{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# M-Estimators for Robust Linear Modeling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.compat import lmap\n",
    "import numpy as np\n",
    "from scipy import stats\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import statsmodels.api as sm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* An M-estimator minimizes the function \n",
    "\n",
    "$$Q(e_i, \\rho) = \\sum_i~\\rho \\left (\\frac{e_i}{s}\\right )$$\n",
    "\n",
    "where $\\rho$ is a symmetric function of the residuals \n",
    "\n",
    "* The effect of $\\rho$ is to reduce the influence of outliers\n",
    "* $s$ is an estimate of scale. \n",
    "* The robust estimates $\\hat{\\beta}$ are computed by the iteratively re-weighted least squares algorithm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* We have several choices available for the weighting functions to be used"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "norms = sm.robust.norms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_weights(support, weights_func, xlabels, xticks):\n",
    "    fig = plt.figure(figsize=(12,8))\n",
    "    ax = fig.add_subplot(111)\n",
    "    ax.plot(support, weights_func(support))\n",
    "    ax.set_xticks(xticks)\n",
    "    ax.set_xticklabels(xlabels, fontsize=16)\n",
    "    ax.set_ylim(-.1, 1.1)\n",
    "    return ax"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Andrew's Wave"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Help on function weights in module statsmodels.robust.norms:\n",
      "\n",
      "weights(self, z)\n",
      "    Andrew's wave weighting function for the IRLS algorithm\n",
      "    \n",
      "    The psi function scaled by z\n",
      "    \n",
      "    Parameters\n",
      "    ----------\n",
      "    z : array_like\n",
      "        1d array\n",
      "    \n",
      "    Returns\n",
      "    -------\n",
      "    weights : ndarray\n",
      "        weights(z) = sin(z/a)/(z/a)     for \\|z\\| <= a*pi\n",
      "    \n",
      "        weights(z) = 0                  for \\|z\\| > a*pi\n",
      "\n"
     ]
    }
   ],
   "source": [
    "help(norms.AndrewWave.weights)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "a = 1.339\n",
    "support = np.linspace(-np.pi*a, np.pi*a, 100)\n",
    "andrew = norms.AndrewWave(a=a)\n",
    "plot_weights(support, andrew.weights, ['$-\\pi*a$', '0', '$\\pi*a$'], [-np.pi*a, 0, np.pi*a]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hampel's 17A"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Help on function weights in module statsmodels.robust.norms:\n",
      "\n",
      "weights(self, z)\n",
      "    Hampel weighting function for the IRLS algorithm\n",
      "    \n",
      "    The psi function scaled by z\n",
      "    \n",
      "    Parameters\n",
      "    ----------\n",
      "    z : array_like\n",
      "        1d array\n",
      "    \n",
      "    Returns\n",
      "    -------\n",
      "    weights : ndarray\n",
      "        weights(z) = 1                            for \\|z\\| <= a\n",
      "    \n",
      "        weights(z) = a/\\|z\\|                        for a < \\|z\\| <= b\n",
      "    \n",
      "        weights(z) = a*(c - \\|z\\|)/(\\|z\\|*(c-b))      for b < \\|z\\| <= c\n",
      "    \n",
      "        weights(z) = 0                            for \\|z\\| > c\n",
      "\n"
     ]
    }
   ],
   "source": [
    "help(norms.Hampel.weights)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "c = 8\n",
    "support = np.linspace(-3*c, 3*c, 1000)\n",
    "hampel = norms.Hampel(a=2., b=4., c=c)\n",
    "plot_weights(support, hampel.weights, ['3*c', '0', '3*c'], [-3*c, 0, 3*c]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Huber's t"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Help on function weights in module statsmodels.robust.norms:\n",
      "\n",
      "weights(self, z)\n",
      "    Huber's t weighting function for the IRLS algorithm\n",
      "    \n",
      "    The psi function scaled by z\n",
      "    \n",
      "    Parameters\n",
      "    ----------\n",
      "    z : array_like\n",
      "        1d array\n",
      "    \n",
      "    Returns\n",
      "    -------\n",
      "    weights : ndarray\n",
      "        weights(z) = 1          for \\|z\\| <= t\n",
      "    \n",
      "        weights(z) = t/\\|z\\|      for \\|z\\| > t\n",
      "\n"
     ]
    }
   ],
   "source": [
    "help(norms.HuberT.weights)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "t = 1.345\n",
    "support = np.linspace(-3*t, 3*t, 1000)\n",
    "huber = norms.HuberT(t=t)\n",
    "plot_weights(support, huber.weights, ['-3*t', '0', '3*t'], [-3*t, 0, 3*t]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Least Squares"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Help on function weights in module statsmodels.robust.norms:\n",
      "\n",
      "weights(self, z)\n",
      "    The least squares estimator weighting function for the IRLS algorithm.\n",
      "    \n",
      "    The psi function scaled by the input z\n",
      "    \n",
      "    Parameters\n",
      "    ----------\n",
      "    z : array_like\n",
      "        1d array\n",
      "    \n",
      "    Returns\n",
      "    -------\n",
      "    weights : ndarray\n",
      "        weights(z) = np.ones(z.shape)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "help(norms.LeastSquares.weights)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "support = np.linspace(-3, 3, 1000)\n",
    "lst_sq = norms.LeastSquares()\n",
    "plot_weights(support, lst_sq.weights, ['-3', '0', '3'], [-3, 0, 3]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Ramsay's Ea"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Help on function weights in module statsmodels.robust.norms:\n",
      "\n",
      "weights(self, z)\n",
      "    Ramsay's Ea weighting function for the IRLS algorithm\n",
      "    \n",
      "    The psi function scaled by z\n",
      "    \n",
      "    Parameters\n",
      "    ----------\n",
      "    z : array_like\n",
      "        1d array\n",
      "    \n",
      "    Returns\n",
      "    -------\n",
      "    weights : ndarray\n",
      "        weights(z) = exp(-a*\\|z\\|)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "help(norms.RamsayE.weights)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsIAAAHXCAYAAAC24TxsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nO3deXCd13nf8d9DgOAOgMDFQpGESGwEQe2GuVOkNkryTKIkEyd2PN6ajOI2TtM0mcRN29RN6okzk6SJJ05UxXEcJ23VNHVqpXEsS9x3idRmkSBAANxAEjsJcMN6n/7xXlxCMEiC5MV6vp8ZjYiLlxdHEnXxxbnnnNfcXQAAAEBoZkz0AAAAAICJQAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBI6RP1hWOxmC9btmyivjwAAAACceTIkTZ3zxv++ISF8LJly3T48OGJ+vIAAAAIhJmdHulxlkYAAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCdNsQNrNvmlmLmX1wk8+bmX3NzOrM7H0zeyz1wwQAAABSazQzwt+S9NwtPv+8pLLEXy9K+vN7HxYAAAAwtm4bwu6+W1LHLS55QdK3PXJQUraZLUrVAAEAAICxkIo1woslnR3ycWPiMQBAivT0D2hnTYv+ev8p1bdemejhAMC0kJ6C57ARHvMRLzR7UdHyCRUVFaXgSwPA9NXV3acdx1v0+rFm7axp1ZWe/uTnSvPn69lVBXpu1SI9sDhTZiO9FAMAbiUVIdwoaemQj5dIOj/She7+sqSXJamqqmrEWAaAkF3ovK7XjzXr9WPNOtjQrr4BV2z+LP3Yw4u0tbJQxXnztLOmVd//oEkv7WrQ13fUa3H2HD1TWaBnVxXqo8sWKj2NA4EAYDRSEcKvSvqimb0iaY2kTne/kILnBYBpz91V03xZrx9t1g+ONeuH5zolScV58/TzG4v1TGWBHl2arRkzbsz4fnb9PH12/TJdvNqrN6qb9drRJv2PN8/oW/tPKWdehp5ema9nVxVqQ2lMs2emTdQ/GgBMeuZ+64lZM/ufkrZIiklqlvSfJM2UJHd/yaL34/5U0ckS1yR93t0P3+4LV1VV+eHDt70MAKadgbjr8KkOvX4sit8zHdckSY8WZWtrZaGeqSxQaf78O3rOqz392lXbqteONml7dYsu9/RrXkaatqzI17MPFOqJFXlaMHvmWPzjAMCkZ2ZH3L3qRx6/XQiPFUIYQEiu9w5oz4lW/eBYs7Yfb1HH1V5lpM3QhtJcPVNZqKdX5is/c3ZKvlZvf1z769v02tFoiUXblR5lpM3Q+tJcPbsqCu3Y/Fkp+VoAMBUQwgAwztqv9GhbYrPbnhOt6u6LK3N2up6syNfWVYV6vDxP82elYoXazQ3EXW+fuajXPmjSa8eadLbjusykj96fo62ronXFS3PmjukYAGCiEcIAMA5Ot1+Nljwcbdbh0x2Ku3Rf1mxtTczErl6eo5kTtJnN3VV94bJeO9qk14426XjTZUnSqvsy9eyqQj27qlDlBfM5gQLAtEMIA8AYcHf98FynfpBYhlDTHMVlReECbV1VqK2VBVp13+Q83uxU21X94FiTvv9Bk94+c0mStCx3bhTFDxTqkSUf3qQHAFMVIQwAKdLbH9fBhvbkMWdNXd2aYdLq5Tl6pjKK36m23KClq1s/OBadQHGgvl39cVf+glnauqpAWysLtbY4VxnpHMsGYGoihAHgHnR192lnTat+cLRJu2padbmnX3Nmpunx8pi2VhbqyYp8LZyXMdHDTInOa33aXtOs1z5o1q7aVl3vG9CCWenaUpGvZyoLtGVFnjI5gQLAFEIIA8AdOn/put6oHn5ziww9VVGgZyoLtLFs+p/Te713QHvr2vT6sSZtq25R+9VezUwzrS3O1dbKAj1dWaBFWXMmepgAcEuEMADchrvreNPl5JKH5M0tYvP0TGWBtq4q0CNLFyot0HWzA3HXO2cu6geJfz8n265Kkh5cnJX897OiYMGkXA8NIGyEMACMoH8grrdOXYzit/rG8WKPLL37m1uEwN1V33olGcXvJDbbLc2Zo2dWRv/euN0zgMmCEAaAhKs9/dpd26rXjzVre02LLl3rU0b6DG0sjemZygI9tTJf+QtSc3OLULR0deuN6ha9fqxJ++rb1dsfV/bcmdGZyZUF2lSWp3ljfGYyANwMIQwgaC2Xu7WtOrq5xd66NvX2x5U1Z6aeSmwAe7ycUEuVoT9obDveos7r/KABYGLdLIR51QcwbdW1XEms923SO2cvyV1asnCOPrWmKLq5xbIc3rofA/Nmpev5Bxfp+QcXqX8grjdPdSTXXW8/3iIz6dGl2XqGpScAJhgzwgCmjYG4692zF5M3t2hIbOZ6YHGmnllZqK2rClRRyGauiXLLzYirCrS1MuzNiADGDksjAExLg8d7vXGsWduON6vtSq/SZySO91pVoKdXFui+bI73moyGHk83eBOPocfTbSiNaU7G9D6eDsD4IIQBTBtNnd3adrxZ26pbtK+uTT39cc2fla4tK/ISN3zIV9YcbvgwlYx0w5LZM2doQ0lMT1cW6KmKfOVnsq4YwN0hhAFMWe6uo+e79EZ1FL+Db6kvWThHT6+MZn1XL8/hFsDTRG9/XIdOtmtbdYveqG5W48XrkqSHlmTpqYpos92q+zJZ4gJg1AhhAFNKd9+ADtS3643qaIPVhc7u5CarpxLxW14wnxia5txdtc1X9EZ1s96obta7iU2Pi7Jm66mV+XpqZYHWFedO+zv8Abg3hDCASa/1co92HI9mAfecaNP1vgHNzUjTprKYnlpZoCcr8hWbP2uih4kJ1HalR9uPt2hb4s/Itd7oz8jG0pieXlmgJyrylbeAPyMAPowQBjDpuLtqmi/rjWPNeqO6Re813pjte3pl9Bb4Wmb7cBPdfQM62HBjCcXguwYPL8lOnlfMLZ8BSIQwgEmip39Ahxo6tK06it9zl6L1nw8vydJTifitXMT6T9wZd9exC13aVh3NFr/XGK0jX5w9R08nllCsKc7RrHR+qAJCRAgDmDAdV3u143iLth1v1u7aNl1JnAiwsTRa8sCJAEi1lq5ubU8ss9lb16buvrjmZaTp8fK85BKKnHkZEz1MAOOEEAYwbtxd9a1X9PqxaHbu7TMXFXcpf8GsxEa3fK0v4YxYjI/uvgHtq2vTG9Ut2n68Wc1dPZph0mNFC5N/Hkvz2XgJTGeEMIAx1dsf15snO7TteHTKw+n2a5KkVfdlJmPjgfuyNIO7hmECubs+ONeVPIXi6PkuSVJRztzoFIoKjuIDpiNCGEDKtXR1a0dNi7Yfb9HeE2262jugjPQZWl+Sm1zywF3dMJld6LyeXFe8r75dvf3REoqNZTE9WZGvJ1awbAeYDghhAPcsHne9f65T249HbzF/cC6aTVuUNVtPVuTryQqWPGDqutbbr/117dpe06IdibOrJenBxVl6IvHn+6HFvKsBTEWEMIC70tXdpz21bdp+vEW7alvUdqU3ub5yMA4qCjmiCtOLu+t40+XED30teiexzj02P0Oby/P11Mp8bSyLKXM2t/IGpgJCGMCoRBvdrmpHIgDeOtWh/rgra85MbS7P05MV+dpcnqeF7LhHQC5e7dWu2tbED4St6rzep/QZpo8uy4mWUFTkqyRvHj8QApMUIQzgpgbP9h2c/TrTEW10qyhckJz1fXRpttLT2EAE9A/E9c7ZS9pWHS2hqGm+LCnacDe4RIgzi4HJhRAG8CHNiXNWtx9v0b666Fa1s9JnaENpLBm/i9noBtxW48Vr2lHTqh2J/5d6+uOam5GmDaU3NtwVZrHhDphIhDAQuIG4673GS8klD4PHRi3OnpOcxVpXwu2MgXtxvXdABxqiNfU7jrcm75y46r7M5BKKh5dkK40Nd8C4IoSBAHVe79OeE9G6xp01req42qu0GaaPDNnoVl7AjQSAseDuqm2+kjxl5cjpaMNdzrwMbSnP0xMV+Xq8PE9Zc9hwB4w1QhgIgLur+sJl7ayNwvft0xfVH3dlz52Z/Ma7uTxP2XPZ6AaMt0vXog13O463aGdtqy5d64t+ML1/oZ5Yka8tK/I4gQUYI4QwME11dfdp34k27axp1a7aVjV1RWefVi7K1BMV0SkPjyxdyFuxwCQyEHe9c+ZitISiplXVF6KlSoWZs7W5PE9PVORpQ2lMCzieDUgJQhiYJgbPN91Z06odNS3JWd8Fs9P1eFmeNq/I05byPO6GBUwhTZ3d2pV4J2fviTZd7ulXemK2eAuzxcA9I4SBKexyd5/21bVpx/EfnfXdsiJPW1bk67EijjcDpoO+gbjePn1RO2tbtXPYbHH0/zuzxcCdIoSBKWTorO/OmhYdGTLru6ksFs0QMesLBOFWs8VPVESzxSsKmC0GboUQBia5wVnfKH5vzPquXJSpJxKzvo8WZWsms75AsIbOFu843qLjTdHNPBZlRWuLmS0GRkYIA5OMu6umObHW9/gIs77l+dq8Ik8FzPoCuImbzRZXLbuxtpjZYoAQBiaFobO+u2pbdaHzxqzvlsQmt8fuX8isL4A71jcQ15HTF5NLqobOFm9ZkafN5fnaUJrLbDGCRAgDE2DorO/OmhYdPpWY9Z2Vro1lseQ3J26/CiDVmC0GbiCEgXFy8Wqv9ta1aXdtq3afaFVzV48kqaJwQbSxhVlfAOPsZrPFBZmz9HhZnh4vz9PG0pgWzuNmO5ieCGFgjPQPxPVe4yXtqm3TrtpWvd94Se5S1pyZ2lgW0+bENxlmfQFMFk2d3dpd26pdJ6LZ4s7rfTKTHl6SrcfL87S5PKaHl3AkI6YPQhhIofOXrkffRGpbta+uTV3d/Zph0iNLo28ij5fn6eEl2dzNDcCkNxB3vdd4Kfma9t7ZS4q7lDk7WsI1OGN8X/aciR4qcNcIYeAedPcN6NDJDu2qiZY71LVckRRtQhn6tmLWXDahAJjaLl3r1b669mQYDx7lWJY/P/mD/prlOZo9M22CRwqMHiEM3AF3V13LFe1KfCN482SHevrjykifoTXLc7Q58c2gLH8+G00ATFvurhMtV5JRfOhkh3r745qVPkNrinP1eFlMm8vzVMprISY5Qhi4jc5rfdpX35ac9R082qw0f35i1jemNctzNSeDWRAAYbreO6BDJ9u1u7ZNu2pbVN96VZJ0X9bs5GzxhhLeHcPkQwgDwwzEXe83Xkq+oL+bWBe3YHa6NpbGki/qi1kXBwAjOje4X6KmVfvq23R5yH6JzeX5erw8pofYL4FJgBAGlNgpfeLGJrdL16Kd0g8tydbmsih+H1nKTmkAuFP9A3G9e/bGprv3z3XKXcqeOzM5ubC5nLtlYmIQwgjS9d4BvXmqQ3tqW7XnRJtqmqOzM/MXzErO+G4sjSmHszMBIKU6hpypvqu2Va2XozPVVxQs0KaymDaV52n1shyWm2FcEMIIQjzuOnahS3tOtGnPiVYdPnVRvQNxZaTNUNWyhclNbhWF3E0JAMaLu+t40+VkFA9/bd5UlqdNZTFVLsrUDJZRYAwQwpi2zl+6rr0n2rT7RKv217er42qvpOhObpvKYtpYxqwDAEwmg+/W7T0RvVs3eKe7nHkZ2lAa06bSmDaWxTi7GClzsxBOn4jBAPfiSk+/Dta3R2+5nWhVQ2LXct6CWdqyIppV2FAaU/4C1qEBwGQ0JyNNmxNrhiWp5XK39tW1aU9tm/bUtekf3zsvSSrJm6dNZdEStrUluZo/i2xBajEjjEmvfyCu9891au+JNu090aa3z1xUf9w1e+YMrVmeG601K8tTeQHnWALAVOfuqm2+oj2J2eJDJ9vV3RdX+gzTY0ULtbEspk1lnEaBO8PSCEwpZ9qvafeJVu090ab99dEtjM2kB+7LSr4IfuT+hZqVznIHAJjOevoHdOTURe2piyZDPjgfnUaROTtd60ti2lQe06bSPBXlzp3ooWISI4QxqXVe79OB+jbtTsz6num4Jik6pH1TWZ42JpY7cLoDAISt42qv9iWieM+JVp1P3PyoKGdu4h3CmNaVxJQ1h5t64AZCGJNK30Bc75y5pL0nWrX7RJveb4xuZjEvI03rSnKT8Vscm8dyBwDAiNxdDW1Xk1F8oL5dV3sHNMOkh5dmJzbd5enRomzN5Hz4oBHCmFDurvrWK9E637q2H32xShyd88hSXqwAAHenL3FTjz21rdpT16b3EncMnT8rXWuLc7QxcRpFSR57SkJDCGPcNXVGu4D31bVpX32bmruiw9Tvz52rjaXRBrd1Jbm8fQUAGBODy+72JCZhTrdHy+4KMmdpQ0m05G5DaUyFWZwyNN0Rwhhzndf6dKChXfvro/itTxxrljMvQ+tLcqMXnJIYGxoAABPiTPs17Ut8jxp67nxJ3rxkFK8tZoJmOiKEkXLdfQM6cvpictb3h+c6FXdpbkaaVi/PSf60XVG4gDsFAQAmlXg8utvd4LuWhxo6dL0vWrL34JJsbSjJ1cbSmB67f6Fmz+SEoqmOEMY9G4i7PjjXmfxp+vCpi+rpj852fGRpdvKn6UeWZisjnXW+AICpo7c/Wl+8t65N++va9M7ZSxqIu2alR7eBHnxX84HFWZxfPAURwrhjg7txB2d8D9S3q6u7X1J0++IofHO1ejl3+wEATC9Xevr15sl27T0RLfkbvA105ux0rUvMFq8v5XSjqYJbLGNUmrsGN7i1a19dm5q6ovMZF2fP0fMPLNKGspjWl+QqNn/WBI8UAICxM39Wup6sKNCTFQWSpNbLPck9MPvq2vXa0WZJ0qKs2VpfEtPGslxtKIkpP5ONd1MJM8KB6+ru08H6du2vb9feujbVtVyRJC2cO1PrE28DbSjNVVHOXH7iBQBA0TumpxMb7/bXtWtffZsuXeuTJJXmz49mi0tytbYkV5mz2Xg3GbA0ApKiDW5vn7mY/Il28EYWc2YmNriVRqc7rCzMZIMbAACjEI+7jl3oSmy8a9ebJ9vV3RfXDJMeWpKdWEaRq8eK2Hg3Ue4phM3sOUl/IilN0jfc/avDPp8l6W8lFSlabvEH7v5Xt3pOQnh89A3E9X5jpw4mjjUb3OCWNnSDW0muHi1ayAY3AABSoKd/QO+cuZTcY/NeY2dy491H7l+o9SW5WleSq4eWcBOp8XLXIWxmaZJqJT0jqVHSW5I+6e7HhlzzW5Ky3P03zSxPUo2kQnfvvdnzEsJjYyDuOna+SwcaojMS3zrZoau9A5KklYsyta44N7HBLUcLeLsGAIAxd7m7T4caOrS/vl0HGtpVfaFL0o3jRtcV52p9SUyV92VyIsUYuZfNcqsl1bl7Q+KJXpH0gqRjQ65xSQssWkQ6X1KHpP57HjVuy91V23xF++ujUx0ONtw42aE0f75+6rElWl+SqzXFucqZlzHBowUAIDwLZs/U05UFeroy2njXcbVXhxqi/Tn769u0s6ZVUnQixZri3OSMcXk+5/CPtdGE8GJJZ4d83ChpzbBr/lTSq5LOS1og6WfdPZ6SEeJD3F0n264mf6o8WN+u9sSdcYpy5upjDy7SupJcrSvOZecqAACTUM68DD3/4CI9/+AiSVJLV3d0Z9a66Hv768eiEyly52VobXEUxetLcrWco9pSbjQhPNK/8eHrKZ6V9K6kJyWVSHrdzPa4e9eHnsjsRUkvSlJRUdGdjzZQZzuu6UBDuw4kfnJs7uqRFB3ZsnlFntYl/idZspBbFwMAMNXkZ87WC48s1guPLJYkNV68pgP1g9/32/VPP7wgSSrInKX1JbHk9/2lOXzfv1ejCeFGSUuHfLxE0czvUJ+X9FWPFhzXmdlJSRWS3hx6kbu/LOllKVojfLeDnu6au7qT0XugoV1nO65LkmLzo58M15fEtK4kV8tyOdIMAIDpZsnCufp41Vx9vGqp3F2n2q8ll0Durm3VP7xzTpK0NGdOcn3xupJcFfBO8B0bTQi/JanMzJZLOifpE5J+btg1ZyQ9JWmPmRVIWiGpIZUDnc7ar/ToYENHMnwbWq9KkrLmzNTa4hz9wsZirSvJVVn+fMIXAICAmJmWx+ZpeWyePrXm/uTeoAP10ab473/QpL873ChJKs6blwzjtcU5yuXmV7c12uPTPibpjxUdn/ZNd/+KmX1Bktz9JTO7T9K3JC1StJTiq+7+t7d6zpBPjei83qdDDe3J5Q6Dt22cPys9uXt0XUmuVi5i9ygAALi5gbir+kJXcsb4zSGnRVUULkjuG1pTnKusOeGeFsUNNSZQ5/U+vXWyQ4dOtutgQ4eOnu9U3KXZM2eo6v6c5CL4BxdnKZ3zBAEAwF3qG4jrh+c6k0ssB+8fYCatui9Ta5fnam1xrj66PCeoMCaEx1HntT69eapDhxradfBku46e75K7lJE+Q48uzU6s883VI0XZmpXOHWYAAMDYGLy5x+ARq++cvaTeRBhXLsrU2uIojFcvy1HW3OkbxoTwGLp0rVdvnuzQwYZo1vfYhRvh+1hRdvIP2SNLs7m1IgAAmDDdfQN69+wlHWyIwvjtMx8O4zXLc7W2OEerl+coe+70uf8AIZxCl6716tDJDh1saNehhg5VN0XhOyt9hh4rWpgI3xw9TPgCAIBJrLtvQO+dvaSDDR2JML6xlGJlYTRjvKY4R2umeBgTwvfg4tUb4XuwoV01zZeT4Vu1bGHip6dcPbw0i6UOAABgyurpH9B7ZzuTzXPk9I0wrijM1NriHK1Znqs1y3O0cArdsZYQvgMdV3v1ZmJj28GGG6c6DG5uW7M8R2tLcvXQEsIXAABMXz39A3q/sVMH66N9T0dOX1R3X3Tz4IrCBcnln5M9jAnhW2i/0pNY4xvFb01zFL5zZqYlZnxztLY4Vw8tyVZGOqc6AACAMPX2x/V+46VkMx05fVHX+24c1za4PHT18lzlTKIwJoSHaPtQ+LartvmKpBvhO/gf8cHFhC8AAMDN9PbH9cNzN9YYHz714TAenExcvXxib/BBCEs63X5Vv/DXh3WiJQrfuRlpqlqWM2TGN0szOccXAADgrkRh/OE1xtcSN/goL5ivb3zmoyrKnTvu47pZCI/mFsvTRmHWbC3NmauffGyx1hZHN7AgfAEAAFIjI32GPnL/Qn3k/oX6pSdKP3SDjyOnL6oga3Ld9jmoGWEAAACE52YzwkyHAgAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSKMKYTN7zsxqzKzOzL50k2u2mNm7ZnbUzHaldpgAAABAaqXf7gIzS5P0dUnPSGqU9JaZverux4Zcky3pzyQ95+5nzCx/rAYMAAAApMJoZoRXS6pz9wZ375X0iqQXhl3zc5K+4+5nJMndW1I7TAAAACC1RhPCiyWdHfJxY+KxocolLTSznWZ2xMw+k6oBAgAAAGPhtksjJNkIj/kIz/MRSU9JmiPpgJkddPfaDz2R2YuSXpSkoqKiOx8tAAAAkCKjmRFulLR0yMdLJJ0f4Zrvu/tVd2+TtFvSw8OfyN1fdvcqd6/Ky8u72zEDAAAA92w0IfyWpDIzW25mGZI+IenVYdd8V9ImM0s3s7mS1kiqTu1QAQAAgNS57dIId+83sy9Kek1SmqRvuvtRM/tC4vMvuXu1mX1f0vuS4pK+4e4fjOXAAQAAgHth7sOX+46PqqoqP3z48IR8bQAAAITDzI64e9Xwx7mzHAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSKMKYTN7zsxqzKzOzL50i+s+amYDZvbTqRsiAAAAkHq3DWEzS5P0dUnPS6qU9Ekzq7zJdb8v6bVUDxIAAABItdHMCK+WVOfuDe7eK+kVSS+McN0vS/o/klpSOD4AAABgTIwmhBdLOjvk48bEY0lmtljST0p6KXVDAwAAAMbOaELYRnjMh338x5J+090HbvlEZi+a2WEzO9za2jraMQIAAAAplz6KaxolLR3y8RJJ54ddUyXpFTOTpFhAf0UAAAy+SURBVJikj5lZv7v/36EXufvLkl6WpKqqquExDQAAAIyb0YTwW5LKzGy5pHOSPiHp54Ze4O7LB39tZt+S9P+GRzAAAAAwmdw2hN2938y+qOg0iDRJ33T3o2b2hcTnWRcMAACAKWc0M8Jy9+9J+t6wx0YMYHf/3L0PCwAAABhb3FkOAAAAQSKEAQAAECRCGAAAAEEihAEAABAkQhgAAABBIoQBAAAQJEIYAAAAQSKEAQAAECRCGAAAAEEihAEAABAkQhgAAABBIoQBAAAQJEIYAAAAQSKEAQAAECRCGAAAAEEihAEAABAkQhgAAABBIoQBAAAQJEIYAAAAQSKEAQAAECRCGAAAAEEihAEAABAkQhgAAABBIoQBAAAQJEIYAAAAQSKEAQAAECRCGAAAAEEihAEAABAkQhgAAABBIoQBAAAQJEIYAAAAQSKEAQAAECRCGAAAAEEihAEAABAkQhgAAABBIoQBAAAQJEIYAAAAQSKEAQAAECRCGAAAAEEihAEAABAkQhgAAABBIoQBAAAQJEIYAAAAQSKEAQAAECRCGAAAAEEihAEAABAkQhgAAABBIoQBAAAQJEIYAAAAQSKEAQAAECRCGAAAAEEihAEAABAkQhgAAABBIoQBAAAQJEIYAAAAQSKEAQAAECRCGAAAAEEihAEAABAkQhgAAABBIoQBAAAQJEIYAAAAQSKEAQAAECRCGAAAAEEaVQib2XNmVmNmdWb2pRE+/ykzez/x134zezj1QwUAAABS57YhbGZpkr4u6XlJlZI+aWaVwy47KWmzuz8k6XclvZzqgQIAAACpNJoZ4dWS6ty9wd17Jb0i6YWhF7j7fne/mPjwoKQlqR0mAAAAkFqjCeHFks4O+bgx8djN/Lykf76XQQEAAABjLX0U19gIj/mIF5o9oSiEN97k8y9KelGSioqKRjlEAAAAIPVGMyPcKGnpkI+XSDo//CIze0jSNyS94O7tIz2Ru7/s7lXuXpWXl3c34wUAAABSYjQh/JakMjNbbmYZkj4h6dWhF5hZkaTvSPq0u9emfpgAAABAat12aYS795vZFyW9JilN0jfd/aiZfSHx+Zck/bakXEl/ZmaS1O/uVWM3bAAAAODemPuIy33HXFVVlR8+fHhCvjYAAADCYWZHRpqk5c5yAAAACBIhDAAAgCARwgAAAAgSIQwAAIAgEcIAAAAIEiEMAACAIBHCAAAACBIhDAAAgCARwgAAAAgSIQwAAIAgEcIAAAAIEiEMAACAIBHCAAAACBIhDAAAgCARwgAAAAgSIQwAAIAgEcIAAAAIEiEMAACAIBHCAAAACBIhDAAAgCARwgAAAAgSIQwAAIAgEcIAAAAIEiEMAACAIBHCAAAACBIhDAAAgCARwgAAAAgSIQwAAIAgEcIAAAAIEiEMAACAIBHCAAAACBIhDAAAgCARwgAAAAgSIQwAAIAgEcIAAAAIEiEMAACAIBHCAAAACBIhDAAAgCARwgAAAAgSIQwAAIAgEcIAAAAIEiEMAACAIBHCAAAACBIhDAAAgCARwgAAAAgSIQwAAIAgEcIAAAAIEiEMAACAIBHCAAAACBIhDAAAgCARwgAAAAgSIQwAAIAgEcIAAAAIEiEMAACAIBHCAAAACBIhDAAAgCARwgAAAAgSIQwAAIAgEcIAAAAIEiEMAACAIBHCAAAACBIhDAAAgCARwgAAAAgSIQwAAIAgjSqEzew5M6sxszoz+9IInzcz+1ri8++b2WOpHyoAAACQOrcNYTNLk/R1Sc9LqpT0STOrHHbZ85LKEn+9KOnPUzxOAAAAIKVGMyO8WlKduze4e6+kVyS9MOyaFyR92yMHJWWb2aIUjxUAAABImdGE8GJJZ4d83Jh47E6vAQAAACaN0YSwjfCY38U1MrMXzeywmR1ubW0dzfgAAACAMTGaEG6UtHTIx0sknb+La+TuL7t7lbtX5eXl3elYAQAAgJQZTQi/JanMzJabWYakT0h6ddg1r0r6TOL0iLWSOt39QorHCgAAAKRM+u0ucPd+M/uipNckpUn6prsfNbMvJD7/kqTvSfqYpDpJ1yR9fuyGDAAAANy724awJLn79xTF7tDHXhrya5f0S6kdGgAAADB2uLMcAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgkQIAwAAIEiEMAAAAIJECAMAACBIhDAAAACCRAgDAAAgSIQwAAAAgmTuPjFf2KxV0ukJ+eJSTFLbBH1tALgXvH4BmKom8vXrfnfPG/7ghIXwRDKzw+5eNdHjAIA7xesXgKlqMr5+sTQCAAAAQSKEAQAAEKRQQ/jliR4AANwlXr8ATFWT7vUryDXCAAAAQKgzwgAAAAjctAphM/tVM3vLzNrNrNvM6szsD80s9ybXf8vMtozzMAHgjpjZUjP7ezPrNLMuM/uOmRVN9LgAhM3MnjWz7WbWZGY9ZtZoZn9nZpUjXPs5M/vyBAzzlqZVCEvKkfQdSZ+T9Jykr0v6F5JeN7MZkmRmP25mTw39TWY228x+82bBDAATxczmStouqULSZyV9WlKZpB1mNm8ixwYgeDmSjkj6oqStkv6dpFWSDprZ/WZWbmb/0szShv4mM/sZM9s4/sP9UdN+jbCZ/aKklyRVufsRM1sp6XckmaRMSe9KelLS30v6E3e/PmGDBYBhzOxXJP2RpBXuXpd4bLmkE5J+w93/aCLHBwBDmdkKSccl/bqkv5L0JUkbJb0vKVtSrqQGSb/j7ucmapyDptuM8EjaE3/vkyR3r3b3j0v6J0lPS/qEpJ9y968ORrCZbTWz75nZBTO7ZmYfmNmvDf+JBgDGwY9LOjgYwZLk7icl7ZP0woSNCgBGluwud+9w99+Q9K8k/Yykn5T0B+7+i4MRbGalZvY3ZnbSzK6bWYOZ/bmZLRyPwU7LGWEzS5eUIekhSX8p6YK7P534XLmk/5z4/OCM8BOS/rekr7n7dTP7gqR5ko5J6pZUJem3JX3d3b80zv84AAJmZk2Svuvuvzjs8T+T9PGRbhkKAOMpMVGYJul+SV+VtF7Sw4omIX9D0mbdmBGOSapTNCN83swel/S8pIOSLkoqlvRbktrdfd1Yjz19rL/AeDOz+ZIuD3noNUkfH/LxSkl/6e5vmNm3FM0M/0dJvyJprqTr7v7SkOczSXsUhfOvm9lvuXt8bP8pACApR9E3h+E6JI3LjAkA3MYhSR9J/LpO0pPu3pJYJnFG0iZF+xuWufuXzexnFQXveXffLWn34BOZ2f7Ec+wxs0fd/Z2xHPiUDOFEnH5omYK79yd+eU3SRyXNlvSopH8v6R/N7Gl373f37w5/PnfvlvT7Q55/kaQvK9pwd58+/O8pX1JTyv5hAOD2RnrrzsZ9FAAwsk8repe9WNHa4NfNbKO710iqkaQo3SLu/r8Gf21mGYnf8xlFM8qzhzzvCkmE8Ag2S9ox7DGTpMRs7eHEY3vN7IeJa39a0itDf4O7f274EydOl3hVUQB/WdGC7+uSfkJRVM8e/nsAYAxdVDQrPNxCjTxTDADjyt2rE788ZGb/LOmUok1yXxhyzbdu8tt/T9IvKzrIYL+id/WXKDoFbMyba6qG8BFFs76jMRjFpaO8vkTRmuBPu/vfDj5oZj82+uEBQMocVXQc0XCVivYxAMCk4e6XzKxOo++uT0j6trv/l8EHEstcx8WUPDXC3S+7++Ghf93i8s2Jv9eP8unnJv7eN/iAmc2U9Km7GCoA3KtXJa01s+LBB8xsmaQNic8BwKRhZgWKzj2/k+7qG/bY51M6qFuYqjPCP8LMsiR9X9J/V3S+pktaLenfSnpP0RT7aFRLOi3pK2Y2oOg/zq+mfMAAMDp/oeiw+u+a2X9Q9Nr2u5LOSvpvEzkwAGEzs3+Q9LaiEyG6JJUraqZ+SX84yqf5vqTPJpay1kn6KUWnToyLaRPCio45q5b0ryUtVvQf4ZSi/xBfc/ee0TyJu/ea2U9I+lNJ31a0M/ubinY9/kXqhw0AN+fuV83sSUn/VdLfKNoPsU3Sv3H3KxM6OAChO6jofOBfU3S61llJOyX9nrufGuVz/LKi17WvJD7+nqRPSnozlQO9mWl5jjAAAABwO1NyjTAAAABwrwhhAAAABIkQBgAAQJAIYQAAAASJEAYAAECQCGEAAAAEiRAGAABAkAhhAAAABIkQBgAAQJD+P6KAolXeZTeLAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "a = .3\n",
    "support = np.linspace(-3*a, 3*a, 1000)\n",
    "ramsay = norms.RamsayE(a=a)\n",
    "plot_weights(support, ramsay.weights, ['-3*a', '0', '3*a'], [-3*a, 0, 3*a]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Trimmed Mean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Help on function weights in module statsmodels.robust.norms:\n",
      "\n",
      "weights(self, z)\n",
      "    Least trimmed mean weighting function for the IRLS algorithm\n",
      "    \n",
      "    The psi function scaled by z\n",
      "    \n",
      "    Parameters\n",
      "    ----------\n",
      "    z : array_like\n",
      "        1d array\n",
      "    \n",
      "    Returns\n",
      "    -------\n",
      "    weights : ndarray\n",
      "        weights(z) = 1             for \\|z\\| <= c\n",
      "    \n",
      "        weights(z) = 0             for \\|z\\| > c\n",
      "\n"
     ]
    }
   ],
   "source": [
    "help(norms.TrimmedMean.weights)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "c = 2\n",
    "support = np.linspace(-3*c, 3*c, 1000)\n",
    "trimmed = norms.TrimmedMean(c=c)\n",
    "plot_weights(support, trimmed.weights, ['-3*c', '0', '3*c'], [-3*c, 0, 3*c]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tukey's Biweight"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Help on function weights in module statsmodels.robust.norms:\n",
      "\n",
      "weights(self, z)\n",
      "    Tukey's biweight weighting function for the IRLS algorithm\n",
      "    \n",
      "    The psi function scaled by z\n",
      "    \n",
      "    Parameters\n",
      "    ----------\n",
      "    z : array_like\n",
      "        1d array\n",
      "    \n",
      "    Returns\n",
      "    -------\n",
      "    weights : ndarray\n",
      "        psi(z) = (1 - (z/c)**2)**2          for \\|z\\| <= R\n",
      "    \n",
      "        psi(z) = 0                          for \\|z\\| > R\n",
      "\n"
     ]
    }
   ],
   "source": [
    "help(norms.TukeyBiweight.weights)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsIAAAHXCAYAAAC24TxsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nO3dd3Sc533m/euHXgcgGlEIEgQrQIqiJIrqtmzHtpRixUmcWNkUe+N19CbOJptsYudkW5JzspuT9saxN361Kd5UxUmU2E5ky92x1UhKokgCYAErGomOQW9zv38AQ8EQKA6Bmbln5vl+ztExMXwwcxGCoYvP/O77NuecAAAAgKDJ8h0AAAAA8IEiDAAAgECiCAMAACCQKMIAAAAIJIowAAAAAokiDAAAgEDK8fXCVVVVrqmpydfLAwAAICBefvnlQedc9erHvRXhpqYmHTt2zNfLAwAAICDM7PJajzMaAQAAgECiCAMAACCQKMIAAAAIJIowAAAAAokiDAAAgECiCAMAACCQKMIAAAAIJIowAAAAAokiDAAAgECiCAMAACCQKMIAAAAIJIowAAAAAokiDAAAgECiCAMAACCQKMIAAAAIJIowAAAAAokiDAAAgECiCAMAACCQKMIAAAAIJIowAAAAAokiDAAAgECiCAMAACCQKMIAAAAIJIowAAAAAokiDAAAgECiCAMAACCQKMIAAAAIJIowAAAAAokiDAAAgECiCAMAACCQblqEzezPzKzfzE7d4PfNzD5uZp1mdsLM7ox/TAAAACC+Yrkj/GlJj7zJ7z8qadfyPx+W9McbjwUAAAAk1k2LsHPu3yQNv8klj0n6C7fkRUnlZlYXr4AAAABAIuTE4TkaJHWt+Lh7+bG+ODw3AASac07n+if0XOeguoanZSY1birUfTuqtKe21Hc8AEhr8SjCtsZjbs0LzT6spfEJbd26NQ4vDQCZ65tnB/T7Xz6r17pGJUlFedlyTpqeX5Qk3b6lTP/53Xv00K5qnzEBIG3Fowh3S2pc8fEWSb1rXeice1LSk5J06NChNcsyAATd1NyC/ss/n9LTr/Roa0WRfv09+/SufZtVGyqQmalvbFrPnrqqP33uon78T4/oB+/cot94bJ+K8+PxIx0AgiMePzU/J+kjZvaUpHskjTnnGIsAgHUYGJ/VT/7ZEXVcDes/vn2nfvbtO5Wfk/0d19SVFeoDD2zX4/ds1Se+1qlPfL1Tp6+G9ecfvFs1pQWekgNA+oll+7S/lfSCpD1m1m1mP2VmT5jZE8uXPCPpgqROSf9H0s8kLC0AZLDhyTn92J+8pIuDk/qzD9ytX3zXnjeU4JXyc7L1S+/aoz/7ybt1YWBS73/yRQ1PziUxMQCkN3POz4TCoUOH3LFjx7y8NgCkmqm5Bb3vUy+os39Cf/6Bu3X/zqpb+vwjF4f143/6klrqQvqb/3CPivIYkwCAKDN72Tl3aPXjnCwHAJ455/SrT59Ue19Yn/qxu265BEvS4e0V+vjjd+hE96h++e9PyNdNDgBIJxRhAPDs/z5/SZ893qtfeuduvW1vzbqf5937avXL796rfz3Zp7890nXzTwCAgKMIA4BHnf0T+q0vnNY79tboZx7eueHn++m3NOuhXVX69c+36ey18TgkBIDMRREGAE8WI04f/ccTKsrL1v/8wduUlbXWtuy3JivL9Ps/fFDF+Tn66D+eUCTCiAQA3AhFGAA8+csXLunlyyP679/XGtdtz6pL8/Vr392iV6+M6m+PXonb8wJApqEIA4AHw5Nz+r0vn9VDu6r0/Qcb4v78P3Bng+5rrtRvf+G0BsZn4/78AJAJKMIA4MEffuWspuYW9d++t1VmGx+JWM3M9Jvfv1+Tc4v6o6+di/vzA0AmoAgDQJJ19k/or166oscPN2rX5tKEvc7OmhI9frhRf/PSFV0anEzY6wBAuqIIA0CS/cGXz6owN1u/8F27E/5a//Edu5SbnaXf/dKZhL8WAKQbijAAJNHZa+N65lSffvL+baoqyU/469WUFuhDD23Xv5zoU0dfOOGvBwDphCIMAEn08a+eU1Futj70YHPSXvNDDzarOC9bf/yN80l7TQBIBxRhAEiSzv4J/evJPv3E/U3aVJyXtNctK8rVj923Tf9yopdZYQBYgSIMAEnyp9++oLzsLH3owe1Jf+2fenC7crKz9KlvclcYAKIowgCQBEMTs3r6lR79wJ1bVJmE2eDVakoL9MOHtujpV3o0OMG+wgAgUYQBICn++qUrml2I6KcebPKW4QP3b9fcYkRPHeG0OQCQKMIAkHCzC4v6ixcu6+E91dpZk7h9g29mZ02JHtpVpb988bLmFyPecgBAqqAIA0CCfantmgYnZvXBB5I/G7zaBx9o0rXwrL546qrvKADgHUUYABLsqaNXtGVToR7aWeU7ih7eXaNtlUX6yxcu+44CAN5RhAEggS4PTeq5ziH9yKFGZWWZ7zjKyjK9/+6tOnJpWBcGJnzHAQCvKMIAkEBPHe1SlknvO9ToO8p1P3hng7KzTP/wcrfvKADgFUUYABJkfjGivz/Wrbfv3azasgLfca6rCRXo4d3V+sdXurXAojkAAUYRBoAE+WpHvwYnZvX44dS5Gxz1vkONuhae1bfODfqOAgDeUIQBIEH+4eVubQ7l6627q31HeYO3761RZXGePnOsy3cUAPCGIgwACTA6Nadvnu3Xe26vV0526v2ozcvJ0nvvaNBXOq5piJPmAARU6v10BoAM8MzJq5pfdHrsYIPvKDf0g3dt0fyi0xfYUxhAQFGEASABPnu8R83VxdpXH/Id5Yb21pZqZ02JPv9ar+8oAOAFRRgA4qx3dFpHLg3r+w82yMz/3sE3Ymb6vgP1OnJpWFfHZnzHAYCkowgDQJx9/rVeOSe95/Z631Fu6ntvr5Nz0r+e7PMdBQCSjiIMAHH22eO9ur2xXE1Vxb6j3NSO6hLtqw8xHgEgkCjCABBHFwcn1d4XTou7wVHfd3u9jneNqmt4yncUAEgqijAAxNGzbUs7MDyyv9Zzkth974E6SdLnuCsMIGAowgAQR188dVUHtpSpobzQd5SYbdlUpION5ddLPAAEBUUYAOKkb2xax7tG9e596XM3OOpd+zbrRPeYekenfUcBgKShCANAnHyp7ZokpWURjmb+Ssc1z0kAIHkowgAQJ188dVU7a0q0s6bEd5RbtqO6RDuqi6+XeQAIAoowAMTB8OScjlwa1iNpeDc46l37avXihSGNTc37jgIASUERBoA4+ErHNS1GXFqORUS9q3WzFiJOXz/T7zsKACQFRRgA4uBrHf2qKyvQ/oaQ7yjrdvuWctWU5utL7eweASAYKMIAsEFzCxF9u3NQD++pkZn5jrNuWVmmd7Zu1jfODGhmftF3HABIOIowAGzQ0UvDmphd0Nv31viOsmHf1bJZU3OLOnJx2HcUAEg4ijAAbNDXTvcrLydLD+ys9B1lw+5trlReTpa+cWbAdxQASDiKMABs0NdP9+ve5koV5eX4jrJhhXnZuq+5Ut9gwRyAAKAIA8AGXBqc1IXBSb0jA8Yioh7eU60Lg5O6PDTpOwoAJBRFGAA24Gunl+6cZsJ8cNTb9iz9WRiPAJDpKMIAsAFfP9OvnTUlaqwo8h0lbpqqitVUWcR+wgAyHkUYANZpem5RL10Y1sO7q31HibuH99TohfNDbKMGIKNRhAFgnY5cGtbcYkQPZWQRrtbsQkQvXBjyHQUAEoYiDADr9O1zA8rLztLhpgrfUeLu3uZK5edk6ZvMCQPIYBRhAFinb50b1KGmTSrMy/YdJe4KcrN1eHuFnj8/6DsKACQMRRgA1mFgfFanr47rwV1VvqMkzP07qnT22oT6wzO+owBAQlCEAWAdnutculP60M7Mmw+OenDnUsl//jxzwgAyE0UYANbhW+cGtakoV/vqQ76jJExrfUjlRbn6difjEQAyE0UYAG6Rc07f7hzQ/TurlJVlvuMkTHaW6f4dlXquc1DOOd9xACDuKMIAcIs6+yd0LTyrh3Zm7nxw1AM7q9Q3NqMLgxy3DCDzUIQB4BZFRwUyeaFc1AM7lv6MzzEeASADUYQB4BZ9+9ygtlcVa8umzDlW+Ua2VRapobyQIgwgI1GEAeAWLCxG9OKFIT2ws9J3lKQwMz24s0rPnx/SYoQ5YQCZhSIMALfgVG9Yk3OLuq8588cioh7YVaXxmQWd7BnzHQUA4ooiDAC34MULS3vqHt6eeccq38j9O5bufjMeASDTUIQB4Ba8dGFIO2tKVF2a7ztK0lSV5Gv35pLrfwkAgExBEQaAGC0sRnT00ojubQ7O3eCoe7ZX6uXLI5pfjPiOAgBxQxEGgBi19YY1Mbuge7YHY6HcSvc0V2hqblFtvWHfUQAgbijCABCjly4ujQbcE8A7wtGZ6JcYjwCQQSjCABCjFy8Mq7m6WDWlBb6jJF1NaYGaq4p15OKw7ygAEDcxFWEze8TMzphZp5l9bI3fLzOzz5vZa2bWZmYfjH9UAPBnMeJ09OKw7m0O3lhE1D3NFTpyaZj9hAFkjJsWYTPLlvRJSY9KapX0uJm1rrrsZyW1O+dul/SwpN8zs7w4ZwUAb9p7wxqfXdA9Ado2bbV7tldqfGZBHX3MCQPIDLHcET4sqdM5d8E5NyfpKUmPrbrGSSo1M5NUImlY0kJckwKAR9Gtw4J8Rzg6J8x4BIBMEUsRbpDUteLj7uXHVvqEpBZJvZJOSvp55xx77ADIGC9dHFJzVbE2h4I3HxxVX16oxorC64sGASDdxVKEbY3HVg+IvVvScUn1kg5K+oSZhd7wRGYfNrNjZnZsYGDglsMCgA+RiNORi8OB3C1itXu2V+rIxWE5x5wwgPQXSxHultS44uMtWrrzu9IHJT3tlnRKuihp7+oncs496Zw75Jw7VF1dvd7MAJBUZ/vHFZ5Z0N1NFOF7tldoZGpe5/onfEcBgA2LpQgflbTLzLYvL4B7v6TPrbrmiqR3SJKZbZa0R9KFeAYFAF+OXhqRJB3aRhGOHibCfsIAMsFNi7BzbkHSRyQ9K6lD0mecc21m9oSZPbF82W9Kut/MTkr6qqSPOucGExUaAJLp5UvDqi7NV2NFoe8o3jVWFKo2VKAjy385AIB0lhPLRc65ZyQ9s+qxT634da+kd8U3GgCkhmOXR3R30yYtbYwTbGamu5o26ZXLFGEA6Y+T5QDgTVwdm1H3yLTuYiziukPbNqlndFp9Y9O+owDAhlCEAeBNHLu8tGfuoW2bPCdJHXctfy1e5q4wgDRHEQaAN3Hs0ogKc7PVWv+GHSEDq6UupMLcbB1jThhAmqMIA8CbOHZ5WAcby5WbzY/LqNzsLN3eWKZXrlCEAaQ3frIDwA1Mzi6oo29ch5oYi1jt0LYKtfWGNTW34DsKAKwbRRgAbuB416gWI+76TCxed9e2TVqMOL3WNeY7CgCsG0UYAG7g6KVhmUl3UoTf4M6t0QVzw56TAMD6UYQB4AZevjyiPZtLFSrI9R0l5ZQV5WpXTYmOsXMEgDRGEQaANSxGnF69Msp88Js4tHywRiTifEcBgHWhCAPAGs71j2tidoH54Ddx59ZNCs8sqHNgwncUAFgXijAArOHVK6OSpIONFOEbOdS0dNoeB2sASFcUYQBYw/EroyovylVTZZHvKCmrqbJIlcV5HKwBIG1RhAFgDa92jehgY7nMzHeUlGVmOthYruNdFGEA6YkiDACrjM/M61z/hO5gLOKmDjaW6/zApMam531HAYBbRhEGgFVOdI/JOeng1nLfUVJe9Gt0onvUcxIAuHUUYQBY5dUrS2/1H9xCEb6ZA8tfo+NXKMIA0g9FGABWOd41qubqYpUVcZDGzZQV5mpHdbFe444wgDREEQaAFZxbOkiD+eDYHWzcpONdo3KOgzUApBeKMACs0D0yraHJOd3BfHDMDm4t1+DEnLpHpn1HAYBbQhEGgBVeic4HN1KEY3XH8tfqeBfjEQDSC0UYAFY43jWqgtws7a0t9R0lbeypLVV+ThZFGEDaoQgDwAqvXhnVgYZy5WTz4zFWudlZuq2hjCIMIO3wkx4Als0uLKq9N8x88DocbCzXqZ4xzS9GfEcBgJhRhAFgWXtvWHOLEeaD1+Hg1nLNLkR0um/cdxQAiBlFGACWRd/av2MrW6fdqoPXF8yNeE4CALGjCAPAslevjKo2VKDasgLfUdJOQ3mhqkry9CpzwgDSCEUYAJad7BnT7Y1lvmOkJTPTwcZyFswBSCsUYQCQNDY9r4uDkzqwhfng9TrYWK4LA5Mam5r3HQUAYkIRBgBJp3rGJEm3NXBHeL0OLh9L/Vo3d4UBpAeKMABIOtFNEd6o6Nfu5PJfKgAg1VGEAUDSyZ5Rba0o0qbiPN9R0lZZUa6aKot0spsiDCA9UIQBQEt3hG/bwt3gjbptSzl3hAGkDYowgMAbmphV98i0DjAWsWEHGsrUMzqtwYlZ31EA4KYowgACL3oHkzvCGxf9GnJXGEA6oAgDCLyTLJSLm331IUnSKeaEAaQBijCAwDvRM6bm6mKVFuT6jpL2Sgty1VxdrBPcEQaQBijCAALvRPco88FxdKChjJ0jAKQFijCAQLsWntG18Kxu40S5uLltS7muhmfUH57xHQUA3hRFGECgRe9c3s5Cubg5wII5AGmCIgwg0E70jCnLpNblRV7YuNa6kMxeP60PAFIVRRhAoJ3oHtWumlIV5eX4jpIxivNztLO6RKe4IwwgxVGEAQSWc04nOVEuIW7bUqYTPWNyzvmOAgA3RBEGEFi9YzMampxjPjgBDjSUaWB8VtfCnDAHIHVRhAEE1omuUUlix4gEiH5NT3SPek4CADdGEQYQWCd6xpSTZdpbW+o7SsZprQspO8vYOQJASqMIAwisk91j2ltXqoLcbN9RMk5hXrZ21ZRQhAGkNIowgEByzulkz5hu40S5hLlt+YQ5FswBSFUUYQCB1D0yrbHpee2nCCfMgS1lGpqcU+8YJ8wBSE0UYQCB1Na79Jb9vnqKcKJEF8ydZMEcgBRFEQYQSKd6wspmoVxC7a0tVU6WccIcgJRFEQYQSG29Y9pZXcJCuQQqyM3WzpoStfWGfUcBgDVRhAEEUltvWPsaQr5jZLx99WVq62XBHIDURBEGEDj94zPqH59lPjgJ9jeENDgxp/5xTpgDkHoowgACJ/pW/f567ggnWvQvG9HFiQCQSijCAAKnbfmQh1aKcMK11C0tRmzrYU4YQOqhCAMInLbesJoqi1RakOs7SsYrLchVU2URC+YApCSKMIDAOdU7xnxwEu2rL1NbH6MRAFIPRRhAoIxNzatreJodI5JoX0NIXcPTGpua9x0FAL4DRRhAoETvTHJHOHmuL5jjrjCAFEMRBhAo7cuzqvtYKJc00a91O3PCAFIMRRhAoJzqGVNtqEBVJfm+owRGVUm+NofyWTAHIOVQhAEESltvmLvBHkRPmAOAVEIRBhAY03OLOj8woX0NzAcn2/76kM4PTGpmftF3FAC4LqYibGaPmNkZM+s0s4/d4JqHzey4mbWZ2TfjGxMANq7jalgRx3ywD631ZVqMOJ2+Ou47CgBcd9MibGbZkj4p6VFJrZIeN7PWVdeUS/rfkt7jnNsn6X0JyAoAG3L9aGXuCCdd9C8fp3oYjwCQOmK5I3xYUqdz7oJzbk7SU5IeW3XNj0p62jl3RZKcc/3xjQkAG9fWM6byolzVlxX4jhI4WzYVqqwwlwVzAFJKLEW4QVLXio+7lx9babekTWb2DTN72cx+Il4BASBe2nrD2l9fJjPzHSVwzEytdSG1s2AOQAqJpQiv9V8Mt+rjHEl3SfoeSe+W9F/NbPcbnsjsw2Z2zMyODQwM3HJYAFiv+cWIzlwdZz7Yo331IZ2+Oq6FxYjvKAAgKbYi3C2pccXHWyT1rnHNF51zk865QUn/Jun21U/knHvSOXfIOXeourp6vZkB4JaduzahucUIO0Z4tL+hTLMLEZ0fmPQdBQAkxVaEj0raZWbbzSxP0vslfW7VNZ+V9JCZ5ZhZkaR7JHXENyoArN+p3ujRytwR9iX6tWc/YQCp4qZF2Dm3IOkjkp7VUrn9jHOuzcyeMLMnlq/pkPRFSSckHZH0J865U4mLDQC3pr03rOK8bG2vLPYdJbCaq0tUkJulUz0smAOQGnJiucg594ykZ1Y99qlVH/+OpN+JXzQAiJ9TPWNqqQspK4uFcr5kZ5n21oa4IwwgZXCyHICMF1k+yKGVsQjv9tWH1N4XlnOr11wDQPJRhAFkvO6RaU3MLqiljiLs2/6GMo3PLKhreNp3FACgCAPIfO19SzOpFGH/WDAHIJVQhAFkvI6+sLJM2rO51HeUwNu9uVTZWXb9LycA4BNFGEDG6+gLq6mqWIV52b6jBF5BbrZ2VBergyIMIAVQhAFkvI6rYcYiUkhLXUjtvRRhAP5RhAFktPGZeXUNT6uVIpwyWupC6h2b0ejUnO8oAAKOIgwgo52+Oi5JaqljPjhVRO/Od/SNe04CIOgowgAyWgc7RqSc6F9KmBMG4BtFGEBG6+gLq7woV7WhAt9RsKymtEBVJfnsHAHAO4owgIzW3jeultqQzDhaOZW01JVyRxiAdxRhABlrMeJ0hh0jUlJrXUjnrk1ofjHiOwqAAKMIA8hYFwcnNTMfYaFcCmqpC2luMaILA5O+owAIMIowgIzFQrnU1bp81HJ7H0ctA/CHIgwgY3X0hZWTZdq1ucR3FKzSXFWsvJwstlAD4BVFGEDG6ugLa0d1ifJzOFo51eRkZ2n35hIWzAHwiiIMIGN19I0zH5zCWmqXjlp2zvmOAiCgKMIAMtLI5JyuhmeYD05hLXUhDU3OaWB81ncUAAFFEQaQkVgol/peXzDHeAQAPyjCADJSO0U45bXULv27YcEcAF8owgAyUkffuKpK8lVdmu87Cm6grChXDeWFLJgD4A1FGEBG6ugLX3/rHamLo5YB+EQRBpBx5hcj6uyfYMeINNBSF9L5gQnNzC/6jgIggCjCADLO+YEJzS1G1Mp8cMprrQsp4qSz15gTBpB8FGEAGYcdI9JH9N8R4xEAfKAIA8g4HX3jysvJUnNVse8ouImtFUUqzstm5wgAXlCEAWScjr6wdm8uUU42P+JSXVaWaU9tqdp7uSMMIPn4rwSAjNPRF76+Ry1SX2t9SB1XOWoZQPJRhAFklP7xGQ1OzDEfnEZa6kIan1lQ98i07ygAAoYiDCCjRGdNKcLpgwVzAHyhCAPIKNEyxdZp6WNvbanMXj8WGwCShSIMIKN09IVVX1agsqJc31EQo6K8HDVVFnNHGEDSUYQBZJSOvjBjEWmotS7EFmoAko4iDCBjzMwv6vzAJEU4DbXUlerK8JTGZ+Z9RwEQIBRhABmjs39CixFHEU5D0X9np69yVxhA8lCEAWSM9utHK5d6ToJbxc4RAHygCAPIGB19YRXmZmtbJUcrp5u6sgKVF+VShAEkFUUYQMbo6AtrT22psrPMdxTcIjNTS21I7SyYA5BEFGEAGcE5p46+ceaD01hLXUhnroa1GOGoZQDJQREGkBH6xmY0Nj2vVuaD01ZLXalm5iO6ODjpOwqAgKAIA8gIHdcXynFHOF2xYA5AslGEAWSEaHnaSxFOW7s2lygnyyjCAJKGIgwgI3T0jWtrRZFK8nN8R8E65edka0d1CUUYQNJQhAFkhKWjlZkPTnctdaUctQwgaSjCANLe1NyCLg5xtHImaKkL6Wp4RiOTc76jAAgAijCAtHfm6ricY6FcJmitZ8EcgOShCANIe9G30lspwmkv+peZdoowgCSgCANIe+19YyrNz9GWTYW+o2CDqkryVV2az5wwgKSgCANIex1949pbVyozjlbOBC11IUYjACQFRRhAWotEnE73hZkPziAtdaXq7J/Q3ELEdxQAGY4iDCCtdY1MaXJukSKcQVrrQppbjOj8wITvKAAyHEUYQFrjaOXMw1HLAJKFIgwgrbX3jSvLpD2bOUwjUzRXFSsvJ4siDCDhKMIA0lpHX1hNVcUqzMv2HQVxkpOdpd2bS9g5AkDCUYQBpLUOFsplpJbapZ0jnHO+owDIYBRhAGkrPDOv7pFpDtLIQC11IQ1NzmlgfNZ3FAAZjCIMIG2dXn7rvKWO+eBMwwlzAJKBIgwgbbFjROZqvb5zBHPCABKHIgwgbXX0hVVelKvaUIHvKIizsqJc1ZcVsHMEgISiCANIWx19YbXUhjhaOUNx1DKARKMIA0hLixGnM9fGGYvIYC11IV0YnNTM/KLvKAAyFEUYQFq6ODipmfkIC+UyWGt9SIsRp3PXOGoZQGJQhAGkJRbKZT6OWgaQaBRhAGmpoy+snCzTrs0lvqMgQbZVFKkoL5st1AAkTExF2MweMbMzZtZpZh97k+vuNrNFM/uh+EUEgDfq6AtrR3WJ8nM4WjlTZWWZ9tSWckcYQMLctAibWbakT0p6VFKrpMfNrPUG1/22pGfjHRIAVuvoG2c+OABa6kJq56hlAAkSyx3hw5I6nXMXnHNzkp6S9Nga1/2cpH+U1B/HfADwBiOTc7oanmE+OABa6kIan1lQz+i07ygAMlAsRbhBUteKj7uXH7vOzBokvVfSp+IXDQDWFn2rvLWeIpzpWpfv+nPCHIBEiKUIr7VT/er3qP5fSR91zr3pZo9m9mEzO2ZmxwYGBmLNCADfoZ0dIwJjTy07RwBInJwYrumW1Lji4y2Selddc0jSU8unO1VJ+m4zW3DO/fPKi5xzT0p6UpIOHTrEwBeAdWnvC6u6NF9VJfm+oyDBSvJztK2yiCIMICFiKcJHJe0ys+2SeiS9X9KPrrzAObc9+msz+7Skf1ldggEgXjr6xtXK3eDAaKnlqGUAiXHT0Qjn3IKkj2hpN4gOSZ9xzrWZ2RNm9kSiAwLASnMLEXX2c7RykLTUhXR5eEqTswu+owDIMLHcEZZz7hlJz6x6bM2Fcc65D2w8FgCsrbN/QvOLjq3TAqSlrlTOSaevjuuubZt8xwGQQThZDkBaib5Fvo8dIwKDo5YBJApFGEBaae8LKz8nS02Vxb6jIEm2bCpUaUEORRhA3FGEAaSVjr6w9tSWKiebH19BYWYsmAOQEPyXBEDacM6poy/MjlQZUxkAABk8SURBVBEB1FJXqtNXxxWJsPMmgPihCANIG1fDMxqZmmfHiABqqQtpam5RV4anfEcBkEEowgDSRgcnygVW9DhtxiMAxBNFGEDa6OgblyTtZeu0wNm9uVRZRhEGEF8UYQBpo703rMaKQoUKcn1HQZIV5GarubpE7RRhAHFEEQaQNjr6wmqpZSwiqFrqQtffFQCAeKAIA0gLU3MLujg0yXxwgLXUlapndFpjU/O+owDIEBRhAGnh9NVxOff6oikEz/UT5q4yHgEgPijCANJCdJEUewgHVytHLQOIM4owgLTQ0RdWaX6Otmwq9B0FntSU5quiOI8iDCBuKMIA0kJ7b1gtdSGZme8o8MTM1FJXyoI5AHFDEQaQ8iIRp9NXx9XC/sGB11Ib0plr41pYjPiOAiADUIQBpLwrw1OamltkxwiopS6kuYWILg5O+o4CIANQhAGkvOsL5dgxIvCifxniYA0A8UARBpDy2vvCyrKlY3YRbDtrSpSbbcwJA4gLijCAlNfRF1ZzdYkKcrN9R4FneTlZ2lFdws4RAOKCIgwg5XX0jbN/MK5rrQtRhAHEBUUYQEobnZpTz+g0C+VwXUtdSP3jsxqamPUdBUCaowgDSGnRWVC2TkNUdNEkc8IANooiDCClcbQyVnt954gxz0kApDuKMICU1t4XVlVJnqpL831HQYqoKM7T5lA+d4QBbBhFGEBK6+jjaGW8UQsL5gDEAUUYQMqaX4zo3LUJFsrhDVrqQursn9DswqLvKADSGEUYQMrq7J/Q3GKE+WC8QUtdSAsRp87+Cd9RAKQxijCAlNXWu/TW9/4GijC+U+vyLiLMCQPYCIowgJTV1jumwtxsba8q8R0FKaapslj5OVnMCQPYEIowgJTV1hPW3rpSZWexUA7fKSc7S3trS9XeSxEGsH4UYQApKRJxau8La189YxFYW2t9mdp6x+Sc8x0FQJqiCANISVeGpzQxu6D99WW+oyBF7W8IKTyzoO6Rad9RAKQpijCAlBRdKLePIowbiH5vtPVywhyA9aEIA0hJp3rHlJNl2l3LQjmsbW/t0vz4qR7mhAGsD0UYQEpq6w1r1+ZS5edk+46CFFWQm61dNSU6xR1hAOtEEQaQcpxzausZY6Ecbmpffdn1MRoAuFUUYQAp51p4VkOTcxRh3NS++pAGxmfVH57xHQVAGqIIA0g50cVPLJTDzexvWPoeYTwCwHpQhAGknOhb3a3cEcZNRL9H2lgwB2AdKMIAUk5b75i2VxWrJD/HdxSkuJL8HG2vKuaOMIB1oQgDSDmnesLcDUbM9tWHWDAHYF0owgBSyujUnHpGpzlRDjHb31Cm7pFpjU7N+Y4CIM1QhAGklPbrJ8pxRxixiX6vcFcYwK2iCANIKaeu7xhBEUZsOGoZwHpRhAGklLbesOrKClRZku87CtJERXGeGsoLOWoZwC2jCANIKW29Ye4G45a11ofYOQLALaMIA0gZU3MLujAwoVYWyuEW7a8v08XBSU3OLviOAiCNUIQBpIyOvnFFHPPBuHX7G0JyTuroYzwCQOwowgBSRnSxU/TYXCBWry+YowgDiB1FGEDKONE9psriPNWXFfiOgjSzOZSvqpI8nephThhA7CjCAFLGye4x3balTGbmOwrSjJlpX32ZTnFHGMAtoAgDSAlTcws61z+uA1vKfUdBmtrfENK5a+OamV/0HQVAmqAIA0gJ7b1hRZx0gPlgrNNtDeVaiDgWzAGIGUUYQEo40b0023nbFoow1ufA8vdO9HsJAG6GIgwgJZzsGdPmUL42h1goh/WpKytQVUk+RRhAzCjCAFLCie5R3dbAfDDWz8x0YEuZTnSP+o4CIE1QhAF4Nz4zrwuDk9ff2gbW68CWMnUOTHDCHICYUIQBeNfWG5ZzzAdj4w5sKZNzYj9hADGhCAPw7mR0oRw7RmCDotvvMScMIBYUYQDenegZU0N5oapK8n1HQZqrKslXQ3mhTnBHGEAMKMIAvDvZPcrdYMTNbQ0smAMQG4owAK/GpuZ1aWiK+WDEzYHGMl0emtLY1LzvKABSHEUYgFenepfewmbHCMTLgeVt+E70cFcYwJuLqQib2SNmdsbMOs3sY2v8/r8zsxPL/zxvZrfHPyqATHSChXKIs+j3EgvmANzMTYuwmWVL+qSkRyW1SnrczFpXXXZR0ludcwck/aakJ+MdFEBmOtkzqq0VRSovyvMdBRmirChXTZVFzAkDuKlY7ggfltTpnLvgnJuT9JSkx1Ze4Jx73jk3svzhi5K2xDcmgEx1onuM+WDE3YEt5dwRBnBTsRThBkldKz7uXn7sRn5K0hc2EgpAMAxOzKp7ZFoHGItAnB3YUqa+sRn1j8/4jgIghcVShG2Nx9yaF5q9TUtF+KM3+P0Pm9kxMzs2MDAQe0oAGen4laW3ru/YuslzEmSa6MEaJ7krDOBNxFKEuyU1rvh4i6Te1ReZ2QFJfyLpMefc0FpP5Jx70jl3yDl3qLq6ej15AWSQ412jys4yFsoh7vbVh5Rl0msUYQBvIpYifFTSLjPbbmZ5kt4v6XMrLzCzrZKelvTjzrmz8Y8JIBMd7xrV3tpSFeZl+46CDFOcn6NdNaV6rYsFcwBu7KZF2Dm3IOkjkp6V1CHpM865NjN7wsyeWL7sv0mqlPS/zey4mR1LWGIAGSEScXqta1QHG8t9R0GGumNruY53jSoSWXOaDwCUE8tFzrlnJD2z6rFPrfj1hyR9KL7RAGSy8wMTGp9doAgjYe7YWq6njnbp4tCkdlSX+I4DIAVxshwAL15loRwS7M7l763o9xoArEYRBuDFq12jKi3IUXNVse8oyFA7qktUmp+jV6+M3PxiAIFEEQbgxfHl+eCsrLV2aAQ2LivLdHBruV7hjjCAG6AIA0i6ydkFnbka1h3MByPB7mgs15mrYU3OLviOAiAFUYQBJN3JnjFFnHRwK0UYiXXH1k2KOHHcMoA1UYQBJF108dLBRhbKIbGiu5K82sWcMIA3oggDSLrjXSPaVlmkiuI831GQ4TYV56m5qpidIwCsiSIMIKmcc3r1yijzwUiag1vL9eqVUTnHwRoAvhNFGEBS9Y7NqH98loM0kDR3bN2kwYlZdY9M+44CIMVQhAEk1bFLw5KkQ00VnpMgKO5cXpT5CvsJA1iFIgwgqY5eGlZJfo721pb6joKA2LO5VIW52cwJA3gDijCApDp2aUR3bC1XTjY/fpAcOdlZur2xTMcuD/uOAiDF8F8iAEkzNj2vM9fGdTdjEUiyw00Vau8Na4KDNQCsQBEGkDSvXBmRc9KhJvYPRnLdvb1CESe9cpk5YQCvowgDSJpjl4aVnWXsGIGku3PrJmVnmY5eYjwCwOsowgCS5uilEe2vD6koL8d3FARMcX6O9tWHdOQiRRjA6yjCAJJidmFRr3WNsm0avDm0rULHu0Y1u7DoOwqAFEERBpAUp3rCml2I6G7mg+HJ4e2bNLsQ0ameMd9RAKQIijCApHh5eeuqu7ZxRxh+RN+NOHKRBXMAllCEASTF0Usj2l5VrOrSfN9REFBVJflqri5mwRyA6yjCABLOOadjl4Z11zbGIuDX4aYKHbs0rEjE+Y4CIAVQhAEkXGf/hEam5pkPhnd3N1UoPLOgs/3jvqMASAEUYQAJ98KFIUnSfc1VnpMg6KKnGh5lGzUAoggDSIIXzg+pobxQjRWFvqMg4BorCrU5lK+XKMIARBEGkGCRiNOLF4Z0b3OlzMx3HAScmene5kq9eGFYzjEnDAQdRRhAQp2+Oq6RqXndv6PSdxRAkvTAjioNTszq7LUJ31EAeEYRBpBQ1+eDKcJIEdHvxefPD3pOAsA3ijCAhHrh/JC2VRapvpz5YKSGxooiba0o0nOdQ76jAPCMIgwgYRYjTi9dHNJ9zdwNRmp5YGelXrowpIXFiO8oADyiCANImLbeMY3PLDAWgZRz344qjc8u6FRv2HcUAB5RhAEkzAvno/sHU4SRWqKLN5/rZE4YCDKKMICE+da5Qe2qKVFNqMB3FOA7VJXka29tKQvmgICjCANIiKm5BR25OKy37q72HQVY0307KnXs0ohm5hd9RwHgCUUYQEK8eGFIc4sRvXUPRRip6cGdVZpdiOjoJU6ZA4KKIgwgIf7t7KAKcrN0d1OF7yjAmu7bUam8nCx9/fSA7ygAPKEIA0iIb54d0H3NlSrIzfYdBVhTUV6O7m2u1DfO9vuOAsATijCAuLs8NKmLg5PMByPlvW1PtS4MTOry0KTvKAA8oAgDiLt/O7v0VvNbKMJIcW/bUyNJ+sYZxiOAIKIIA4i7b54dUGNFobZXFfuOAryppqpiNVUW6RtnGI8AgogiDCCuZuYX9fz5Ib1lV7XMzHcc4KYe3lOj588PsY0aEEAUYQBx9cL5IU3NLeq7Wjf7jgLE5G17azS7ENELF4Z8RwGQZBRhAHH1pfZrKs7Lvn6ELZDq7tleocLcbH2tg/EIIGgowgDiJhJx+nL7NT28p0b5OWybhvRQkJutt+yu0pfaryoScb7jAEgiijCAuHm1a1SDE7N61z7GIpBeHt1fp2vhWR3vHvUdBUASUYQBxM2X268pJ8v08PKWVEC6eNveGuVmm549ddV3FABJRBEGEBfOOT3bdlX3NleqrDDXdxzglpQV5ur+HVX6YttVOcd4BBAUFGEAcdHWG9bFwUl9z4E631GAdXlkf60uD03p9NVx31EAJAlFGEBcfO61XuVkmR7dX+s7CrAu72zdrOws0+df6/UdBUCSUIQBbFgk4vQvr/XqLburVV6U5zsOsC5VJfl6cGeVPnu8l90jgICgCAPYsJevjKh3bEbvub3edxRgQ957R4N6Rqd17PKI7ygAkoAiDGDDPnu8R/k5WZwmh7T3rn2bVZSXrX96tcd3FABJQBEGsCHTc4v67PFePbK/ViX5Ob7jABtSlJejd++r1b+e6NXM/KLvOAASjCIMYEO+cKpP4zMLev/dW31HAeLi++9oUHhmQV9uv+Y7CoAEowgD2JCnjnapqbJI9zZX+I4CxMWDO6vUWFGov3rxsu8oABKMIgxg3c4PTOjIxWH9yN1bZWa+4wBxkZ1l+tHD2/TSxWGdvcaewkAmowgDWLe/fOGycrJMP3hXg+8oQFz98KEtysvO4q4wkOEowgDWZWRyTn93tEuPHWxQTWmB7zhAXFWW5Ot7DtTp6Vd6FJ6Z9x0HQIJQhAGsy1+9eFnT84v68FuafUcBEuLfP7BdE7ML+ssXuCsMZCqKMIBbNjO/qE8/f0lv31ujPbWlvuMACXHbljK9bU+1/uRbFzQ5u+A7DoAEoAgDuGV/9eJlDU3O6ae5G4wM93Pv2KWRqXlmhYEMRREGcEvGpuf1ia936qFdVbqnudJ3HCCh7ty6SQ/tqtL/928XNDbNrDCQaSjCAG7Jx796TmPT8/rYo3t9RwGS4qOP7NXI1Jz+8CvnfEcBEGcUYQAxe61rVH/+3EU9fnir9tWX+Y4DJMX+hjI9fnir/u8Ll3Sie9R3HABxRBEGEJOJ2QX94meOq7o0n7vBCJyPvnuvakrz9Qt/d1xTcyycAzJFTEXYzB4xszNm1mlmH1vj983MPr78+yfM7M74RwXgy8JiRL/0meO6ODipP/iRgwoV5PqOBCRVWVGufu99t+vi4KR+4anjWow435EAxMFNi7CZZUv6pKRHJbVKetzMWldd9qikXcv/fFjSH8c5JwBPZuYX9Qt/d1zPtl3Tr31Pq+7fUeU7EuDF/Tur9N+/t1Vfar+mn3/qVU3PLfqOBGCDcmK45rCkTufcBUkys6ckPSapfcU1j0n6C+eck/SimZWbWZ1zri/uiTdgbiGyoXPj3QZvADit/wk2/tob/PwNBtjIZ2/0z77RP/1GXn/jX/eNfv7GnqBzYEJ/8q2Lujg4qV99dK9+6sHtGwsEpLkPPLBdMwsR/fYXT6utN6wPPbRdrXUh5WYzaQjEYvfmUuXlpM7/X2Ipwg2SulZ83C3pnhiuaZCUUkV4cGJW3/tH3/YdA0gru2pK9Dcfukf37+ROMCBJT7x1h1rrQvqtZzr0a/90ynccIK08/7G3q7680HeM62IpwrbGY6tvM8Vyjczsw1oandDWrVtjeOn4qijO0//5iUMbeo61/qC39PkbfIKNfL5tNL3HT7cNfuG8/3vbQIKNv/b6VZXma1dNyYa//kCmecvuaj24s0oXBid0eWiKmWEgRhXFeb4jfIdYinC3pMYVH2+R1LuOa+Sce1LSk5J06NChpP/UKMjN1jtbNyf7ZQEAGSgry7SzplQ7azhmHEhXsQxpHJW0y8y2m1mepPdL+tyqaz4n6SeWd4+4V9JYqs0HAwAAACvd9I6wc27BzD4i6VlJ2ZL+zDnXZmZPLP/+pyQ9I+m7JXVKmpL0wcRFBgAAADYultEIOeee0VLZXfnYp1b82kn62fhGAwAAABIndfavAAAAAJKIIgwAAIBAoggDAAAgkCjCAAAACCSKMAAAAAKJIgwAAIBAoggDAAAgkCjCAAAACCSKMAAAAAKJIgwAAIBAoggDAAAgkCjCAAAACCSKMAAAAAKJIgwAAIBAoggDAAAgkCjCAAAACCSKMAAAAAKJIgwAAIBAoggDAAAgkCjCAAAACCSKMAAAAAKJIgwAAIBAoggDAAAgkCjCAAAACCSKMAAAAAKJIgwAAIBAoggDAAAgkCjCAAAACCSKMAAAAAKJIgwAAIBAoggDAAAgkMw55+eFzQYkXfby4lKVpEFPrw0AG8HPLwDpyufPr23OuerVD3orwj6Z2THn3CHfOQDgVvHzC0C6SsWfX4xGAAAAIJAowgAAAAikoBbhJ30HAIB14ucXgHSVcj+/AjkjDAAAAAT1jjAAAAACLqOKsJn9JzM7amZDZjZjZp1m9ntmVnmD6z9tZg8nOSYA3BIzazSzfzCzMTMLm9nTZrbVdy4AwWZm7zazr5nZVTObNbNuM/uMmbWuce0HzOx/eIj5pjKqCEuqkPS0pA9IekTSJyX9e0lfNrMsSTKz95jZO1Z+kpkVmNlHb1SYAcAXMyuS9DVJeyX9pKQfl7RL0tfNrNhnNgCBVyHpZUkfkfQuSb8qaZ+kF81sm5ntNrP/x8yyV36Smf2wmT2Y/LhvlPEzwmb205I+JemQc+5lM2uR9BuSTFJI0nFJb5f0D5L+0Dk37S0sAKxiZj8v6fcl7XHOdS4/tl3SOUm/4pz7fZ/5AGAlM9sj6bSk/yzpzyV9TNKDkk5IKpdUKemCpN9wzvX4yhmVaXeE1zK0/L/zkuSc63DOvU/Sv0r6Lknvl/QDzrn/tbIEm9l7zew5M5tYfivyiJm9J+npAQTdeyS9GC3BkuScuyjpOUmPeUsFAGu73rucc8POuV+R9DOSfljSeyX9rnPup1eWYDO73cz+aXm0ddrMzpjZryYjbE4yXiTZzCxHUp6kA5J+XdJXnXMnln9v9/JjeZK+qqU7wk+b2d9L+rhzbtrMfk7SxyX9s5beipyQdKekpiT/UQBgn6TPrvF4m6T3JTkLALzB8uhDtqRtkv6XpKuSnjKzTZJ+RdJbJX1GS3eEf9nM3qulO8K9ZnZY0jckdUr6T5K6tTT+dSAZ2TOuCJtZiaTxFQ89q+/8j0WLpD91zn3FzD6tpTvD/1XSz0sqMrNcSb8l6Z+ccz+w6nkAINkqJI2s8fiwpE1JzgIAa3lJ0l3Lv+6U9HbnXP/ymMQVSQ9paX1Dk3Puf5jZj0hqltQr6Xe1dBf5Xufc1PJzfC1ZwdOyCJuZaelvHtc55xaWfzkl6W5JBZLukPRrkj5vZt/lnFtwzr3hzopzbkbSby8/9yOSSpSCmz4DCKy1FnNY0lMAwNp+XEvrrpq1NBv8ZTN70Dl3RtIZSVqqbkucc3+3/FiRpAck/c6KEpxUaVmEtXSL/eurHjNJcs5FJB1bfuzbZnZy+dofkvTUyk9wzn1gjeeO7hzRHa+wALABI1q6K7zaJq19pxgAkso517H8y5fM7AuSLmlpkdwTK6759BqfuklL69W8da50LcIva+mubyyipXhnjNcPLv9vg6RTtxIKABKgTUtzwqu1SmpPchYAeFPOuVEz61RsvWtEUkRLncuLtNw1wjk37pw7tvKfN7n8rcv/ez7Gp39eS4vjPryhkAAQH5+TdK+ZNUcfMLMmLb2d+DlPmQBgTWa2WUv7nt+0dy2PQ3xb0o+ZWWGis60lY/YRNrMySV+U9Nda2l/TSTos6Re1NKh9j3NuNsbn+oikP9LS4Rx/raXFdwclzTjn/ij+6QFgbcuHZrwmaVrSf9HSz7bflFQq6YBzbsJjPAABZmb/JOkVLe0RHJa0W0s7P9RKOuycOxvDc9wt6ZuSzkr6PS2NSTRLOuic+7kERX/99TOoCOdL+mMtbdrcIGlBSzMqn9HStmjjN/7sNZ/vhyT9spa275iX1CHpN51z/xLH2ABwU8vHKf+BpHdqaT3EVyX9gnPuks9cAILNzD6qpf2Bd2hpW9ouLW2F9j9v5eeTmd2hpcPOHpKUL+mypD93zv12nCO/8bUzpQgDAAAAtyItZ4QBAACAjaIIAwAAIJAowgAAAAgkijAAAAACiSIMAACAQKIIAwAAIJAowgAAAAgkijAAAAACiSIMAACAQPr/AfV7K5y+bTNWAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "c = 4.685\n",
    "support = np.linspace(-3*c, 3*c, 1000)\n",
    "tukey = norms.TukeyBiweight(c=c)\n",
    "plot_weights(support, tukey.weights, ['-3*c', '0', '3*c'], [-3*c, 0, 3*c]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Scale Estimators"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Robust estimates of the location"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = np.array([1, 2, 3, 4, 500])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* The mean is not a robust estimator of location"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "102.0"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* The median, on the other hand, is a robust estimator with a breakdown point of 50%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.0"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.median(x)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Analogously for the scale\n",
    "* The standard deviation is not robust"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "199.00251254695254"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x.std()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Median Absolute Deviation\n",
    "\n",
    "$$ median_i |X_i - median_j(X_j)|) $$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Standardized Median Absolute Deviation is a consistent estimator for $\\hat{\\sigma}$\n",
    "\n",
    "$$\\hat{\\sigma}=K \\cdot MAD$$\n",
    "\n",
    "where $K$ depends on the distribution. For the normal distribution for example,\n",
    "\n",
    "$$K = \\Phi^{-1}(.75)$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6744897501960817"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats.norm.ppf(.75)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[  1   2   3   4 500]\n"
     ]
    }
   ],
   "source": [
    "print(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.482602218505602"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sm.robust.scale.mad(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.4142135623730951"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array([1,2,3,4,5.]).std()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* The default for Robust Linear Models is MAD\n",
    "* another popular choice is Huber's proposal 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(12345)\n",
    "fat_tails = stats.t(6).rvs(40)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "kde = sm.nonparametric.KDEUnivariate(fat_tails)\n",
    "kde.fit()\n",
    "fig = plt.figure(figsize=(12,8))\n",
    "ax = fig.add_subplot(111)\n",
    "ax.plot(kde.support, kde.density);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0688231044810875 1.3471633229698652\n"
     ]
    }
   ],
   "source": [
    "print(fat_tails.mean(), fat_tails.std())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(0.0688231044810875, 1.3471633229698652)\n"
     ]
    }
   ],
   "source": [
    "print(stats.norm.fit(fat_tails))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(6, 0.03900835312789366, 1.0563837144431252)\n"
     ]
    }
   ],
   "source": [
    "print(stats.t.fit(fat_tails, f0=6))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.04048984333271795 1.1557140047569665\n"
     ]
    }
   ],
   "source": [
    "huber = sm.robust.scale.Huber()\n",
    "loc, scale = huber(fat_tails)\n",
    "print(loc, scale)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.115335001165415"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sm.robust.mad(fat_tails)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0483916565928972"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sm.robust.mad(fat_tails, c=stats.t(6).ppf(.75))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.115335001165415"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sm.robust.scale.mad(fat_tails)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Duncan's Occupational Prestige data - M-estimation for outliers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.graphics.api import abline_plot\n",
    "from statsmodels.formula.api import ols, rlm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "prestige = sm.datasets.get_rdataset(\"Duncan\", \"carData\", cache=True).data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "            type  income  education  prestige\n",
      "accountant  prof      62         86        82\n",
      "pilot       prof      72         76        83\n",
      "architect   prof      75         92        90\n",
      "author      prof      55         90        76\n",
      "chemist     prof      64         86        90\n",
      "minister    prof      21         84        87\n",
      "professor   prof      64         93        93\n",
      "dentist     prof      80        100        90\n",
      "reporter      wc      67         87        52\n",
      "engineer    prof      72         86        88\n"
     ]
    }
   ],
   "source": [
    "print(prestige.head(10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(12,12))\n",
    "ax1 = fig.add_subplot(211, xlabel='Income', ylabel='Prestige')\n",
    "ax1.scatter(prestige.income, prestige.prestige)\n",
    "xy_outlier = prestige.loc['minister', ['income','prestige']]\n",
    "ax1.annotate('Minister', xy_outlier, xy_outlier+1, fontsize=16)\n",
    "ax2 = fig.add_subplot(212, xlabel='Education',\n",
    "                           ylabel='Prestige')\n",
    "ax2.scatter(prestige.education, prestige.prestige);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               prestige   R-squared:                       0.828\n",
      "Model:                            OLS   Adj. R-squared:                  0.820\n",
      "Method:                 Least Squares   F-statistic:                     101.2\n",
      "Date:                Mon, 24 Feb 2020   Prob (F-statistic):           8.65e-17\n",
      "Time:                        22:49:06   Log-Likelihood:                -178.98\n",
      "No. Observations:                  45   AIC:                             364.0\n",
      "Df Residuals:                      42   BIC:                             369.4\n",
      "Df Model:                           2                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "==============================================================================\n",
      "                 coef    std err          t      P>|t|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "Intercept     -6.0647      4.272     -1.420      0.163     -14.686       2.556\n",
      "income         0.5987      0.120      5.003      0.000       0.357       0.840\n",
      "education      0.5458      0.098      5.555      0.000       0.348       0.744\n",
      "==============================================================================\n",
      "Omnibus:                        1.279   Durbin-Watson:                   1.458\n",
      "Prob(Omnibus):                  0.528   Jarque-Bera (JB):                0.520\n",
      "Skew:                           0.155   Prob(JB):                        0.771\n",
      "Kurtosis:                       3.426   Cond. No.                         163.\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "ols_model = ols('prestige ~ income + education', prestige).fit()\n",
    "print(ols_model.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accountant            0.303900\n",
      "pilot                 0.340920\n",
      "architect             0.072256\n",
      "author                0.000711\n",
      "chemist               0.826578\n",
      "minister              3.134519\n",
      "professor             0.768277\n",
      "dentist              -0.498082\n",
      "reporter             -2.397022\n",
      "engineer              0.306225\n",
      "undertaker           -0.187339\n",
      "lawyer               -0.303082\n",
      "physician             0.355687\n",
      "welfare.worker       -0.411406\n",
      "teacher               0.050510\n",
      "conductor            -1.704032\n",
      "contractor            2.043805\n",
      "factory.owner         1.602429\n",
      "store.manager         0.142425\n",
      "banker                0.508388\n",
      "bookkeeper           -0.902388\n",
      "mail.carrier         -1.433249\n",
      "insurance.agent      -1.930919\n",
      "store.clerk          -1.760491\n",
      "carpenter             1.068858\n",
      "electrician           0.731949\n",
      "RR.engineer           0.808922\n",
      "machinist             1.887047\n",
      "auto.repairman        0.522735\n",
      "plumber              -0.377954\n",
      "gas.stn.attendant    -0.666596\n",
      "coal.miner            1.018527\n",
      "streetcar.motorman   -1.104485\n",
      "taxi.driver           0.023322\n",
      "truck.driver         -0.129227\n",
      "machine.operator      0.499922\n",
      "barber                0.173805\n",
      "bartender            -0.902422\n",
      "shoe.shiner          -0.429357\n",
      "cook                  0.127207\n",
      "soda.clerk           -0.883095\n",
      "watchman             -0.513502\n",
      "janitor              -0.079890\n",
      "policeman             0.078847\n",
      "waiter               -0.475972\n",
      "Name: student_resid, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "infl = ols_model.get_influence()\n",
    "student = infl.summary_frame()['student_resid']\n",
    "print(student)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "minister      3.134519\n",
      "reporter     -2.397022\n",
      "contractor    2.043805\n",
      "Name: student_resid, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print(student.loc[np.abs(student) > 2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dfb_Intercept      0.144937\n",
      "dfb_income        -1.220939\n",
      "dfb_education      1.263019\n",
      "cooks_d            0.566380\n",
      "standard_resid     2.849416\n",
      "hat_diag           0.173058\n",
      "dffits_internal    1.303510\n",
      "student_resid      3.134519\n",
      "dffits             1.433935\n",
      "Name: minister, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print(infl.summary_frame().loc['minister'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                    student_resid   unadj_p  sidak(p)\n",
      "minister                 3.134519  0.003177  0.133421\n",
      "reporter                -2.397022  0.021170  0.618213\n",
      "contractor               2.043805  0.047433  0.887721\n",
      "insurance.agent         -1.930919  0.060428  0.939485\n",
      "machinist                1.887047  0.066248  0.954247\n",
      "store.clerk             -1.760491  0.085783  0.982331\n",
      "conductor               -1.704032  0.095944  0.989315\n",
      "factory.owner            1.602429  0.116738  0.996250\n",
      "mail.carrier            -1.433249  0.159369  0.999595\n",
      "streetcar.motorman      -1.104485  0.275823  1.000000\n",
      "carpenter                1.068858  0.291386  1.000000\n",
      "coal.miner               1.018527  0.314400  1.000000\n",
      "bartender               -0.902422  0.372104  1.000000\n",
      "bookkeeper              -0.902388  0.372122  1.000000\n",
      "soda.clerk              -0.883095  0.382334  1.000000\n",
      "chemist                  0.826578  0.413261  1.000000\n",
      "RR.engineer              0.808922  0.423229  1.000000\n",
      "professor                0.768277  0.446725  1.000000\n",
      "electrician              0.731949  0.468363  1.000000\n",
      "gas.stn.attendant       -0.666596  0.508764  1.000000\n",
      "auto.repairman           0.522735  0.603972  1.000000\n",
      "watchman                -0.513502  0.610357  1.000000\n",
      "banker                   0.508388  0.613906  1.000000\n",
      "machine.operator         0.499922  0.619802  1.000000\n",
      "dentist                 -0.498082  0.621088  1.000000\n",
      "waiter                  -0.475972  0.636621  1.000000\n",
      "shoe.shiner             -0.429357  0.669912  1.000000\n",
      "welfare.worker          -0.411406  0.682918  1.000000\n",
      "plumber                 -0.377954  0.707414  1.000000\n",
      "physician                0.355687  0.723898  1.000000\n",
      "pilot                    0.340920  0.734905  1.000000\n",
      "engineer                 0.306225  0.760983  1.000000\n",
      "accountant               0.303900  0.762741  1.000000\n",
      "lawyer                  -0.303082  0.763360  1.000000\n",
      "undertaker              -0.187339  0.852319  1.000000\n",
      "barber                   0.173805  0.862874  1.000000\n",
      "store.manager            0.142425  0.887442  1.000000\n",
      "truck.driver            -0.129227  0.897810  1.000000\n",
      "cook                     0.127207  0.899399  1.000000\n",
      "janitor                 -0.079890  0.936713  1.000000\n",
      "policeman                0.078847  0.937538  1.000000\n",
      "architect                0.072256  0.942750  1.000000\n",
      "teacher                  0.050510  0.959961  1.000000\n",
      "taxi.driver              0.023322  0.981507  1.000000\n",
      "author                   0.000711  0.999436  1.000000\n"
     ]
    }
   ],
   "source": [
    "sidak = ols_model.outlier_test('sidak')\n",
    "sidak.sort_values('unadj_p', inplace=True)\n",
    "print(sidak)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                    student_resid   unadj_p  fdr_bh(p)\n",
      "minister                 3.134519  0.003177   0.142974\n",
      "reporter                -2.397022  0.021170   0.476332\n",
      "contractor               2.043805  0.047433   0.596233\n",
      "insurance.agent         -1.930919  0.060428   0.596233\n",
      "machinist                1.887047  0.066248   0.596233\n",
      "store.clerk             -1.760491  0.085783   0.616782\n",
      "conductor               -1.704032  0.095944   0.616782\n",
      "factory.owner            1.602429  0.116738   0.656653\n",
      "mail.carrier            -1.433249  0.159369   0.796844\n",
      "streetcar.motorman      -1.104485  0.275823   0.999436\n",
      "carpenter                1.068858  0.291386   0.999436\n",
      "coal.miner               1.018527  0.314400   0.999436\n",
      "bartender               -0.902422  0.372104   0.999436\n",
      "bookkeeper              -0.902388  0.372122   0.999436\n",
      "soda.clerk              -0.883095  0.382334   0.999436\n",
      "chemist                  0.826578  0.413261   0.999436\n",
      "RR.engineer              0.808922  0.423229   0.999436\n",
      "professor                0.768277  0.446725   0.999436\n",
      "electrician              0.731949  0.468363   0.999436\n",
      "gas.stn.attendant       -0.666596  0.508764   0.999436\n",
      "auto.repairman           0.522735  0.603972   0.999436\n",
      "watchman                -0.513502  0.610357   0.999436\n",
      "banker                   0.508388  0.613906   0.999436\n",
      "machine.operator         0.499922  0.619802   0.999436\n",
      "dentist                 -0.498082  0.621088   0.999436\n",
      "waiter                  -0.475972  0.636621   0.999436\n",
      "shoe.shiner             -0.429357  0.669912   0.999436\n",
      "welfare.worker          -0.411406  0.682918   0.999436\n",
      "plumber                 -0.377954  0.707414   0.999436\n",
      "physician                0.355687  0.723898   0.999436\n",
      "pilot                    0.340920  0.734905   0.999436\n",
      "engineer                 0.306225  0.760983   0.999436\n",
      "accountant               0.303900  0.762741   0.999436\n",
      "lawyer                  -0.303082  0.763360   0.999436\n",
      "undertaker              -0.187339  0.852319   0.999436\n",
      "barber                   0.173805  0.862874   0.999436\n",
      "store.manager            0.142425  0.887442   0.999436\n",
      "truck.driver            -0.129227  0.897810   0.999436\n",
      "cook                     0.127207  0.899399   0.999436\n",
      "janitor                 -0.079890  0.936713   0.999436\n",
      "policeman                0.078847  0.937538   0.999436\n",
      "architect                0.072256  0.942750   0.999436\n",
      "teacher                  0.050510  0.959961   0.999436\n",
      "taxi.driver              0.023322  0.981507   0.999436\n",
      "author                   0.000711  0.999436   0.999436\n"
     ]
    }
   ],
   "source": [
    "fdr = ols_model.outlier_test('fdr_bh')\n",
    "fdr.sort_values('unadj_p', inplace=True)\n",
    "print(fdr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                    Robust linear Model Regression Results                    \n",
      "==============================================================================\n",
      "Dep. Variable:               prestige   No. Observations:                   45\n",
      "Model:                            RLM   Df Residuals:                       42\n",
      "Method:                          IRLS   Df Model:                            2\n",
      "Norm:                          HuberT                                         \n",
      "Scale Est.:                       mad                                         \n",
      "Cov Type:                          H1                                         \n",
      "Date:                Mon, 24 Feb 2020                                         \n",
      "Time:                        22:49:06                                         \n",
      "No. Iterations:                    18                                         \n",
      "==============================================================================\n",
      "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "Intercept     -7.1107      3.879     -1.833      0.067     -14.713       0.492\n",
      "income         0.7015      0.109      6.456      0.000       0.489       0.914\n",
      "education      0.4854      0.089      5.441      0.000       0.311       0.660\n",
      "==============================================================================\n",
      "\n",
      "If the model instance has been used for another fit with different fit parameters, then the fit options might not be the correct ones anymore .\n"
     ]
    }
   ],
   "source": [
    "rlm_model = rlm('prestige ~ income + education', prestige).fit()\n",
    "print(rlm_model.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accountant            1.000000\n",
      "pilot                 1.000000\n",
      "architect             1.000000\n",
      "author                1.000000\n",
      "chemist               1.000000\n",
      "minister              0.344596\n",
      "professor             1.000000\n",
      "dentist               1.000000\n",
      "reporter              0.441669\n",
      "engineer              1.000000\n",
      "undertaker            1.000000\n",
      "lawyer                1.000000\n",
      "physician             1.000000\n",
      "welfare.worker        1.000000\n",
      "teacher               1.000000\n",
      "conductor             0.538445\n",
      "contractor            0.552262\n",
      "factory.owner         0.706169\n",
      "store.manager         1.000000\n",
      "banker                1.000000\n",
      "bookkeeper            1.000000\n",
      "mail.carrier          0.690764\n",
      "insurance.agent       0.533499\n",
      "store.clerk           0.618656\n",
      "carpenter             0.935848\n",
      "electrician           1.000000\n",
      "RR.engineer           1.000000\n",
      "machinist             0.570360\n",
      "auto.repairman        1.000000\n",
      "plumber               1.000000\n",
      "gas.stn.attendant     1.000000\n",
      "coal.miner            0.963821\n",
      "streetcar.motorman    0.832870\n",
      "taxi.driver           1.000000\n",
      "truck.driver          1.000000\n",
      "machine.operator      1.000000\n",
      "barber                1.000000\n",
      "bartender             1.000000\n",
      "shoe.shiner           1.000000\n",
      "cook                  1.000000\n",
      "soda.clerk            1.000000\n",
      "watchman              1.000000\n",
      "janitor               1.000000\n",
      "policeman             1.000000\n",
      "waiter                1.000000\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print(rlm_model.weights)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hertzprung Russell data for Star Cluster CYG 0B1 - Leverage Points"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Data is on the luminosity and temperature of 47 stars in the direction of Cygnus."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "dta = sm.datasets.get_rdataset(\"starsCYG\", \"robustbase\", cache=True).data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib.patches import Ellipse\n",
    "fig = plt.figure(figsize=(12,8))\n",
    "ax = fig.add_subplot(111, xlabel='log(Temp)', ylabel='log(Light)', title='Hertzsprung-Russell Diagram of Star Cluster CYG OB1')\n",
    "ax.scatter(*dta.values.T)\n",
    "# highlight outliers\n",
    "e = Ellipse((3.5, 6), .2, 1, alpha=.25, color='r')\n",
    "ax.add_patch(e);\n",
    "ax.annotate('Red giants', xy=(3.6, 6), xytext=(3.8, 6),\n",
    "            arrowprops=dict(facecolor='black', shrink=0.05, width=2),\n",
    "            horizontalalignment='left', verticalalignment='bottom',\n",
    "            clip_on=True, # clip to the axes bounding box\n",
    "            fontsize=16,\n",
    "     )\n",
    "# annotate these with their index\n",
    "for i,row in dta.loc[dta['log.Te'] < 3.8].iterrows():\n",
    "    ax.annotate(i, row, row + .01, fontsize=14)\n",
    "xlim, ylim = ax.get_xlim(), ax.get_ylim()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image\n",
    "Image(filename='star_diagram.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/numpy/core/fromnumeric.py:2495: FutureWarning: Method .ptp is deprecated and will be removed in a future version. Use numpy.ptp instead.\n",
      "  return ptp(axis=axis, out=out, **kwargs)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y = dta['log.light']\n",
    "X = sm.add_constant(dta['log.Te'], prepend=True)\n",
    "ols_model = sm.OLS(y, X).fit()\n",
    "abline_plot(model_results=ols_model, ax=ax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rlm_mod = sm.RLM(y, X, sm.robust.norms.TrimmedMean(.5)).fit()\n",
    "abline_plot(model_results=rlm_mod, ax=ax, color='red')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Why? Because M-estimators are not robust to leverage points."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "infl = ols_model.get_influence()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10    0.194103\n",
       "19    0.194103\n",
       "29    0.198344\n",
       "33    0.194103\n",
       "Name: hat_diag, dtype: float64"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "h_bar = 2*(ols_model.df_model + 1 )/ols_model.nobs\n",
    "hat_diag = infl.summary_frame()['hat_diag']\n",
    "hat_diag.loc[hat_diag > h_bar]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    student_resid   unadj_p  sidak(p)\n",
      "16      -2.049393  0.046415  0.892872\n",
      "13      -2.035329  0.047868  0.900286\n",
      "33       1.905847  0.063216  0.953543\n",
      "18      -1.575505  0.122304  0.997826\n",
      "1        1.522185  0.135118  0.998911\n",
      "3        1.522185  0.135118  0.998911\n",
      "21      -1.450418  0.154034  0.999615\n",
      "17      -1.426675  0.160731  0.999735\n",
      "29       1.388520  0.171969  0.999859\n",
      "14      -1.374733  0.176175  0.999889\n",
      "35       1.346543  0.185023  0.999933\n",
      "34      -1.272159  0.209999  0.999985\n",
      "28      -1.186946  0.241618  0.999998\n",
      "20      -1.150621  0.256103  0.999999\n",
      "44       1.134779  0.262612  0.999999\n",
      "39       1.091886  0.280826  1.000000\n",
      "19       1.064878  0.292740  1.000000\n",
      "6       -1.026873  0.310093  1.000000\n",
      "30      -1.009096  0.318446  1.000000\n",
      "22      -0.979768  0.332557  1.000000\n",
      "8        0.961218  0.341695  1.000000\n",
      "5        0.913802  0.365801  1.000000\n",
      "11       0.871997  0.387943  1.000000\n",
      "12       0.856685  0.396261  1.000000\n",
      "46      -0.833923  0.408829  1.000000\n",
      "10       0.743920  0.460879  1.000000\n",
      "42       0.727179  0.470968  1.000000\n",
      "15      -0.689258  0.494280  1.000000\n",
      "43       0.688272  0.494895  1.000000\n",
      "7        0.655712  0.515424  1.000000\n",
      "40      -0.646396  0.521381  1.000000\n",
      "26      -0.640978  0.524862  1.000000\n",
      "25      -0.545561  0.588123  1.000000\n",
      "37       0.472819  0.638680  1.000000\n",
      "32       0.472819  0.638680  1.000000\n",
      "38       0.462187  0.646225  1.000000\n",
      "0        0.430686  0.668799  1.000000\n",
      "31       0.341726  0.734184  1.000000\n",
      "36       0.318911  0.751303  1.000000\n",
      "4        0.307890  0.759619  1.000000\n",
      "9        0.235114  0.815211  1.000000\n",
      "41       0.187732  0.851950  1.000000\n",
      "2       -0.182093  0.856346  1.000000\n",
      "23      -0.156014  0.876736  1.000000\n",
      "27      -0.147406  0.883485  1.000000\n",
      "24       0.065195  0.948314  1.000000\n",
      "45       0.045675  0.963776  1.000000\n"
     ]
    }
   ],
   "source": [
    "sidak2 = ols_model.outlier_test('sidak')\n",
    "sidak2.sort_values('unadj_p', inplace=True)\n",
    "print(sidak2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    student_resid   unadj_p  fdr_bh(p)\n",
      "16      -2.049393  0.046415   0.764747\n",
      "13      -2.035329  0.047868   0.764747\n",
      "33       1.905847  0.063216   0.764747\n",
      "18      -1.575505  0.122304   0.764747\n",
      "1        1.522185  0.135118   0.764747\n",
      "3        1.522185  0.135118   0.764747\n",
      "21      -1.450418  0.154034   0.764747\n",
      "17      -1.426675  0.160731   0.764747\n",
      "29       1.388520  0.171969   0.764747\n",
      "14      -1.374733  0.176175   0.764747\n",
      "35       1.346543  0.185023   0.764747\n",
      "34      -1.272159  0.209999   0.764747\n",
      "28      -1.186946  0.241618   0.764747\n",
      "20      -1.150621  0.256103   0.764747\n",
      "44       1.134779  0.262612   0.764747\n",
      "39       1.091886  0.280826   0.764747\n",
      "19       1.064878  0.292740   0.764747\n",
      "6       -1.026873  0.310093   0.764747\n",
      "30      -1.009096  0.318446   0.764747\n",
      "22      -0.979768  0.332557   0.764747\n",
      "8        0.961218  0.341695   0.764747\n",
      "5        0.913802  0.365801   0.768599\n",
      "11       0.871997  0.387943   0.768599\n",
      "12       0.856685  0.396261   0.768599\n",
      "46      -0.833923  0.408829   0.768599\n",
      "10       0.743920  0.460879   0.770890\n",
      "42       0.727179  0.470968   0.770890\n",
      "15      -0.689258  0.494280   0.770890\n",
      "43       0.688272  0.494895   0.770890\n",
      "7        0.655712  0.515424   0.770890\n",
      "40      -0.646396  0.521381   0.770890\n",
      "26      -0.640978  0.524862   0.770890\n",
      "25      -0.545561  0.588123   0.837630\n",
      "37       0.472819  0.638680   0.843682\n",
      "32       0.472819  0.638680   0.843682\n",
      "38       0.462187  0.646225   0.843682\n",
      "0        0.430686  0.668799   0.849556\n",
      "31       0.341726  0.734184   0.892552\n",
      "36       0.318911  0.751303   0.892552\n",
      "4        0.307890  0.759619   0.892552\n",
      "9        0.235114  0.815211   0.922751\n",
      "41       0.187732  0.851950   0.922751\n",
      "2       -0.182093  0.856346   0.922751\n",
      "23      -0.156014  0.876736   0.922751\n",
      "27      -0.147406  0.883485   0.922751\n",
      "24       0.065195  0.948314   0.963776\n",
      "45       0.045675  0.963776   0.963776\n"
     ]
    }
   ],
   "source": [
    "fdr2 = ols_model.outlier_test('fdr_bh')\n",
    "fdr2.sort_values('unadj_p', inplace=True)\n",
    "print(fdr2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Let's delete that line"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "l =  ax.lines[-1]\n",
    "l.remove()\n",
    "del l"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weights = np.ones(len(X))\n",
    "weights[X[X['log.Te'] < 3.8].index.values - 1] = 0\n",
    "wls_model = sm.WLS(y, X, weights=weights).fit()\n",
    "abline_plot(model_results=wls_model, ax=ax, color='green')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* MM estimators are good for this type of problem, unfortunately, we do not yet have these yet. \n",
    "* It's being worked on, but it gives a good excuse to look at the R cell magics in the notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "yy = y.values[:,None]\n",
    "xx = X['log.Te'].values[:,None]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Note**: The R code and the results in this notebook has been converted to markdown so that R is not required to build the documents. The R results in the notebook were computed using R 3.5.1 and robustbase 0.93."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "```ipython\n",
    "%load_ext rpy2.ipython\n",
    "\n",
    "%R library(robustbase)\n",
    "%Rpush yy xx\n",
    "%R mod <- lmrob(yy ~ xx);\n",
    "%R params <- mod$coefficients;\n",
    "%Rpull params\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "```ipython\n",
    "%R print(mod)\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "```\n",
    "Call:\n",
    "lmrob(formula = yy ~ xx)\n",
    " \\--> method = \"MM\"\n",
    "Coefficients:\n",
    "(Intercept)           xx  \n",
    "     -4.969        2.253  \n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-4.969387980288108 2.2531613477892365\n"
     ]
    }
   ],
   "source": [
    "params = [-4.969387980288108, 2.2531613477892365]  # Computed using R\n",
    "print(params[0], params[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "abline_plot(intercept=params[0], slope=params[1], ax=ax, color='red')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Exercise: Breakdown points of M-estimator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(12345)\n",
    "nobs = 200\n",
    "beta_true = np.array([3, 1, 2.5, 3, -4])\n",
    "X = np.random.uniform(-20,20, size=(nobs, len(beta_true)-1))\n",
    "# stack a constant in front\n",
    "X = sm.add_constant(X, prepend=True) # np.c_[np.ones(nobs), X]\n",
    "mc_iter = 500\n",
    "contaminate = .25 # percentage of response variables to contaminate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_betas = []\n",
    "for i in range(mc_iter):\n",
    "    y = np.dot(X, beta_true) + np.random.normal(size=200)\n",
    "    random_idx = np.random.randint(0, nobs, size=int(contaminate * nobs))\n",
    "    y[random_idx] = np.random.uniform(-750, 750)\n",
    "    beta_hat = sm.RLM(y, X).fit().params\n",
    "    all_betas.append(beta_hat)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_betas = np.asarray(all_betas)\n",
    "se_loss = lambda x : np.linalg.norm(x, ord=2)**2\n",
    "se_beta = lmap(se_loss, all_betas - beta_true)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Squared error loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.4450294873068618"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array(se_beta).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 2.99711706,  0.99898147,  2.49909344,  2.99712918, -3.99626521])"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_betas.mean(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 3. ,  1. ,  2.5,  3. , -4. ])"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "beta_true"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.236091328675419e-05"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "se_loss(all_betas.mean(0) - beta_true)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.2rc2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}
