StockPricePrediction.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import sklearn as sk\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn import metrics\n",
    "from sklearn.model_selection import train_test_split\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# !pip install pandas_datareader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "#from pandas_datareader import data, wb\n",
    "import pandas as pd\n",
    "# the line below is the fix for is_list_like lub\n",
    "pd.core.common.is_list_like = pd.api.types.is_list_like\n",
    "\n",
    "import pandas_datareader as web\n",
    "import datetime\n",
    "start = datetime.datetime(2010, 1, 1)\n",
    "end = datetime.datetime(2020, 1, 30)\n",
    "stockList = [\"PYPL\", \"CTSH\", \"TSLA\"]\n",
    "# data.head(-10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# data.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Finding how long in future, the model can predict with prediction score > 70% for stock:  PYPL\n",
      "For 10th day in future, prediction score is 98.73% for stock PYPL\n",
      "For 20th day in future, prediction score is 98.08% for stock PYPL\n",
      "For 30th day in future, prediction score is 97.64% for stock PYPL\n",
      "For 40th day in future, prediction score is 97.19% for stock PYPL\n",
      "For 50th day in future, prediction score is 96.46% for stock PYPL\n",
      "For 60th day in future, prediction score is 96.97% for stock PYPL\n",
      "For 70th day in future, prediction score is 96.2% for stock PYPL\n",
      "For 80th day in future, prediction score is 95.8% for stock PYPL\n",
      "For 90th day in future, prediction score is 94.58% for stock PYPL\n",
      "For 100th day in future, prediction score is 95.4% for stock PYPL\n",
      "For 110th day in future, prediction score is 94.67% for stock PYPL\n",
      "For 120th day in future, prediction score is 93.93% for stock PYPL\n",
      "For 130th day in future, prediction score is 94.05% for stock PYPL\n",
      "For 140th day in future, prediction score is 93.58% for stock PYPL\n",
      "For 150th day in future, prediction score is 94.81% for stock PYPL\n",
      "For 160th day in future, prediction score is 94.59% for stock PYPL\n",
      "For 170th day in future, prediction score is 93.07% for stock PYPL\n",
      "For 180th day in future, prediction score is 93.52% for stock PYPL\n",
      "For 190th day in future, prediction score is 93.41% for stock PYPL\n",
      "For 200th day in future, prediction score is 92.45% for stock PYPL\n",
      "For 210th day in future, prediction score is 92.77% for stock PYPL\n",
      "For 220th day in future, prediction score is 92.48% for stock PYPL\n",
      "For 230th day in future, prediction score is 89.41% for stock PYPL\n",
      "For 240th day in future, prediction score is 90.33% for stock PYPL\n",
      "For 250th day in future, prediction score is 88.2% for stock PYPL\n",
      "For 260th day in future, prediction score is 87.66% for stock PYPL\n",
      "For 270th day in future, prediction score is 88.8% for stock PYPL\n",
      "For 280th day in future, prediction score is 85.44% for stock PYPL\n",
      "For 290th day in future, prediction score is 87.99% for stock PYPL\n",
      "For 300th day in future, prediction score is 87.65% for stock PYPL\n",
      "For 310th day in future, prediction score is 88.15% for stock PYPL\n",
      "For 320th day in future, prediction score is 88.95% for stock PYPL\n",
      "For 330th day in future, prediction score is 88.45% for stock PYPL\n",
      "For 340th day in future, prediction score is 87.83% for stock PYPL\n",
      "For 350th day in future, prediction score is 85.79% for stock PYPL\n",
      "For 360th day in future, prediction score is 83.81% for stock PYPL\n",
      "For 370th day in future, prediction score is 83.95% for stock PYPL\n",
      "For 380th day in future, prediction score is 84.23% for stock PYPL\n",
      "For 390th day in future, prediction score is 81.83% for stock PYPL\n",
      "For 400th day in future, prediction score is 80.72% for stock PYPL\n",
      "For 410th day in future, prediction score is 77.93% for stock PYPL\n",
      "For 420th day in future, prediction score is 80.77% for stock PYPL\n",
      "For 430th day in future, prediction score is 80.82% for stock PYPL\n",
      "For 440th day in future, prediction score is 83.35% for stock PYPL\n",
      "For 450th day in future, prediction score is 80.99% for stock PYPL\n",
      "For 460th day in future, prediction score is 81.13% for stock PYPL\n",
      "For 470th day in future, prediction score is 81.73% for stock PYPL\n",
      "For 480th day in future, prediction score is 82.14% for stock PYPL\n",
      "For 490th day in future, prediction score is 77.86% for stock PYPL\n",
      "For 500th day in future, prediction score is 81.91% for stock PYPL\n",
      "For 510th day in future, prediction score is 78.96% for stock PYPL\n",
      "For 520th day in future, prediction score is 74.36% for stock PYPL\n",
      "For 530th day in future, prediction score is 72.74% for stock PYPL\n",
      "Finding how long in future, the model can predict with prediction score > 70% for stock:  CTSH\n",
      "For 10th day in future, prediction score is 97.59% for stock CTSH\n",
      "For 20th day in future, prediction score is 96.34% for stock CTSH\n",
      "For 30th day in future, prediction score is 94.27% for stock CTSH\n",
      "For 40th day in future, prediction score is 92.65% for stock CTSH\n",
      "For 50th day in future, prediction score is 91.67% for stock CTSH\n",
      "For 60th day in future, prediction score is 91.16% for stock CTSH\n",
      "For 70th day in future, prediction score is 90.24% for stock CTSH\n",
      "For 80th day in future, prediction score is 89.97% for stock CTSH\n",
      "For 90th day in future, prediction score is 88.06% for stock CTSH\n",
      "For 100th day in future, prediction score is 87.45% for stock CTSH\n",
      "For 110th day in future, prediction score is 85.93% for stock CTSH\n",
      "For 120th day in future, prediction score is 85.22% for stock CTSH\n",
      "For 130th day in future, prediction score is 83.48% for stock CTSH\n",
      "For 140th day in future, prediction score is 83.56% for stock CTSH\n",
      "For 150th day in future, prediction score is 83.41% for stock CTSH\n",
      "For 160th day in future, prediction score is 82.42% for stock CTSH\n",
      "For 170th day in future, prediction score is 82.48% for stock CTSH\n",
      "For 180th day in future, prediction score is 82.05% for stock CTSH\n",
      "For 190th day in future, prediction score is 81.87% for stock CTSH\n",
      "For 200th day in future, prediction score is 82.0% for stock CTSH\n",
      "For 210th day in future, prediction score is 83.84% for stock CTSH\n",
      "For 220th day in future, prediction score is 82.36% for stock CTSH\n",
      "For 230th day in future, prediction score is 81.25% for stock CTSH\n",
      "For 240th day in future, prediction score is 81.86% for stock CTSH\n",
      "For 250th day in future, prediction score is 81.18% for stock CTSH\n",
      "For 260th day in future, prediction score is 80.13% for stock CTSH\n",
      "For 270th day in future, prediction score is 82.14% for stock CTSH\n",
      "For 280th day in future, prediction score is 81.13% for stock CTSH\n",
      "For 290th day in future, prediction score is 81.91% for stock CTSH\n",
      "For 300th day in future, prediction score is 81.49% for stock CTSH\n",
      "For 310th day in future, prediction score is 81.13% for stock CTSH\n",
      "For 320th day in future, prediction score is 80.16% for stock CTSH\n",
      "For 330th day in future, prediction score is 81.3% for stock CTSH\n",
      "For 340th day in future, prediction score is 81.24% for stock CTSH\n",
      "For 350th day in future, prediction score is 79.22% for stock CTSH\n",
      "For 360th day in future, prediction score is 78.17% for stock CTSH\n",
      "For 370th day in future, prediction score is 78.59% for stock CTSH\n",
      "For 380th day in future, prediction score is 78.39% for stock CTSH\n",
      "For 390th day in future, prediction score is 78.41% for stock CTSH\n",
      "For 400th day in future, prediction score is 75.44% for stock CTSH\n",
      "For 410th day in future, prediction score is 75.51% for stock CTSH\n",
      "For 420th day in future, prediction score is 72.34% for stock CTSH\n",
      "For 430th day in future, prediction score is 70.65% for stock CTSH\n",
      "Finding how long in future, the model can predict with prediction score > 70% for stock:  TSLA\n",
      "For 10th day in future, prediction score is 97.6% for stock TSLA\n",
      "For 20th day in future, prediction score is 94.47% for stock TSLA\n",
      "For 30th day in future, prediction score is 91.78% for stock TSLA\n",
      "For 40th day in future, prediction score is 90.91% for stock TSLA\n",
      "For 50th day in future, prediction score is 88.48% for stock TSLA\n",
      "For 60th day in future, prediction score is 86.84% for stock TSLA\n",
      "For 70th day in future, prediction score is 85.15% for stock TSLA\n",
      "For 80th day in future, prediction score is 84.14% for stock TSLA\n",
      "For 90th day in future, prediction score is 79.92% for stock TSLA\n",
      "For 100th day in future, prediction score is 83.39% for stock TSLA\n",
      "For 110th day in future, prediction score is 83.57% for stock TSLA\n",
      "For 120th day in future, prediction score is 83.7% for stock TSLA\n",
      "For 130th day in future, prediction score is 81.4% for stock TSLA\n",
      "For 140th day in future, prediction score is 84.04% for stock TSLA\n",
      "For 150th day in future, prediction score is 81.66% for stock TSLA\n",
      "For 160th day in future, prediction score is 78.42% for stock TSLA\n",
      "For 170th day in future, prediction score is 81.27% for stock TSLA\n",
      "For 180th day in future, prediction score is 81.15% for stock TSLA\n",
      "For 190th day in future, prediction score is 81.04% for stock TSLA\n",
      "For 200th day in future, prediction score is 78.91% for stock TSLA\n",
      "For 210th day in future, prediction score is 77.7% for stock TSLA\n",
      "For 220th day in future, prediction score is 77.09% for stock TSLA\n",
      "For 230th day in future, prediction score is 76.05% for stock TSLA\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For 240th day in future, prediction score is 74.6% for stock TSLA\n",
      "For 250th day in future, prediction score is 75.89% for stock TSLA\n",
      "For 260th day in future, prediction score is 73.88% for stock TSLA\n",
      "For 270th day in future, prediction score is 74.06% for stock TSLA\n",
      "For 280th day in future, prediction score is 75.06% for stock TSLA\n",
      "For 290th day in future, prediction score is 73.39% for stock TSLA\n",
      "For 300th day in future, prediction score is 73.47% for stock TSLA\n",
      "For 310th day in future, prediction score is 70.9% for stock TSLA\n",
      "\n",
      "\n",
      "Result:\n",
      "Our model can predict with prediction score > 70% as shown below for each stock\n",
      "PYPL: 531\n",
      "CTSH: 432\n",
      "TSLA: 312\n"
     ]
    }
   ],
   "source": [
    "result = dict()\n",
    "\n",
    "for stock in stockList:\n",
    "    data = web.DataReader(stock,'yahoo', start, end)\n",
    "    data.columns = [x.strip() for x in data.columns]\n",
    "    data.columns = [x.replace(' ', '_') for x in data.columns]\n",
    "    \n",
    "    efficiency = 1.0\n",
    "    number = 0\n",
    "    \n",
    "    print(\"Finding how long in future, the model can predict with prediction score > 70% for stock: \",stock)\n",
    "    \n",
    "    while efficiency > 0.70:\n",
    "        number += 1\n",
    "        pd.concat([data.Close, data.Close.shift(-number)], axis=1)\n",
    "        def valid(col):\n",
    "            return \"Close\" not in col\n",
    "        testData = data.dropna()\n",
    "\n",
    "        # we remove the last observation (last row) because there is no tomorrow for the last row to predict!\n",
    "        X = testData[[x for x in testData.columns if valid(x)]].iloc[0:-number]\n",
    "\n",
    "        # for the dependent variable, we remove the NA, so that removes the last row, too\n",
    "        y = testData.Close.shift(-number).dropna()\n",
    "        X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42)\n",
    "            \n",
    "        # Create a model \n",
    "        rf_model = RandomForestRegressor()\n",
    "\n",
    "        # Train the model\n",
    "        rf_model.fit(X_train, y_train)\n",
    "        efficiency = rf_model.score(X_test, y_test)\n",
    "        if number%10 == 0:\n",
    "            print(\"For\",number,\"\\bth day in future, prediction score is\", round(efficiency*100,2),\"\\b% for stock\",stock)\n",
    "    \n",
    "    result.update({stock:number-1})\n",
    "\n",
    "print(\"\\n\\nResult:\")\n",
    "print(\"Our model can predict with prediction score > 70% as shown below for each stock\")\n",
    "for stock in result:\n",
    "    print(stock+\":\",result[stock])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Result:\n",
      "Our model can predict with prediction score above 70% for number of days as shown below for each stock\n",
      "PYPL: 531\n",
      "CTSH: 432\n",
      "TSLA: 312\n"
     ]
    }
   ],
   "source": [
    "print(\"Result:\")\n",
    "print(\"Our model can predict with prediction score above 70% for number of days as shown below for each stock\")\n",
    "for stock in result:\n",
    "    print(stock+\":\",result[stock])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}