Linear  AR Model with SK-Learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols = ['return', 'log_volume', 'volatility']\n",
    "Y, train = X['log_volume'], X['train']\n",
    "X_reg = X.drop(columns=['train'] +cols)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.19997338230088157"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "M = LinearRegression()\n",
    "M.fit(X_reg[train], Y[train])\n",
    "M.score(X_reg[~train], Y[~train]) # the test result is not good. R^2=0.2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Adding `day_of_week`\n",
    "X_day = pd.merge(X_reg, \n",
    "                 pd.get_dummies(apple['day_of_week']),\n",
    "                 on='Date')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.17872110527570284"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "M.fit(X_day[train], Y[train])\n",
    "M.score(X_day[~train], Y[~train])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Linear AR using statmodels\n",
    "Here is only for the purpose of practicing using the statsmodels  OLS method. A better way for time searies AR model is to use statsmodels SARIMA model. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "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>coef</th>\n",
       "      <th>std err</th>\n",
       "      <th>t</th>\n",
       "      <th>P&gt;|t|</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>intercept</th>\n",
       "      <td>0.0010</td>\n",
       "      <td>0.008</td>\n",
       "      <td>0.129</td>\n",
       "      <td>0.898</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_1</th>\n",
       "      <td>-0.0114</td>\n",
       "      <td>0.008</td>\n",
       "      <td>-1.465</td>\n",
       "      <td>0.143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>log_volume_1</th>\n",
       "      <td>0.4693</td>\n",
       "      <td>0.011</td>\n",
       "      <td>42.573</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>volatility_1</th>\n",
       "      <td>0.0734</td>\n",
       "      <td>0.009</td>\n",
       "      <td>8.030</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_2</th>\n",
       "      <td>-0.0164</td>\n",
       "      <td>0.008</td>\n",
       "      <td>-2.106</td>\n",
       "      <td>0.035</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>log_volume_2</th>\n",
       "      <td>0.0729</td>\n",
       "      <td>0.012</td>\n",
       "      <td>6.053</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>volatility_2</th>\n",
       "      <td>-0.0376</td>\n",
       "      <td>0.009</td>\n",
       "      <td>-4.118</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_3</th>\n",
       "      <td>0.0028</td>\n",
       "      <td>0.008</td>\n",
       "      <td>0.364</td>\n",
       "      <td>0.715</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>log_volume_3</th>\n",
       "      <td>0.0566</td>\n",
       "      <td>0.012</td>\n",
       "      <td>4.696</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>volatility_3</th>\n",
       "      <td>-0.0240</td>\n",
       "      <td>0.009</td>\n",
       "      <td>-2.626</td>\n",
       "      <td>0.009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_4</th>\n",
       "      <td>0.0088</td>\n",
       "      <td>0.008</td>\n",
       "      <td>1.137</td>\n",
       "      <td>0.255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>log_volume_4</th>\n",
       "      <td>0.0602</td>\n",
       "      <td>0.012</td>\n",
       "      <td>4.999</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>volatility_4</th>\n",
       "      <td>-0.0171</td>\n",
       "      <td>0.009</td>\n",
       "      <td>-1.875</td>\n",
       "      <td>0.061</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_5</th>\n",
       "      <td>-0.0002</td>\n",
       "      <td>0.008</td>\n",
       "      <td>-0.023</td>\n",
       "      <td>0.982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>log_volume_5</th>\n",
       "      <td>0.0650</td>\n",
       "      <td>0.011</td>\n",
       "      <td>6.096</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>volatility_5</th>\n",
       "      <td>-0.0252</td>\n",
       "      <td>0.009</td>\n",
       "      <td>-2.782</td>\n",
       "      <td>0.005</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                coef  std err       t  P>|t|\n",
       "intercept     0.0010    0.008   0.129  0.898\n",
       "return_1     -0.0114    0.008  -1.465  0.143\n",
       "log_volume_1  0.4693    0.011  42.573  0.000\n",
       "volatility_1  0.0734    0.009   8.030  0.000\n",
       "return_2     -0.0164    0.008  -2.106  0.035\n",
       "log_volume_2  0.0729    0.012   6.053  0.000\n",
       "volatility_2 -0.0376    0.009  -4.118  0.000\n",
       "return_3      0.0028    0.008   0.364  0.715\n",
       "log_volume_3  0.0566    0.012   4.696  0.000\n",
       "volatility_3 -0.0240    0.009  -2.626  0.009\n",
       "return_4      0.0088    0.008   1.137  0.255\n",
       "log_volume_4  0.0602    0.012   4.999  0.000\n",
       "volatility_4 -0.0171    0.009  -1.875  0.061\n",
       "return_5     -0.0002    0.008  -0.023  0.982\n",
       "log_volume_5  0.0650    0.011   6.096  0.000\n",
       "volatility_5 -0.0252    0.009  -2.782  0.005"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "design = MS(X_reg.columns) #add \"intercept=False\" if no intercept.\n",
    "X_tr = design.fit_transform(X_reg[train]) \n",
    "\n",
    "y = Y[train]\n",
    "model = sm.OLS(y, X_tr)\n",
    "results = model.fit()\n",
    "summarize(results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "def calc_r2_MSE(yhat, y):\n",
    "    \"\"\"\n",
    "    Calculate the R^2 and MSE\n",
    "\n",
    "    INPUT: \n",
    "    yhat: predicted value vectors\n",
    "    y: ground-truth y-values\n",
    "    \n",
    "    OUTPUT: tuple (r2, mse)\n",
    "    \"\"\"\n",
    "    r2 = 1- np.sum((yhat-y)**2)/np.sum((y-np.mean(y))**2)\n",
    "    mse = np.mean((y-yhat)**2)\n",
    "    return (r2, mse)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.19997338230088169, 0.2226826900933754)"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_te = design.transform(X_reg[~train])\n",
    "yhat = results.predict(X_te)\n",
    "y_test = Y[~train]\n",
    "#Calculate r^2\n",
    "\n",
    "calc_r2_MSE(yhat, y_test)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Variable Selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define the scoring function\n",
    "def nCp(sigma2, estimator, X, Y):\n",
    "    \"Negative Cp statistic\"\n",
    "    n, p = X.shape\n",
    "    Yhat = estimator.predict(X)\n",
    "    RSS = np.sum((Y - Yhat)**2)\n",
    "    return -(RSS + 2 * p * sigma2) / n "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.627736921302886"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "design = MS(X_reg.columns).fit(X_reg)\n",
    "Y_ = np.array(Y)\n",
    "X_ = design.transform(X_reg)\n",
    "sigma2 = OLS(Y_,X_).fit().scale  #.scale: RSE: residual standard error estimating sigma^2\n",
    "sigma2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('log_volume_1',\n",
       " 'log_volume_2',\n",
       " 'log_volume_3',\n",
       " 'log_volume_4',\n",
       " 'log_volume_5',\n",
       " 'return_1',\n",
       " 'return_2',\n",
       " 'volatility_1',\n",
       " 'volatility_2',\n",
       " 'volatility_3',\n",
       " 'volatility_4',\n",
       " 'volatility_5')"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "neg_Cp = partial(nCp, sigma2) #freeze sigma2 in the function nCp\n",
    "strategy = Stepwise.first_peak(design, # note have to use design as defined above which is an estimator instance. \n",
    "                               direction='forward',\n",
    "                               max_terms=len(design.terms))\n",
    "hitters_Cp = sklearn_selected(OLS,\n",
    "                               strategy,\n",
    "                               scoring=neg_Cp)#default scoring MSE, will choose all variables\n",
    "hitters_Cp.fit(X_reg, Y) # the same as hitters_Cp.fit(Hitters.drop('Salary', axis=1), Y), as strategy specify the design matrix\n",
    "hitters_Cp.selected_state_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('log_volume_1',\n",
       " 'log_volume_2',\n",
       " 'log_volume_3',\n",
       " 'log_volume_4',\n",
       " 'log_volume_5',\n",
       " 'return_1',\n",
       " 'return_2',\n",
       " 'volatility_1',\n",
       " 'volatility_2',\n",
       " 'volatility_3',\n",
       " 'volatility_4',\n",
       " 'volatility_5')"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Backward Selection\n",
    "strategy = Stepwise.first_peak(design, # note have to use design as defined above which is an estimator instance. \n",
    "                               direction='backward',\n",
    "                               initial_terms= design.terms,\n",
    "                               max_terms=len(design.terms))\n",
    "hitters_Cp = sklearn_selected(OLS,\n",
    "                               strategy,\n",
    "                               scoring=neg_Cp)#default scoring MSE, will choose all variables\n",
    "hitters_Cp.fit(X_reg, Y) # the same as hitters_Cp.fit(Hitters.drop('Salary', axis=1), Y), as strategy specify the design matrix\n",
    "hitters_Cp.selected_state_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ridge regression using SKLearning Pipeline and Cross-Validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-5 {color: black;}#sk-container-id-5 pre{padding: 0;}#sk-container-id-5 div.sk-toggleable {background-color: white;}#sk-container-id-5 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-5 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-5 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-5 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-5 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-5 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-5 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-5 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-5 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-5 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-5 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-5 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-5 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-5 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-5 div.sk-item {position: relative;z-index: 1;}#sk-container-id-5 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-5 div.sk-item::before, #sk-container-id-5 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-5 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-5 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-5 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-5 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-5 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-5 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-5 div.sk-label-container {text-align: center;}#sk-container-id-5 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-5 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-5\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GridSearchCV(cv=KFold(n_splits=5, random_state=0, shuffle=True),\n",
       "             estimator=Pipeline(steps=[(&#x27;scaler&#x27;, StandardScaler()),\n",
       "                                       (&#x27;ridge&#x27;,\n",
       "                                        ElasticNet(alpha=0.01, l1_ratio=0))]),\n",
       "             param_grid={&#x27;ridge__alpha&#x27;: array([9.91526673e+07, 7.85767931e+07, 6.22707648e+07, 4.93485161e+07,\n",
       "       3.91078551e+07, 3.09923064e+07, 2.45608728e+07, 1.94640716e+07,\n",
       "       1.54249439e+07, 1.22240043e+07, 9.68...\n",
       "       2.08706860e+00, 1.65396618e+00, 1.31073991e+00, 1.03873896e+00,\n",
       "       8.23182863e-01, 6.52358341e-01, 5.16982829e-01, 4.09700051e-01,\n",
       "       3.24680284e-01, 2.57303572e-01, 2.03908681e-01, 1.61594142e-01,\n",
       "       1.28060594e-01, 1.01485831e-01, 8.04257854e-02, 6.37360595e-02,\n",
       "       5.05097372e-02, 4.00281029e-02, 3.17215870e-02, 2.51388152e-02,\n",
       "       1.99220811e-02, 1.57879086e-02, 1.25116476e-02, 9.91526673e-03])},\n",
       "             scoring=&#x27;neg_mean_squared_error&#x27;)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-5\" type=\"checkbox\" ><label for=\"sk-estimator-id-5\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GridSearchCV</label><div class=\"sk-toggleable__content\"><pre>GridSearchCV(cv=KFold(n_splits=5, random_state=0, shuffle=True),\n",
       "             estimator=Pipeline(steps=[(&#x27;scaler&#x27;, StandardScaler()),\n",
       "                                       (&#x27;ridge&#x27;,\n",
       "                                        ElasticNet(alpha=0.01, l1_ratio=0))]),\n",
       "             param_grid={&#x27;ridge__alpha&#x27;: array([9.91526673e+07, 7.85767931e+07, 6.22707648e+07, 4.93485161e+07,\n",
       "       3.91078551e+07, 3.09923064e+07, 2.45608728e+07, 1.94640716e+07,\n",
       "       1.54249439e+07, 1.22240043e+07, 9.68...\n",
       "       2.08706860e+00, 1.65396618e+00, 1.31073991e+00, 1.03873896e+00,\n",
       "       8.23182863e-01, 6.52358341e-01, 5.16982829e-01, 4.09700051e-01,\n",
       "       3.24680284e-01, 2.57303572e-01, 2.03908681e-01, 1.61594142e-01,\n",
       "       1.28060594e-01, 1.01485831e-01, 8.04257854e-02, 6.37360595e-02,\n",
       "       5.05097372e-02, 4.00281029e-02, 3.17215870e-02, 2.51388152e-02,\n",
       "       1.99220811e-02, 1.57879086e-02, 1.25116476e-02, 9.91526673e-03])},\n",
       "             scoring=&#x27;neg_mean_squared_error&#x27;)</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-6\" type=\"checkbox\" ><label for=\"sk-estimator-id-6\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;scaler&#x27;, StandardScaler()),\n",
       "                (&#x27;ridge&#x27;, ElasticNet(alpha=0.01, l1_ratio=0))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-7\" type=\"checkbox\" ><label for=\"sk-estimator-id-7\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">StandardScaler</label><div class=\"sk-toggleable__content\"><pre>StandardScaler()</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-8\" type=\"checkbox\" ><label for=\"sk-estimator-id-8\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ElasticNet</label><div class=\"sk-toggleable__content\"><pre>ElasticNet(alpha=0.01, l1_ratio=0)</pre></div></div></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "GridSearchCV(cv=KFold(n_splits=5, random_state=0, shuffle=True),\n",
       "             estimator=Pipeline(steps=[('scaler', StandardScaler()),\n",
       "                                       ('ridge',\n",
       "                                        ElasticNet(alpha=0.01, l1_ratio=0))]),\n",
       "             param_grid={'ridge__alpha': array([9.91526673e+07, 7.85767931e+07, 6.22707648e+07, 4.93485161e+07,\n",
       "       3.91078551e+07, 3.09923064e+07, 2.45608728e+07, 1.94640716e+07,\n",
       "       1.54249439e+07, 1.22240043e+07, 9.68...\n",
       "       2.08706860e+00, 1.65396618e+00, 1.31073991e+00, 1.03873896e+00,\n",
       "       8.23182863e-01, 6.52358341e-01, 5.16982829e-01, 4.09700051e-01,\n",
       "       3.24680284e-01, 2.57303572e-01, 2.03908681e-01, 1.61594142e-01,\n",
       "       1.28060594e-01, 1.01485831e-01, 8.04257854e-02, 6.37360595e-02,\n",
       "       5.05097372e-02, 4.00281029e-02, 3.17215870e-02, 2.51388152e-02,\n",
       "       1.99220811e-02, 1.57879086e-02, 1.25116476e-02, 9.91526673e-03])},\n",
       "             scoring='neg_mean_squared_error')"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ridge = skl.ElasticNet(alpha= 0.01, l1_ratio=0) #note the difference from ElasticNet.path() \n",
    "# alpha =0.01 is just temporarily set here. \n",
    "scaler = StandardScaler(with_mean=True,  with_std=True)\n",
    "pipe = Pipeline(steps=[('scaler', scaler), ('ridge', ridge)])\n",
    "\n",
    "K = 5\n",
    "kfold = skm.KFold(K,\n",
    "                  random_state=0,\n",
    "                  shuffle=True)\n",
    "\n",
    "lambdas = 10**np.linspace(8, -2, 100) / Y[train].std() #[yw] why lambdas is chosen to be normalized by Y.std()\n",
    "param_grid = {'ridge__alpha': lambdas}\n",
    "grid = skm.GridSearchCV(pipe,\n",
    "                        param_grid,\n",
    "                        cv=kfold,\n",
    "                        scoring='neg_mean_squared_error')\n",
    "grid.fit(X_reg[train], Y[train])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.009915266726338525,)"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid.best_params_['ridge__alpha'], "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.20046702306162334, 0.22254528809935445)"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yhat = grid.best_estimator_.predict(X_reg[~train])\n",
    "calc_r2_MSE(yhat,Y[~train])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ridge_fig, ax = subplots(figsize=(5,5))\n",
    "ax.errorbar(-np.log(lambdas),\n",
    "            -grid.cv_results_['mean_test_score'],\n",
    "            yerr=grid.cv_results_['std_test_score'] / np.sqrt(K))\n",
    "# ax.set_ylim([50000,250000])\n",
    "ax.set_xlabel('$-\\log(\\lambda)$', fontsize=20)\n",
    "ax.set_ylabel('Cross-validated MSE', fontsize=20);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-6 {color: black;}#sk-container-id-6 pre{padding: 0;}#sk-container-id-6 div.sk-toggleable {background-color: white;}#sk-container-id-6 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-6 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-6 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-6 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-6 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-6 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-6 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-6 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-6 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-6 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-6 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-6 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-6 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-6 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-6 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-6 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-6 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-6 div.sk-item {position: relative;z-index: 1;}#sk-container-id-6 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-6 div.sk-item::before, #sk-container-id-6 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-6 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-6 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-6 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-6 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-6 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-6 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-6 div.sk-label-container {text-align: center;}#sk-container-id-6 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-6 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-6\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GridSearchCV(cv=KFold(n_splits=5, random_state=0, shuffle=True),\n",
       "             estimator=Pipeline(steps=[(&#x27;scaler&#x27;, StandardScaler()),\n",
       "                                       (&#x27;ridge&#x27;,\n",
       "                                        ElasticNet(alpha=0.01, l1_ratio=0))]),\n",
       "             param_grid={&#x27;ridge__alpha&#x27;: array([9.91526673e+07, 7.85767931e+07, 6.22707648e+07, 4.93485161e+07,\n",
       "       3.91078551e+07, 3.09923064e+07, 2.45608728e+07, 1.94640716e+07,\n",
       "       1.54249439e+07, 1.22240043e+07, 9.68...\n",
       "       2.08706860e+00, 1.65396618e+00, 1.31073991e+00, 1.03873896e+00,\n",
       "       8.23182863e-01, 6.52358341e-01, 5.16982829e-01, 4.09700051e-01,\n",
       "       3.24680284e-01, 2.57303572e-01, 2.03908681e-01, 1.61594142e-01,\n",
       "       1.28060594e-01, 1.01485831e-01, 8.04257854e-02, 6.37360595e-02,\n",
       "       5.05097372e-02, 4.00281029e-02, 3.17215870e-02, 2.51388152e-02,\n",
       "       1.99220811e-02, 1.57879086e-02, 1.25116476e-02, 9.91526673e-03])})</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-9\" type=\"checkbox\" ><label for=\"sk-estimator-id-9\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GridSearchCV</label><div class=\"sk-toggleable__content\"><pre>GridSearchCV(cv=KFold(n_splits=5, random_state=0, shuffle=True),\n",
       "             estimator=Pipeline(steps=[(&#x27;scaler&#x27;, StandardScaler()),\n",
       "                                       (&#x27;ridge&#x27;,\n",
       "                                        ElasticNet(alpha=0.01, l1_ratio=0))]),\n",
       "             param_grid={&#x27;ridge__alpha&#x27;: array([9.91526673e+07, 7.85767931e+07, 6.22707648e+07, 4.93485161e+07,\n",
       "       3.91078551e+07, 3.09923064e+07, 2.45608728e+07, 1.94640716e+07,\n",
       "       1.54249439e+07, 1.22240043e+07, 9.68...\n",
       "       2.08706860e+00, 1.65396618e+00, 1.31073991e+00, 1.03873896e+00,\n",
       "       8.23182863e-01, 6.52358341e-01, 5.16982829e-01, 4.09700051e-01,\n",
       "       3.24680284e-01, 2.57303572e-01, 2.03908681e-01, 1.61594142e-01,\n",
       "       1.28060594e-01, 1.01485831e-01, 8.04257854e-02, 6.37360595e-02,\n",
       "       5.05097372e-02, 4.00281029e-02, 3.17215870e-02, 2.51388152e-02,\n",
       "       1.99220811e-02, 1.57879086e-02, 1.25116476e-02, 9.91526673e-03])})</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-10\" type=\"checkbox\" ><label for=\"sk-estimator-id-10\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;scaler&#x27;, StandardScaler()),\n",
       "                (&#x27;ridge&#x27;, ElasticNet(alpha=0.01, l1_ratio=0))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-11\" type=\"checkbox\" ><label for=\"sk-estimator-id-11\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">StandardScaler</label><div class=\"sk-toggleable__content\"><pre>StandardScaler()</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-12\" type=\"checkbox\" ><label for=\"sk-estimator-id-12\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ElasticNet</label><div class=\"sk-toggleable__content\"><pre>ElasticNet(alpha=0.01, l1_ratio=0)</pre></div></div></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "GridSearchCV(cv=KFold(n_splits=5, random_state=0, shuffle=True),\n",
       "             estimator=Pipeline(steps=[('scaler', StandardScaler()),\n",
       "                                       ('ridge',\n",
       "                                        ElasticNet(alpha=0.01, l1_ratio=0))]),\n",
       "             param_grid={'ridge__alpha': array([9.91526673e+07, 7.85767931e+07, 6.22707648e+07, 4.93485161e+07,\n",
       "       3.91078551e+07, 3.09923064e+07, 2.45608728e+07, 1.94640716e+07,\n",
       "       1.54249439e+07, 1.22240043e+07, 9.68...\n",
       "       2.08706860e+00, 1.65396618e+00, 1.31073991e+00, 1.03873896e+00,\n",
       "       8.23182863e-01, 6.52358341e-01, 5.16982829e-01, 4.09700051e-01,\n",
       "       3.24680284e-01, 2.57303572e-01, 2.03908681e-01, 1.61594142e-01,\n",
       "       1.28060594e-01, 1.01485831e-01, 8.04257854e-02, 6.37360595e-02,\n",
       "       5.05097372e-02, 4.00281029e-02, 3.17215870e-02, 2.51388152e-02,\n",
       "       1.99220811e-02, 1.57879086e-02, 1.25116476e-02, 9.91526673e-03])})"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_r2 = skm.GridSearchCV(pipe, \n",
    "                           param_grid,\n",
    "                           cv=kfold) #default scoring=R^2 \n",
    "grid_r2.fit(X_reg[train], Y[train])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "r2_fig, ax = subplots(figsize=(8,8))\n",
    "ax.errorbar(-np.log(lambdas),\n",
    "            grid_r2.cv_results_['mean_test_score'],\n",
    "            yerr=grid_r2.cv_results_['std_test_score'] / np.sqrt(K))\n",
    "ax.set_xlabel('$-\\log(\\lambda)$', fontsize=20)\n",
    "ax.set_ylabel('Cross-validated $R^2$', fontsize=20);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.20046702306162334, 0.22254528809935445)"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yhat = grid_r2.best_estimator_.predict(X_reg[~train])\n",
    "calc_r2_MSE(yhat,Y[~train])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fast Cross-Validation for Solution Paths using ElasticNetCV()\n",
    "The ridge, lasso, and elastic net can be efficiently fit along a sequence of $\\lambda$ values, creating what is known as a *solution path* or *regularization path*. Hence there is specialized code to fit\n",
    "such paths, and to choose a suitable value of $\\lambda$ using cross-validation. Even with\n",
    "identical splits the results will not agree *exactly* with our `grid`\n",
    "above because the standardization of each feature  in `grid` is carried out on each fold,\n",
    "while in `pipeCV` below it is carried out only once.\n",
    "Nevertheless, the results are similar as the normalization\n",
    "is relatively stable across folds."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "lambdas = 10**np.linspace(8, -2, 100) / Y[train].std() #[yw] why lambdas is chosen to be normalized by Y.std()\n",
    "\n",
    "ridgeCV = skl.ElasticNetCV(alphas=lambdas, # ElasticNetCV accepts a sequence of alphas\n",
    "                           l1_ratio=0,#ridge\n",
    "                           cv=kfold)\n",
    "pipeCV = Pipeline(steps=[('scaler', scaler), # scaling is done once, and then Cross-val\n",
    "                         ('ridge', ridgeCV)])\n",
    "pipeCV.fit(X_reg[train], Y[train])\n",
    "tuned_ridge = pipeCV.named_steps['ridge']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ridgeCV_fig, ax = subplots(figsize=(5,5))\n",
    "ax.errorbar(-np.log(lambdas),\n",
    "            tuned_ridge.mse_path_.mean(1),\n",
    "            yerr=tuned_ridge.mse_path_.std(1) / np.sqrt(K))\n",
    "ax.axvline(-np.log(tuned_ridge.alpha_), c='k', ls='--')\n",
    "# ax.set_ylim([50000,250000])\n",
    "ax.set_xlabel('$-\\log(\\lambda)$', fontsize=20)\n",
    "ax.set_ylabel('Cross-validated MSE', fontsize=20);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.009915266726338525, 0.641632551165717)"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tuned_ridge.alpha_, np.min(tuned_ridge.mse_path_.mean(1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([-1.10707633e-02,  4.65838410e-01,  7.61502479e-02, -1.63727462e-02,\n",
       "         7.62264990e-02, -3.69897970e-02,  2.84660521e-03,  5.77802098e-02,\n",
       "        -2.44335566e-02,  8.96324322e-03,  6.11548207e-02, -1.75038681e-02,\n",
       "        -6.27834721e-05,  6.58906951e-02, -2.53912218e-02]),\n",
       " 0.0009161973337824278)"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tuned_ridge.coef_, tuned_ridge.intercept_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.19934681975422808, 0.2228570900824681)"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yhat = tuned_ridge.predict(X_reg[~train])\n",
    "calc_r2_MSE(yhat,Y[~train])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fast Cross Validation for a solution path for  Lasso using ElasticNetCV()\n",
    "We now ask whether the lasso can yield\n",
    "either a more accurate or a more interpretable model than ridge\n",
    "regression. In order to fit a lasso model, we once again use the\n",
    "`ElasticNetCV()`  function; however, this time we use the argument\n",
    "`l1_ratio=1`. Other than that change, we proceed just as we did in\n",
    "fitting a ridge model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0036485643518026282"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lassoCV = skl.ElasticNetCV(n_alphas=100, #test 100 alpha values\n",
    "                           l1_ratio=1,\n",
    "                           cv=kfold)\n",
    "pipeCV = Pipeline(steps=[('scaler', scaler),\n",
    "                         ('lasso', lassoCV)])\n",
    "pipeCV.fit(X_reg[train], Y[train])\n",
    "tuned_lasso = pipeCV.named_steps['lasso']\n",
    "tuned_lasso.alpha_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "lassoCV_fig, ax = subplots(figsize=(5,5))\n",
    "ax.errorbar(-np.log(tuned_lasso.alphas_),\n",
    "            tuned_lasso.mse_path_.mean(1),\n",
    "            yerr=tuned_lasso.mse_path_.std(1) / np.sqrt(K))\n",
    "ax.axvline(-np.log(tuned_lasso.alpha_), c='k', ls='--')\n",
    "# ax.set_ylim([50000,250000])\n",
    "ax.set_xlabel('$-\\log(\\lambda)$', fontsize=20)\n",
    "ax.set_ylabel('Cross-validated MSE', fontsize=20);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.6414242348883088,\n",
       " array([-0.00833788,  0.47544532,  0.06772294, -0.01251405,  0.06875301,\n",
       "        -0.03318071,  0.        ,  0.05420293, -0.01944543,  0.00609821,\n",
       "         0.05792846, -0.01250511,  0.        ,  0.06306209, -0.02051067]))"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.min(tuned_lasso.mse_path_.mean(1)), tuned_lasso.coef_, "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.20118664970669153, 0.22234498426740507)"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yhat = tuned_lasso.predict(X_reg[~train])\n",
    "calc_r2_MSE(yhat,Y[~train])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Calculate the Lasso coefficient along a path using Lasso.path()\n",
    "This can be done using ElasticNet.path() as we did befor for ridge. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "lambdas, soln_array = skl.Lasso.path(X_reg[train], # this can be done using ElasticNet.path\n",
    "                                    Y[train],\n",
    "                                    l1_ratio=1,\n",
    "                                    n_alphas=100)[:2]\n",
    "soln_path = pd.DataFrame(soln_array.T,\n",
    "                         columns=X_reg.columns,\n",
    "                         index=-np.log(lambdas))\n",
    "# soln_path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "path_fig, ax = subplots(figsize=(8,8))\n",
    "soln_path.plot(ax=ax, legend=False)\n",
    "ax.legend(loc='upper left')\n",
    "ax.set_xlabel('$-\\log(\\lambda)$', fontsize=20)\n",
    "ax.set_ylabel('Standardized coefficiients', fontsize=20);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## PCA Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.25981628, 0.12324188])"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca = PCA(n_components=2)\n",
    "linreg = skl.LinearRegression()\n",
    "pipe = Pipeline([('scaler', scaler), \n",
    "                 ('pca', pca),\n",
    "                 ('linreg', linreg)])\n",
    "pipe.fit(X_reg[train], Y[train])\n",
    "pipe.named_steps['linreg'].coef_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.22912884, 0.0931941 ])"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pipe.named_steps['pca'].explained_variance_ratio_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.01110809230837384, 0.2752522295441419)"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "yhat = pipe.predict(X_reg[~train])\n",
    "calc_r2_MSE(yhat, Y[~train])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "param_grid = {'pca__n_components': range(1, len(X_reg.columns))}\n",
    "grid = skm.GridSearchCV(pipe,\n",
    "                        param_grid,\n",
    "                        cv=kfold,\n",
    "                        scoring='neg_mean_squared_error')\n",
    "grid.fit(X_reg[train], Y[train]);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.19588674326582645, 0.22382018196381334)"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yhat = grid.best_estimator_.predict(X_reg[~train])\n",
    "calc_r2_MSE(yhat, Y[~train])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pcr_fig, ax = subplots(figsize=(8,8))\n",
    "n_comp = param_grid['pca__n_components']\n",
    "ax.errorbar(n_comp,\n",
    "            -grid.cv_results_['mean_test_score'],\n",
    "            grid.cv_results_['std_test_score'] / np.sqrt(K))\n",
    "ax.set_ylabel('Cross-validated MSE', fontsize=20)\n",
    "ax.set_xlabel('# principal components', fontsize=20)\n",
    "ax.set_xticks(n_comp[::2]);\n",
    "# ax.set_ylim([50000,250000]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fit a null model with only intercept\n",
    "The CV score is provided for each possible number of components from\n",
    "1 to 19 inclusive. The `PCA()` method complains\n",
    "if we try to fit an intercept only with `n_components=0`\n",
    "so we also compute the MSE for just the null model with\n",
    "these splits."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6416328996418488"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Xn = np.zeros((X_reg.shape[0], 1)) \n",
    "cv_null = skm.cross_validate(linreg,\n",
    "                             X_reg[train],\n",
    "                             Y[train],\n",
    "                             cv=kfold,\n",
    "                             scoring='neg_mean_squared_error')\n",
    "-cv_null['test_score'].mean()\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " ### Partial Least Squres"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-7 {color: black;}#sk-container-id-7 pre{padding: 0;}#sk-container-id-7 div.sk-toggleable {background-color: white;}#sk-container-id-7 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-7 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-7 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-7 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-7 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-7 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-7 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-7 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-7 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-7 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-7 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-7 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-7 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-7 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-7 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-7 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-7 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-7 div.sk-item {position: relative;z-index: 1;}#sk-container-id-7 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-7 div.sk-item::before, #sk-container-id-7 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-7 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-7 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-7 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-7 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-7 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-7 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-7 div.sk-label-container {text-align: center;}#sk-container-id-7 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-7 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-7\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>PLSRegression()</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-13\" type=\"checkbox\" checked><label for=\"sk-estimator-id-13\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">PLSRegression</label><div class=\"sk-toggleable__content\"><pre>PLSRegression()</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "PLSRegression()"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pls = PLSRegression(n_components=2, \n",
    "                    scale=True) # standarsize the data \n",
    "pls.fit(X_reg[train], Y[train]) #X has no- intercept"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-8 {color: black;}#sk-container-id-8 pre{padding: 0;}#sk-container-id-8 div.sk-toggleable {background-color: white;}#sk-container-id-8 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-8 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-8 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-8 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-8 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-8 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-8 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-8 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-8 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-8 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-8 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-8 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-8 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-8 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-8 div.sk-item {position: relative;z-index: 1;}#sk-container-id-8 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-8 div.sk-item::before, #sk-container-id-8 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-8 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-8 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-8 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-8 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-8 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-8 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-8 div.sk-label-container {text-align: center;}#sk-container-id-8 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-8 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-8\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GridSearchCV(cv=KFold(n_splits=5, random_state=0, shuffle=True),\n",
       "             estimator=PLSRegression(),\n",
       "             param_grid={&#x27;n_components&#x27;: range(1, 20)},\n",
       "             scoring=&#x27;neg_mean_squared_error&#x27;)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-14\" type=\"checkbox\" ><label for=\"sk-estimator-id-14\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GridSearchCV</label><div class=\"sk-toggleable__content\"><pre>GridSearchCV(cv=KFold(n_splits=5, random_state=0, shuffle=True),\n",
       "             estimator=PLSRegression(),\n",
       "             param_grid={&#x27;n_components&#x27;: range(1, 20)},\n",
       "             scoring=&#x27;neg_mean_squared_error&#x27;)</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-15\" type=\"checkbox\" ><label for=\"sk-estimator-id-15\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: PLSRegression</label><div class=\"sk-toggleable__content\"><pre>PLSRegression()</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-16\" type=\"checkbox\" ><label for=\"sk-estimator-id-16\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">PLSRegression</label><div class=\"sk-toggleable__content\"><pre>PLSRegression()</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "GridSearchCV(cv=KFold(n_splits=5, random_state=0, shuffle=True),\n",
       "             estimator=PLSRegression(),\n",
       "             param_grid={'n_components': range(1, 20)},\n",
       "             scoring='neg_mean_squared_error')"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "param_grid = {'n_components':range(1, 20)}\n",
    "grid = skm.GridSearchCV(pls,\n",
    "                        param_grid,\n",
    "                        cv=kfold,\n",
    "                        scoring='neg_mean_squared_error')\n",
    "grid.fit(X_reg[train], Y[train])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.20047839646997379, 0.22254212237822596)"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yhat = grid.best_estimator_.predict(X_reg[~train])\n",
    "calc_r2_MSE(yhat.squeeze(), Y[~train])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pls_fig, ax = subplots(figsize=(8,8))\n",
    "n_comp = param_grid['n_components']\n",
    "ax.errorbar(n_comp,\n",
    "            -grid.cv_results_['mean_test_score'],\n",
    "            grid.cv_results_['std_test_score'] / np.sqrt(K))\n",
    "ax.set_ylabel('Cross-validated MSE', fontsize=20)\n",
    "ax.set_xlabel('# principal components', fontsize=20)\n",
    "ax.set_xticks(n_comp[::2]);\n",
    "# ax.set_ylim([50000,250000]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## RF Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-13 {color: black;}#sk-container-id-13 pre{padding: 0;}#sk-container-id-13 div.sk-toggleable {background-color: white;}#sk-container-id-13 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-13 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-13 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-13 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-13 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-13 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-13 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-13 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-13 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-13 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-13 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-13 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-13 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-13 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-13 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-13 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-13 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-13 div.sk-item {position: relative;z-index: 1;}#sk-container-id-13 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-13 div.sk-item::before, #sk-container-id-13 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-13 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-13 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-13 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-13 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-13 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-13 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-13 div.sk-label-container {text-align: center;}#sk-container-id-13 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-13 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-13\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>RandomForestRegressor(max_features=&#x27;sqrt&#x27;, n_estimators=500, random_state=0)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-21\" type=\"checkbox\" checked><label for=\"sk-estimator-id-21\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomForestRegressor</label><div class=\"sk-toggleable__content\"><pre>RandomForestRegressor(max_features=&#x27;sqrt&#x27;, n_estimators=500, random_state=0)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "RandomForestRegressor(max_features='sqrt', n_estimators=500, random_state=0)"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf = RF(max_features='sqrt', random_state=0, n_estimators=500,)\n",
    "rf.fit(X_reg[train], Y[train])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.14855483869998498, 0.23699473841329446)"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = subplots(figsize=(5,5))[1]\n",
    "y_hat_rf = rf.predict(X_reg[~train])\n",
    "\n",
    "ax.scatter(y_hat_rf, Y[~train], alpha=0.5)\n",
    "plt.axis('equal')\n",
    "calc_r2_MSE(y_hat_rf, Y[~train])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "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>importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>log_volume_1</th>\n",
       "      <td>0.249185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>log_volume_2</th>\n",
       "      <td>0.101215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>log_volume_3</th>\n",
       "      <td>0.072206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>log_volume_4</th>\n",
       "      <td>0.065246</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>log_volume_5</th>\n",
       "      <td>0.063621</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>volatility_1</th>\n",
       "      <td>0.062642</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_1</th>\n",
       "      <td>0.053319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_2</th>\n",
       "      <td>0.043337</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>volatility_2</th>\n",
       "      <td>0.043110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_3</th>\n",
       "      <td>0.042316</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_4</th>\n",
       "      <td>0.041987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>volatility_3</th>\n",
       "      <td>0.041654</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>return_5</th>\n",
       "      <td>0.040617</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>volatility_4</th>\n",
       "      <td>0.040059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>volatility_5</th>\n",
       "      <td>0.039487</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              importance\n",
       "log_volume_1    0.249185\n",
       "log_volume_2    0.101215\n",
       "log_volume_3    0.072206\n",
       "log_volume_4    0.065246\n",
       "log_volume_5    0.063621\n",
       "volatility_1    0.062642\n",
       "return_1        0.053319\n",
       "return_2        0.043337\n",
       "volatility_2    0.043110\n",
       "return_3        0.042316\n",
       "return_4        0.041987\n",
       "volatility_3    0.041654\n",
       "return_5        0.040617\n",
       "volatility_4    0.040059\n",
       "volatility_5    0.039487"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature_names = list(X_reg.columns)\n",
    "feature_imp = pd.DataFrame(\n",
    "    {'importance':rf.feature_importances_},\n",
    "    index=feature_names)\n",
    "feature_imp.sort_values(by='importance', ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Gradient Boosting Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-14 {color: black;}#sk-container-id-14 pre{padding: 0;}#sk-container-id-14 div.sk-toggleable {background-color: white;}#sk-container-id-14 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-14 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-14 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-14 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-14 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-14 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-14 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-14 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-14 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-14 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-14 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-14 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-14 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-14 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-14 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-14 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-14 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-14 div.sk-item {position: relative;z-index: 1;}#sk-container-id-14 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-14 div.sk-item::before, #sk-container-id-14 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-14 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-14 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-14 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-14 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-14 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-14 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-14 div.sk-label-container {text-align: center;}#sk-container-id-14 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-14 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-14\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GradientBoostingRegressor(learning_rate=0.001, n_estimators=5000,\n",
       "                          random_state=0)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-22\" type=\"checkbox\" checked><label for=\"sk-estimator-id-22\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GradientBoostingRegressor</label><div class=\"sk-toggleable__content\"><pre>GradientBoostingRegressor(learning_rate=0.001, n_estimators=5000,\n",
       "                          random_state=0)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "GradientBoostingRegressor(learning_rate=0.001, n_estimators=5000,\n",
       "                          random_state=0)"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "boost = GBR(n_estimators=5000,\n",
    "                   learning_rate=0.001,\n",
    "                   max_depth=3,\n",
    "                   random_state=0)\n",
    "boost.fit(X_reg[train], Y[train])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_error = np.zeros_like(boost.train_score_)\n",
    "for idx, y_ in enumerate(boost.staged_predict(X_reg[~train])):\n",
    "   test_error[idx] = np.mean((Y[~train] - y_)**2)\n",
    "\n",
    "plot_idx = np.arange(boost.train_score_.shape[0])\n",
    "ax = subplots(figsize=(5,5))[1]\n",
    "ax.plot(plot_idx,\n",
    "        boost.train_score_,\n",
    "        'b',\n",
    "        label='Training')\n",
    "ax.plot(plot_idx,\n",
    "        test_error,\n",
    "        'r',\n",
    "        label='Test')\n",
    "ax.legend();\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.17726474416964144, 0.22900350557069668)"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_hat_boost = boost.predict(X_reg[~train]);\n",
    "calc_r2_MSE(y_hat_boost, Y[~train])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bayesian Additive Regression Trees"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-10 {color: black;}#sk-container-id-10 pre{padding: 0;}#sk-container-id-10 div.sk-toggleable {background-color: white;}#sk-container-id-10 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-10 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-10 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-10 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-10 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-10 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-10 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-10 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-10 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-10 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-10 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-10 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-10 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-10 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-10 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-10 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-10 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-10 div.sk-item {position: relative;z-index: 1;}#sk-container-id-10 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-10 div.sk-item::before, #sk-container-id-10 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-10 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-10 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-10 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-10 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-10 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-10 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-10 div.sk-label-container {text-align: center;}#sk-container-id-10 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-10 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-10\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>BART(burnin=5, ndraw=15, random_state=0)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-18\" type=\"checkbox\" checked><label for=\"sk-estimator-id-18\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">BART</label><div class=\"sk-toggleable__content\"><pre>BART(burnin=5, ndraw=15, random_state=0)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "BART(burnin=5, ndraw=15, random_state=0)"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bart_ = BART(random_state=0, burnin=5, ndraw=15) #num_trees=200, max_states=100\n",
    "# ndraw: number of iterations or samples to draw from the posterior distribution after the burn-in \n",
    "bart_.fit(X_reg[train], Y[train])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.13185137162609006, 0.24164404994820135)"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "# yhat_test = bart_.predict(X_reg[~train].astype(np.float32))\n",
    "yhat_test = bart_.predict(np.float32(X_reg[~train]))\n",
    "calc_r2_MSE(yhat_test, Y[~train])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "log_volume_1    32.800000\n",
       "log_volume_2    25.200000\n",
       "log_volume_3    24.200000\n",
       "log_volume_4    22.800000\n",
       "return_1        17.666667\n",
       "volatility_1    17.666667\n",
       "log_volume_5    16.533333\n",
       "return_5        15.333333\n",
       "return_4        14.933333\n",
       "volatility_3    14.066667\n",
       "volatility_4    13.733333\n",
       "volatility_2    12.733333\n",
       "volatility_5    12.733333\n",
       "return_3        12.533333\n",
       "return_2        12.200000\n",
       "dtype: float64"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "var_inclusion = pd.Series(bart_.variable_inclusion_.mean(0),\n",
    "                               index=X_reg.columns)\n",
    "var_inclusion.sort_values(ascending=False)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### RNN Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For an input shape `(5,3)`, each row represents a lagged version of the three variables.\n",
    "The `nn.RNN()` layer also expects the first row of each\n",
    "observation to be earliest in time, so we must reverse the current order."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Reshaping the data\n",
    "ordered_cols = []\n",
    "for lag in range(5,0,-1):\n",
    "    for col in cols:\n",
    "        ordered_cols.append('{0}_{1}'.format(col, lag))\n",
    "X = X.reindex(columns=ordered_cols)\n",
    "# X.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_rnn = X.to_numpy().reshape((-1,5,3))\n",
    "# X_rnn.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "class NYSEModel(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(NYSEModel, self).__init__()\n",
    "        self.rnn = nn.RNN(3, #number of features for each time step\n",
    "                          12, #hidden size\n",
    "                          batch_first=True) #batch size as the first dim\n",
    "        self.dense = nn.Linear(12, 1)\n",
    "        self.dropout = nn.Dropout(0.1)\n",
    "    def forward(self, x):\n",
    "        val, h_n = self.rnn(x)\n",
    "        val = self.dense(self.dropout(val[:,-1])) #extract the last time step from RNN\n",
    "        return torch.flatten(val)\n",
    "nyse_model = NYSEModel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "datasets = []\n",
    "for mask in [train, ~train]:\n",
    "    X_rnn_t = torch.tensor(X_rnn[mask].astype(np.float32))\n",
    "    Y_t = torch.tensor(Y[mask].astype(np.float32))\n",
    "    datasets.append(TensorDataset(X_rnn_t, Y_t))\n",
    "nyse_train, nyse_test = datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10493, 296)"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(nyse_train), len(nyse_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(torch.Size([296, 5, 3]), torch.Size([296]))"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nyse_test.tensors[0].shape, nyse_test.tensors[1].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "===================================================================================================================\n",
       "Layer (type:depth-idx)                   Input Shape               Output Shape              Param #\n",
       "===================================================================================================================\n",
       "NYSEModel                                [296, 5, 3]               [296]                     --\n",
       "├─RNN: 1-1                               [296, 5, 3]               [296, 5, 12]              204\n",
       "├─Dropout: 1-2                           [296, 12]                 [296, 12]                 --\n",
       "├─Linear: 1-3                            [296, 12]                 [296, 1]                  13\n",
       "===================================================================================================================\n",
       "Total params: 217\n",
       "Trainable params: 217\n",
       "Non-trainable params: 0\n",
       "Total mult-adds (Units.MEGABYTES): 0.31\n",
       "===================================================================================================================\n",
       "Input size (MB): 0.02\n",
       "Forward/backward pass size (MB): 0.14\n",
       "Params size (MB): 0.00\n",
       "Estimated Total Size (MB): 0.16\n",
       "==================================================================================================================="
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "summary(nyse_model,\n",
    "        input_data=X_rnn_t,\n",
    "        col_names=['input_size',\n",
    "                   'output_size',\n",
    "                   'num_params'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_num_workers = rec_num_workers()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "nyse_dm = SimpleDataModule(nyse_train,\n",
    "                           nyse_test,\n",
    "                           num_workers=min(4, max_num_workers),\n",
    "                           validation=nyse_test,\n",
    "                           batch_size=64)\n",
    "                           "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 minibatch input size: torch.Size([64]) model minibatch output size: torch.Size([64])\n"
     ]
    }
   ],
   "source": [
    "# Example the minibatch size of y and the output size in each minibatch\n",
    "for idx, (x, y) in enumerate(nyse_dm.train_dataloader()):\n",
    "    out = nyse_model(x)\n",
    "    print(idx, \"minibatch input size:\", y.size(), \"model minibatch output size:\", out.size())\n",
    "    \n",
    "    if idx ==0:\n",
    "        break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "nyse_optimizer = RMSprop(nyse_model.parameters(),\n",
    "                         lr=0.001)\n",
    "nyse_module = SimpleModule.regression(nyse_model,\n",
    "                                      optimizer=nyse_optimizer,\n",
    "                                      metrics={'r2':R2Score()})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "GPU available: True (cuda), used: True\n",
      "TPU available: False, using: 0 TPU cores\n",
      "IPU available: False, using: 0 IPUs\n",
      "HPU available: False, using: 0 HPUs\n",
      "You are using a CUDA device ('NVIDIA GeForce RTX 4090') that has Tensor Cores. To properly utilize them, you should set `torch.set_float32_matmul_precision('medium' | 'high')` which will trade-off precision for performance. For more details, read https://pytorch.org/docs/stable/generated/torch.set_float32_matmul_precision.html#torch.set_float32_matmul_precision\n",
      "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n",
      "\n",
      "  | Name  | Type      | Params\n",
      "------------------------------------\n",
      "0 | model | NYSEModel | 217   \n",
      "1 | loss  | MSELoss   | 0     \n",
      "------------------------------------\n",
      "217       Trainable params\n",
      "0         Non-trainable params\n",
      "217       Total params\n",
      "0.001     Total estimated model params size (MB)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 19: 100%|██████████| 164/164 [00:00<00:00, 168.95it/s, v_num=17]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "`Trainer.fit` stopped: `max_epochs=20` reached.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 19: 100%|██████████| 164/164 [00:00<00:00, 166.27it/s, v_num=17]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing DataLoader 0: 100%|██████████| 5/5 [00:00<00:00, 405.02it/s] \n",
      "────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────\n",
      "Runningstage.testing metric      DataLoader 0\n",
      "────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────\n",
      "        test_loss           0.22700099647045135\n",
      "         test_r2            0.18445897102355957\n",
      "────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[{'test_loss': 0.22700099647045135, 'test_r2': 0.18445897102355957}]"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nyse_trainer = Trainer(deterministic=False,\n",
    "                       max_epochs=20,\n",
    "                       callbacks=[ErrorTracker()])\n",
    "nyse_trainer.fit(nyse_module,\n",
    "                 datamodule=nyse_dm)\n",
    "nyse_trainer.test(nyse_module,\n",
    "                  datamodule=nyse_dm)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Build a linear and non-linear AR model using Torch "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "datasets = []\n",
    "for mask in [train, ~train]:\n",
    "    X_day_t = torch.tensor(\n",
    "                   np.asarray(X_day[mask]).astype(np.float32))\n",
    "    Y_t = torch.tensor(np.asarray(Y[mask]).astype(np.float32))\n",
    "    datasets.append(TensorDataset(X_day_t, Y_t))\n",
    "day_train, day_test = datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([10493, 20])"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# day_train.tensors[0].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "day_dm = SimpleDataModule(day_train,\n",
    "                          day_test,\n",
    "                          num_workers=min(4, max_num_workers),\n",
    "                          validation=day_test,\n",
    "                          batch_size=64)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10789, 20)"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_day.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "class NonLinearARModel(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(NonLinearARModel, self).__init__()\n",
    "        self._forward = nn.Sequential(nn.Flatten(), #flatten a multi-dim tensor into a 1-d tensor while keeping the batch size\n",
    "                                      nn.Linear(20, 32), \n",
    "                                      nn.ReLU(),\n",
    "                                      nn.Dropout(0.5),\n",
    "                                      nn.Linear(32, 1))\n",
    "    def forward(self, x):\n",
    "        return torch.flatten(self._forward(x))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "nl_model = NonLinearARModel()\n",
    "nl_optimizer = RMSprop(nl_model.parameters(),\n",
    "                           lr=0.001)\n",
    "nl_module = SimpleModule.regression(nl_model,\n",
    "                                        optimizer=nl_optimizer,\n",
    "                                        metrics={'r2':R2Score()})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "GPU available: True (cuda), used: True\n",
      "TPU available: False, using: 0 TPU cores\n",
      "IPU available: False, using: 0 IPUs\n",
      "HPU available: False, using: 0 HPUs\n",
      "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n",
      "\n",
      "  | Name  | Type             | Params\n",
      "-------------------------------------------\n",
      "0 | model | NonLinearARModel | 705   \n",
      "1 | loss  | MSELoss          | 0     \n",
      "-------------------------------------------\n",
      "705       Trainable params\n",
      "0         Non-trainable params\n",
      "705       Total params\n",
      "0.003     Total estimated model params size (MB)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sanity Checking: 0it [00:00, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 19: 100%|██████████| 164/164 [00:00<00:00, 194.34it/s, v_num=18]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "`Trainer.fit` stopped: `max_epochs=20` reached.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 19: 100%|██████████| 164/164 [00:00<00:00, 190.82it/s, v_num=18]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing DataLoader 0: 100%|██████████| 5/5 [00:00<00:00, 416.50it/s] \n",
      "────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────\n",
      "Runningstage.testing metric      DataLoader 0\n",
      "────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────\n",
      "        test_loss           0.22326382994651794\n",
      "         test_r2            0.19788551330566406\n",
      "────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[{'test_loss': 0.22326382994651794, 'test_r2': 0.19788551330566406}]"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nl_trainer = Trainer(deterministic=False,\n",
    "                         max_epochs=20,\n",
    "                         callbacks=[ErrorTracker()])\n",
    "nl_trainer.fit(nl_module, datamodule=day_dm)\n",
    "nl_trainer.test(nl_module, datamodule=day_dm) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "class LinearARModel(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(LinearARModel, self).__init__()\n",
    "        self._forward = nn.Sequential(nn.Flatten(), #flatten a multi-dim tensor into a 1-d tensor while keeping the batch size\n",
    "                                    #   nn.Linear(20, 32), \n",
    "                                    #   nn.ReLU(),\n",
    "                                    #   nn.Dropout(0.5),\n",
    "                                      nn.Linear(20, 1))\n",
    "    def forward(self, x):\n",
    "        return torch.flatten(self._forward(x))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "linearAR_model = LinearARModel()\n",
    "linearAR_optimizer = RMSprop(linearAR_model.parameters(),\n",
    "                           lr=0.001)\n",
    "linearAR_module = SimpleModule.regression(linearAR_model,\n",
    "                                        optimizer=linearAR_optimizer,\n",
    "                                        metrics={'r2':R2Score()})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "GPU available: True (cuda), used: True\n",
      "TPU available: False, using: 0 TPU cores\n",
      "IPU available: False, using: 0 IPUs\n",
      "HPU available: False, using: 0 HPUs\n",
      "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n",
      "\n",
      "  | Name  | Type          | Params\n",
      "----------------------------------------\n",
      "0 | model | LinearARModel | 21    \n",
      "1 | loss  | MSELoss       | 0     \n",
      "----------------------------------------\n",
      "21        Trainable params\n",
      "0         Non-trainable params\n",
      "21        Total params\n",
      "0.000     Total estimated model params size (MB)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sanity Checking: 0it [00:00, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 19: 100%|██████████| 164/164 [00:00<00:00, 223.67it/s, v_num=19]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "`Trainer.fit` stopped: `max_epochs=20` reached.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 19: 100%|██████████| 164/164 [00:00<00:00, 219.73it/s, v_num=19]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing DataLoader 0: 100%|██████████| 5/5 [00:00<00:00, 322.04it/s]\n",
      "────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────\n",
      "Runningstage.testing metric      DataLoader 0\n",
      "────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────\n",
      "        test_loss           0.22887705266475677\n",
      "         test_r2            0.1777191162109375\n",
      "────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[{'test_loss': 0.22887705266475677, 'test_r2': 0.1777191162109375}]"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "linearAR_trainer = Trainer(deterministic=False,\n",
    "                         max_epochs=20,\n",
    "                         callbacks=[ErrorTracker()])\n",
    "linearAR_trainer.fit(linearAR_module, datamodule=day_dm)\n",
    "linearAR_trainer.test(linearAR_module, datamodule=day_dm) "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "tf",
   "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": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}