time series with machine learning.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import statsmodels.api as sm\n",
    "from datetime import date,timedelta\n",
    "from sklearn.ensemble import GradientBoostingRegressor\n",
    "from sklearn import linear_model\n",
    "\n",
    "filter_date = date.today()-timedelta(2)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\dnishimoto.BOISE\\AppData\\Local\\Continuum\\anaconda3\\envs\\deeplearning\\lib\\site-packages\\ipykernel_launcher.py:61: MatplotlibDeprecationWarning: \n",
      "The 'normed' kwarg was deprecated in Matplotlib 2.1 and will be removed in 3.1. Use 'density' instead.\n",
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZcAAAEWCAYAAACqitpwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAcH0lEQVR4nO3de5hddX3v8ffHBAgoFw2hRxNkYhP1AD2iRtTaeqwUC8dLaAs1aiu1tGg11tb2aUN7oD0caeWpp3iBqihY5KhgqbZjTY1H8VKoAoNQISCPY8Qy4CXcAggBAt/zx1pTN8NMZidZO5Nh3q/n2c+s9Vu/32//9s6e+WTdfjtVhSRJXXrcTA9AkvTYY7hIkjpnuEiSOme4SJI6Z7hIkjpnuEiSOme4SAOSpJIsm+lxbKskQ+3Y58/0WDR7GS6aE5LcmOS+JHcnuTPJvyV5U5JOfgeSfDnJb3fR14R+/UOvWclw0VzyyqraGzgIeCfwJ8A5MzukHWfwaFdkuGjOqapNVTUMvBo4PsmhAEn2SPKuJP+R5IdJPpBkz3bbE5P8c5KNSe5ol5e0204Dfh44M8k9Sc7sebpfTPLtts1ZSdK2WZbkK0k2Jbk1yYVTDPer7c87275fmOQ3k1ya5IwktwN/0fb5W0mub59rXZKDxjtp937eNMVY5rWv+9YkG4CXd/NOay4zXDRnVdXlwBhNMACcDjwdOAxYBiwGTmm3PQ74CM1ez1OB+4Az237+DPhXYHVVPaGqVvc8zSuA5wHPAn4N+KW2/H8DnweeCCwB3jfFMF/c/tyv7ftr7frzgQ3AAcBpSY4B/hT4FWBRO55PTOhrqrH8Trvt2cAK4NgpxiL1zXDRXHcL8KT2f/G/A/xBVd1eVXcDfwmsAqiq26rqH6rq3nbbacB/76P/d1bVnVX1H8CXaIIL4EGaoHpKVW2uqku2ddxV9b6q2lJV9wFvBP6qqq6vqi3t2A/r3XvZylh+DXh3Vd1UVbcDf7WNY5EexXDRXLcYuJ3mf/t7AVe2J/zvBD7XlpNkryQfTPK9JHfRHK7aL8m8afr/Qc/yvcAT2uU/BgJcnmR9kt/axnHfNGH9IOA9PWO/ve1/cR9jecqE/r63jWORHsUTgZqzkjyP5o/vJcCtNIe6Dqmqmyep/ofAM4DnV9UPkhwGXEXzBxxgm6YXr6of0OwpkeTngC8k+WpVjU6sOlUXE9ZvAk6rqo9tyzha3wcO7Fl/6nb0IT2Cey6ac5Lsk+QVwAXA/62qa6rqYeBDwBlJDmjrLU4yfl5ib5rwuTPJk4A/n9DtD4GnbcMYjhu/IAC4gyYsHpqk6kbg4T76/gBwUpJD2v73TXJcn8P5JPB7SZYkeSKwps920pQMF80ln0lyN83/8v8M+BvgDT3b/wQYBb7eHvr6As3eCsC7gT1p9nC+TnPIrNd7gGPbK7He28dYngdcluQeYBh4W1V9d2KlqrqX5vzOpe0hrxdM1llVfZrmgoQL2rFfCxzdxzigCdV1wL8D3wA+1Wc7aUrxy8IkSV1zz0WS1DnDRZLUOcNFktQ5w0WS1Lk5cZ/L/vvvX0NDQzM9DEmaVa688spbq2rR9rSdE+EyNDTEyMjITA9DkmaVJNs9W4OHxSRJnTNcJEmdM1wkSZ2bE+dcJGmue/DBBxkbG2Pz5s2P2rZgwQKWLFnCbrvt1tnzGS6SNAeMjY2x9957MzQ0RPslpABUFbfddhtjY2MsXbq0s+fzsJgkzQGbN29m4cKFjwgWgCQsXLhw0j2aHWG4SNIcMTFYpivfEYaLJKlzhoskqXOe0J/G0JrPzvQQdrob3/nymR6CpAGoqkkPgQ3ie73cc5GkOWDBggXcdtttjwqS8avFFixY0OnzueciSXPAkiVLGBsbY+PGjY/aNn6fS5cMF0maA3bbbbdO72OZjofFJEmdM1wkSZ0zXCRJnTNcJEmdG2i4JDkqyQ1JRpOsmWT7HkkubLdflmSoLT88ydXt49+T/HK/fUqSZt7AwiXJPOAs4GjgYOA1SQ6eUO0E4I6qWgacAZzell8LrKiqw4CjgA8mmd9nn5KkGTbIPZfDgdGq2lBVDwAXACsn1FkJnNcuXwQckSRVdW9VbWnLFwDjd/3006ckaYYNMlwWAzf1rI+1ZZPWacNkE7AQIMnzk6wHrgHe1G7vp0/a9icmGUkyMtlNQ5KkwRlkuEw2h/PECWymrFNVl1XVIcDzgJOSLOizT9r2Z1fViqpasWjRom0YtiRpRw0yXMaAA3vWlwC3TFUnyXxgX+D23gpVdT3wY+DQPvuUJM2wQYbLFcDyJEuT7A6sAoYn1BkGjm+XjwUurqpq28wHSHIQ8Azgxj77lCTNsIHNLVZVW5KsBtYB84Bzq2p9klOBkaoaBs4Bzk8ySrPHsqpt/nPAmiQPAg8Db66qWwEm63NQr0GStH0GOnFlVa0F1k4oO6VneTNw3CTtzgfO77dPSdKuxTv0JUmdM1wkSZ0zXCRJnTNcJEmdM1wkSZ0zXCRJnTNcJEmdM1wkSZ0zXCRJnTNcJEmdM1wkSZ0zXCRJnTNcJEmdM1wkSZ0zXCRJnTNcJEmdM1wkSZ0zXCRJnTNcJEmdM1wkSZ0zXCRJnTNcJEmdM1wkSZ0baLgkOSrJDUlGk6yZZPseSS5st1+WZKgtPzLJlUmuaX++tKfNl9s+r24fBwzyNUiStt38QXWcZB5wFnAkMAZckWS4qq7rqXYCcEdVLUuyCjgdeDVwK/DKqrolyaHAOmBxT7vXVdXIoMYuSdoxg9xzORwYraoNVfUAcAGwckKdlcB57fJFwBFJUlVXVdUtbfl6YEGSPQY4VklShwYZLouBm3rWx3jk3scj6lTVFmATsHBCnV8Frqqq+3vKPtIeEjs5SSZ78iQnJhlJMrJx48YdeR2SpG00yHCZ7I9+bUudJIfQHCp7Y8/211XVzwA/3z5+Y7Inr6qzq2pFVa1YtGjRNg1ckrRjBhkuY8CBPetLgFumqpNkPrAvcHu7vgT4NPD6qvrOeIOqurn9eTfwcZrDb5KkXcggw+UKYHmSpUl2B1YBwxPqDAPHt8vHAhdXVSXZD/gscFJVXTpeOcn8JPu3y7sBrwCuHeBrkCRth4GFS3sOZTXNlV7XA5+sqvVJTk3yqrbaOcDCJKPA24Hxy5VXA8uAkydccrwHsC7JN4GrgZuBDw3qNUiSts/ALkUGqKq1wNoJZaf0LG8Gjpuk3TuAd0zR7XO7HKMkqXveoS9J6pzhIknqnOEiSeqc4SJJ6pzhIknqnOEiSeqc4SJJ6pzhIknqnOEiSeqc4SJJ6pzhIknqnOEiSeqc4SJJ6pzhIknqnOEiSeqc4SJJ6pzhIknqnOEiSeqc4SJJ6pzhIknqnOEiSeqc4SJJ6pzhIknqXN/hkmTPJM8Y5GAkSY8NfYVLklcCVwOfa9cPSzLcR7ujktyQZDTJmkm275Hkwnb7ZUmG2vIjk1yZ5Jr250t72jy3LR9N8t4k6e+lSpJ2ln73XP4COBy4E6CqrgaGttYgyTzgLOBo4GDgNUkOnlDtBOCOqloGnAGc3pbfCryyqn4GOB44v6fN+4ETgeXt46g+X4MkaSfpN1y2VNWmbez7cGC0qjZU1QPABcDKCXVWAue1yxcBRyRJVV1VVbe05euBBe1ezpOBfarqa1VVwEeBY7ZxXJKkAes3XK5N8lpgXpLlSd4H/Ns0bRYDN/Wsj7Vlk9apqi3AJmDhhDq/ClxVVfe39cem6ROAJCcmGUkysnHjxmmGKknqUr/h8lbgEOB+4BPAXcDvT9NmsnMhtS11khxCc6jsjdvQZ1NYdXZVraiqFYsWLZpmqJKkLs3vp1JV3Qv8Wfvo1xhwYM/6EuCWKeqMJZkP7AvcDpBkCfBp4PVV9Z2e+kum6VOSNMP6Cpckn+HRewibgBHgg1W1eZJmVwDLkywFbgZWAa+dUGeY5oT914BjgYurqpLsB3wWOKmqLh2vXFXfT3J3khcAlwGvB97Xz2uQJO08/R4W2wDcA3yofdwF/BB4erv+KO05lNXAOuB64JNVtT7JqUle1VY7B1iYZBR4OzB+ufJqYBlwcpKr28cB7bbfBT4MjALfAf6l3xcrSdo5+tpzAZ5dVS/uWf9Mkq9W1YuTrJ+qUVWtBdZOKDulZ3kzcNwk7d4BvGOKPkeAQ/sctyRpBvS757IoyVPHV9rl/dvVBzoflSRpVut3z+UPgUuSfIfmiq2lwJuTPJ6f3KciSRLQ/9Via5MsB55JEy7f6jmJ/+5BDU6SNDv1u+cCzVQrzwAWAP8tCVX10cEMS5I0m/V7KfKfAy+hmSNsLc18YZfQTL8iSdIj9HtC/1jgCOAHVfUG4FnAHgMblSRpVus3XO6rqoeBLUn2AX4EPG1ww5IkzWb9nnMZae+a/xBwJc0NlZcPbFSSpFmt36vF3twufiDJ52imvf/m4IYlSZrN+v0myi+OL1fVjVX1zd4ySZJ6bXXPJckCYC9g/yRP5CdT3u8DPGXAY5MkzVLTHRZ7I833tjyF5lzLeLjcRfMVxpIkPcpWw6Wq3gO8J8lbq8qp7SVJfen3hP77kvwsMNTbxjv0JUmT6fcO/fOBnwauBh5qiwvv0JckTaLf+1xWAAdX1aTfVy9JUq9+79C/FvgvgxyIJOmxo989l/2B65JcDtw/XlhVr5q6iSRpruo3XP5ikIOQJD229Hu12FeSHAQsr6ovJNkLmDfYoUmSZqt+p3/5HeAi4INt0WLgHwc1KEnS7NbvCf23AC+iuTOfqvo2cMCgBiVJmt36DZf7q+qB8ZUk82nuc5Ek6VH6DZevJPlTYM8kRwJ/D3xmcMOSJM1m/YbLGmAjcA3NZJZrgf85XaMkRyW5IclokjWTbN8jyYXt9suSDLXlC5N8Kck9Sc6c0ObLbZ9Xtw8Pz0nSLqbfS5H3BM6tqg8BJJnXlt07VYO2zlnAkcAYcEWS4aq6rqfaCcAdVbUsySrgdODVwGbgZODQ9jHR66pqpM+xS5J2sn73XL5IEybj9gS+ME2bw4HRqtrQnq+5AFg5oc5K4Lx2+SLgiCSpqh9X1SU0ISNJmmX6DZcFVXXP+Eq7vNc0bRYDN/Wsj7Vlk9apqi3AJmBhH+P5SHtI7OQkmaxCkhOTjCQZ2bhxYx9dSpK60m+4/DjJc8ZXkjwXuG+aNpP90Z94hVk/dSZ6XVX9DPDz7eM3JqtUVWdX1YqqWrFo0aJpupQkdanfcy5vA/4+yS3t+pNpzo1szRhwYM/6EuCWKeqMtZc37wvcvrVOq+rm9ufdST5Oc/jNqf8laRcybbgkeRywO/BM4Bk0exvfqqoHp2l6BbA8yVLgZmAV8NoJdYaB44GvAccCF29tWv82gParqluT7Aa8gunP/UiSdrJpw6WqHk7yf6rqhTRT7/elqrYkWQ2so5mH7NyqWp/kVGCkqoaBc4Dzk4zS7LGsGm+f5EZgH2D3JMcALwO+B6xrg2UeTbB8qN8xSZJ2jn4Pi30+ya8Cn9qWLwyrqrU098T0lp3Ss7wZOG6KtkNTdPvcfp9fkjQz+g2XtwOPBx5Kch/NobGqqn0GNjJJ0qzV75T7ew96IJKkx45+p9xPkl9PcnK7fmCSwwc7NEnSbNXvfS5/C7yQn1ztdQ/N1C6SJD1Kv+dcnl9Vz0lyFUBV3ZFk9wGOS5I0i/W75/JgOxFlASRZBDw8sFFJkma1fsPlvcCngQOSnAZcAvzlwEYlSZrV+r1a7GNJrgSOoLkM+Ziqun6gI5MkzVpbDZckC4A3Actovijsg+3sxZIkTWm6w2LnAStoguVo4F0DH5Ekadab7rDYwe309iQ5B7h88EOSJM120+25/OfMxx4OkyT1a7o9l2cluatdDrBnu+7cYpKkKW01XKpq3s4aiCTpsaPf+1wkSeqb4SJJ6pzhIknqnOEiSeqc4SJJ6pzhIknqnOEiSeqc4SJJ6pzhIknq3EDDJclRSW5IMppkzSTb90hyYbv9siRDbfnCJF9Kck+SMye0eW6Sa9o2702SQb4GSdK2G1i4tF+LfBbNVP0HA69JcvCEaicAd1TVMuAM4PS2fDNwMvBHk3T9fuBEYHn7OKr70UuSdsQg91wOB0arakNVPQBcAKycUGclzXfGAFwEHJEkVfXjqrqEJmT+U5InA/tU1deqqoCPAscM8DVIkrbDIMNlMXBTz/pYWzZpnXZK/03Awmn6HJumTwCSnJhkJMnIxo0bt3HokqQdMchwmexcSG1Hne2qX1VnV9WKqlqxaNGirXQpSeraIMNlDDiwZ30JcMtUdZLMB/YFbp+mzyXT9ClJmmGDDJcrgOVJlibZHVgFDE+oMwwc3y4fC1zcnkuZVFV9H7g7yQvaq8ReD/xT90OXJO2I6b6JcrtV1ZYkq4F1wDzg3Kpan+RUYKSqhoFzgPOTjNLssawab5/kRmAfYPckxwAvq6rrgN8F/g7YE/iX9iFJ2oUMLFwAqmotsHZC2Sk9y5uB46ZoOzRF+QhwaHejlCR1zTv0JUmdG+ieizRbDK357EwPYae78Z0vn+kh7FT+G+9c7rlIkjpnuEiSOme4SJI6Z7hIkjpnuEiSOme4SJI6Z7hIkjpnuEiSOme4SJI6Z7hIkjpnuEiSOme4SJI6Z7hIkjpnuEiSOme4SJI6Z7hIkjpnuEiSOme4SJI6Z7hIkjpnuEiSOme4SJI6Z7hIkjo30HBJclSSG5KMJlkzyfY9klzYbr8syVDPtpPa8huS/FJP+Y1JrklydZKRQY5fkrR95g+q4yTzgLOAI4Ex4Iokw1V1XU+1E4A7qmpZklXA6cCrkxwMrAIOAZ4CfCHJ06vqobbdL1TVrYMauyRpxwxyz+VwYLSqNlTVA8AFwMoJdVYC57XLFwFHJElbfkFV3V9V3wVG2/4kSbPAIMNlMXBTz/pYWzZpnaraAmwCFk7TtoDPJ7kyyYlTPXmSE5OMJBnZuHHjDr0QSdK2GWS4ZJKy6rPO1tq+qKqeAxwNvCXJiyd78qo6u6pWVNWKRYsW9TtmSVIHBhkuY8CBPetLgFumqpNkPrAvcPvW2lbV+M8fAZ/Gw2WStMsZ2Al94ApgeZKlwM00J+hfO6HOMHA88DXgWODiqqokw8DHk/wNzQn95cDlSR4PPK6q7m6XXwacOsDXMCcNrfnsTA9BO4H/zhqkgYVLVW1JshpYB8wDzq2q9UlOBUaqahg4Bzg/ySjNHsuqtu36JJ8ErgO2AG+pqoeS/BTw6eacP/OBj1fV5wb1GiRJ22eQey5U1Vpg7YSyU3qWNwPHTdH2NOC0CWUbgGd1P1JJUpe8Q1+S1DnDRZLUOcNFktQ5w0WS1DnDRZLUOcNFktQ5w0WS1DnDRZLUOcNFktQ5w0WS1DnDRZLUOcNFktQ5w0WS1DnDRZLUOcNFktQ5w0WS1DnDRZLUOcNFktQ5w0WS1DnDRZLUOcNFktQ5w0WS1DnDRZLUOcNFktS5gYZLkqOS3JBkNMmaSbbvkeTCdvtlSYZ6tp3Ult+Q5Jf67VOSNPMGFi5J5gFnAUcDBwOvSXLwhGonAHdU1TLgDOD0tu3BwCrgEOAo4G+TzOuzT0nSDBvknsvhwGhVbaiqB4ALgJUT6qwEzmuXLwKOSJK2/IKqur+qvguMtv3106ckaYbNH2Dfi4GbetbHgOdPVaeqtiTZBCxsy78+oe3idnm6PgFIciJwYrt6T5IbtuM17Ar2B26d6UHswnx/puZ7s3WP+fcnp+9Q8/2Bg7a38SDDJZOUVZ91piqfbE9rYp9NYdXZwNlbG+BskGSkqlbM9Dh2Vb4/U/O92Trfn61r35+h7W0/yMNiY8CBPetLgFumqpNkPrAvcPtW2vbTpyRphg0yXK4AlidZmmR3mhP0wxPqDAPHt8vHAhdXVbXlq9qryZYCy4HL++xTkjTDBnZYrD2HshpYB8wDzq2q9UlOBUaqahg4Bzg/ySjNHsuqtu36JJ8ErgO2AG+pqocAJutzUK9hFzHrD+0NmO/P1Hxvts73Z+t26P1Js6MgSVJ3vENfktQ5w0WS1DnDZReS5MAkX0pyfZL1Sd7Wlj8pyf9L8u325xNneqwzpZ2p4aok/9yuL22nDvp2O5XQ7jM9xpmSZL8kFyX5VvsZeqGfnUaSP2h/p65N8okkC+byZyfJuUl+lOTanrJJPytpvLedcuubSZ7Tz3MYLruWLcAfVtV/BV4AvKWd3mYN8MWqWg58sV2fq94GXN+zfjpwRvve3EEzpdBc9R7gc1X1TOBZNO/TnP/sJFkM/B6woqoOpbkYaBVz+7PzdzRTa/Wa6rNyNM0Vu8tpbkx/fz9PYLjsQqrq+1X1jXb5bpo/Dot55DQ55wHHzMwIZ1aSJcDLgQ+36wFeSjN1EMzt92Yf4MU0V2BSVQ9U1Z342Rk3H9izvZ9uL+D7zOHPTlV9leYK3V5TfVZWAh+txteB/ZI8ebrnMFx2Ue0M0c8GLgN+qqq+D00AAQfM3Mhm1LuBPwYebtcXAndW1ZZ2vXeaoLnmacBG4CPtYcMPJ3k8fnaoqpuBdwH/QRMqm4Ar8bMz0VSflcmm8pr2vTJcdkFJngD8A/D7VXXXTI9nV5DkFcCPqurK3uJJqs7Va+vnA88B3l9VzwZ+zBw8BDaZ9tzBSmAp8BTg8TSHeiaaq5+d6WzX75nhsotJshtNsHysqj7VFv9wfDe0/fmjmRrfDHoR8KokN9LMhv1Smj2Z/dpDHTC3pwMaA8aq6rJ2/SKasPGzA78IfLeqNlbVg8CngJ/Fz85EU31WtmvaLcNlF9KeQzgHuL6q/qZnU+80OccD/7SzxzbTquqkqlrSTqS3imaqoNcBX6KZOgjm6HsDUFU/AG5K8oy26AiaGS7m/GeH5nDYC5Ls1f6Ojb83fnYeaarPyjDw+vaqsRcAm8YPn22Nd+jvQpL8HPCvwDX85LzCn9Kcd/kk8FSaX5Tjqmriybg5I8lLgD+qqlckeRrNnsyTgKuAX6+q+2dyfDMlyWE0FzvsDmwA3kDzH8g5/9lJ8r+AV9NckXkV8Ns05w3m5GcnySeAl9BMq/9D4M+Bf2SSz0obyGfSXF12L/CGqhqZ9jkMF0lS1zwsJknqnOEiSeqc4SJJ6pzhIknqnOEiSeqc4SJtpyQPJbm6nW3335O8Pcl2/U61Mxq/uWf9JeMzP0uzkeEibb/7quqwqjoEOBL4HzT3C2yP/YA3T1tLmiUMF6kDVfUjmunIV7d3Ms9L8tdJrmi/A+ON0Mwbl+SLSb6R5JokK9su3gn8dLsn9Ndt2RN6vp/lY+3NbCR5Z5Lr2n7ftdNfrNSH+dNXkdSPqtrQHhY7gGaixE1V9bwkewCXJvk8zeyyv1xVdyXZH/h6kmGaSSYPrarD4D9nIXg2cAjNPE6XAi9Kch3wy8Azq6qS7LeTX6bUF/dcpG6NzyD7Mpr5mK6mmb5nIc2XLQX4yyTfBL5AMwXJT03R1+VVNVZVDwNXA0PAXcBm4MNJfoVmOg5pl+Oei9SRdp6zh2hmkw3w1qpaN6HObwKLgOdW1YPtLM8Lpuiyd56rh4D5VbUlyeE0ky+uAlbTzBAt7VLcc5E6kGQR8AHgzGom7FsH/G77FQokeXr75V370nwvzYNJfgE4qO3ibmDvPp7nCcC+VbUW+H3gsO5fjbTj3HORtt+e7WGv3Whm2z0fGP+qhA/THMb6RnsifiPN18Z+DPhMkhGaQ13fAqiq25JcmuRa4F+Az07xnHsD/5RkAc3e0R8M4oVJO8pZkSVJnfOwmCSpc4aLJKlzhoskqXOGiySpc4aLJKlzhoskqXOGiySpc/8f08m2i91p9FYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\dnishimoto.BOISE\\AppData\\Local\\Continuum\\anaconda3\\envs\\deeplearning\\lib\\site-packages\\ipykernel_launcher.py:70: MatplotlibDeprecationWarning: \n",
      "The 'normed' kwarg was deprecated in Matplotlib 2.1 and will be removed in 3.1. Use 'density' instead.\n",
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df = pd.read_csv(\"https://raw.githubusercontent.com/nytimes/covid-19-data/master/us-counties.csv\",encoding = \"ISO-8859-1\")\n",
    "\n",
    "df['state_county']=df['state']+\" \"+df['county']\n",
    "filter=df[\"state\"].isin(['Colorado','Idaho','Utah'])\n",
    "\n",
    "#filter=df[\"state\"].isin(['Utah'])\n",
    "df=df[filter]\n",
    "filter=df[\"date\"]==str(filter_date)\n",
    "df=df[filter]\n",
    "\n",
    "filter=df.groupby(['state_county'])['deaths'].max()>10\n",
    "subset_df=filter.loc[filter.values==True]\n",
    "result=pd.merge(df,subset_df,how='inner',on=['state_county'])\n",
    "#print(result.head(5))\n",
    "df=result[result['deaths_y']==True]\n",
    "df=df.sort_values(['county'],ascending=True)\n",
    "\n",
    "#print(df.head(25))\n",
    "counties=df['state_county']\n",
    "counties_index=np.arange(0,len(counties))\n",
    "#print(counties[counties_index])\n",
    "dates=np.array(df['date'])\n",
    "deaths=np.array(df['deaths_x'])\n",
    "cases=np.array(df['cases'])\n",
    "mortality_rate=deaths/cases\n",
    "\n",
    "#print(type(counties))\n",
    "#print(deaths)\n",
    "#print(morality_rate)\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "ax.scatter(counties_index,deaths,label='deaths')\n",
    "ax.set_xlabel(\"County\")\n",
    "ax.set_ylabel(\"Deaths\")\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "model = linear_model.LinearRegression()\n",
    "model.fit(cases.reshape(-1,1), morality_rate)\n",
    "\n",
    "SEED=1\n",
    "gbt=GradientBoostingRegressor(n_estimators=300, max_depth=1, random_state=SEED)\n",
    "gbt.fit(cases.reshape(-1,1), morality_rate)  \n",
    "\n",
    "prediction_space=np.linspace(min(cases),max(cases)).reshape(-1,1)\n",
    "\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "ax.scatter(cases,deaths,color='red',label='deaths')\n",
    "ax.set_xlabel('Cases')\n",
    "ax.set_ylabel('Deaths')\n",
    "ax2=ax.twinx()\n",
    "ax2.scatter(cases,mortality_rate,color='green',label='morality')\n",
    "ax2.plot(prediction_space,model.predict(prediction_space),c=\"yellow\")\n",
    "ax2.plot(prediction_space,gbt.predict(prediction_space),c=\"purple\")\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "ax.hist(deaths,bins=5, normed=1)\n",
    "ax.set_xlabel('Deaths')\n",
    "ax.set_ylabel('Percentage')\n",
    "ax.set_title('Deaths trend')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "ax.hist(mortality_rate,bins=40, normed=1)\n",
    "ax.set_xlabel('Morality rate')\n",
    "ax.set_ylabel('Percentage')\n",
    "ax.set_title('Mortality rate trend')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}