Chapter 5.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "# get project dir\n",
    "# import standard libs\n",
    "from IPython.display import display\n",
    "from IPython.core.debugger import set_trace as bp\n",
    "from pathlib import PurePath, Path\n",
    "import sys\n",
    "import time\n",
    "from collections import OrderedDict as od\n",
    "import re\n",
    "import os\n",
    "import json\n",
    "\n",
    "pp = PurePath(Path.cwd()).parts[:]\n",
    "pdir = PurePath(*pp)\n",
    "data_script_dir = pdir / 'src' / 'data'\n",
    "bars_script_dir = pdir / 'src' / 'features'\n",
    "sys.path.append(data_script_dir.as_posix())\n",
    "sys.path.append(bars_script_dir.as_posix())\n",
    "viz_dir = pdir / 'reports' / 'figures'\n",
    "data_dir = pdir / 'data'\n",
    "\n",
    "# import python scientific stack\n",
    "import pandas as pd\n",
    "pd.set_option('display.max_rows', 100)\n",
    "from dask import dataframe as dd\n",
    "from dask.diagnostics import ProgressBar\n",
    "from multiprocessing import cpu_count\n",
    "pbar = ProgressBar()\n",
    "pbar.register()\n",
    "import numpy as np\n",
    "import scipy.stats as stats\n",
    "import statsmodels.api as sm\n",
    "from numba import jit\n",
    "import math\n",
    "\n",
    "# import visual tools\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.gridspec as gridspec\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "\n",
    "plt.style.use('seaborn-talk')\n",
    "plt.style.use('bmh')\n",
    "#plt.rcParams['font.family'] = 'DejaVu Sans Mono'\n",
    "plt.rcParams['font.size'] = 9.5\n",
    "plt.rcParams['font.weight'] = 'medium'\n",
    "plt.rcParams['figure.figsize'] = 10,7\n",
    "blue, green, red, purple, gold, teal = sns.color_palette('colorblind', 6)\n",
    "\n",
    "# import util libs\n",
    "# from tqdm import tqdm, tqdm_notebook\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "from utils import *\n",
    "from bars import *\n",
    "from labelling import *\n",
    "from mpEngine import *\n",
    "from sampleWeights import *\n",
    "from ffd import *\n",
    "\n",
    "RANDOM_STATE = 777\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [5.1] Generate a time series from an IID Gaussian random process. This is a memory-less, stationary series:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fed8ec8a470>"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "np.random.seed(0)\n",
    "\n",
    "N = 252 * 10\n",
    "s = pd.Series(np.random.randn(N)) # returna sample from the gaussian distribution\n",
    "s.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### a) Compute the ADF statistics on this series. What is the p-value?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ADF stat:  -50.803321802760166  p-value:  0.0\n"
     ]
    }
   ],
   "source": [
    "res = sm.tsa.stattools.adfuller(s)\n",
    "print('ADF stat: ',res[0],' p-value: ',res[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### b) Compute the cumulative sum of the observations. This is a non-stationary series without memory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fed8e95d3c8>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cmsm = pd.Series(s).cumsum()\n",
    "cmsm.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### i) What is the order of integration of this cumulative series?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "===========================\n",
      "order: 0, pval:0.5704444806659957\n",
      "===========================\n",
      "order: 1, pval:0.0\n",
      "===========================\n",
      "order: 2, pval:7.347529850655246e-30\n",
      "===========================\n",
      "order: 3, pval:0.0\n",
      "===========================\n",
      "order: 4, pval:0.0\n"
     ]
    }
   ],
   "source": [
    "orders = [0,1,2,3,4]\n",
    "p_val = lambda s: sm.tsa.stattools.adfuller(s)[1]\n",
    "\n",
    "for o in orders:\n",
    "    diff_ = np.diff(cmsm, o)\n",
    "    print('=' * 27) \n",
    "    print(f'order: {o}, pval:{p_val(diff_)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In statistics, the order of integration, denoted I(d), of a time series is a summary statistic, which reports the minimum number of differences required to obtain a covariance-stationary series.\n",
    "\n",
    "In this example, the order of integration is 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### ii) Compute the ADF stat on this series, What is the p-value?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The p-value of the cumulative series:  0.5704444806659957\n"
     ]
    }
   ],
   "source": [
    "print(\"The p-value of the cumulative series: \", p_val(cmsm))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### c) Differentiate the series twice. What is the p-value of this over-differentiated series?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "p-value of the original series:  0.0 p-value of the cum.series:  7.347529850655246e-30\n"
     ]
    }
   ],
   "source": [
    "diff_ = np.diff(s,2)\n",
    "cum_diff_ = np.diff(cmsm, 2)\n",
    "print(\"p-value of the original series: \", p_val(diff_), \"p-value of the cum.series: \", p_val(cum_diff_))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [5.2] Generate a time series that follows a sinusoidal function. This is a stationary series with memory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fed8e9105c0>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAD9CAYAAABTJWtQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvqOYd8AAAIABJREFUeJztvXmYXNdZ5/99a69eqrqru9Td2roltSzJm7ySOLYTB0HiJBCGxBliEpIwecDAMxPgIRCYwINhfhC2H/PL8GMIgTBZSWYgCVsWJ5gsckJsR7JlW5Ylt9Tqlnqt7qqurWuvM3/cpUulWu6tuufeU3Xv53nqkbrurarzvWd5z3nPe84hxhgcHBwcHBwAwGV1AhwcHBwcxMExCg4ODg4OKo5RcHBwcHBQcYyCg4ODg4OKYxQcHBwcHFQ8ViegW775zW8yv9/f0WfL5TI8np5/BLqwm2ZHb/9jN81G6d3e3t44ceJEtP79nn+Sfr8fR48e7eizhUIBnRqUXsVumh29/Y/dNBul9/Tp0wuN3re1+2h1ddXqJJiO3TQ7evsfu2nmrdfWRsHr9VqdBNOxm2ZHb/9jN8289draKITDYauTYDp20+zo7X/sppm3XlsbhY2NDauTYDp20+zo7X/sppm3XlsbBbv1MAD7aXb09j9209xzIwUiejsRnSSiFBGVNdx/FxE9RUTbRHSRiN5pdJqaUSwWzfopYbCbZkdv/2M3zbz18hgpJAD8TwC/1O5GIgoD+AqAzwMYBfBzAD5CRPdwSNd15HI5M35GKOym2dHb/9hNM2+9hq9TYIw9BgBE9ICG298CIAfgj5i0h/fXieiLAH4WwL9r+b1yuYy5uTn170gkgkgkoimtk5OTmu6r58xyGl94IYbldAEzIwH8x+MTODw+0NF3mU2nmkWhVKniiy/E8MTlLVQYw917Q3jbrRMY9Lkb3t/revXCW+9yqoDPPruK87FtRAa8+JGj47h3Jgwi4vq7rbBLHjPGcHJ+C//8YgHJ75/DkegA/sNNURwaM7btsXrx2nEAp9m1hzqcBvBTWr9gfX0dDz30kPr3I488gkcffRSrq6sYHByE2+1GKpVCNBpFPB4HYwzRaBRra2vI5XKIRqPIZDKYmJhALBYDESESiSAWiyEUCqFSqSCbzWJychKrq6v42mIRf3c+o/7eQiKPJy5v4X2v2o0bB/Lw+XwYHh7G5uYmRkdHkcvlkM/n1c8HAgEEg0EkEgmMjY0hnU6jWCyq14PBIHw+H5LJJMbHx5FMJlEqldTr7TQNDQ0BQFNNuVwO4+Pj12jyer0Ih8PY2NhAOBxGsVhELpdTr4uiaT2Rwp+/kMeFjbz6/F/eyOGbc5v4wKuiGGCF6zRdvHgR+/fvF1aTnrKnJZ8ymQx8Ph8XTZeybvzpkxsoVKTqejmRx+mlNH7kyCh+dB8QGh7moqldPiWTSfj9/p7KJ71lL18o4AsLDI+9vKWW/cuJPFCt4OEbAh1pagbxOmRHHin8K2OsqeEhoo8B8DDG3l3z3k8D+CBjbFbL75w8eZIFAgH1bz0jhZWVFUxNTWm6FwD+9eU4/uhbCyAAP3n7JO7ZH8ZXL2ziX85twEXA7z94CHfsCWn+PivQq1kUKlWG3/jqHJ5dzmDXkBc//8q9CHhc+OiTS5hP5HFgNID/7803IOi9dsTQq3o7hZfe+XgO7/unCyiUq3j1gRG89ZZdeGk9i79+ahmlKsN77pzCT95uTY/dDnn88e8v42+fXYPfTXjb0SHcfWgK37qUwJtvjGJ3qLPVzadPnz514sSJu+rft3qkkAYwU/feCICU1i/weDyYndVkP65jWO7ZaGE9U8SHv3MFAPBf7t2HHzk2DgC4ITqAIZ8bnzuzhj/59iL+6q3HmroyRECPZpH4h7MxPLucwWjQgz9502FMDksV4f+NDuB9/3QB84k8/tf3V/AL9+y95nO9qrdTeOgtVxn+6FsLKJSr+MFDo/i1B6bhIsKxXYOYCvnx21+7hE+cWsFde0O4IWq+G7Xf8/jcehafO7MGFwG//cMHcXTEhaGhQRzbNcjl96wOST0D4Pa6926X3+fO5uam5ns/8r2rKJSruG9mRDUICu++cwpHogPYyJbwqdMrRifTUPRoFoXNbAkf//4yAOCX79+vGgQAGPJ78MEfnIGLgH96MYaLm9vXfrYH9XYDD73Sc81hYsiHX7xvH1w18wev3B/GW26OggH4s+9eQdWC4337OY+rjOHDTyyiyoC33bILd+0NcdfLIyTVTUQBAD7574D8ajQT9UUAA0T0q0TkI6ITkCafP2p0uhoxOjqq6b4LG9t44nISfo8LP3/Pnuuuu12EX7x3HwDgX85tYDNbMjSdRqJVs0h87swqChWGe6fDeOX+62O0D40N4MdujKLKgE+evnZfmF7U2w1G682VKvjcs2sAgF+4Z+917jkA+Kk7phAZ8OB8bBtPLmoe5BtGP+fxE/NbuBTPY3zQi5+6Q3KR8dbLY6TwU5Aiih4D4Jb/nwMwTUT3E1GGiPYDAGNsC8AbAbwNQBLAXwH4OcaYpsijbtEa2vW/z0iV4kePjSM66Gt4z+z4AO6bGUGxwvB/nlszLI1G02vhe5vZEr780iYIwLvubO43fvvxCfjchH9fSF4zWug1vd1itN5/ObeBrXwZR6IDeOX+xvNlAz43fuLWCQDA3z67Cl7zlM3o1zxmjOFTz0idnIePT8DnkZpr3noNNwqMsY8zxqjB6zJj7CRjbIgxtlhz/9OMsR9gjAUZYwcZY582Ok3NyOfzbe9ZShbwxPwWvC7CW2/e1fLed9wuVYzHLmwiV6oYkkaj0aJZJL5yfgOlKsO9M2EciASb3jc64MUbj0puvX84G1Pf7zW93WKk3ipj+Odz0pYK77h9smXY6RuOjmMkII0WXljLGpYGLfRrHj+7ksFCIo+xAS9ef2RMfZ+3XqvnFCxFS3zzYxc2wQA8cGgUY4Otdyc8NDaAmyYGsV2q4t8uJgxKpbH0Ukx3pcrw5fOS//RHj113Fsh1vPlGySh889IWskXJKPeSXiMwUu/ppTRW00VMDPlw997WUXUBjwtvkBuur7xk7l5E/ZrHX5YN8huOjMHn3mmqeeu1tVFoty95pcrw9ZfjAIAHayx1K5RJ6C+dE3OTrl7ae/6pKylsZEvYE/Lj+O7WsdUAsDccwPGpIRTKVTw+J+VbL+k1AiP1/otcht94dAxuV/vFaUod+fb8FtKFtjvcGEY/5nEiV8J3FpJwEfCGo9e2Pc55ChypXd/QiFNLKWxul7A75MfNE9rCv+6fGcGgz425zRyubIk3rG2nWSSUhv0NR8euiXhpxRvlCvSvsjHvJb1GYJTeVL6MJxelRun1N2jrEE2F/Lh99zCKFYZvXdpq/wGD6Mc8/ubFBMpVacV+/Twmb722NgrBYHMfNQB8Q3YBve5wRPMyfp/HhXunpQiZb10Sz4XUTrMo5EoVPLmYBAA8cFB7tMUr94fh97jwUmwba+liz+g1CqP0fnchiQoDbts9jMiA9kNdTsxKefXtefPKfj/m8bfnJaP6g7PXl33eem1tFBKJ5gW3VKnie3J43WsOjuj63gcOSRn5jYsJ0yMx2tFKs0g8dSWFQoXh2K4B7BpqHPHViKDXjVfuk/zfJ+cTPaPXKIzSqzTqrz6gr+zfMx2Gx0V4biWDRM6c0Ox+y+ONbBFn17Lwugmv2Hd9CDZvvbY2CmNjzYfFZ1YyyBYrmB4NYE9Y33Dttt3DCAc8uJIsSPuTCEQrzSKh9JRefUB/TPb9shH/1vxWz+g1CiP0pvJlPLOUhouAe2f0GYVhvwe37x5GlQHfuZzsOi1a6Lc8fkJ+bnfvDWGgwe4IvPXa2iik0+mm174rZ4ziCtKDx0VqTPdTV8xfzNOKVppFoViuqs/tfp09VQD4gX2SC+l8bBuLMfN82yJgRP5+b1FyHR2fkjo3enm1bJS/u2DOs++FMq2HJ9QOUeOyz1uvrY1Cs8Mqqozhu4tSxtyns6ekcLfswnjyijm9Ja30woEkz69mUChXcTAS1OU6Ugh4XLhtSopWemZlu83d/YUR+fv9q5JBvqeDDhEA/IBc9s+sZJAvV7tOTzt6oUxrJVus4IW1DFy08xzr6cVDdnqGZvG+8/Ec4ttljA96cWiss0mdO/eE4CbgxbWsqeF57eiFmG6lUbq7SaXQgvLZuYy9ini3+VupMpxaknqid+/tbKO50aAXN4wPoFRheG6Ffy++F8q0Vp5ZTqPKgBt3DWLI33iU5qxT4EizeF+lUty5Z7jjw0MGfW7cPDmEKgNOXRVneNsLMd1PX+2uUQJ2jMLp5QzKVbEm+3nSbf5e2NhGulDB1LCv4y2ZAeAuOe+evsK/7PdCmdaK0iG6s8ViQWedAkeahXadlo3CHXu625JXGf49fVWceQXRw/fW0kUsbuUx4HV1tTXw1LAfe8N+5MoML5q87YKVdJu/SqN0195QV6ep3W1i2Re9TGuFMaZ2IFutIHdCUjni813vry6Wq3hhVTpZ7fbd3RkFxaicWUkLE5raSLNIfH9JakRu2z0Mr7u74qk0TKeWxDHKvOk2f2uNQjccjQ5i2O/GcqqAlVShq+9qh+hlWitXkgWsZYoIBzyYHW/e8PPWa2ujkExePwn8wloGxQrD7FgQI0Hti3YacSASxLDfjfVMCatpMSbDGmkWiTPLO667brltSvqO51cybe7sH7rJ32yxgvOxbbgJOD7VfluRVrhdhFsnpe84w/n5i16mtfJMjYei1Qp+3nptbRTGx8eve88o1xEAuMi8iqGVRppFgTGG51clV8+tXTZKAHDz5CAIwPnYtilRMCLQTf6+uJZFlQGHxwcaxsfrRcnD51b5ln2Ry7QenpefU7uyz1uvrY1Cw5GCgY0SABzfveNCEgGRe1Ur6SI2t0sIBzzYP9L9/i7Dfg/2hTwoVRnOrdtjXqGb/NXaKGlF+R7eIzWRy7RWGGOq2/qWydbP3xkpcKRUunYZfqFcxYWNbRCAmyYMMgpyxTiznBFiXqFes0gojdLNE4NdTXLWcsOI1OO1iwupm/x9XmOjpJWZ0SCGfG6sZYpY4+g+FblMa2U5VUA8V0Y44MG+cOuoL956bW0U6uN9z8e2Ua4yHIgEMWjA8BkApkcDCAc82NguYTll/byCyDHdSsN9i0E9VQD4gQPSOQyiuO9402n+FspVnI8pHSJjDoR3uwg3T0rf9dwqv5GyyGVaK8/JHopbJtt3iJx1Chypj/c9uyb3VCeNqRSANK+gVLIX161vmESO6X7O4J4qAOyiDAjAS+tZFG0wr9Bp/p5bz6JcZTg4Fmy6aKoTlDm15zgaZZHLtFa0uo4AZ50CVwYHr238lfkEo1xHCjfK8fbn1qzfcqFesyisZ4pYTRcx4HXhYItjN/Wya2QY+0cDKFUZLsb78yzfWjrNX3U+wUCDDOyM+s5yXCsiapnWg+o61fD8eeu1tVFwu3dcRJUqw4vyZKSRIwUAOKaOFKyf7KzVLBJKo3HjxKCmU7604na7cSwqG2UBnj9vOs1fZYHfTQaX/YORILxuwtVkgdt2L6KWaa1sZPV1iHjrtbVRSKV2FjUtJPLIFiuYGPJdd9JRt9wwPgA3AZcTOWzLZwdbRa1mkXgpJhuFLlYxNyKVSuHYrgEA9jAKneRvlTGcj0mj2G5WkTfC63bh8Jj0/JXfMBpRy7RWzq1Lz+XoLm0dIt56bW0UotGdw+CVXvyNBk2y1eL3uHBobABVBpzfsNaFVKtZJM7XVAwjiUaj6ne+tG69+443neTvcqqATLGCyIAH4zpOWdPKUc5GWdQyrZXzcofoaHRA0/289draKMTjcfX/ejNGL8fUeQVre6u1mkWhXGWY25Qa7BvGjX3+8Xgc+0cCGPC6sJaR1kH0M53kr2Isj0aNCwWuRS37nIyCiGVaD8oI6khUW4eIt15bG4XadQMXdGaMXm6cEMOFIcJaiXrm4zkUKwx7Qn6EOjjUpRWMMbhdhCOysX+pz11IneSv4rpTevRGoxiF87FtLuVPxDKtlUqV4cKGYpS1PX/eem1tFJRhWK5UwcJWHi4CDnZ4fkI7jtb0lqwsxCIOtXd6SsY3SoreHRdSfxuFTvJXb09VL9FBLyJBD9KFCpY4bI4nYpnWypVkHrlSFRNDPoxqdN057iOOrK2tAQAubuZQZdIKzICHzyOZHPJhJOBBqlDBasa6RWyKZpFQGmqj5xOAHb1Kb/XFPp9X0Ju/xXIVFzdzIBjvulMgoms6RUYjYpnWSicdIt56bW0UhoakmGBl+Majp6pAROp2uC9bONmsaBYJniMFRa/y3XOb26j2sLuhHXrz92I8h3KVYf9IwLBV/I1QXFM8IpBELNNaUQIs9JR93nptbRQUlIJ6mFNPSUH5/rmN/l9EpZVssYLFrTw8LsIhAxet1TMa9GJ80ItcqYplzvv79xI7ozTOZX/MKfuNUOZzeLnuOsHWRiGTkVYRXuDYU61FMQpWjhQUzaLw8sY2GKRFTj4OrrtavUrD9HIfN0x681cpi7xcRwqz8vdfjOdQMfh4VNHKtFYK5Srm4zm4CDjc4lCdenjrtbVRmJiYQKZQxlKqAK+bcIBjTxWobZT4RGFoYWJiwpLfbcbcptRA66kUeqjVq7jv5ixeK8ITvfmrPP9ZzkYhHPBg15AXhXIVS0ljR2qilWmtzMdzqDBg30gAQa921x1vvbY2CrFYTO01HooE4TFwe4VG7BryIuR3I1WoIJa1Jl4+FotZ8rvNuCSvTzg0xqdRqtU7qxjlzf41Cnryt1iuYlGOuuPdIQL4PX/RyrRWlL24ZnVGPPLWa2ujQETqJDPv+QTl95TfuWBRb5XH4qRuUHqqhziFAtfqPayOFHI9HdveCj35ezmRR5UBe8MBblF3tcyqc2rGln3RyrRWLtZ0SPXAW6+tjUIkEsGlDq11p/CqGFqJRCKW/G4jzOip1uodG/BiJOBBpmhtWDBP9OTvnDpKM6fsH5Z/R+kIGIVIZVoPF+Py89fZIeWt19ZGIRaL4ZLaU+U/UgB2eqtWTXaKNNQ2o6daq7c2LLhfo2D05O9FzqO0emZrAi2MDAsWqUxrpVJluBTPA9A/UnDcRxwJDA7hSlLqqU6Pdn8msBasnmwOhUKm/2YzLprQU63XuxMa2Z/zCnryVzEKZo2Sxwa8GA16sF2qYtXA4zlFKtNaWUoVUChXER306t7ahbdeWxuFhURB7an6TfCpAsDksA9DPje28mVLNmerVKzdursW3vMJwPV61d5qn042a83fSs2hQ2aNkoGdyWYjjbJIZVorOwZZ/7PnrdfWRuFCTIr3PRgxZ5QASC4M5SCNeXn4aCbZrDh7/5jRU63Xq/xWv7qPtObvstxTHR/0ImzwJoStUFf1GzivIFKZ1sqlLkbJvPXa2ijEK34A/DbBa8YB2QjNJ8xvmEQ55FzyqfLvqdbrnRj2YcDrwla+jESu/7bR1pq/cya7jhSURnDewKNRRSnTelCefydtD2+9tjYKF9akE4yMPBNYCwcixlcMrYhyyPlyqoC8CT3Ver0uIsyMSs//sgUjNd5ozV9lPqcT90U3HBg1vuyLUqa1whjryijz1mtbo8AYw5WM5Jsz06cKWGsUvF7jT9bqBDXyhbNBbqR3Rh6pXbLg+fNGa/6aMZ/TiN0hP3xuQixbMuzMZlHKtFbiuTKS+TIGfW5MDOk/+pe3XsONAhG5ieiPiShGRGki+jwRjTe59wEiYkSUqXl91+g0NWI9U0KuzBAOeBAJmudTBYCZ0QAIwOJWAaVK1dTfDofDpv5eM3YmOfk2So30KiPDyxa473ijNX+VDonZrlO3i9RIv8sJY0ZqopRprahRd5FgRwvReOvlMVL4dQA/BuAVAPbK732qxf0VxthQzetVHNJ0HUov8WCHGdMNQa8bUyE/ylWGqwbvA9OOjY0NU3+vGZdNapQa6Z0ZtW6inzda8ncrV0IiV0bQ68JkBz3VbjHahSRKmdbKpS7LPm+9PLrIPwvgdxljlwCAiH4NwBwRzTDGLhv9Y+VyGXNzc+rfkUhE04o/s3qqzTgYCWA5VcB8PGfKvjMKovSqlF6i0kDzopFeZaL/ckLasdPNec8rM9GSvwvqsw9YskXETMTYOR1RyrRWFN0HOlwbxVuvoUaBiMIA9gM4pbzHGLtIRCkAtwK43OBjbiK6AsArf+6/MsbOaP3N9fV1PPTQQ+rfjzzyCB599FGsrq5icHAQbrcbqVQK0WgU8XgcjDFEo1G8cEWytrt8FSwsLGBiYgKxWAxEhEgkglgshlAohEqlgmw2i8nJSayursLr9SIcDmNjYwPhcBjFYhG5XE697vP5MDw8jM3NTYyOjiKXyyGfz6vXA4EAgsEgRl3SCOH5Kxs45E2r14PBIHw+H5LJJMbHx5FMJlEqldTrrTStra2pB3BkMpmGmqrVKsrlMhdNiUQCY2NjSKfTKBaLTTWtbiawlinC4yKUE6vYKA91palVPi0vLwPANZriq6uIBFyI56v4/kvzOLZ3vGtNRudTp2WvVCohmUy2zKfTF1eksu9nWFhYMF3TYEkK8HhpdQvLy+6uy14+n0cymeyZfJqLpQEA/kISqZRPd9krlaSouW41NYOMXFVLRPsALAI4yBibr3l/AcAHGWOfrrt/EsAEgLMAhgB8ANJI4xbG2LKW3zx58iQLBHYsrtaRwnv+z1ksp4r4y7ccNbWnrvDE/BZ+9/F53L03hN978JBpv7uwsIDp6WnTfq8R59az+MV/uoCDkSA+8pajXH+rmd4PfvUinr6awm+dOID7D4xwTYOZaMnfDz+xiC+9tImff+Ue/PjNu0xK2Q6J7RJ+4m9fwIDXhS++69auRysilGmtVKoMb/7EGZQqDF98160dnXZnlN7Tp0+fOnHixF317xs9p5CW/60f34wASNXfzBhbZYydYYyVGWNbjLHfABAH8AatP+jxeDA7O6u+tBgExhh+6PAYXrV/GPtGzFu4VotVEUgixHRfrnFf8KaZXmXBohURYDzRkr9mue6aMToghSFvl6pYz3S/VkSEMq2V5VQBpQrDriFvx8ef9tQ6BcbYFqSRwh3Ke0R0EEAIwHMav6YKgKujk4jwztsn8dNHvNzPUGjGVMgHv8eFje0SUnljQvO0IEJMtxL1M2PCSvJmemcsDAvmSbv8ZYyZapSbYeQCThHKtFaMMMi9uE7howA+QEQHiCgE4A8BPNZokpmIfpCIZonIRURDRPQoJHfSYxzSdR0+n/mRFwouInWiyczQSCs1KygTbWb0VJvpVSNgDAqLFIV2+buxXUK2WEE44MGIyaHYtRgZgSRCmdaK2iHqwiDz1svDKPwBgH8G8DSAJQBuAO8EACJ6BxHVHjB6HMDjkNxOlwC8EsAPM8aucEjXdQwPD5vxM005YMEeSFZrBoAFuWKYsTNtM737RvxwE7CSKiBX6r0N1ZrRLn93DLI1kUcKRrpPRSjTWlkwYKTAW6/hRoExVmGMvZ8xNs4YG2aMvYUxtiFf+wxjbKjm3v/OGJtmjA0yxnYxxh5kjD1tdJqasbm5adZPNWS/PJ+xsGWeUbBaczJfRjxXRsDj6mg1p16a6fW6Xdg3EgDDTkXtB9rlrxE9VSPYcR91/+ytLtN6MMJ1x1uvbbe5AIDR0VFLf1/pKS+a2ChZrbl2lOAyoafaSq/SWzXTKPOmXf4qjdK0RZPMCtOjQRCAq1v5rlf1W12mtVKsVHFVPr+lmwAX3nptbRRyOWsnGRWjYGajZLVmsyc5W+lVR2p9NFJol7/KSKHThVNGEfC4MDnsQ4VJETndYHWZ1srVLen8lqlhf1fnt/DWa2ujkM9b2xiMD3gx4HUhmS9jy6RtnK3WbOYkM9Ba77RsFBb7aKTQSm+lytRRqVknDbbCKPep1WVaK0a57njrtbVRsDq+mYjUimFWw2S15stb5vq0W+ndP9p/I4VWelfTBRQqDOODXgz5rYs8UjDKfWp1mdaKOsnc5WLZnlqn0GuIEN88bXLDZKVmxphhFUMrrfTuDvnhcRHWMsW+iUBqpXdegPUJtRg1UhChHmvBKNdpL65T6Blqt8ewCrNHClZqjm+XkS5UMOx3m7ZdeSu9HhdhT1g6fe+KybvV8qKVXqtXMtejlP0rXZZ9EeqxFi4bFIrNW6+tjUIwaH3lMHpv+XZYqXle9amat115O73qvEKfuJBa6V0QJBxVQTUKyQIq1c73YBOhHrcjV6pgJS1tArkn5O/qu3jrtbVRSCQSVicB0yNSBps1UrBSs6Jx2sT9ptrptWKtCE9a6VV65Pst2u+rngGfG9FBL0oVhtV05yM1EepxO65sSfr2hP3wurtrdnnrtbVRGBsbszoJiA55EfC4kMiVTdkDyUrNSqO0b6S7npIe2und32cjhWZ6KzUHOlm1CWQjjAjLFqEet8PIDhFvvbY2Cul0uv1NnHGZHIFkpebFLfMbpXZ6rVgrwpNmetczRRQrDGMDne/OyYN9BqwVEaEet2OnQ9R92eet19ZGoVgsWp0EADuhkWbMK1ip2Qr3RTu9e8J+uEgK1yyWzT0vmwfN9C6qz968UZoWpg2YbBalHrfiStK4589br62NgijxzTMmjhSs0pzKl7GVl/Y8Gh/0mva77fT63C7sDvlRZTD9vGweNNNrZE/VSKYNmNMRpR63Qh0lh7t//s46BY6IEt9s5iIqqzQrPaW9Yb8pex4paNHbT5PNzfQqjZIok8wK+9QOUQHVDk+BFKUeN6NcZepWHnsNeP7OOgWOiBLKZuZ2C1ZpvmJRo6RFbz9td9FMr2KURRsphAIejAY9KJSrWM905hYRpR43YzVdQLnKMDHkQ6CLPY8UnJBUjohyOMeuIR/8bsLmdgmZAt8IJKs0L1rkvtCit5+2u2iklzG2M6dggPvCaLoNtBClHjdj0eCou148ZKdnSCaTVicBAOB20U4UBufeqlWarYqR16K3n0YKjfRu5aWV5ANeFyID1u95VE+3eyCJUo+bccXgqDveem1tFMbHx61Ogso+NQqD72SnVZp33BfmRr9o0bt3JAACsJTsfm9/q2mkt7ZRsvK0tWZ0O6cjUj1uhDpSMGiUxltDqjh+AAAgAElEQVSvrY2CSD2MfQaE5mnBCs3FchWr6SJcJG1CZyZa9Bq5t7/VNNK7KNhK5nq6HamJVI8bYfQo2RkpcKRUMucMAy3sVzdm42sUrNC8lNo5XMTX5RJ/vWjVa9ZIjTeN9Fo1StNK7WFHrIMIJJHqcT2MMXWzRaOeP2+9tjYKIsU3m9UoWaHZyj13tOrdZ5JR5k0jvaLteVTPSNCDYb8b26Uq4jn9gRYi1eN64rkyskVpZ+CRgDHzOc46BY6IFN+8J+QHAVhJF7j6ta3QvGhwT0kPWvWa5b7jTSO9oruPiEj1t3fy/EWqx/VcqZlPMGo+x1mnwJHBwUGrk6Dik/3aVQaspPgtY7dCs5WrabXq3as0Sj2+qrleb65UwXqmBI+LMDUspvsI2OkwdGIURKrH9fDYBJK3XlsbBbdbnI3BgJrVnRxdGFZottJ9oVVvbaPUiV9bFOr1Klt37An54XaJF3mkoIwUOtlqRLR6XAuPleS89draKKRSKauTcA17w533lrRituYqYzVDaPN7qlr1jgS682uLQr1eK7Yr74S9I53P6YhWj2vhsZKct15bG4VoNGp1Eq5B9WtzdGGYrTmWKaFQYYgEPZYcFq9Vb61f+2oPzyvU67VqJbleduYU9Jd90epxLTzmc3jrtbVRiMfjVifhGrqZbNOK2Zqt3nNHj969fXBec71eUTfCq2cq5IebpHMf8jq3MBetHitsFyvYyJbgdREmhozbmoK3XlsbBdF8x2b4tc3WbHVPVY/efohAqtdrtVHWisdFmAr5wSCtLNeDaPVY4Wpq5whOI+dzeOu1tVEQbdhphl/bbM1WzicA+vTu68KvLQq1eitVhiUlHNii56+HTtfqiFaPFZS9nIwepTnuI46sra1ZnYRrMMOvbbZmK47grEWP3m782qJQq1fZsjk66EXQK26EjkKnq/pFq8cKvEZpvPXa2igMDQ1ZnYTr2Omt8mmYzNZs9WpaPXq78WuLQq3eXplPUFAaT71hqSLWY6C27Bs7SuOt19ZGQUTMmGw2i9ojOKMmHsHZKd34tUVE9JXM9ezto7IP1OxOK+AZFq2wtVHIZDJWJ+E6uonX1oKZmms3YrNqy2a9ejvtrYpCrV5Rz2VuRm30l56jOUWsx5Uqw1LNRLOR8NZra6MwMTFhdRKug7df20zNIvSU9OrdZ8ICQp7U6l3k5L7gRSjgwUhAOppzI6t9J1AR6/FKzRGcRs/n8NZra6MQi8WsTsJ1KH7tNU5+bTM1i+C+0KvXjAWEPFH0XrNlcw+5L/Z2sAeSiPV452Aj4w0yb722NgoinkLlcZF6EA0Pv7aZmkVwX+jV2+tzOorehLxl85DPjZGgeEdwNmNfBxsTiliPea7P4a3X1kYhEolYnYSGqBvjcXAhmalZmVOw0n2hV2+nfm1RUPReFWA+pxN25nS0G2UR63HtltlGw1uvrY2CiMNOoLOKoRWzNBcrO0dwTpl8BGctevWGAh6EO/Bri4KiV+lQ7O0h1xHQ2ZyOiPWYZ4fIcR9xJBQKWZ2EhvCc7DRL81LSuiM4a+lEbzd7+1uNolf0Izib0cmqZtHqMWOsZk7BeKPMW6+tjUKlUrE6CQ3hOdlplmZRGqVO9Hazt7/VKHqvChD51QkTQz54XYSN7RK2i9ryTrR6vJUrI6PM5xh0BGctvPUabhSIyE1Ef0xEMSJKE9HniWi8xf0PEtFZIsoR0QtE9Dqj09SMbDZr1k/pQvFrX93KG+7XNkuzKI1SJ3p3jHLvjRQUvb2yEV49bhdht1L+U9qMsmj1mPf6HN56eYwUfh3AjwF4BYC98nufanQjER0E8AUAHwIQlv/9IhHNcEjXdYh64Pew34PRoAeFCkMsY6xf2yzNSsXYa3Gj1IneXl6rMDk5iUK5ijVlPmfYuC2bzUJvBJho9Zh3KDBvvTyMws8C+EPG2CXGWBLArwF4sElD/24Apxhjn2aMFRljnwFwWn5fE+VyGXNzc+pLz17jIh/4vROaZ2zDZJZmxae63+LdOTvR2+lunSKwurqKpWQBDMDukB9eC+dzOkXvnI5o9Zh3KDZvvYY6vIgoDGA/gFPKe4yxi0SUAnArgMt1Hzlee6/Mafl9Tayvr+Ohhx5S/37kkUfw6KOPYnV1FYODg3C73UilUohGo4jH42CMIRqNYm1tDcViEZubm8hkMpiYmEAsFgMRIRKJIBaLIRQKoVKpIJvNYnJyEqurq/B6vQiHw9jY2EA4HEaxWEQul1Ov+3w+DA8PY3NzE6Ojo8jlcsjn8+r1QCCAYDCIRCKBsbExpNNpFItF9XowGITP58OIWxohvLCwjmgloV5vp0nZLKuZplKphI2NDa6aUqkUFhLbUoakY1h3bcPn8yGZTGJ8fBzJZBKlUskwTa3yKZ1OY2trS5emQnwFHhewsV3C+YuXsWdivGk+WaGpVT5Vq1U8c/EqAGBXgLCwsKC77FmtaaAsbeNwfiWBlRVqW/YqlQoWFhaE0TS3JqXfX0xha8tveBtRKBSwtbXVtaZmkJEHNhDRPgCLAA4yxuZr3l8A8EHG2Kfr7n8cwBOMsd+uee93ANzLGPshLb958uRJFgjsWORIJKI5jjebzWJwcFDTvWbzhRfW8ZHvLeFHjo3jfffuM+x7zdC8kS3iJz97FuGAB3/3zlu4/lY7OtX7M58/h4VEHv//fziCG8YHOKSMD9lsFl+8kMYnT63gbbfsws+8Yo/VSdLN+VgW/+UfL+DAaAB/+dZjbe8XrR6/63+fxWq6iL9+6BiX1fxG6T19+vSpEydO3FX/vtFjy7T8b7ju/REAjU6bTuu4tyEejwezs7PqS8/Cjo2NDc33mg2vlbVmaL4i0MEunert1ZXNGxsbQqwk7wZlbcXVVAGVavtOq0j1WJnPcRPUnQmMhrdeQ40CY2wL0kjhDuU9eTI5BOC5Bh85U3uvzO3y+9wJh+vtkTjwipU3Q7OSZhEWTnWqt1fXKoTD4WtWM/cigz43xga8KFUY1rPFtveLVI+XU9J8zlTID4+BR3DWwlsvj1mojwL4ABEdIKIQgD8E8Bhj7HKDez8J4C4iepiIvET0MIA7AXyCQ7quo1hsX+CsYteQDz43IZ4rI1Mw7mhOMzTz3AxML53q7dW1CoVCQYjdabtlr44IMJHqMc/tLRR46+VhFP4AwD8DeBrAEgA3gHcCABG9g4jUzcAZYxcBvAXAb0JyGf0mgB9vYkAMJ5fLmfEzHeEi2jl0xMCGyQzNIsXId6q3V0cKy1vbyJerCAc8CHFYOGUWeiLARKrHi0n+HSLeeg03CoyxCmPs/YyxccbYMGPsLYyxDfnaZxhjQ3X3f5UxdhNjLCj/+zWj09QM0eKb69nPoWEyQ7MZvSWtdKpXr19bFIp+aQsEEeZzukFJv5b9v0Sqx2bM5/TiOoWeQbT45np2ekvGGQXemnOlCmLZEjwuwqQAC6c61avXry0K565Kk5AijNK6Qc9IQaR6bEaHiLdeWxsFn8/6RqsVSsFaNNB9xFvzkpzWPSE/3Jwm2vTQjd5edCGty56FvT0/UtC+eFOUeswYU+egeD5/3nptbRSGh4etTkJL9nMYKfDWLMpGeArd6FWNcg+tbF7PS66uXh8pRIe88LsJCQ2BFqLU443tkinzObz12toobG5uWp2EluwJ+0GQwtxKFWOO5uStWbTIl2708nDf8eZqDx7B2QgXkbpvVrtAC1Hq8c58At8OEW+9tjYKo6OjViehJX6PCxPDPlQZsJIyxq/NW7O6RkGQkUI3enlM9PMkV6ognq/CK8h8TrdoDUsVpR6b1SHirdfWRkGkULZm7LgwjGmYeGtW3UeC9FS70bt/xNhnzxtllLBbkPmcbtF6XrMo9Xin7PPtEPVcSGovkc+LX9nV3qpBu6Xy1FytmWgTxafdjd6xAS8GvC6kChUk88YtIOSFWe4Ls9DqvhOlHvM8ba0W3nptbRREim9uhtF+bZ6a1zNFFCsMkQEPBn1ubr+jh270EpH6/HthtNAv8wkKSoeo3bMXpR6btWjTWafAEZHim5th9NGcPDWLNskMdK+3l4yCaPM53bI3HNAUaCFCPd4uVrCRLcHrIkwM8Z3PcdYpcKR2y21RqQ1LNWKbc56aRZtPALrX20unsIn4/LvB73FhUg60WG5xNKcI9Vg5OnR3mP98Dm+9tjYKwWDQ6iS0JRzwIOR3Y7tUxeZ290dz8tQsok+7W729Mtks4nyOEew8/+ZGQYR6bObWLrz12tooJBIJq5OgCSOPh+SpWcRGqVu9vXI0pzKfE/aRMPM5RqDFfSdCPb5qwkZ4Crz12toojI2NWZ0ETRjZW+WpWaSN8BS61bs75IebgLVMEfmyMQsIeWBW5IvZaDEKItRjM8s+b722NgrpdLr9TQKg+rUNCEvlpTlTKCOeK8PvJkSHvFx+oxO61etxEfYoO6YK7EJSyka0v2yCpgWEItRjJX08jt+sh7deWxsFkQ7naIWRYam8NCvRUXvCAbhInIVTRug10ijz4qo8UtgVFOfZG0FtoEW1SaCF1fW4UmXqRPMeEzYi7MVDdnoGUeKb27FPw2SbVnhpFnGSGTBGr5bJTqtRDNaxPeMWp8RYhv0ejAY9KFQY1jONG0Or6/F6pohShWFswGvKfI6zToEjIsQ3a2FiyAevm7C5XUK2WOnqu3hpFnXhlBF6e2GtgmKU/cWkxSkxnnZzalbXY7N3BnbWKXBEhFA2LbhdhL0h7SdRtYKXZlFHCkbo5bGFuZGka+Zz9owOWp0cw2k3Ura6Hpu9aNMJSeWIKIdzaMGo3iovzYsCRh4BxuhVDN1SUsyjOdVnPxJAwC+WUTaCdkbZ6nqsPH+zDjZyDtnhSDLZO0Pt/QbFy/PQXKpUsZQqgCBeSKQReoNeN8YHvShVGVbT4gUnLCakRml6NNBTZVor7SKQrNa8ID//mVFzRiy89draKIyP986knFFHQ/LQvJQqoMqAqZAPfo9YRcoovSKvbF6oCYfspTKtlXbP3krNjDE1XftHzekQ8dYrVg02Gat7GHrQurd8O3hoVnqqZsRo68UovSLPKyzWGIVeKtNaqd3CfCt3/VYvVmqO58rIFCsY8rkRCfI7grMWZ6TAkVKp+72EzEKJf15K5lHuwq/NQ7PSU50W0CgYpVfktQqKUZgeDfRUmdbKtVuYX98pslJzreuOTFqfw1uvrY2C1fHNegh63ZgY8qHCgJUWO0a2g4dmxadq1vBZD0bpFdV9tF2sYD0jbdk8NezvqTKth1bP30rNCyauZFZw1ilwxOr4Zr3sM+AUNh6ad0YK4oX4GqW3dmM8I7YwNwqlLOyVt2zutTKtlVZzalZqrh0pmIWzToEjg4O9FdNtxHnNRmsuVxmWTNwhUi9G6R0NejDkcyNTrCCRE+dozvpRWq+Vaa20GilYqdmKkQJvvbY2Cm53b20xbMR2F0ZrXk4VUK4yTAz5EPSK9zyN0ktEQrqQFuvmc3qtTGul1bO3UnPtfI5Z8NZra6OQSqWsToIujIiAMVqzyJFHgLF6jQoLNpL6kUKvlWmtTA374XURYtkScqVrt3qxSvNWroRkvowBrwvjA+btDMxbr62NQjQatToJupiRK/7lRPMdI9thtOYFC3pKejBSr5EbExpF/Uih18q0Vtwuwm71aNRrn79VmmtDgc2KPAL467W1UYjH41YnQRehgAeRoAeFchVrHa6sNVqzFcNnPRipV2l4F7Zyhn1nN+TLVaymi3CRdBgQ0HtlWg/NXEhWaV6wYJIZ4K/X1kZBpCgSrUzLS+kvJzpzYRiteUFw95GRepVtDC7HxXAfXd3KgwHYE/LD65aqci+Waa00MwpWaV60YJIZ4K/X1kahF4faMxHFhdRZb9VIzZUqU0MiRTUKRurdNSStrN3Kl5FosLLWbBqN0nqxTGtFHanVdYis0myV69RxH3FkbW3N6iToZmZkZ16hE4zUvJqWDhcZHzTncJFOMFIvEamjhfqGyQoahUP2YpnWitIhmq/rEFml2aogC956bW0UhoaGrE6CbmYi3TVKRmqun+QUEaPzeHq0O6NsJI0WTvVimdbK3nAAHhdhNV28JgLJCs2pvHyGhceFXUPmbt3NW6+tjUIvUhuWavXe/sqEq4jbW/BCiQCbj1s/2WzFwikr8bhI3YPKaqN8RX32fqHOJDcCWxuFTCZjdRJ0M+iT9kAqVRmWOtgDyUjNok8yA8bncbcjNaMolKtYThXgomvPsOjFMq0H5fnXGgUrNF+20CDz1mtrozAxMWF1EjpCcRd00jAZqXlejsI5YNLhIp1gdB7vrBXJWRrps7CVR5VJLhWfe6ca92qZ1or6/GtGalZoVn7firLPW6+tjUIsFrM6CR1R2zDpxSjN5SpTh9AzAruPjM7j0aAX4YAH26UqYlnrIpB2GqVrn32vlmmtHFBHCjtl3wrNl5QOUcR8o8Bbr62NgpmrEI1kpou1CkZpvprMo1RlmBz2YUDQyCOATx53Y5SN4pJiFOoapV4t01rZmdPZKftma2aMqXl/IGJ+h4i3XkONAhENENHfEFGCiLaI6GNE1NSUEtF7iKhKRJma12eNTFMrIpGIWT9lKN24j4zS3AuuI4BPHouwiG2+SU+1V8u0VqSNF69dK2K25o3tEtKFCob9boyZuOeRAm+9Ro8UPgzgqPy6AcAxAH/a5jOXGGNDNa+HDU5TU3p1qL1/JACC1FsvVqq6PmuU5vm4dT0lPfDI4wNdLiA0AuW3ZyL2ch9Ja0WuDQs2W/N8zXyCFSOznnEfySOCdwL4LcbYGmNsHcBvAXg3EXFrOcrlMubm5tSXnn1BQqEQr2Rxxe9xYXfIjyqDepaBVozSPN/EfSEaPPJYGSnMWxSBlMiVkMhJu3NO1MXI92qZ1sPOSE0qg2ZrbjZKMwveeo08afoIgACAUzXvnQYQhDRqeK7J5/YR0SqAEoDvAPgNxti81h9dX1/HQw89pP79yCOP4NFHH8Xq6ioGBwfhdruRSqUQjUYRj8fBGEM0GsXa2hqq1Sqq1SoymQwmJiYQi8VARIhEIojFYgiFQqhUKshms5icnMTq6iq8Xi/C4TA2NjYQDodRLBaRy+XU6z6fD8PDw9jc3MTo6ChyuRzy+bx6PRAIIBgMIpFIYGxsDOl0GsViUb0eDAbh8/mQTCYxPj6OZDKJUqmkXlc0Rf1VLAE4NbcEz5Rf1aQsbGmmiTFmiKaXY1JYnD+fwNpa0RBNzfKpnaZW+bS6ugoiMjSfXBmpp7aYyOPS/GVM7IqaqunJ8ysAgD1DHlxZXLxGU6VSQSqV4lr2eGjSU/bCkIzBC1c28KooQ6FQQCqVMk3T2atpAECIZZFIJExvI8rlMoioa03NIC1hdUT0cQDvbnHL7wF4DMC3AbiY/KVE5AJQAXA/Y+yJBt97EJJhmgOwC8AfALgPwHHGWLZtwgCcPHmSBQI7A5FIJKLZ57awsIDp6WlN94rGx7+/jL99dg0/cXwC7717t+bPGaE5UyjjLZ96Hl434Z/efRxul7iTm7zy+J2fewHrmRI+9tCxa9YJmMEXXljHR763hDceHcMv3bf/mmu9XKa18uxyGr/25Tkc2zWAD7/5iOmaH/n8Ocwn8vjwm2/AsV3mn/pmlN7Tp0+fOnHixF3172sdKfxnAO9vcX0b0mgAAMIAtmr+DwANT4VgjF2q+XOViH4GQBLAKwE8riVhHo8Hs7OzWm69jl4+5PzQ2AAA4NKmPr+2EZoVX+70SEBogwDwy+OZ0SDWMyXMJ3KmG4X5FjHyvVymtVJ/roiZmkuVKq7ILlurQrF569U0p8AYyzDGNlq8tgGcB5AHcEfNR28HkANwQWN6mPwypaXp5UPOD41JDcLFzW1dnzNCsxIOeVDw+QSAXx4faBAaaRatfNq9XKa1MhL0YjToQa5UxVqmaKrmq0np+NmpYeuOn+Wt17CJZsZYDsCnAfwuEe0iol0AfhfAJxljDWsOEb2JiPaSRATAnwPYAPA9o9LVCq/X/HAyo5gc9mHA60I8V0Z8W/siKiM0K6GYMz1gFHjlsTJSm9vQZ5S7pVJlWGgRI9/LZVoP6mR/PGeqZhECLHjrNTok9RchjQqU13kAv6xcJKL/SkRna+5/AMBTADIAzgIYA/DDjDFTNjMJh8PtbxIUFxEOyqOFSzo2ZzNC885IQexwVIBfHs+OKyM1c8NSV9IFFCoM4wNeDPuv9/72cpnWw6xc9uc2cqZqFsEo8NZrqFFgjG0zxv4TY2xEfr1XHkEo13+fMXZTzd+/yhjbzRgbZIxNMcYeYoxpdTV1zcbGhlk/xYVDEam3qqdh6lZztXY1p+AL1wB+ebw75EfQ68LGdglbJh64o8whNWuUer1Ma0UxynOb26ZqVre3sHBrF956bb3NRa/3qjqZV+hW80qqiO1SFZEBD0YtWM2pF1557CJS51TmTBwtvCz/ltIo1tPrZVors4r7btO8kQJjDC/L7sLZ8QFTfrMRPTVS6DWKxaLVSeiKHaOgvVHqVvOcbIAOj1lXKfTAM487ef7dosxhNHv+vV6mtbIn7EfA48JGtoTNtDnPP75dxla+jEGfG1PD5h6sUwvvPLa1UcjlrD8opRumRwNwkxQRUXsSVSu61TwnQE9JDzzzWJ1s1hkB1imMMXVU0myk0OtlWiu1I7ULMU1LmrrmZTmfZ8es2d5CgXce29oo9HpMt8/twvRoAAzad0ztVrPqvhgTfz4B4JvHsyaPFGLZEpL5Mob97uu2t1Do9TKtB8Uwxpk5ZVEdpVncIRJinUK/0g8x3QfH9E02d6O51qdqdcXQCs88VkZqSzpGat0wp6Gn2g9lWivKSO3F5a02dxqDKB2inlmn0Iv4fNb5BY3iUETfZHM3mtcz0pbB4YAH0UHxJ5kBvnksjdSCYNAXFtwpcxtKo9TcIPdDmdaK0jhfyfA3yIA4rlPeeWxrozA8PGx1ErpGmezUGgHTjWZRfKp64J3HZrqQ1JFCi0apH8q0VqZHA/C4CGvZCveR2lauhFi2hIDHhT0hP9ffagfvPLa1Udjc3LQ6CV1zeHwABCl+XcvZCt1oFqWnpAfeeXyoZhEVb5TfONxkkhnojzKtldo5Nb17gOlF6XQdGgtavt8X7zy2tVEYHR21OgldM+hzY99IAKUqU1dbtqIbzUrFONwjk8wA/zxW5lYucN7uIrFdwsZ2CUGvdJZGM/qhTOtBGanxfv478znWd4h457GtjUK/hO8diUoF9XysfcXoRnMvjhR45/Hs+ABcJJ2ExtOFobjuDkWCcLVw3fVLmdbKkai0dfVLGsp+N2gZpZmFE5LKkXzeujN2jUQxCloqRqea1zNFxHNlDFm8cEcvvPM44HHhYCSIKgNe5uhCemldylslr5vRL2VaK0fVDhHftQpKh0uEqDveeWxro9AvMd1H5d7S+fX2FaNTzS/J331010DPTDID5uTxjlHm1zAp393uUJd+KdNaORAJwucmLKeKSObLXH5jc7uEtUwRA14X9pt8dkYjnHUKHOmXmO4DkQC8LsKVZAGZQuuK0anmc+vaGiXRMCOPj8rPROnNG02VMfW7j7Z5/v1SprXidhGmh6VzDXiNFpQO0ZHogOWTzICzToErtcd49jJet0uNgmk34dap5nNyo9RrRsGMPD7KeaSwlCwgU6xgbMDbdn1Iv5RpPRwekybeeRnlc+ooWYyyzzuPbW0UgkHrJ42MQplwazfZ3InmUqWqTnS282mLhhl5vG8kgAGvvDlb1vhttHdGae1dd/1UprVybJd0EP05De7TTlA6RDcKYhR457GtjUIikbA6CYahdbK5E80XN3MoVRj2hf0ND3YRGTPy2EXEdV5Bq+sI6K8yrZVdHmni9XxsG1XGDP3uSpWpG+6JMlLgnce2NgpjY2NWJ8Ewju2S94FZy4K1qBidaO7V+QTAvDw+yjE08pzGSWagv8q0VmZ3RxEZ8CBTrGApWTD0u+fjORQqDLtDfoQDYnSIeOexrY1COp22OgmGsTvkx2jQg2S+jKstKkYnmpWG7thE7xkFs/JYeTZnV409STZXqmA+noOLtIVD9lOZ1komk8Ex2Si/aLALqdZ1Jwq889jWRqGfDiQhItwyKflWn2/RMOnVzBjD2TXp+5SK10uYlcc3TQyCILkwCuX2241o5dx6FlUmba8Q8LSvrv1UprVSLBZxs1L2V4w1yi+siTdKdg7Z4Ui/xXRrMQp6Na9miljPlDDsd2Mm0nuRLWbl8bDfgwORIEpVpoYwGsFzciN3q5y37ei3Mq2FyclJ3DolPZ/nDBypMcZUI3OLxudvBs46BY70W0y3FqOgV/NzNZWi1fYKomJmHisNU6vnrxfVKExp2xmz38q0FlZXV3EwEsSgz43VdBHrGWN60supIja2SwgHPJgeFadD5KxT4Ei/he/NRAIY8rmxnilhLd24YujVrDRKx6fE6Snpwcw8VoyyUb3VQrmK87FtEICbJ7W5L/qtTGshGJR2Lr1Zntd5ziAX0nMrku/+lslBoTpETkgqR/rtQBIXkdp4NOut6tW801PtTaNgZh7fIj/7c2tZlDRsY96Oc+tZlKoMB8eCmkOB+61Ma0HRfIvBIzXFuGsdpZmFc8gOR5LJpNVJMBylt3pmpXGEgh7Nq+kC1jJFDPvdOBDpzR6omXk8EvRi/0gAhQozZCtnvfMJQH+W6XYomtWRmgEjBcYYzgg6Suadx7Y2CuPj41YnwXBu2y31ak4tpRuuV9CjWakUN/fofAJgfh4rDcipq92HDT67LH2HnlFaP5bpdiiaD48PIOh1YSlV6HpeYTlVxEZWCrAQaT4B4J/HtjYK/dirOjQWxEjAg41sCQtb12+xq0fz01dSAIA794g1fNaD2Xl8194QAOD7V1NdfU+mUMaL61m4acfQa6Efy3Q7FM0eF+F2+Vk93eXzVz5/x55h4TpEzkiBI6WS8fvUWFjXv9cAAA/sSURBVI2LCHftlSrG969cXzG0aq5UGU4tST3Vu+WGrhcxO49v2z0Er4twPrbd1VbOp5fSqDLgpokhDPrcmj/Xj2W6HbWa794nldWnG5R9PTx1RWp4RSz7vPPY1kahX2O6ld7q0w1cGFo1v7ieRbZYwd6wH1MWH1TeDWbncdDrxs2Tg2AATnXRW1V6qnft0zdK69cy3YpazUoj/sxyuuPJ/kK5qs5LiGgUnHUKHOnXmO479gyDALywmrnuiEitmpWeltLz6lWsyOO7u3QhMcZUo6C3UerXMt2KWs27hnyYHg0gV6ri7FpniwjPrKRRrDDMjgUxOtB6q3IrcNYpcGRwUJyl60YyEvTihugASjUuIAWtmp9clIbPPyBgT0kPVuTxXft2RmqVqv5dO1/ezCG+XcbYgBcHdUZ99WuZbkW9ZsWQPtWhC+kpwTtEvPPY1kbB7dbuq+017p8ZAQCcnN+65n0tmq9s5TGfyGPQ51Zjv3sVK/J4eiSAvWE/kvly09DgVpy8JG2NfO9MWPfRp/1cpptRr/me6TAAqey32jG4EZUqwxNynblXrkOiwTuPbW0UUqnuJqNE5v6DUoH+3mLymg3atGj+tlwpXjUdhs/d20XEijwmIrz6gPT8v3Vpq83d18IYw7fk5698hx76uUw3o17zTRODGB/wYi1T1L2V+dm1DOK5MqaGfTg8JubaHN553Ns1vkui0ajVSeDG1LAfh8eDyJWq1/i2tWj+ttxTfc1BMXtKerAqj19zcBQA8J3LWyjrcCG9vJnDarqI0aAHN03oH6X1c5luRr1mF5HaKfrmJX0H0ihG/NUHRnSP0syCdx7b2ijE43Grk8CV1xyQGqZvXNypGO00X07kMJ/IY8jnVmO+exmr8nhmNIB9YT9ShQqeWdLuQvrGnJTe+2ZGOjokvt/LdCMaaX5ANsrfvrSleV6nVKmqo+T75c+LCO88trVR0Otv7DVeOzsKFwHfXUgisS3FNrfT/OWXNgFIowRvj7uOAOvymIhwYjYCAPjySxuaPlMsV/H1l6UK//obOjtdq9/LdCMaaT4aHcDUsA+b2yXNUWD/vpBEMl/GzGhAWNcRwD+Pe7/Wd0G/D7Wjgz68Yl8Y5SrDYy9LjX0rzYVyFf8qN0pvOtof2yVYmccPHhmDm4B/X0xiI9t+24XvLGwhVajg0FgQh8c7a5T6vUw3opFmIlLL8L+c02aUvyR3iN54dFxY1xHguI+4sra2ZnUSuPOmY1KP80vnNlGpspaa/20ujkyxgiPRAcxqOPqxF7AyjyMDXtw3M4Iq22lwmsEYwz+cjQEA3nBkrONGyQ5lup5mml9/ZAxeN+GpKymspFqf3byQyOGZ5TR8bsIPzYrrOgL457GtjcLQUG+HW2rhzj0h7An5sZYp4vG5eFPN5SrDZ89Ihe3Hb+qf3qbVefxm+Vn+49kYMoXm2148u5zBufVthPxu/PDhSMe/Z7VeK2imORzw4LUHR8EAfPbZ1g3p38rXX3fDGIY0blNuFbzz2NZGwQ64XYR33C4ti//MM6soVRr7I792YROr6SL2hv1q5IxD99wyOYTjU0PIFCv4++fXG95TZQyfOLUCAHjrLbsQ9NpvrQEvHr5tEi4CvvbyJq4mr98gEgDm4zl861ICHhfh7ccnTE6heNjaKGQyxh7yLSqvPTSK/SMBrKSL+Nzzseuub+VK+NjTywCAn7pjqqOoF1ERIY/fc+cUAODvnlvHQiJ33fXHLsTx4noWo0EP3nxjd6M0EfSaTSvNe8J+vP6GMVQZ8OEnrqBaN0lbZQx/9p0rqDLgjUfHsGtI/EOKeOexoUaBiN5HRE8S0TYRzWn8zINEdJaIckT0AhG9zsg0tWJiwh69AreL8Ev37QMB+Orl/DXRGOUqw4e+sYB0oYLbdw/jgT5Ym1CLCHl80+QQ3nBkDKUqw4e+cRnZ4s5+VJc2c/iLf78KAHjkFXt07YjaCBH0mk07ze+9ezdGAh6cWcmobiKFT5xawQtrkkFWjLfo8M5jo0cKywD+CMDvabmZiA4C+AKADwEIy/9+kYhmDE5XQ2Kx63vN/crNk0P4j8cnUGXA73z9Ej7//DpOL6XwG1+ZwzPLaYwEPPiVV+8XOuqiE0TJ4599xR7sCflxKZ7H+7/0Mp5cTOKr5zfxq19+GflyFSdmR/HaQ9277UTRaybtNIcCHvzy/ftBAD55agV//t2reG4ljf/xxBV89tk1uAj41ddMCz+XoMA7jw19CoyxvwcAInqPxo+8G8Apxtin5b8/Q0Q/J7//O1q+oFwuY25uZ1ASiUQQiWibqOu3BrAdP33XFFY3k/jW1Tz+8skl9f2RgAf/z4OHemLorBdR8njQ58aH3nAIH/jyHC5u5vBbX7ukXrtnfxi/dJ8xBlkUvWaiRfM902H80v378T+eWMQ/vhjDP74oNaxuAn7l1dPqdvO9AO88tto0Hgdwqu690/L7mlhfX8dDDz2k/v3II4/g0UcfxerqKgYHB+F2u5FKpRCNRhGPx8EYQzQaxdraGrxeLzY3N5HJZDAxMYFYLAYiQiQSQSwWQygUQqVSQTabxeTkJFZXV+H1ehEOh7GxsYFwOIxisYhcLqde9/l8GB4exubmJkZHR5HL5ZDP59XrgUAAwWAQiUQCY2NjSKfTKBaL6vVgMAifz4dkMonx8XEkk0mUSiX1ejtNSmRCM03vvW0Ut+8u4vFLKRThwfQgwxtnh7AnyLCwsNCTmlrlUz6fx9bWljCa/vsbZ/Dppxfx8haDz0O4c9yNH7ttCusrS4aUvUAggIWFhZ7Lp27qk8/nw8LCQltNr947jvCrIvjKxQwSJRcivip+7OgoDkYICwsLQmlqlU9utxtbW1td51MzSMvqOCL6OKTeezN+jzH2mzX3vwfAbzLGZtt87+MAnmCM/XbNe78D4F7G2A+1TRiAkydPskBg5wxVPSOFhYUFTE9Pa7q3X7CbZkdv/2M3zUbpPX369KkTJ07cVf++1pHCfwbw/hbX9W1FuEMa0lxCLSMANG8D6PF4MDvb0vY0JRTqnSGjUdhNs6O3/7GbZt56NRkFxlgGAI84qDMAXlv33u0AHufwW9dRqVTa39Rn2E2zo7f/sZtm3nqNDkn1EFEAgFf6kwLy3834JIC7iOhhIvIS0cMA7gTwCSPT1YxstrPj+noZu2l29PY/dtPMW6/RIam/CSAH4KMADsr/V1frENE7iEgdcTDGLgJ4i/y5lPzvjzPGLhucrobY/ZBzO+Do7X/sppm3XkONAmPsUcYY1b9qrn+GMTZU95mvMsZuYowF5X+/ZmSaWmH3Q87tgKO3/7GbZt56bbvNRTwex0c/+lFbHUpiN82O3v7HbprN0Gtro/AXf/EXtilMgP00O3r7H7tpNkOvbY2Cg4ODg8P1aFq8JjKPP/54DMCC3s+Vy2V3PB6fiEQiax6PxxYxbXbT7Ojtf+ym2WC90ydOnLhuW96eNwoODg4ODsbhuI8cHBwcHFQco+Dg4ODgoOIYBQcHBwcHFccoODg4ODioOEbBwcHBwUHFMQoODg4ODiqOUXBwcHBwUHGMgoODg4ODimMUHBwcHBxUbGkUiMhNRH9MRDEiShPR54lo3Op0dQIRfZyISkSUqXn9Qt097yKii0S0TURPEtGdddfvIqKn5OsXieid5qpoDRG9nYhOElGKiMoNrj9IRGeJKEdELxDR6+quzxLRvxJRloiuEtGv1F0fIKK/IaIEEW0R0ceIKMhbVzNa6SWiB4iI1eX3d+vu6TW9fyjnX4qIlonor4goUndPV2WYiHYR0Rfk+h6Tf9Oy9q+dZiJ6DxFV6/L5s3XfwUczY8x2LwAfBHAB0kFAYQCfB/AVq9PVoZaPA/jrFtfvA5AF8DoAfgC/BmANQEi+HgYQA/AB+foPQzp69R6rtdVoeD2AhwH8JwDlumsHIZ0R/k4APgDvkPXOyNfdAM4B+DMAAwDuALAO4CdqvuOvAHwXwASAXfL//0JQvQ/Uv1d3vRf1/j6kY3i9AKIAvgLgH40swwC+DuAL8r0H5fr/AYE1vwfAXIvPc9NsyQOx+gVpA7331vx9CABTGpJeeqG9UfgEgE/V/E0AFgG8W/77p+W/qeaeTwH4X1Zra6DlugYRwO8AOFn33kkAvy3//7WQjMZQzfX/BuAb8v+DkE4HPFFz/YT8mYCAeq97r+56z+qtSc+bACRr/u6qDAM4INfvQzXX3wtg3mqtLTS3MwrcNNvOfUREYQD7AZxS3mPSsaApALdala4ueSsRxYnoguwWqz3d7jiu1coAPCO/r1w/Lb+vcLrmuuhco0+mNv3HAVxgjGWaXD8CIFD3HachNZ43GJ5aY3AT0RUiWiWiLxFRbV71g94TAJ6r+bvbMnwcUoN7se76DBGFDE57p9RrBoB9ch5fIaLPEdGBmmvcNNvOKABQHkiy7v2tmmu9xJ8BOApgHMCPA3gNJPeAwjBaa213XXS61Tcs/1t7j/J/EZ/BSwBug9QTPAqpIfk3ItotX+9pvUT0VgA/A+AXa942Io8bXQfE1fxtALcA2A3gbgB5AF8nokH5OjfNdjQKafnfcN37I5BGCz0FY+wUY2yNMVZljJ0F8MsAHiIiv3xLGq21trsuOt3qa1QelP8L9wwYY6uMsTOMsTJjbIsx9hsA4gDeIN/Ss3qJ6G2QOjRvZoydrrlkRB43uq5cs4xmmhljlxhjF+R6vQrJaOwG8Er5Fm6abWcUGGNbkHxxdyjvEdFBSNazfvjWi1Tlf0n+9wyu1UqQeppnaq7fXvcdt9dcF51r9MnUpv8MgBtqelj1189D6oXdUXc9B2lirheo4tr87jm9RPTTAP4SwI8yxr5Rd7nbMnwGQFiu57XXLzPG6nvTptFGcz1MftXmMx/NVk+wWDSp80FIleMAJGPwdwC+anW6OtTydgAj8v8PQ4ok+XzN9fsgRSWcgBSd835cG7kxAimK4Vfl6ycgXvSRG5If/HUAyvL/A3IFOQRpkvRhSJEcD6Nx9NGHIfnNb5P1v73m+/8KwBOQInF2yf//iKB6fxDALKQO3RCARyG5Bfb1sN73AdgEcHeT612XYUiROH8v1/cDcv3/dYE1vwnAXjnPI5CMxwLkAAKemi15IFa/5IrzJwA2IA2lvgBg3Op0dajlm5DcB1kA8wD+VKksNfe8C8AlSL3BpwDcWXf9bvn9nHzfO63WVZe+92Cnp1T7mpGvPwjgrJz+swBeV/f5WQCPQzIeywDeX3d9AMDfyI3rFoCPAQiKqBeSe3BBzu91AF+tb1h6UC8DUJIbNfVlZBmGZPy+INf3DQB/BMAlqmYAfyznXRbAity432CGZuc4TgcHBwcHFdvNKTg4ODg4NMcxCg4ODg4OKo5RcHBwcHBQcYyCg4ODg4OKYxQcHBwcHFQco+Dg4ODgoOIYBQcHBwcHFccoODg4ODio/F8lSfPmeLT3LAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "np.random.seed(0)\n",
    "rand = np.random.random(N)\n",
    "idx = np.linspace(0, 10, N)\n",
    "s = pd.Series(1 * np.sin(2.* idx + .5))\n",
    "s.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### a) Compute the ADF stat on this series. What is the p-value?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ADF stat:  -8053074074508.672  p-value:  0.0\n"
     ]
    }
   ],
   "source": [
    "res = sm.tsa.stattools.adfuller(s)\n",
    "print('ADF stat: ',res[0],' p-value: ',res[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### b) Shift every observation by the same positive value. Compute the cumulative sum of the observations. This is a non-stationary series with memory."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fed8e91ce80>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "s_ = (s+1).cumsum().rename('fake_close').to_frame()\n",
    "s_.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### i) Compute the ADF Statistics on this series. What is the p-value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ADF stat:  -0.09055556893818328  p-value:  0.9503711480907152\n"
     ]
    }
   ],
   "source": [
    "res = sm.tsa.stattools.adfuller(s_['fake_close'])\n",
    "print('ADF stat: ',res[0],' p-value: ',res[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### ii) Apply an expanding facdiff, with $\\tau = 1E-2$. For what minimum $d$ value do you get  a p-value below 5%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols = ['adfStat','pVal','lags','nObs','95% conf']#,'corr']\n",
    "out = pd.DataFrame(columns=cols)\n",
    "for d in np.linspace(0,0.4,40):\n",
    "    df0 = fracDiff(s_,d)\n",
    "    df0 = sm.tsa.stattools.adfuller(df0['fake_close'],maxlag=1,regression='c',autolag=None)\n",
    "    out.loc[d]=list(df0[:4])+[df0[4]['5%']]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>adfStat</th>\n",
       "      <th>pVal</th>\n",
       "      <th>lags</th>\n",
       "      <th>nObs</th>\n",
       "      <th>95% conf</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.000000</th>\n",
       "      <td>2.833609</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2517.0</td>\n",
       "      <td>-2.862689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.010256</th>\n",
       "      <td>3.823719</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1715.0</td>\n",
       "      <td>-2.863227</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.020513</th>\n",
       "      <td>-8.815686</td>\n",
       "      <td>1.937221e-14</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1224.0</td>\n",
       "      <td>-2.863904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.030769</th>\n",
       "      <td>-3.358019</td>\n",
       "      <td>1.247600e-02</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1009.0</td>\n",
       "      <td>-2.864409</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.041026</th>\n",
       "      <td>6.753537</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>897.0</td>\n",
       "      <td>-2.864768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.051282</th>\n",
       "      <td>11.535198</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>832.0</td>\n",
       "      <td>-2.865020</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.061538</th>\n",
       "      <td>12.725323</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>794.0</td>\n",
       "      <td>-2.865187</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.071795</th>\n",
       "      <td>12.177383</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>772.0</td>\n",
       "      <td>-2.865291</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.082051</th>\n",
       "      <td>11.004925</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>762.0</td>\n",
       "      <td>-2.865340</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.092308</th>\n",
       "      <td>9.741210</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>759.0</td>\n",
       "      <td>-2.865355</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.102564</th>\n",
       "      <td>8.588915</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>762.0</td>\n",
       "      <td>-2.865340</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.112821</th>\n",
       "      <td>7.605360</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>771.0</td>\n",
       "      <td>-2.865296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.123077</th>\n",
       "      <td>6.652378</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>783.0</td>\n",
       "      <td>-2.865238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.133333</th>\n",
       "      <td>5.657688</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>798.0</td>\n",
       "      <td>-2.865169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.143590</th>\n",
       "      <td>4.523025</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>817.0</td>\n",
       "      <td>-2.865084</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.153846</th>\n",
       "      <td>3.151188</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>838.0</td>\n",
       "      <td>-2.864995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.164103</th>\n",
       "      <td>1.457237</td>\n",
       "      <td>9.973593e-01</td>\n",
       "      <td>1.0</td>\n",
       "      <td>862.0</td>\n",
       "      <td>-2.864899</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.174359</th>\n",
       "      <td>-0.583977</td>\n",
       "      <td>8.745616e-01</td>\n",
       "      <td>1.0</td>\n",
       "      <td>888.0</td>\n",
       "      <td>-2.864800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.184615</th>\n",
       "      <td>-3.051783</td>\n",
       "      <td>3.033343e-02</td>\n",
       "      <td>1.0</td>\n",
       "      <td>917.0</td>\n",
       "      <td>-2.864697</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.194872</th>\n",
       "      <td>-5.830686</td>\n",
       "      <td>3.981801e-07</td>\n",
       "      <td>1.0</td>\n",
       "      <td>947.0</td>\n",
       "      <td>-2.864597</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.205128</th>\n",
       "      <td>-9.017084</td>\n",
       "      <td>5.910639e-15</td>\n",
       "      <td>1.0</td>\n",
       "      <td>980.0</td>\n",
       "      <td>-2.864494</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.215385</th>\n",
       "      <td>-12.378299</td>\n",
       "      <td>5.096355e-23</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1014.0</td>\n",
       "      <td>-2.864395</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.225641</th>\n",
       "      <td>-15.820122</td>\n",
       "      <td>1.022483e-28</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1050.0</td>\n",
       "      <td>-2.864297</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.235897</th>\n",
       "      <td>-19.072210</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1088.0</td>\n",
       "      <td>-2.864200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.246154</th>\n",
       "      <td>-21.786048</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1128.0</td>\n",
       "      <td>-2.864106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.256410</th>\n",
       "      <td>-23.634447</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1169.0</td>\n",
       "      <td>-2.864016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.266667</th>\n",
       "      <td>-24.544472</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1211.0</td>\n",
       "      <td>-2.863930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.276923</th>\n",
       "      <td>-24.599767</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1255.0</td>\n",
       "      <td>-2.863846</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.287179</th>\n",
       "      <td>-24.008199</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1300.0</td>\n",
       "      <td>-2.863766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.297436</th>\n",
       "      <td>-22.960331</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1346.0</td>\n",
       "      <td>-2.863690</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.307692</th>\n",
       "      <td>-21.620436</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1392.0</td>\n",
       "      <td>-2.863619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.317949</th>\n",
       "      <td>-19.924991</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1440.0</td>\n",
       "      <td>-2.863549</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.328205</th>\n",
       "      <td>-18.069507</td>\n",
       "      <td>2.606750e-30</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1488.0</td>\n",
       "      <td>-2.863484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.338462</th>\n",
       "      <td>-16.272623</td>\n",
       "      <td>3.475655e-29</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1536.0</td>\n",
       "      <td>-2.863424</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.348718</th>\n",
       "      <td>-14.849403</td>\n",
       "      <td>1.786659e-27</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1585.0</td>\n",
       "      <td>-2.863365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.358974</th>\n",
       "      <td>-14.315459</td>\n",
       "      <td>1.164973e-26</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1633.0</td>\n",
       "      <td>-2.863312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.369231</th>\n",
       "      <td>-14.873774</td>\n",
       "      <td>1.648274e-27</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1681.0</td>\n",
       "      <td>-2.863261</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.379487</th>\n",
       "      <td>-16.563112</td>\n",
       "      <td>1.899402e-29</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1728.0</td>\n",
       "      <td>-2.863214</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.389744</th>\n",
       "      <td>-19.240339</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1775.0</td>\n",
       "      <td>-2.863170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.400000</th>\n",
       "      <td>-22.597921</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1821.0</td>\n",
       "      <td>-2.863128</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            adfStat          pVal  lags    nObs  95% conf\n",
       "0.000000   2.833609  1.000000e+00   1.0  2517.0 -2.862689\n",
       "0.010256   3.823719  1.000000e+00   1.0  1715.0 -2.863227\n",
       "0.020513  -8.815686  1.937221e-14   1.0  1224.0 -2.863904\n",
       "0.030769  -3.358019  1.247600e-02   1.0  1009.0 -2.864409\n",
       "0.041026   6.753537  1.000000e+00   1.0   897.0 -2.864768\n",
       "0.051282  11.535198  1.000000e+00   1.0   832.0 -2.865020\n",
       "0.061538  12.725323  1.000000e+00   1.0   794.0 -2.865187\n",
       "0.071795  12.177383  1.000000e+00   1.0   772.0 -2.865291\n",
       "0.082051  11.004925  1.000000e+00   1.0   762.0 -2.865340\n",
       "0.092308   9.741210  1.000000e+00   1.0   759.0 -2.865355\n",
       "0.102564   8.588915  1.000000e+00   1.0   762.0 -2.865340\n",
       "0.112821   7.605360  1.000000e+00   1.0   771.0 -2.865296\n",
       "0.123077   6.652378  1.000000e+00   1.0   783.0 -2.865238\n",
       "0.133333   5.657688  1.000000e+00   1.0   798.0 -2.865169\n",
       "0.143590   4.523025  1.000000e+00   1.0   817.0 -2.865084\n",
       "0.153846   3.151188  1.000000e+00   1.0   838.0 -2.864995\n",
       "0.164103   1.457237  9.973593e-01   1.0   862.0 -2.864899\n",
       "0.174359  -0.583977  8.745616e-01   1.0   888.0 -2.864800\n",
       "0.184615  -3.051783  3.033343e-02   1.0   917.0 -2.864697\n",
       "0.194872  -5.830686  3.981801e-07   1.0   947.0 -2.864597\n",
       "0.205128  -9.017084  5.910639e-15   1.0   980.0 -2.864494\n",
       "0.215385 -12.378299  5.096355e-23   1.0  1014.0 -2.864395\n",
       "0.225641 -15.820122  1.022483e-28   1.0  1050.0 -2.864297\n",
       "0.235897 -19.072210  0.000000e+00   1.0  1088.0 -2.864200\n",
       "0.246154 -21.786048  0.000000e+00   1.0  1128.0 -2.864106\n",
       "0.256410 -23.634447  0.000000e+00   1.0  1169.0 -2.864016\n",
       "0.266667 -24.544472  0.000000e+00   1.0  1211.0 -2.863930\n",
       "0.276923 -24.599767  0.000000e+00   1.0  1255.0 -2.863846\n",
       "0.287179 -24.008199  0.000000e+00   1.0  1300.0 -2.863766\n",
       "0.297436 -22.960331  0.000000e+00   1.0  1346.0 -2.863690\n",
       "0.307692 -21.620436  0.000000e+00   1.0  1392.0 -2.863619\n",
       "0.317949 -19.924991  0.000000e+00   1.0  1440.0 -2.863549\n",
       "0.328205 -18.069507  2.606750e-30   1.0  1488.0 -2.863484\n",
       "0.338462 -16.272623  3.475655e-29   1.0  1536.0 -2.863424\n",
       "0.348718 -14.849403  1.786659e-27   1.0  1585.0 -2.863365\n",
       "0.358974 -14.315459  1.164973e-26   1.0  1633.0 -2.863312\n",
       "0.369231 -14.873774  1.648274e-27   1.0  1681.0 -2.863261\n",
       "0.379487 -16.563112  1.899402e-29   1.0  1728.0 -2.863214\n",
       "0.389744 -19.240339  0.000000e+00   1.0  1775.0 -2.863170\n",
       "0.400000 -22.597921  0.000000e+00   1.0  1821.0 -2.863128"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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": [
    "f,ax=plt.subplots()\n",
    "out['pVal'].plot(ax=ax, marker='X')\n",
    "ax.axhline(out['95% conf'].mean(),lw=1,color='r',ls='dotted')\n",
    "ax.set_title('min d with thresh=0.01')\n",
    "ax.set_xlabel('d values')\n",
    "ax.set_ylabel('p-value');\n",
    "display(out)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### iii) Apply FFD with $\\tau = 1E-5$. For what minimum $d$ value do you get a p-value below $5\\%$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#cols = ['adfStat','pVal','lags','nObs','95% conf']#,'corr']\n",
    "#out = pd.DataFrame(columns=cols)\n",
    "#for d in np.linspace(0,0.4,40):\n",
    "#    try:\n",
    "#        df0 = fracDiff_FFD(s_,d,thres=1e-5)\n",
    "#        df0 = sm.tsa.stattools.adfuller(df0['fake_close'],maxlag=1,regression='c',autolag=None)\n",
    "#        out.loc[d]=list(df0[:4])+[df0[4]['5%']]\n",
    "#    except Exception as e:\n",
    "#        print(f'd: {d}, error: {e}')\n",
    "#        continue\n",
    "\n",
    "#f,ax=plt.subplots()\n",
    "#out['pVal'].plot(ax=ax, marker='X')\n",
    "#ax.axhline(out['95% conf'].mean(),lw=1,color='r',ls='dotted')\n",
    "#ax.set_title('min d with thresh=0.0001')\n",
    "#ax.set_xlabel('d values')\n",
    "#ax.set_ylabel('p-value');\n",
    "#display(out)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [5.3] Take the series from exercise 2.b:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### (a) Fit the series to a sine function. What is the R-squared?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "## fitting function taken from stackoverflow## fitti \n",
    "##   https://stackoverflow.com/questions/16716302/how-do-i-fit-a-sine-curve-to-my-data-with-pylab-and-numpy#16716964\n",
    "#import numpy, scipy.optimize\n",
    "\n",
    "#def fit_sin(tt, yy):\n",
    "#    '''Fit sin to the input time sequence, and return fitting parameters \"amp\", \"omega\", \"phase\", \"offset\", \"freq\", \"period\" and \"fitfunc\"'''\n",
    "#    tt = numpy.array(tt)\n",
    "#    yy = numpy.array(yy)\n",
    "#    ff = numpy.fft.fftfreq(len(tt), (tt[1]-tt[0]))   # assume uniform spacing\n",
    "#    Fyy = abs(numpy.fft.fft(yy))\n",
    "#    guess_freq = abs(ff[numpy.argmax(Fyy[1:])+1])   # excluding the zero frequency \"peak\", which is related to offset\n",
    "#    guess_amp = numpy.std(yy) * 2.**0.5\n",
    "#    guess_offset = numpy.mean(yy)\n",
    "#    guess = numpy.array([guess_amp, 2.*numpy.pi*guess_freq, 0., guess_offset])\n",
    "\n",
    "#    def sinfunc(t, A, w, p, c):  return A * numpy.sin(w*t + p) + c\n",
    "#    popt, pcov = scipy.optimize.curve_fit(sinfunc, tt, yy, p0=guess)\n",
    "#    A, w, p, c = popt\n",
    "#    f = w/(2.*numpy.pi)\n",
    "#    fitfunc = lambda t: A * numpy.sin(w*t + p) + c\n",
    "#    return {\"amp\": A, \"omega\": w, \"phase\": p, \"offset\": c, \"freq\": f, \"period\": 1./f, \"fitfunc\": fitfunc, \"maxcov\": numpy.max(pcov), \"rawres\": (guess,popt,pcov)}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "#res = fit_sin(s_.index.values, s_.values.ravel())\n",
    "#res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "#xx = s_.index.values\n",
    "#yy = s_.values.ravel()\n",
    "\n",
    "#plt.plot(xx, yy, \"-k\", label=\"y\", linewidth=2)\n",
    "##plt.plot(tt, yynoise, \"ok\", label=\"y with noise\")\n",
    "#plt.plot(xx, res[\"fitfunc\"](xx), \"r-\", label=\"y fit curve\", linewidth=2)\n",
    "#plt.legend(loc=\"best\")\n",
    "#plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "#slope, intercept, r_value, p_value, std_err = scipy.stats.linregress(yy, res[\"fitfunc\"](xx))\n",
    "#r_value**2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### (b) Apply FFD$(d=1)$. Fit the series to a sine function. What is the R-squared?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "#cols = ['adfStat','pVal','lags','nObs','95% conf']#,'corr']\n",
    "#out = pd.DataFrame(columns=cols)\n",
    "##df1 = fracDiff_FFD(s_,d=1)\n",
    "#df1 = sm.tsa.stattools.adfuller(df0['fake_close'],maxlag=1,regression='c',autolag=None)\n",
    "#out.loc[d]=list(df0[:4])+[df0[4]['5%']]\n",
    "#df1.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "#xx = df1.index.values\n",
    "#yy = df1.values.ravel()\n",
    "\n",
    "#res = fit_sin(xx, yy)\n",
    "#slope, intercept, r_value, p_value, std_err = scipy.stats.linregress(yy, res[\"fitfunc\"](xx))\n",
    "#r_value**2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [5.4] Take dollar bar series on E-mini S&P 500 futures. Using the code in Snippet 5.3, for some d in [0,2], compute fracDiff_FFD(fracDiff_FFD(series,d). What do you get? Why?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-------------------------------------------------------------------------------\n",
      "dataframe information\n",
      "-------------------------------------------------------------------------------\n",
      "                      price     bid     ask   size      v          dv\n",
      "dates                                                                \n",
      "2018-10-05 15:58:23  116.19  116.19  116.20   1354   1354   157321.26\n",
      "2018-10-05 15:59:20  116.18  116.17  116.19    300    300    34854.00\n",
      "2018-10-05 15:59:55  116.19  116.18  116.20    800    800    92952.00\n",
      "2018-10-05 15:59:59  116.20  116.18  116.19   2000   2000   232400.00\n",
      "2018-10-05 16:00:00  116.22  116.20  116.21  42884  42884  4983978.48\n",
      "--------------------------------------------------\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 50766 entries, 2009-09-28 09:46:35 to 2018-10-05 16:00:00\n",
      "Data columns (total 6 columns):\n",
      "price    50766 non-null float64\n",
      "bid      50766 non-null float64\n",
      "ask      50766 non-null float64\n",
      "size     50766 non-null int64\n",
      "v        50766 non-null int64\n",
      "dv       50766 non-null float64\n",
      "dtypes: float64(4), int64(2)\n",
      "memory usage: 2.7 MB\n",
      "None\n",
      "-------------------------------------------------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "def dask_resample(ser, freq='L'):\n",
    "    dds = dd.from_pandas(ser, chunksize=len(ser)//100)\n",
    "    tdf = (dds\n",
    "           .resample(freq)\n",
    "           .mean()\n",
    "           .dropna()\n",
    "          ).compute()\n",
    "    return tdf\n",
    "\n",
    "infp=PurePath(data_dir/'processed'/'clean_IVE_fut_prices.parq')\n",
    "df = pd.read_parquet(infp)\n",
    "df = df.assign(dates = df.index)\n",
    "#dv_rs = dask_resample(df, '1s')\n",
    "#cprint(dv_rs)\n",
    "\n",
    "dbars = dollar_bar_df(df, 'dv', 1_000_000)\n",
    "dbars = dbars.drop(['dates'], axis = 1)\n",
    "cprint(dbars)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Bug: [5.4] getWeights_FFD cannot recognize negative $d$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [5.5] Take the dollar bar series on E-mini S&P 500 futures"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### (a) Form a new series as a cumulative sum of log-prices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-------------------------------------------------------------------------------\n",
      "dataframe information\n",
      "-------------------------------------------------------------------------------\n",
      "                             price\n",
      "dates                             \n",
      "2018-10-05 15:58:23  225860.438048\n",
      "2018-10-05 15:59:20  225865.193189\n",
      "2018-10-05 15:59:55  225869.948416\n",
      "2018-10-05 15:59:59  225874.703729\n",
      "2018-10-05 16:00:00  225879.459214\n",
      "--------------------------------------------------\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 49996 entries, 2009-09-28 09:46:35 to 2018-10-05 16:00:00\n",
      "Data columns (total 1 columns):\n",
      "price    49996 non-null float64\n",
      "dtypes: float64(1)\n",
      "memory usage: 781.2 KB\n",
      "None\n",
      "-------------------------------------------------------------------------------\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fed8c05eda0>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = np.log(dbars.price).cumsum()\n",
    "x = x[~x.index.duplicated()]\n",
    "cprint(x)\n",
    "x.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### (b) Apply FFD, with $\\tau = 1E-5$. Determine for what minimum $d \\in [0,2]$ the new series is stationary "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_optimal_ffd(data, end, interval, t=1e-5):\n",
    "    \n",
    "    cols = ['adfStat','pVal','lags','nObs','95% conf']#,'corr']\n",
    "    out = pd.DataFrame(columns=cols)\n",
    "    \n",
    "    for d in np.linspace(0,end,interval):\n",
    "        try:\n",
    "            \n",
    "            dfx = fracDiff_FFD(data.to_frame(),d,thres=t)\n",
    "            dfx = sm.tsa.stattools.adfuller(dfx['price'], maxlag=1,regression='c',autolag=None)\n",
    "            out.loc[d]=list(dfx[:4])+[dfx[4]['5%']]\n",
    "        except Exception as e:\n",
    "            print(f'{d} error: {e}')\n",
    "            \n",
    "    return out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "out = get_optimal_ffd(x, 2, 10)"
   ]
  },
  {
   "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></th>\n",
       "      <th>adfStat</th>\n",
       "      <th>pVal</th>\n",
       "      <th>lags</th>\n",
       "      <th>nObs</th>\n",
       "      <th>95% conf</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.000000</th>\n",
       "      <td>67.156387</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>49994.0</td>\n",
       "      <td>-2.861598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.222222</th>\n",
       "      <td>9.564391</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>46871.0</td>\n",
       "      <td>-2.861602</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.444444</th>\n",
       "      <td>2.210799</td>\n",
       "      <td>0.998891</td>\n",
       "      <td>1.0</td>\n",
       "      <td>48802.0</td>\n",
       "      <td>-2.861599</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.666667</th>\n",
       "      <td>0.442605</td>\n",
       "      <td>0.983020</td>\n",
       "      <td>1.0</td>\n",
       "      <td>49560.0</td>\n",
       "      <td>-2.861598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.888889</th>\n",
       "      <td>-0.487582</td>\n",
       "      <td>0.894417</td>\n",
       "      <td>1.0</td>\n",
       "      <td>49860.0</td>\n",
       "      <td>-2.861598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.111111</th>\n",
       "      <td>-1.227327</td>\n",
       "      <td>0.661731</td>\n",
       "      <td>1.0</td>\n",
       "      <td>49910.0</td>\n",
       "      <td>-2.861598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.333333</th>\n",
       "      <td>-1.514305</td>\n",
       "      <td>0.526423</td>\n",
       "      <td>1.0</td>\n",
       "      <td>49908.0</td>\n",
       "      <td>-2.861598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.555556</th>\n",
       "      <td>-2.583721</td>\n",
       "      <td>0.096409</td>\n",
       "      <td>1.0</td>\n",
       "      <td>49928.0</td>\n",
       "      <td>-2.861598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.777778</th>\n",
       "      <td>-4.831333</td>\n",
       "      <td>0.000047</td>\n",
       "      <td>1.0</td>\n",
       "      <td>49951.0</td>\n",
       "      <td>-2.861598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.000000</th>\n",
       "      <td>-268.417713</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>49992.0</td>\n",
       "      <td>-2.861598</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             adfStat      pVal  lags     nObs  95% conf\n",
       "0.000000   67.156387  1.000000   1.0  49994.0 -2.861598\n",
       "0.222222    9.564391  1.000000   1.0  46871.0 -2.861602\n",
       "0.444444    2.210799  0.998891   1.0  48802.0 -2.861599\n",
       "0.666667    0.442605  0.983020   1.0  49560.0 -2.861598\n",
       "0.888889   -0.487582  0.894417   1.0  49860.0 -2.861598\n",
       "1.111111   -1.227327  0.661731   1.0  49910.0 -2.861598\n",
       "1.333333   -1.514305  0.526423   1.0  49908.0 -2.861598\n",
       "1.555556   -2.583721  0.096409   1.0  49928.0 -2.861598\n",
       "1.777778   -4.831333  0.000047   1.0  49951.0 -2.861598\n",
       "2.000000 -268.417713  0.000000   1.0  49992.0 -2.861598"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The min. D that makes the fracdiff stationary 1.7777777777777777\n"
     ]
    }
   ],
   "source": [
    "min_ffd = out[out.pVal < 0.05].iloc[0].name\n",
    "print(\"The min. D that makes the fracdiff stationary\",min_ffd)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### (c) Compute the correlation of the fracdiff series to the original (untransformed) series"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-------------------------------------------------------------------------------\n",
      "dataframe information\n",
      "-------------------------------------------------------------------------------\n",
      "                        price\n",
      "2018-10-05 15:58:23 -0.000430\n",
      "2018-10-05 15:59:20 -0.000086\n",
      "2018-10-05 15:59:55  0.000086\n",
      "2018-10-05 15:59:59  0.000086\n",
      "2018-10-05 16:00:00  0.000172\n",
      "--------------------------------------------------\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 49994 entries, 2009-09-28 09:55:26 to 2018-10-05 16:00:00\n",
      "Data columns (total 1 columns):\n",
      "price    49994 non-null float64\n",
      "dtypes: float64(1)\n",
      "memory usage: 781.2 KB\n",
      "None\n",
      "-------------------------------------------------------------------------------\n",
      "\n"
     ]
    },
    {
     "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>price</th>\n",
       "      <th>original</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>price</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.000013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>original</th>\n",
       "      <td>-0.000013</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             price  original\n",
       "price     1.000000 -0.000013\n",
       "original -0.000013  1.000000"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfx2 = fracDiff_FFD(x.to_frame(),2)\n",
    "cprint(dfx2)\n",
    "joined = dfx2.join(x.rename('original'),how='left')\n",
    "joined.corr()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### (d) Apply Engel-Granger cointegration test on the original and fracdiff series. Are they cointegrated? Why?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "p-value:  0.0\n",
      "They are cointegrated\n"
     ]
    }
   ],
   "source": [
    "coint_pval = sm.tsa.stattools.coint(joined.price, joined.original)[1]\n",
    "print(\"p-value: \", coint_pval)\n",
    "if coint_pval < 0.05:\n",
    "    print(\"They are cointegrated\")\n",
    "else:\n",
    "    print(\"They are not cointegrated\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### (e) Apply a Jarque-Bera normality test on the fracdiff series."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "p-value:  0.0\n",
      "The series is not normally distributed\n"
     ]
    }
   ],
   "source": [
    "jb_pval = stats.jarque_bera(dfx2)[1]\n",
    "print(\"p-value: \", jb_pval)\n",
    "if jb_pval < 0.05:\n",
    "    print(\"The series is not normally distributed\")\n",
    "else:\n",
    "    print(\"The series is normally distributed\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [5.6] Take the fracdiff series from exercise 5"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### (a) Apply a CUSUM filter (Chapter 2), where h is twice the standard deviation of the series."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-------------------------------------------------------------------------------\n",
      "dataframe information\n",
      "-------------------------------------------------------------------------------\n",
      "                        price\n",
      "2018-10-05 15:58:23 -0.000430\n",
      "2018-10-05 15:59:20 -0.000086\n",
      "2018-10-05 15:59:55  0.000086\n",
      "2018-10-05 15:59:59  0.000086\n",
      "2018-10-05 16:00:00  0.000172\n",
      "--------------------------------------------------\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 49994 entries, 2009-09-28 09:55:26 to 2018-10-05 16:00:00\n",
      "Data columns (total 1 columns):\n",
      "price    49994 non-null float64\n",
      "dtypes: float64(1)\n",
      "memory usage: 781.2 KB\n",
      "None\n",
      "-------------------------------------------------------------------------------\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.9400002590332018"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfx2 = dfx2[~dfx2.index.duplicated()]\n",
    "cprint(dfx2)\n",
    "ffd_std = dfx2.std()[0]\n",
    "ffd_std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getTEvents2(gRaw, h, symmetric = True, isReturn = False):\n",
    "    \"\"\"\n",
    "    Symmetric CUSUM Filter\n",
    "    Sample a bar t iff S_t >= h at which point S_t is reset\n",
    "    Multiple events are not triggered by gRaw hovering around a threshold level\n",
    "    It will require a full run of length h for gRaw to trigger an event\n",
    "    \n",
    "    Two arguments:\n",
    "        gRaw: the raw time series we wish to filter (gRaw), e.g. return\n",
    "        h: threshold\n",
    "        \n",
    "    Return:\n",
    "        pd.DatatimeIndex.append(tEvents): \n",
    "    \"\"\"\n",
    "    tEvents, sPos, sNeg = [], 0, 0\n",
    "    if isReturn:\n",
    "        diff = gRaw\n",
    "    else:\n",
    "        diff = gRaw.diff()\n",
    "    if symmetric:\n",
    "        if np.shape(h) == ():\n",
    "\n",
    "            for i in diff.index[1:]:\n",
    "                tmp = diff.loc[i].mean()\n",
    "                sPos, sNeg = max(0,sPos+tmp), min(0,sNeg+tmp)\n",
    "                if sNeg < -h and tradableHour(i):\n",
    "                    sNeg = 0; tEvents.append(i)\n",
    "                elif sPos > h and tradableHour(i):\n",
    "                    sPos = 0; tEvents.append(i)\n",
    "        else:\n",
    "            for i in diff.index[1:]:\n",
    "                tmp = diff.loc[i].mean()\n",
    "                sPos, sNeg = max(0,sPos+tmp), min(0,sNeg+tmp)\n",
    "                if sNeg < -h[i] and tradableHour(i):\n",
    "                    sNeg = 0; tEvents.append(i)\n",
    "                elif sPos > h[i] and tradableHour(i):\n",
    "                    sPos = 0; tEvents.append(i)\n",
    "    else:\n",
    "        if np.shape(h) == ():\n",
    "\n",
    "            for i in diff.index[1:]:\n",
    "                tmp = diff.loc[i].mean()\n",
    "                sAbs = sAbs+tmp\n",
    "                if sAbs > h and tradableHour(i):\n",
    "                    sNeg = 0; tEvents.append(i)\n",
    "                \n",
    "        else:\n",
    "            for i in diff.index[1:]:\n",
    "                tmp = diff.loc[i].mean()\n",
    "                sAbs = sAbs+tmp\n",
    "                if sAbs > h[i] and tradableHour(i):\n",
    "                    sNeg = 0; tEvents.append(i)\n",
    "            \n",
    "    return pd.DatetimeIndex(tEvents)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DatetimeIndex(['2009-11-03 11:20:59', '2009-11-03 11:21:01',\n",
       "               '2009-11-03 11:22:04', '2009-12-03 13:13:31',\n",
       "               '2009-12-03 14:24:15', '2009-12-03 14:50:50',\n",
       "               '2009-12-04 09:54:02', '2009-12-04 10:18:41',\n",
       "               '2010-04-22 12:23:34', '2010-04-22 12:33:19',\n",
       "               ...\n",
       "               '2018-09-21 09:55:07', '2018-09-25 14:23:14',\n",
       "               '2018-09-25 14:34:53', '2018-09-25 14:53:26',\n",
       "               '2018-10-02 11:39:06', '2018-10-02 11:49:10',\n",
       "               '2018-10-02 12:12:54', '2018-10-04 13:44:24',\n",
       "               '2018-10-04 13:54:45', '2018-10-04 14:01:44'],\n",
       "              dtype='datetime64[ns]', length=1633, freq=None)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "tEvents = getTEvents2(dfx2,h=ffd_std*2)\n",
    "display(tEvents)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-------------------------------------------------------------------------------\n",
      "dataframe information\n",
      "-------------------------------------------------------------------------------\n",
      "                        dbars  frac_diff_feat\n",
      "dates                                        \n",
      "2018-10-02 11:49:10  116.4980       -4.754773\n",
      "2018-10-02 12:12:54  116.5699        0.000617\n",
      "2018-10-04 13:44:24  116.1600        4.754977\n",
      "2018-10-04 13:54:45  116.0337       -4.755023\n",
      "2018-10-04 14:01:44  116.0100       -0.000204\n",
      "--------------------------------------------------\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 1675 entries, 2009-11-03 11:20:59 to 2018-10-04 14:01:44\n",
      "Data columns (total 2 columns):\n",
      "dbars             1675 non-null float64\n",
      "frac_diff_feat    1675 non-null float64\n",
      "dtypes: float64(2)\n",
      "memory usage: 39.3 KB\n",
      "None\n",
      "-------------------------------------------------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "dbars_feat = dbars.price.loc[tEvents]\n",
    "frac_diff_feat = dfx2.loc[tEvents]\n",
    "ftMtx = (pd.DataFrame()\n",
    "         .assign(dbars=dbars_feat,\n",
    "                 frac_diff_feat=frac_diff_feat)\n",
    "         .drop_duplicates().dropna())\n",
    "cprint(ftMtx)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2018-10-27 17:32:56.096388 14.29% applyPtSlOnT1 done after 0.0 minutes. Remaining 0.03 minutes.\r",
      "2018-10-27 17:32:56.196749 28.57% applyPtSlOnT1 done after 0.01 minutes. Remaining 0.02 minutes.\r",
      "2018-10-27 17:32:56.197178 42.86% applyPtSlOnT1 done after 0.01 minutes. Remaining 0.01 minutes.\r",
      "2018-10-27 17:32:56.217753 57.14% applyPtSlOnT1 done after 0.01 minutes. Remaining 0.01 minutes.\r",
      "2018-10-27 17:32:56.218798 71.43% applyPtSlOnT1 done after 0.01 minutes. Remaining 0.0 minutes.\r",
      "2018-10-27 17:32:56.220014 85.71% applyPtSlOnT1 done after 0.01 minutes. Remaining 0.0 minutes.\r",
      "2018-10-27 17:32:56.220751 100.0% applyPtSlOnT1 done after 0.01 minutes. Remaining 0.0 minutes.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-------------------------------------------------------------------------------\n",
      "dataframe information\n",
      "-------------------------------------------------------------------------------\n",
      "                     t1      trgt\n",
      "2018-10-02 11:49:10 NaT  0.020142\n",
      "2018-10-02 12:12:54 NaT  0.020067\n",
      "2018-10-04 13:44:24 NaT  0.019922\n",
      "2018-10-04 13:54:45 NaT  0.019802\n",
      "2018-10-04 14:01:44 NaT  0.019685\n",
      "--------------------------------------------------\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 1619 entries, 2010-05-03 14:16:41 to 2018-10-04 14:01:44\n",
      "Data columns (total 2 columns):\n",
      "t1      1613 non-null datetime64[ns]\n",
      "trgt    1619 non-null float64\n",
      "dtypes: datetime64[ns](1), float64(1)\n",
      "memory usage: 37.9 KB\n",
      "None\n",
      "-------------------------------------------------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "def getDailyVol2(close,span0=100):\n",
    "    # daily vol reindexed to close\n",
    "    df0=close.index.searchsorted(close.index-pd.Timedelta(days=1))\n",
    "    #bp()\n",
    "    df0=df0[df0>0]\n",
    "    #bp()\n",
    "    df0=(pd.Series(close.index[df0-1],\n",
    "                   index=close.index[close.shape[0]-df0.shape[0]:]))\n",
    "    #bp()\n",
    "    try:\n",
    "        df0=close.loc[df0.index]/close.loc[df0.values].values-1 # daily rets\n",
    "    except Exception as e:\n",
    "        print(e)\n",
    "        print('adjusting shape of close.loc[df0.index]')\n",
    "        cut = close.loc[df0.index].shape[0] - close.loc[df0.values].shape[0]\n",
    "        # I dont't think it make sense\n",
    "        df0=close.loc[df0.index].iloc[:-cut]/close.loc[df0.values].values-1\n",
    "    df0=df0.ewm(span=span0).std().rename('dailyVol')\n",
    "    return df0\n",
    "\n",
    "\n",
    "ftMtx = ftMtx[~ftMtx.index.duplicated()]\n",
    "dailyVol = getDailyVol2(ftMtx.dbars)\n",
    "t1 = addVerticalBarrier(tEvents, ftMtx.dbars, hour=120)\n",
    "\n",
    "ptsl = [1,1]\n",
    "#ptsl = [daily]\n",
    "target=dailyVol*2\n",
    "# select minRet\n",
    "minRet = 0.01\n",
    "# get cpu count - 1\n",
    "cpus = cpu_count() - 1\n",
    "events = getEvents(ftMtx.dbars,tEvents,ptsl,target,minRet,cpus,t1=t1)\n",
    "cprint(events)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2018-10-27 17:32:56.704921 100.0% mpNumCoEvents done after 0.0 minutes. Remaining 0.0 minutes..\n",
      "2018-10-27 17:32:56.990270 100.0% mpSampleTW done after 0.0 minutes. Remaining 0.0 minutes..\n",
      "2018-10-27 17:32:57.436798 100.0% mpSampleW done after 0.0 minutes. Remaining 0.0 minutes..\n"
     ]
    }
   ],
   "source": [
    "close=ftMtx.dbars\n",
    "out = get_Concur_Uniqueness(close,events,cpus)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dropped label:  0.0 0.0012353304508956147\n",
      "-------------------------------------------------------------------------------\n",
      "dataframe information\n",
      "-------------------------------------------------------------------------------\n",
      "                          ret  bin\n",
      "2018-10-02 11:39:06 -0.002536 -1.0\n",
      "2018-10-02 11:49:10 -0.004189 -1.0\n",
      "2018-10-02 12:12:54 -0.004803 -1.0\n",
      "2018-10-04 13:44:24 -0.001291 -1.0\n",
      "2018-10-04 13:54:45 -0.000204 -1.0\n",
      "--------------------------------------------------\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 1617 entries, 2010-05-03 14:16:41 to 2018-10-04 13:54:45\n",
      "Data columns (total 2 columns):\n",
      "ret    1617 non-null float64\n",
      "bin    1617 non-null float64\n",
      "dtypes: float64(2)\n",
      "memory usage: 37.9 KB\n",
      "None\n",
      "-------------------------------------------------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "labels = getBins(events, ftMtx.dbars)\n",
    "clean_labels = dropLabels(labels)\n",
    "cprint(clean_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### (d) Fit a bagging classifier of decision trees where:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### (i) The observed features are bootstrapped using the sequential method from chapter 4.\n",
    "Note: must use multiprocessing of some kind as seqBootstrap is very slow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numba as nb\n",
    "@nb.njit\n",
    "def func(arr,i):\n",
    "    col = arr[i]\n",
    "    mask = np.where(col>0)\n",
    "    return np.mean(col[mask])\n",
    "\n",
    "@nb.njit\n",
    "def njit_getAvgUniqueness(indM):\n",
    "    # Average uniqueness from indicator matrix\n",
    "    c=indM.sum(axis=1).reshape(-1,1) # concurrency\n",
    "    u=np.divide(indM,c) # uniqueness\n",
    "    avgU = np.zeros(len(u.T)) # avg. uniqueness\n",
    "    i = 0\n",
    "    for i in range(len(u.T)):\n",
    "        avgU[i] = func(u.T,i)\n",
    "        i+=1\n",
    "    return avgU\n",
    "\n",
    "@nb.jit\n",
    "def jit_seqBootstrap(indM,sLength=None):\n",
    "    # Generate a sample via sequential bootstrap\n",
    "    if sLength is None:sLength=indM.shape[1]\n",
    "    phi=[]\n",
    "    while len(phi)<sLength:\n",
    "        avgU=pd.Series()\n",
    "        for i in indM:\n",
    "            indM_=indM[phi+[i]] # reduce indM\n",
    "            avgU.loc[i]=njit_getAvgUniqueness(indM_.values)[-1]\n",
    "        prob=avgU/avgU.sum() # draw prob\n",
    "        phi+=[np.random.choice(indM.columns,p=prob)]\n",
    "    return phi\n",
    "#------------------------\n",
    "\n",
    "def split_t1(t1, partitions):\n",
    "    return np.array_split(t1, partitions)\n",
    "\n",
    "def mp_func(indM):\n",
    "    # jit funcs about 2x as fast\n",
    "    phi = jit_seqBootstrap(indM)\n",
    "    seqU = njit_getAvgUniqueness(indM[phi].values).mean()\n",
    "    #phi = seqBootstrap(indM)\n",
    "    #seqU= getAvgUniqueness(indM[phi])\n",
    "    return seqU\n",
    "\n",
    "def main_mp(t1, partitions=100, cpus=8):\n",
    "    jobs = []\n",
    "    splits = split_t1(t1,partitions=100)\n",
    "    for part_t1 in splits:\n",
    "        indM = getIndMatrix(part_t1.index, part_t1)\n",
    "        job = {'func':mp_func,'indM':indM}\n",
    "        jobs.append(job)\n",
    "    if cpus==1: out= processJobs_(jobs)\n",
    "    else: out= processJobs(jobs,numThreads=cpus)\n",
    "    return pd.DataFrame(out)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2018-10-27 17:33:05.791147 100.0% mp_func done after 0.12 minutes. Remaining 0.0 minutes.\n"
     ]
    }
   ],
   "source": [
    "seqUs = main_mp(t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get avg uniqueness for bootstrapping\n",
    "avgU = seqUs.mean()[0]\n",
    "avgU1 = out['tW'].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### (ii) On each bootstrapped sample, sample weights are determined using the techniques from Chapter 4\n",
    "Note: alternative implementations are welcome."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier, RandomForestRegressor, BaggingClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluate(X,y,clf):\n",
    "    from sklearn import metrics\n",
    "    # The random forest model by itself\n",
    "    y_pred_rf = clf.predict_proba(X)[:, 1]\n",
    "    y_pred = clf.predict(X)\n",
    "    fpr_rf, tpr_rf, _ = metrics.roc_curve(y, y_pred_rf)\n",
    "    print(metrics.classification_report(y, y_pred))\n",
    "\n",
    "    plt.figure(figsize=(9,6))\n",
    "    plt.plot([0, 1], [0, 1], 'k--')\n",
    "    plt.plot(fpr_rf, tpr_rf, label='clf')\n",
    "    plt.xlabel('False positive rate')\n",
    "    plt.ylabel('True positive rate')\n",
    "    plt.title('ROC curve')\n",
    "    plt.legend(loc='best')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2010-05-03 14:16:41   -1.0\n",
       "2010-05-03 14:27:25   -1.0\n",
       "2010-05-03 14:57:03   -1.0\n",
       "2010-06-22 10:00:00   -1.0\n",
       "2010-06-22 10:19:06   -1.0\n",
       "2010-06-22 11:20:49   -1.0\n",
       "2010-07-06 15:15:51    1.0\n",
       "2010-07-06 15:36:44    1.0\n",
       "2010-07-07 10:16:53    1.0\n",
       "2010-08-02 15:47:58   -1.0\n",
       "2010-08-03 10:03:41   -1.0\n",
       "2010-08-03 10:48:54   -1.0\n",
       "2010-08-04 12:59:55   -1.0\n",
       "2010-08-04 14:50:03   -1.0\n",
       "2010-08-04 14:56:36   -1.0\n",
       "2010-08-31 15:20:53    1.0\n",
       "2010-08-31 15:35:45    1.0\n",
       "2010-08-31 15:35:46    1.0\n",
       "2010-09-01 11:33:42    1.0\n",
       "2010-09-01 11:33:43    1.0\n",
       "2010-09-01 11:37:51    1.0\n",
       "2010-09-07 12:44:44    1.0\n",
       "2010-09-07 13:56:46    1.0\n",
       "2010-09-07 14:54:49    1.0\n",
       "2010-09-08 09:45:53    1.0\n",
       "2010-10-05 10:24:21    1.0\n",
       "2010-10-05 10:24:23    1.0\n",
       "2010-10-05 10:24:25    1.0\n",
       "2010-11-02 11:30:52    1.0\n",
       "2010-11-02 12:47:47    1.0\n",
       "2010-11-02 13:48:10    1.0\n",
       "2010-11-08 10:03:02    1.0\n",
       "2010-11-08 10:35:44    1.0\n",
       "2010-11-08 11:02:11    1.0\n",
       "2010-11-11 12:33:18    1.0\n",
       "2010-11-11 12:59:16    1.0\n",
       "2010-11-11 13:36:26    1.0\n",
       "2010-11-12 10:53:52    1.0\n",
       "2010-11-12 11:04:27    1.0\n",
       "2010-11-12 11:33:09    1.0\n",
       "2011-01-06 10:24:49    1.0\n",
       "2011-01-06 10:59:15    1.0\n",
       "2011-01-06 11:09:35    1.0\n",
       "2011-01-07 15:33:45    1.0\n",
       "2011-01-07 15:36:04    1.0\n",
       "2011-01-07 15:47:40    1.0\n",
       "2011-01-10 12:01:59    1.0\n",
       "2011-01-10 12:05:59    1.0\n",
       "2011-01-10 12:28:06    1.0\n",
       "2011-01-26 10:45:43    1.0\n",
       "                      ... \n",
       "2018-07-31 14:17:45   -1.0\n",
       "2018-07-31 14:29:15   -1.0\n",
       "2018-07-31 14:37:51   -1.0\n",
       "2018-07-31 14:54:22   -1.0\n",
       "2018-07-31 14:58:28   -1.0\n",
       "2018-07-31 15:02:45   -1.0\n",
       "2018-07-31 15:49:19   -1.0\n",
       "2018-08-01 09:43:40   -1.0\n",
       "2018-08-01 09:57:47   -1.0\n",
       "2018-08-13 10:50:02    1.0\n",
       "2018-08-13 11:07:31    1.0\n",
       "2018-08-13 11:18:34    1.0\n",
       "2018-08-15 14:21:45    1.0\n",
       "2018-08-15 14:25:17    1.0\n",
       "2018-08-15 14:28:46    1.0\n",
       "2018-08-16 15:16:08    1.0\n",
       "2018-08-16 15:21:58    1.0\n",
       "2018-08-16 15:27:47    1.0\n",
       "2018-08-16 15:29:46    1.0\n",
       "2018-08-16 15:33:17    1.0\n",
       "2018-08-16 15:37:26    1.0\n",
       "2018-08-29 12:00:21   -1.0\n",
       "2018-08-29 12:41:21   -1.0\n",
       "2018-08-29 12:56:59   -1.0\n",
       "2018-09-04 14:17:20    1.0\n",
       "2018-09-04 14:33:39    1.0\n",
       "2018-09-04 14:52:41    1.0\n",
       "2018-09-05 09:56:43    1.0\n",
       "2018-09-05 10:02:13    1.0\n",
       "2018-09-05 10:12:08    1.0\n",
       "2018-09-12 12:06:19    1.0\n",
       "2018-09-12 12:25:45    1.0\n",
       "2018-09-12 13:08:15    1.0\n",
       "2018-09-13 10:16:00    1.0\n",
       "2018-09-13 10:16:39    1.0\n",
       "2018-09-13 10:28:22    1.0\n",
       "2018-09-14 12:03:46    1.0\n",
       "2018-09-14 12:08:21    1.0\n",
       "2018-09-14 12:40:42    1.0\n",
       "2018-09-20 15:45:38   -1.0\n",
       "2018-09-21 09:49:04   -1.0\n",
       "2018-09-21 09:55:07   -1.0\n",
       "2018-09-25 14:23:14   -1.0\n",
       "2018-09-25 14:34:53   -1.0\n",
       "2018-09-25 14:53:26   -1.0\n",
       "2018-10-02 11:39:06   -1.0\n",
       "2018-10-02 11:49:10   -1.0\n",
       "2018-10-02 12:12:54   -1.0\n",
       "2018-10-04 13:44:24   -1.0\n",
       "2018-10-04 13:54:45   -1.0\n",
       "Name: bin, Length: 1617, dtype: float64"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trgt = clean_labels.bin\n",
    "trgt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-------------------------------------------------------------------------------\n",
      "dataframe information\n",
      "-------------------------------------------------------------------------------\n",
      "                        dbars  frac_diff_feat        tW         w  bin\n",
      "dates                                                                 \n",
      "2018-10-02 11:39:06  116.3050        4.756302  0.265476  0.504378 -1.0\n",
      "2018-10-02 11:49:10  116.4980       -4.754773  0.290000  0.038143 -1.0\n",
      "2018-10-02 12:12:54  116.5699        0.000617  0.237500  0.362243 -1.0\n",
      "2018-10-04 13:44:24  116.1600        4.754977  0.205556  0.442645 -1.0\n",
      "2018-10-04 13:54:45  116.0337       -4.755023  0.183333  0.098369 -1.0\n",
      "--------------------------------------------------\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 1617 entries, 2010-05-03 14:16:41 to 2018-10-04 13:54:45\n",
      "Data columns (total 5 columns):\n",
      "dbars             1617 non-null float64\n",
      "frac_diff_feat    1617 non-null float64\n",
      "tW                1617 non-null float64\n",
      "w                 1617 non-null float64\n",
      "bin               1617 non-null float64\n",
      "dtypes: float64(5)\n",
      "memory usage: 75.8 KB\n",
      "None\n",
      "-------------------------------------------------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# model data# model  \n",
    "data = ftMtx.join(out,how='left').join(trgt,how='left').dropna()\n",
    "data = data.drop(['t1'],axis = 1)\n",
    "X = data.iloc[:,:-1].values\n",
    "y = data.iloc[:,-1].values.reshape(-1,1)\n",
    "cprint(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=RANDOM_STATE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "base_clf = DecisionTreeClassifier(criterion='entropy',max_features='auto',\n",
    "                                  class_weight='balanced')\n",
    "bc = BaggingClassifier(base_estimator=base_clf,n_estimators=1000,\n",
    "                       max_samples=avgU1,max_features=1.,random_state=RANDOM_STATE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "fit  =  bc.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "        -1.0       0.57      0.25      0.35       184\n",
      "         1.0       0.66      0.89      0.76       302\n",
      "\n",
      "   micro avg       0.65      0.65      0.65       486\n",
      "   macro avg       0.62      0.57      0.55       486\n",
      "weighted avg       0.63      0.65      0.60       486\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "evaluate(X_test,y_test,fit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "        -1.0       0.59      0.28      0.38       184\n",
      "         1.0       0.67      0.88      0.76       302\n",
      "\n",
      "   micro avg       0.65      0.65      0.65       486\n",
      "   macro avg       0.63      0.58      0.57       486\n",
      "weighted avg       0.64      0.65      0.62       486\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "rf_clf  =  RandomForestClassifier(n_estimators=1,\n",
    "                                class_weight='balanced_subsample',\n",
    "                                criterion='entropy', \n",
    "                                bootstrap=False)\n",
    "bc_rf = BaggingClassifier(base_estimator=rf_clf, n_estimators=1000,\n",
    "                          max_samples=avgU1, max_features=1., \n",
    "                          random_state=RANDOM_STATE)\n",
    "\n",
    "fit = bc_rf.fit(X_train,y_train)\n",
    "evaluate(X_test,y_test,fit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "        -1.0       0.50      0.20      0.28       184\n",
      "         1.0       0.64      0.88      0.74       302\n",
      "\n",
      "   micro avg       0.62      0.62      0.62       486\n",
      "   macro avg       0.57      0.54      0.51       486\n",
      "weighted avg       0.59      0.62      0.57       486\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "stdsc = StandardScaler()\n",
    "X_train_std = stdsc.fit_transform(X_train)\n",
    "X_test_std = stdsc.fit_transform(X_test)\n",
    "\n",
    "\n",
    "rf_clf  =  RandomForestClassifier(n_estimators=1,\n",
    "                                class_weight='balanced_subsample',\n",
    "                                criterion='entropy', \n",
    "                                bootstrap=False)\n",
    "bc_rf = BaggingClassifier(base_estimator=rf_clf, n_estimators=1000,\n",
    "                          max_samples=avgU1, max_features=1., \n",
    "                          random_state=RANDOM_STATE)\n",
    "fit = bc_rf.fit(X_train_std,y_train)\n",
    "evaluate(X_test_std, y_test, fit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "        -1.0       0.49      0.22      0.30       184\n",
      "         1.0       0.64      0.86      0.74       302\n",
      "\n",
      "   micro avg       0.62      0.62      0.62       486\n",
      "   macro avg       0.57      0.54      0.52       486\n",
      "weighted avg       0.58      0.62      0.57       486\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "base_clf = DecisionTreeClassifier(criterion='entropy',max_features='auto',\n",
    "                                  class_weight='balanced')\n",
    "bc = BaggingClassifier(base_estimator=base_clf,n_estimators=1000,\n",
    "                       max_samples=avgU1,max_features=1.,random_state=RANDOM_STATE)\n",
    "fit  =  bc.fit(X_train_std, y_train)\n",
    "evaluate(X_test_std, y_test, fit)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [default]",
   "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}