Notes.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Notes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## selection of variable"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Principal Variables\n",
    "iterative search of variables that covariates more with Y response vector. After the first PV is found, the matrix is reduced to find the next one.\n",
    "\n",
    "KW: supervised methods\n",
    "\n",
    "\n",
    "Nørgaard, L., Saudland, A., Wagner, J., Nielsen, J. P., Munck, L., & Engelsen, S. B. (2000). Interval Partial Least-Squares Regression ( i PLS): A Comparative Chemometric Study with an Example from Near-Infrared Spectroscopy. Applied Spectroscopy, 54(3), 413–419. http://doi.org/10.1366/0003702001949500"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Forward stepwise selection\n",
    "Apply univariate statistic to all variables and validate prediction on test set for all of them. Variable with lower RMSEP (error on the test/validation set) is chosen. All two-variable models are then build and evaluated, until the inclusion of new variable doesn't affect RMSEP anymore. For the selection of variable care must be taken not to overfit and independant validation test is often required.\n",
    "\n",
    "KW: supervised methods\n",
    "\n",
    "\n",
    "Nørgaard, L., Saudland, A., Wagner, J., Nielsen, J. P., Munck, L., & Engelsen, S. B. (2000). Interval Partial Least-Squares Regression ( i PLS): A Comparative Chemometric Study with an Example from Near-Infrared Spectroscopy. Applied Spectroscopy, 54(3), 413–419. http://doi.org/10.1366/0003702001949500"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Recursive weighted Regression\n",
    "the loadings are used to weight the X matrix recursively until convergence. Only the number of variable is chosen arbitrarily.\n",
    "\n",
    "KW: supervised methods\n",
    "\n",
    "Rinnan, Å., Andersson, M., Ridder, C., & Engelsen, S. B. (2014). Recursive weighted partial least squares (rPLS): An efficient variable selection method using PLS. Journal of Chemometrics, 28(5), 439–447. http://doi.org/10.1002/cem.2582"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Modeling"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### interval PCA / PLS\n",
    "the spectra is split into interval (fixed or variable) size and models are build for each of them. Interval with best results are attributed to intervals of the spectra of most interest for the global model.\n",
    "\n",
    "Remark: how does it behave with very small interval (literature uses interval twice as big as regular interval for PLS). How dows it compare with univariate variable selection?\n",
    "\n",
    "Nørgaard, L., Saudland, A., Wagner, J., Nielsen, J. P., Munck, L., & Engelsen, S. B. (2000). Interval Partial Least-Squares Regression ( i PLS): A Comparative Chemometric Study with an Example from Near-Infrared Spectroscopy. Applied Spectroscopy, 54(3), 413–419. http://doi.org/10.1366/0003702001949500"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "254.46900494077323"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import scipy as sp\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline \n",
    "R = 5*1.8 #km\n",
    "A = np.pi*np.power(R,2)\n",
    "A\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## PCA\n",
    "\n",
    "### number of components\n",
    "\n",
    "Determining the number of components in PCA analysis is crucial because it allows to distinguis the informative variance from the noise. It is expected that the former is larger than the latter and therefore the first eignevalues of the covariance matrix (the variances of the first components) are larger than the following ones.\n",
    "\n",
    "The Tracy-Widom probability density function describes the distribution of the largest eigenvalues for completely random covariance matrices. Therefore it provides a test ot whether the data have structure (informative variables) or not.\n",
    "\n",
    "Saccenti, E., & Timmerman, M. E. (2016). Approaches to sample size determination for multivariate data: Applications to PCA and PLS-DA of omics data. Journal of Proteome Research, 15(8), 2379–2393. http://doi.org/10.1021/acs.jproteome.5b01029"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "no funciono"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAEICAYAAABYoZ8gAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n4hcRSTBK/CIiCUaJX0QkwSjxi4gkmP8D70OFlGflhdAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x108035ba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAEICAYAAABYoZ8gAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10b24eac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "ename": "NameError",
     "evalue": "name 'pd' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-2-019b224d814b>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     23\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshow\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     24\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 25\u001b[0;31m \u001b[0mdf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mDataFrame\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mA\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     26\u001b[0m \u001b[0mdf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdensity\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtitle\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'limiting density of eigenvalues'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'pd' is not defined"
     ]
    }
   ],
   "source": [
    "n = 2000\n",
    "p = 500\n",
    "\n",
    "S = np.matrix( (np.random.normal(0,1,n*p).reshape(n,p)) )\n",
    "C = np.cov(np.transpose(S)) \n",
    "\n",
    "S = np.subtract(S, np.mean( S, axis=0 ))\n",
    "#np.round( np.mean( S, axis=0 ) )\n",
    "C2 = np.divide( np.matmul(np.transpose(S),S), n )\n",
    "#C = C - np.diag(C)\n",
    "#C = np.identity(p)\n",
    "#C[0,0] = 10\n",
    "#C[1,1] = 8\n",
    "A = np.linalg.eig(C2)\n",
    "\n",
    "plt.title('sorted eigenvalues')\n",
    "#plt.plot(np.arange(C.shape[0]), A[0])\n",
    "plt.plot(np.arange(C.shape[0]), sorted(A[0],reverse=True))\n",
    "plt.show()\n",
    "\n",
    "plt.title('limiting histogram of eigenvalues')\n",
    "plt.hist(sorted(np.real(A[0])),bins=50)\n",
    "plt.show()\n",
    "\n",
    "df = pd.DataFrame(A[0])\n",
    "df.plot.density(title='limiting density of eigenvalues')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 360,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.98327283381558295"
      ]
     },
     "execution_count": 360,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.var(S[:,1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 314,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.552713678800501e-17"
      ]
     },
     "execution_count": 314,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(S[:,1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 379,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAW4AAAD8CAYAAABXe05zAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAHw1JREFUeJzt3Xl0XOWd5vHvr7Ra+1ayZEuybMs2Nl6wLYwNDiEQCCZ7\nh3QIWcg2zkkyTJbOyWSZyZz0THpJdzIJp5N0ExI6ZAESBpoAAUI7mC3YRt7lRV4lS7JkydrlRbJU\n7/xR14pwLKuEJdWt0vM5p45L996q+l1d8fDWe997X3POISIisSMQ7QJERGRsFNwiIjFGwS0iEmMU\n3CIiMUbBLSISYxTcIiIxRsEtIhJjFNwiIjFGwS0iEmMSJ+JNCwoKXHl5+US8tYhIXNq6detJ51ww\nkm0nJLjLy8upqqqaiLcWEYlLZlYX6bbqKhERiTEKbhGRGKPgFhGJMQpuEZEYo+AWEYkxCm4RkRij\n4BYRiTG+Cu57NhzkhQOt0S5DRMTXfBXcP954mJcPKrhFRC7FV8GdGDAGQ9GuQkTE30YNbjNbYGY7\nhj26zewLE1FMQoIxEFJyi4hcyqj3KnHO1QBXAZhZAtAIPDYRxQTMcG4i3llEJH6MtavkJuCwcy7i\nm6GMqRiDkJJbROSSxhrcdwAPXmyFma03syozq2ptfWMnGM2MkHJbROSSIg5uM0sG3gX89mLrnXP3\nOucqnXOVwWBEt5T9y2IMnFrcIiKXNJYW9zpgm3PuxEQVY5i6SkRERjGW4P4gI3STjJdwi3siP0FE\nJPZFFNxmlg7cDDw6kcWoj1tEZHQRTV3mnDsF5E9wLQQC6uMWERmNr66cDJj6uEVERuPD4I52FSIi\n/uar4DZdgCMiMipfBbcueRcRGZ3PglstbhGR0fgsuHVyUkRkNL4Kbo3jFhEZna+CW/cqEREZnc+C\nWy1uEZHR+Cy4dXJSRGQ0vgpu9XGLiIzOV8EdMAgpuUVELslXwZ0YCDCo4BYRuSRfBXdCQLO8i4iM\nxlfBnZhgDKjFLSJySf4K7oAxMKjgFhG5FH8Fd0JALW4RkVH4K7gDxsCg+rhFRC7FX8GdoFElIiKj\n8VdwB4xzGlUiInJJkc7ynmNmj5jZfjPbZ2ZrJqKYxIAxqJOTIiKXFNEs78APgGecc7ebWTKQNiHF\nJAQ4p64SEZFLGjW4zSwbuB74GIBzrh/on5BiAqY+bhGRUUTSVTIbaAXuN7PtZnafmaVPRDGJCcY5\njSoREbmkSII7EVgB/Ng5txw4BXz1wo3MbL2ZVZlZVWtr6xsqRi1uEZHRRRLcDUCDc26z9/MjhIP8\ndZxz9zrnKp1zlcFg8A0Vk5gQ0JWTIiKjGDW4nXPNQL2ZLfAW3QTsnYhiEnWTKRGRUUU6quRu4Ffe\niJIjwMcnpJhAgJAL35M7ELCJ+AgRkZgXUXA753YAlRNcC4kJ4bAeCDmSFdwiIhfluysnAXWXiIhc\ngr+COyFczrkBnaAUERmJr4J7WlICAGfODUa5EhER//JVcKclh4P7dP9AlCsREfEvXwX3tKHgVotb\nRGQkvgru8y1udZWIiIzMl8GtFreIyMh8FdzTksLDys+oj1tEZES+Cm61uEVERqfgFhGJMb4K7vOj\nSs4ouEVERuSr4E5LDvdxq8UtIjIyXwV3QsBITgxw+pxOToqIjMRXwQ3hfm51lYiIjMx/wZ2UwKk+\nBbeIyEh8F9zBzBRaes5GuwwREd/yXXCX5KVR23Yq2mWIiPiW74K7KCuVtt7+aJchIuJbvgvuvPRk\nTvcP6gSliMgIfBfchZkpAOrnFhEZge+Cu8AL7pO9fVGuRETEnyKa5d3MaoEeYBAYcM5N2Izv0zNT\nATjRreAWEbmYiILb8xbn3MkJq8QzM3caAPXtpyf6o0REYpLvukqypyVRmJlCzYmeaJciIuJLkQa3\nA/5gZlvNbP3FNjCz9WZWZWZVra2tl1XUgqJMDii4RUQuKtLgXuucWwGsAz5nZtdfuIFz7l7nXKVz\nrjIYDF5WUXMK0qk7eRrn3GW9j4hIPIoouJ1zjd6/LcBjwKqJLKokN42evgG6z+gugSIiFxo1uM0s\n3cwyzz8HbgGqJ7Ko0jzvBGWHTlCKiFwokhb3dOBlM9sJbAGecs49M5FFleSmARpZIiJyMaMOB3TO\nHQGWTUItQ+YGM0hODPBabQfrlhRP5keLiPie74YDQnjuyTVz8nm+piXapYiI+I4vgxvgxisKOXry\nFMfa1F0iIjKcb4N78cwsAA639ka5EhERf/FtcFcEMwkYvHDg8i7mERGJN74N7uy0JErz0tjf3B3t\nUkREfMW3wQ1w7dwC9h7vJhTSFZQiIuf5OriXl+bQfXZA/dwiIsP4Orivnx8kOSHAz16pjXYpIiK+\n4evgLspO5f2VJTz82jE6T2sCYRER8HlwA7xvZQkhB7/cVBftUkREfMH3wb2iLJe1FQU8XFWv27yK\niBADwQ3wzmXF1LefYduxzmiXIiISdTER3G9fOoPkhAC/eLU22qWIiERdTAR3Rkoia+cV8J/7Wjjd\nr8kVRGRqi4ngBvj09XPo7Rvgub0nol2KiEhUxUxwX12ex4zsVB7b3hjtUkREoipmgjsQMG6vLGVj\nTSu7G7qiXY6ISNTETHADfHLtbNKTE/jeczUaGigiU1ZMBXf2tCS+dMsCnq9p5acvH412OSIiURFT\nwQ3wievKua4in394ej8v6l7dIjIFRRzcZpZgZtvN7MmJLCiCOvjxh1cyuyCd//Ef1Zzq0/BAEZla\nxtLi/jywb6IKGYus1CS+9a4rqe84zXee2R/tckREJlVEwW1mJcDbgfsmtpzIXVtRwF1rynlgUx0v\nHVSXiYhMHZG2uL8PfAUITWAtY/blty1gTkE6X3lkF609fdEuR0RkUowa3Gb2DqDFObd1lO3Wm1mV\nmVW1tk5OCzgjJZHvf2A5nafPcdfPtnCmf3BSPldEJJoiaXFfB7zLzGqBh4AbzeyXF27knLvXOVfp\nnKsMBoPjXObIlpRk8y93LmdfczffeGy3xneLSNwbNbidc19zzpU458qBO4A/Ouc+POGVjcFNC6fz\nxbfO59HtjXzz8T0KbxGJa4nRLmC83H1jBT1nz/GTl46Sm57Ml26eH+2SREQmxJiC2zm3Edg4IZVc\nJjPj67ctpPP0Oe7ZcJBgRjIfWVMe7bJERMZd3LS4IRzef/9XS2g71c//fHwPfQMhPn7dbBICFu3S\nRETGTcxd8j6axIQAP/rQCm66opD/89Q+7vzJJnp1daWIxJG4C26A1KQE7rurku+8bylbattZ/0AV\n/QO+GoIuIvKGxWVwQ7jb5K+vLuXb71nCnw638eH7NtN15ly0yxIRuWxxG9zn3XlNGd99/zKq6tq5\n6bsv8Fpte7RLEhG5LHEf3ADvW1nC459bS2ZqIh/96RYe2nJMY71FJGZNieCG8BWWD396NUtKsvnq\no7u58yebqWs7Fe2yRETGbMoEN0BhZioP/ZfV/N17l1Dd2MXbvv8i979yVK1vEYkpUyq4ITzp8J3X\nlPHcl97Mmjn5fOuJvdxx7yZO9urugiISG6ZccJ9XlJ3KfXddzf9+z2Kq6jp4yz9t5IFXa9X6FhHf\nm7LBDZAQMD6yehbPfuFNLCvN4ZuP7+FTP6+iprkn2qWJiIxoSgf3eRWFmTzwiVX8zc3z2Xy0ndvu\neYm/+/0+zWcpIr6k4PYEAsbdN83jpa+8hfevLOHeF49w03df4NebjxEKqftERPxDwX2B3PRk/uF9\nS/nNp9dQnJPK1x/bzbt/+Ao76zujXZqICKDgHtGq2Xk8+plr+b8fWEZDx2ne/cNX+NB9m3j1cFu0\nSxORKc4mYhRFZWWlq6qqGvf3jZbevgF+vbmO+146SktPHzdeUcjHri3nuooC3TJWRMaFmW11zlVG\ntK2CO3Jnzw1y/yu1/GjjIXrODlCcncp7ls/kzlVllOalRbs8EYlhCu4JdvbcIBv2tfDI1npePHiS\nkHOsLMvlw6tnsW5JESmJCdEuUURijIJ7EjV1neGRqgYe3d7I0ZOnCGamcPeNFbx72Uyy05KiXZ6I\nxAgFdxSEQo6XD53kng0HqarrICFg3LakmE+tnc3SkmzM1BcuIiMbS3DH1ZyT0RQIGNfPD/KmeQVs\nrevgmepmHni1jid2HmfB9Ew++abZvHPpDKYlqxtFRC7PqC1uM0sFXgRSCAf9I865/3Wp10zFFvfF\ntJ/q59k9zfzs5aMcbOklLTmBWxcX8ZHVs1helhvt8kTER8a1q8TC3/HTnXO9ZpYEvAx83jm3aaTX\nKLhfb2AwxJaj7Tyx6zhP7Gyit2+ApSXZfHBVGe9dPpPUJLXCRaa6CevjNrM0wsH9Gefc5pG2U3CP\nrLdvgMe2NfCLTXUcONFLRkoit1w5nfdcNVPjwkWmsHEPbjNLALYCFcAPnXP//SLbrAfWA5SVla2s\nq6sbU9FTjXOOPx1u4/EdjTxd3UzP2QGmZ6Xw3uUl3L6yhIrCjGiXKCKTaCJb3DnAY8DdzrnqkbZT\ni3tszp4b5I/7W3h0WwPP17QyGHJcUZTJusXFvHNZMbML0jUqRSTOTehwQDP7JnDaOffPI22j4H7j\nWrrP8tTuJp7a1cTWYx04B2V5adyyaDq3XFnEylm56k4RiUPjfXIyCJxzznWa2TTgD8A/OueeHOk1\nCu7xcbzzDBv2neC5fS1sOtxG/2CIgoxk3rpwOrcuLuJN84IKcZE4Md7BvRT4OZBA+G6Cv3HO/e2l\nXqPgHn+9fQNsrGnh2T0neH5/C719A8zMmcYtV05n3eJiKmflElCIi8QsXTkZ5/oGBnlu7wke29bI\nS4dO0j8QIpiZwlsXFnLDgkKunZtPZqoutxeJJQruKaS3b4Bnq5v54/4WNta0cKp/kISAsaIsh+vn\nBbl+fpDFM7PVpSLicwruKap/IMT2Yx28eLCVFw+cZHdjFwA5aUm8eX6Qa+fm85YFhRRmpUa5UhG5\nkIJbAGjr7ePlQyd54UArL9S00naqH4CVs3K5YX6QayvyWVqSQ1KCJkISiTYFt/wF5xw1J3r4w54T\nbNh3gl2NXTgHyYkBVpbl8pYrgqytCLKgKFPdKiJRoOCWUbX19rHlaDtVdR28cugk+5t7AEhPTuDq\n2XmsKMvlmtl5LC3J0R0NRSaBbusqo8rPSGHdkmLWLSkGwhNCbDrSxta6DjYfaeeFA604B4kBY2lJ\nNjcsKGT1nHwWFmdqxIpIlKnFLRfVebqf12o72FrXwaYjbeyo7xxaNzeYzqrZ+ayek8eq2XkUZ0+L\nYqUi8UFdJTLu2nr72NnQyZ7GbrYd66CqtoOevgEASvOmcc3sfFaV57F4ZjYVhRkkJ+qEp8hYKLhl\nwg2GHPuautl8tJ3NR9p4rbadjtPnAMielsTyshxWzc6jclaeuldEIqDglkkXCjmOnDzFnuNdvFDT\nyp7j3dSc6BlaPzeYzvKyXJaX5XBVaQ4LpmeSqGGIIkMU3OILrT197GropLqxm92NnWw71km7N5Y8\nLTmBpSXZrCjLpbI8lzVzCjR6RaY0Bbf4knOOurbT7KjvZPuxDrYd62RvUzeDIUdCwFhYnMnaiqA3\nDDGb/IyUaJcsMmkU3BIzzvQPUlXXzpaj4cfWug4GQuG/yZk501gyM5sFRZksmpFF5axchbnELY3j\nlpgxLTmBN80L8qZ5QQBO9w+ws76L3Y2d7Groorqxi2f3NnO+fTGvMINVs/NYVpLDohlZzJueQUqi\nulhkalGLW3zv7LlBqhu72FIbbpVX1XbQ6w1FTEow5k/PZHlZDstLc7m6PI/SvGma6k1ijrpKJK6F\nQo669tPsOd7FnuPd7G7oYkd951CY56Qlsag4i4VDj0wqCtUyF39TV4nEtUDAmF2QzuyCdN6xdAYQ\nHld+qKWX12rbqW7sYl9TN7/cVEffQAgIX7pfUZgxFOQLi7O4oiiLYKb6zCX2KLglLiQEjAVFmSwo\nyhxaNhhyHD15in1N3UOPVw+38dj2xqFtCjJSWFicyaLiLBbNyGJZSQ6z8tPU1SK+puCWuJXgtbIr\nCjN457IZQ8s7TvWzr6mbvU3d7GvqYX9zN/e/Ukv/YLh1XpCRwlWl2Vw5I5uFxZnMn55JeX665vQU\n31Bwy5STm57MtRUFXFtRMLTs3GCIgyd62Xasg211Hexo6GTD/pah0SzpyQksmhHuM188M5v50zOZ\nlZdGbnpylPZCprJIZnkvBR4ApgMOuNc594NLvUYnJyUenOkf5GBLD/ube9jT2EX18W4ONPcM3VwL\noCgrlaUl2SwrzWFpSTZLZmaTk6Ywl7Eb75OTA8DfOOe2mVkmsNXMnnPO7b2sKkV8blpyAktLclha\nkgOVpcCf78ly9OQpak+eovp4F7sbuvjD3hNDr5uVnxZ+3cxslpZks3hmNukp+nIr42fUvybnXBPQ\n5D3vMbN9wExAwS1TTmBYv/lwXWfOUd3Yxc6GTnY3dLGtroMndh4HwAwqghlD92W5coZufSuXZ0zN\nADMrB5YDmyeiGJFYlT0tiesqCrhuWL95a0/f0BWgO+s7eWZPMw9X1QPh4YlzguksKMriyhlZLJmZ\nzZUzstTNIhGJ+AIcM8sAXgC+7Zx79CLr1wPrAcrKylbW1dWNZ50iMS8Uchxu7WVvUzcHTvRQ09zD\nvqYeGjvPDG1TmjeNxTPC3SuLZ2azrER95lPFuF85aWZJwJPAs8657422vU5OikSu41R/+ArQxi6q\nj3exp7GL2rbTQ+tn5aexoiyXRcVZzC5IZ9GMLIqzUzXWPM6Ma3Bb+K/j50C7c+4Lkbypglvk8nSd\nOceexi52NHSyra6TXQ2dtPT0Da3PnpbkXTgUHmu+oCiTucEMnQSNYeMd3GuBl4DdQMhb/HXn3O9H\neo2CW2R8OefoPjPAwZYe7+KhHvY2dVPT3M3Zc+H/LAMGJblpzAmmMzcYPoE6uyCdxTOzyVCg+964\nDgd0zr0M6DuZSBSZGdlpSVSW51FZnje0/Pxl/Ydawv3lh1t7OdJ6ik1H2oYC3QwKM1Moz0/37mue\nx+KZWZTmpulq0BiluwOKxKFQyNHYeYZDrb1UN3RxrP00B1t62T+shZ6RksjC4kyWleSwrDSHZSU5\nuiVuFOm2riJyUf0DIfY3d7P3ePheLdXeFaH93l0Uc9OSWFKSw7KSbJZ6/xZmpUa56qlBt3UVkYtK\nTgz8+WpQT/9AiAMnetjZ0Mmu+vBFRD/aeJJBbwq585f1r5yVy8Li8KxDRVka1RJNCm6RKS45MTA0\nbvxD14SXnekfZG9TFzvru9jV0MmO+s7XXdafmZrI/Onh0SxXleQwb3r4XuepSZqsYjKoq0REItJ+\nqp/9Td0cbu1lf3MPB1t62dfUTc/Z8E23EgPGohlZrCrPo7I8lxVluepmGQP1cYvIpDg/jVxNcw+7\nGjqpqutgR33nUJ95WV4ai2dmcf28INfMyadck1SMSH3cIjIphk8jd+viIgD6BgbZe7ybqtoOth3r\nYPuxTn6/uxkIT1JxdXkuK2flsmZuPguLsjQk8Q1QcIvIuEpJTGB5WS7Ly3KB8MVD4flAO3ittp0t\nR9t5ujoc5NnTklg1O4/Vc/JZPSdPQR4hBbeITCgzY970TOZNz+TOa8oAaO46yyuHTrLlaDubjrbx\nnHfiMy89mRVluSwvy2F5aQ5LSrLJTE2KZvm+pD5uEYm6451nePVwG3863Mb2Yx0cOXlqaF1FYQaV\ns3JZVprD1eV5zA2mx2U/uU5OikhM6zzdz4768KQU2+s7ea22fWj0SlFWKmvm5rO2ooAbryiMm3k/\ndXJSRGJaTloyNywo5IYFhcCfR6+8eriNVw6f5KWDrTy2vZGAwaIZWdwwv5BbFxexqHhq9JGrxS0i\nMScUcuxu7GLD/hY2H2njtdp2Qg7y05O5fn6QGxYEuX5eMKZa42pxi0hcCwQsfGOs0vCl+01dZ9h8\npJ0XDrTywoFwa9wMlpXkcMuV03n7kmJm5adHuerxoxa3iMSVUMixq7GLjTUtPF/Tys76TgAWFmex\nbnER6xYXMW96ZpSr/Es6OSki4mnoOM0z1c08Xd3M1roOAOYG01m3uJg7VpVSkpsW5QrDFNwiIhdx\novssz+5p5undzWw+2kbAjPetKGH9m+cwN5gR1doU3CIiozjeeYYfbTzEb6sa6BsIsbaigL++upTb\nFheRmBCY9HoU3CIiEWrt6ePXm4/xyLZ66tvPUJydyodXz+Kua8snda5OBbeIyBiFQo4N+1v49z8d\n5ZVDbczMmcZXbl3Au5bNmJQrNccS3JP/fUBExIcCAePmRdP51adW8+tPXUNOWhKff2gH/+2hHbT1\n9kW7vNcZNbjN7Gdm1mJm1ZNRkIhItF1bUcAT/3Utn79pHs9UN/H2e15m85G2aJc1JJIW978Dt05w\nHSIivhIIGF+8eT6Pf24tKUkB7vjJJj77q600d52NdmmjB7dz7kWgfRJqERHxnUUzsvjd59byuRsq\neH5/K2+/5yU21rREtaZx6+M2s/VmVmVmVa2treP1tiIiUZedlsSX37aAJ+6+jrz0ZD52/2t877kD\nhELjP7gjEuMW3M65e51zlc65ymAwOF5vKyLiGxWFmTxx91puX1nCPRsOcveD2+kbGJz0OjSqRERk\nDFKTEvin25fy5Vvm89TuJn5b1TDpNSi4RUTGyMz47A0VlOencd9LR+jtG5jUz49kOOCDwKvAAjNr\nMLNPTnxZIiL+FggYf/9XS6lrP823n9o3uZ892gbOuQ8654qdc0nOuRLn3E8nozAREb9bMzefO64u\n49FtDfScPTdpn6uuEhGRy/CBq0vpGwjx4JZjk/aZCm4RkctwVWkO187N594Xj07a8EAFt4jIZfrA\n1aWc7O1jR0PnpHyegltE5DJdOSMLgPr205PyeQpuEZHLVJQ9DYCmSbqPiYJbROQyZaQkkpmaSFPn\nmUn5PAW3iMg4KM5OVYtbRCSWFGVPo7lbwS0iEjOyUhPpPTs5l74ruEVExkFacgKn+yfnToGTN4Wx\niEgcu6o0l8AkTCoMCm4RkXFx5zVl3HlN2aR8lrpKRERijIJbRCTGKLhFRGKMgltEJMYouEVEYoyC\nW0Qkxii4RURijIJbRCTGmHPjP9WOmbUCdW/w5QXAyXEsJxZon+PfVNtf0D6P1SznXDCSDSckuC+H\nmVU55yqjXcdk0j7Hv6m2v6B9nkjqKhERiTEKbhGRGOPH4L432gVEgfY5/k21/QXt84TxXR+3iIhc\nmh9b3CIicgm+CW4zu9XMaszskJl9Ndr1XA4zKzWz581sr5ntMbPPe8vzzOw5Mzvo/ZvrLTczu8fb\n911mtmLYe93lbX/QzO6K1j5FwswSzGy7mT3p/TzbzDZ7+/WwmSV7y1O8nw9568uHvcfXvOU1Zva2\n6OxJ5Mwsx8weMbP9ZrbPzNbE83E2sy96f9PVZvagmaXG43E2s5+ZWYuZVQ9bNm7H1cxWmtlu7zX3\nmI1xBgbnXNQfQAJwGJgDJAM7gUXRrusy9qcYWOE9zwQOAIuA7wBf9ZZ/FfhH7/ltwNOAAauBzd7y\nPOCI92+u9zw32vt3if3+EvBr4Env598Ad3jP/xX4jPf8s8C/es/vAB72ni/yjn0KMNv7m0iI9n6N\nss8/Bz7lPU8GcuL1OAMzgaPAtGHH92PxeJyB64EVQPWwZeN2XIEt3rbmvXbdmOqL9i/I24k1wLPD\nfv4a8LVo1zWO+/c4cDNQAxR7y4qBGu/5vwEfHLZ9jbf+g8C/DVv+uu389ABKgA3AjcCT3h/kSSDx\nwmMMPAus8Z4netvZhcd9+HZ+fADZXpDZBcvj8jh7wV3vBVGid5zfFq/HGSi/ILjH5bh66/YPW/66\n7SJ5+KWr5PwfxHkN3rKY5309XA5sBqY755q8Vc3AdO/5SPsfS7+X7wNfAULez/lAp3Pu/LTXw2sf\n2i9vfZe3fSztL4Rbi63A/V4X0X1mlk6cHmfnXCPwz8AxoInwcdtK/B/n88bruM70nl+4PGJ+Ce64\nZGYZwP8DvuCc6x6+zoX/VxsXQ3rM7B1Ai3Nua7RrmWSJhL9O/9g5txw4Rfgr9JA4O865wLsJ/w9r\nBpAO3BrVoqIk2sfVL8HdCJQO+7nEWxazzCyJcGj/yjn3qLf4hJkVe+uLgRZv+Uj7Hyu/l+uAd5lZ\nLfAQ4e6SHwA5ZnZ+QurhtQ/tl7c+G2gjdvb3vAagwTm32fv5EcJBHq/H+a3AUedcq3PuHPAo4WMf\n78f5vPE6ro3e8wuXR8wvwf0aMM87O51M+ETG76Jc0xvmnSH+KbDPOfe9Yat+B5w/s3wX4b7v88s/\n6p2dXg10eV/JngVuMbNcr7Vzi7fMV5xzX3POlTjnygkfuz865z4EPA/c7m124f6e/z3c7m3vvOV3\neKMRZgPzCJ/E8SXnXDNQb2YLvEU3AXuJ0+NMuItktZmleX/j5/c3ro/zMONyXL113Wa22vs9fnTY\ne0Um2icAhnXQ30Z49MVh4BvRrucy92Ut4a9Ru4Ad3uM2wv17G4CDwH8Ced72BvzQ2/fdQOWw9/oE\ncMh7fDza+xbBvt/An0eVzCH8H+Qh4LdAirc81fv5kLd+zrDXf8P7PdQwxjPtUdrfq4Aq71j/B+HR\nA3F7nIFvAfuBauAXhEeGxN1xBh4k3I9/jvA3q0+O53EFKr3f4WHgX7jgBPdoD105KSISY/zSVSIi\nIhFScIuIxBgFt4hIjFFwi4jEGAW3iEiMUXCLiMQYBbeISIxRcIuIxJj/D5vOgmO9aFonAAAAAElF\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11f01f080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n = 5\n",
    "p = 5\n",
    "D = []\n",
    "L = 10000\n",
    "\n",
    "for i in range(0, L):\n",
    "    #S = np.matrix( (sp.randn(n,p)) )\n",
    "    S = np.matrix( (np.random.normal(0,1,n*p).reshape(n,p)) )\n",
    "    S = np.subtract(S, np.mean( S, axis=0 ))\n",
    "    C = np.divide( np.matmul(np.transpose(S),S), n )\n",
    "    A = np.linalg.eig(C)\n",
    "    D.append(np.max(A[0]))\n",
    "\n",
    "\n",
    "plt.plot(np.arange(L), sorted(D,reverse=True))\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 380,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD8CAYAAAB5Pm/hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1234c2ba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(D,bins=100)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 381,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x121262d68>"
      ]
     },
     "execution_count": 381,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAD8CAYAAACYebj1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhYKIiIxRKIiIyJj/D3+Hmtpb69JNAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x123de08d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df = pd.DataFrame(D)\n",
    "df.plot(kind='density')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 391,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "C = np.identity(p)\n",
    "C[0,0] = 10\n",
    "C[1,1] = 8"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 397,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 10.,   0.,   0.,   0.,   0.],\n",
       "       [  0.,   8.,   0.,   0.,   0.],\n",
       "       [  0.,   0.,   1.,   0.,   0.],\n",
       "       [  0.,   0.,   0.,   1.,   0.],\n",
       "       [  0.,   0.,   0.,   0.,   1.]])"
      ]
     },
     "execution_count": 397,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "C"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 398,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([ 10.,   8.,   1.,   1.,   1.]), array([[ 1.,  0.,  0.,  0.,  0.],\n",
       "        [ 0.,  1.,  0.,  0.,  0.],\n",
       "        [ 0.,  0.,  1.,  0.,  0.],\n",
       "        [ 0.,  0.,  0.,  1.,  0.],\n",
       "        [ 0.,  0.,  0.,  0.,  1.]]))"
      ]
     },
     "execution_count": 398,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.linalg.eig(C)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 401,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0, 1, 2],\n",
       "       [3, 4, 5],\n",
       "       [6, 7, 8]])"
      ]
     },
     "execution_count": 401,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "C = np.arange(0,9).reshape(3,3)\n",
    "C"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 403,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([  1.33484692e+01,  -1.34846923e+00,  -1.15433316e-15]),\n",
       " array([[ 0.16476382,  0.79969966,  0.40824829],\n",
       "        [ 0.50577448,  0.10420579, -0.81649658],\n",
       "        [ 0.84678513, -0.59128809,  0.40824829]]))"
      ]
     },
     "execution_count": 403,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "A = np.linalg.eig(C)\n",
    "A"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 407,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([  1.12097436,   6.61079673,  12.1006191 ])"
      ]
     },
     "execution_count": 407,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.matmul( A[1], np.transpose(A[0]) )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 390,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAECCAYAAADesWqHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ncLkSZfuMG0AAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x122af0c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "matrix([[  6.05680942,   6.35746886,  -3.65037618,   2.22960854,\n",
       "           1.06565521],\n",
       "        [  6.35746886,   8.81753445,  -4.94913105,   3.29954631,\n",
       "          -0.21422307],\n",
       "        [ -3.65037618,  -4.94913105,   5.84039731,  -2.4141185 ,\n",
       "          -0.76359265],\n",
       "        [  2.22960854,   3.29954631,  -2.4141185 ,   1.86817163,\n",
       "           2.07191094],\n",
       "        [  1.06565521,  -0.21422307,  -0.76359265,   2.07191094,\n",
       "          10.04302941]])"
      ]
     },
     "execution_count": 390,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plt.matshow(np.matmul(np.transpose(S),S))\n",
    "plt.show()\n",
    "np.matmul(np.transpose(S),S)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 385,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.230687041939861"
      ]
     },
     "execution_count": 385,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.percentile(D,50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 310,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAD8CAYAAABw1c+bAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nGINfkhpj8EtSYwx+SWqMwS9Jjfk/EeUgCRyVL/IAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x108dec710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(sorted(np.real(A[0])),bins=50)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "40.000000000000007"
      ]
     },
     "execution_count": 249,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.power( np.sqrt(n)+np.sqrt(p), 2 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 328,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": [
    "mu_np = np.power( np.sqrt(n-1) + np.sqrt(p), 2)\n",
    "sigma_np = np.multiply( np.sqrt(n-1) + np.sqrt(p), np.power(1/np.sqrt(n-1)+1/np.sqrt(p),1/3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 329,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-251.81718334607544"
      ]
     },
     "execution_count": 329,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(A[0][0] - mu_np) / sigma_np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 330,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.9689203201800956"
      ]
     },
     "execution_count": 330,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "A[0][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.2432954748161231"
      ]
     },
     "execution_count": 240,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.std(S[:,0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nBzEW3zyhiZ2oAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x124f5bf98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(S[:,0],bins=50)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD8CAYAAABn919SAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAADxpJREFUeJzt3X+IZWd9x/H3xzWpErVJzHTZJtluoCFWSjVlSCuKWNdY\na8Tdgg0GkW0bGARtI7aYrUKDbYWVgj9oi3UxtiukxjQx7FJ/1O2aYP3D6CaummRjk4YEd9nsrj+C\nxkJl9ds/5qSOydy9587cu/fOM+8XDPecc8+d+TDsfuaZ557zTKoKSdLa94xpB5AkjYeFLkmNsNAl\nqREWuiQ1wkKXpEZY6JLUCAtdkhoxtNCTXJbk0JKPHyR5e5Lzk+xP8mD3eN6ZCCxJWl5GubEoyQbg\nKPBbwFuB71XVriQ7gfOq6vrJxJQkDTNqob8auKGqXprkW8ArqupYkk3AnVV12elef8EFF9SWLVtW\nFViS1pu77777O1U1N+y8Z474ed8IfKLb3lhVx7rtx4CNy70gyQKwALB582YOHjw44peUpPUtyaN9\nzuv9pmiSs4HXA//61OdqcZi/7FC/qnZX1XxVzc/NDf0BI0laoVGucvk94J6qOt7tH++mWugeT4w7\nnCSpv1EK/Rp+Nt0CsA/Y0W3vAPaOK5QkaXS9Cj3JOcCVwKeWHN4FXJnkQeBV3b4kaUp6vSlaVT8C\nnv+UY98Ftk4ilCRpdN4pKkmNsNAlqREWuiQ1wkKXpEaMeqeopNPYsvPTA597ZNdVZzCJ1iNH6JLU\nCAtdkhphoUtSIyx0SWqEhS5JjbDQJakRFrokNcJCl6RGWOiS1AgLXZIaYaFLUiNcy0VagdOt2SJN\niyN0SWqEhS5JjbDQJakRFrokNcJCl6RG9Cr0JOcmuTXJA0kOJ3lJkvOT7E/yYPd43qTDSpIG6ztC\n/xDwuap6AfAi4DCwEzhQVZcCB7p9SdKUDC30JL8IvBy4EaCqflxVjwPbgD3daXuA7ZMKKUkars8I\n/RLgJPBPSb6W5KNJzgE2VtWx7pzHgI3LvTjJQpKDSQ6ePHlyPKklSU/Tp9CfCfwm8OGquhz4EU+Z\nXqmqAmq5F1fV7qqar6r5ubm51eaVJA3Qp9CPAEeq6q5u/1YWC/54kk0A3eOJyUSUJPUxtNCr6jHg\n20ku6w5tBe4H9gE7umM7gL0TSShJ6qXv4lx/AtyU5GzgYeCPWPxhcEuSa4FHgasnE1GS1EevQq+q\nQ8D8Mk9tHW8cSdJKeaeoJDXCQpekRljoktQIC12SGmGhS1IjLHRJaoSFLkmNsNAlqREWuiQ1wkKX\npEb0XctFmklbdn56pPMf2XXVhJJI0+cIXZIaYaFLUiMsdElqhIUuSY2w0CWpERa6JDXCQpekRljo\nktQIbyySpmzQzVHeBKVROUKXpEZY6JLUiF5TLkkeAX4I/AQ4VVXzSc4HPglsAR4Brq6q708mpta7\nUddskdajUUbov1NVL66q+W5/J3Cgqi4FDnT7kqQpWc2UyzZgT7e9B9i++jiSpJXqe5VLAZ9PUsBH\nqmo3sLGqjnXPPwZsXO6FSRaABYDNmzevMq60fnj1i0bVt9BfVlVHk/wSsD/JA0ufrKrqyv5puvLf\nDTA/P7/sOZKk1es15VJVR7vHE8DtwBXA8SSbALrHE5MKKUkabmihJzknyXOf3AZeDdwL7AN2dKft\nAPZOKqQkabg+Uy4bgduTPHn+v1TV55J8FbglybXAo8DVk4spSRpmaKFX1cPAi5Y5/l1g6yRCSZJG\n552iktQIC12SGmGhS1IjLHRJaoSFLkmNsNAlqREWuiQ1wj9Bp3Vl1AWvXIdda4kjdElqhIUuSY1w\nykUzxSkOaeUcoUtSIyx0SWqEhS5JjbDQJakRFrokNcKrXCS8ukZtcIQuSY2w0CWpERa6JDXCQpek\nRljoktSI3le5JNkAHASOVtXrklwC3Aw8H7gbeHNV/XgyMaW1zytpNGmjjNCvAw4v2X8f8IGq+lXg\n+8C14wwmSRpNr0JPchFwFfDRbj/AK4Fbu1P2ANsnEVCS1E/fEfoHgXcCP+32nw88XlWnuv0jwIXL\nvTDJQpKDSQ6ePHlyVWElSYMNLfQkrwNOVNXdK/kCVbW7quaran5ubm4ln0KS1EOfN0VfCrw+yWuB\nZwHPAz4EnJvkmd0o/SLg6ORiSpKGGTpCr6q/qKqLqmoL8EbgC1X1JuAO4A3daTuAvRNLKUkaajXX\noV8PvCPJQyzOqd84nkiSpJUYabXFqroTuLPbfhi4YvyRJEkr4Z2iktQIC12SGmGhS1IjLHRJaoSF\nLkmNsNAlqREWuiQ1YqTr0CXNrkHrrT+y66oznETT4ghdkhphoUtSI5xykdYY/5SdBnGELkmNsNAl\nqREWuiQ1wkKXpEZY6JLUCAtdkhphoUtSIyx0SWqEhS5JjbDQJakRQws9ybOSfCXJ15Pcl+Q93fFL\nktyV5KEkn0xy9uTjSpIG6bOWy/8Cr6yqJ5KcBXwpyWeBdwAfqKqbk/wjcC3w4QlmVUNcj0Qav6Ej\n9Fr0RLd7VvdRwCuBW7vje4DtE0koSeql1xx6kg1JDgEngP3AfwOPV9Wp7pQjwIWTiShJ6qPX8rlV\n9RPgxUnOBW4HXtD3CyRZABYANm/evJKMWgP8aznS9I10lUtVPQ7cAbwEODfJkz8QLgKODnjN7qqa\nr6r5ubm5VYWVJA3W5yqXuW5kTpJnA1cCh1ks9jd0p+0A9k4qpCRpuD5TLpuAPUk2sPgD4Jaq+rck\n9wM3J/kb4GvAjRPMKUkaYmihV9U3gMuXOf4wcMUkQkmSRuedopLUCP9ItCbKG4ikM8cRuiQ1wkKX\npEZY6JLUCAtdkhphoUtSI7zKRVqnXH+nPY7QJakRFrokNcJCl6RGWOiS1AgLXZIaYaFLUiO8bFFq\nnAukrR+O0CWpERa6JDXCQpekRljoktQIC12SGuFVLurNqyXWBxftWrscoUtSI4YWepKLk9yR5P4k\n9yW5rjt+fpL9SR7sHs+bfFxJ0iB9RuingD+rqhcCvw28NckLgZ3Agaq6FDjQ7UuSpmRooVfVsaq6\np9v+IXAYuBDYBuzpTtsDbJ9USEnScCPNoSfZAlwO3AVsrKpj3VOPARvHmkySNJLehZ7kOcBtwNur\n6gdLn6uqAmrA6xaSHExy8OTJk6sKK0karFehJzmLxTK/qao+1R0+nmRT9/wm4MRyr62q3VU1X1Xz\nc3Nz48gsSVpGn6tcAtwIHK6q9y95ah+wo9veAewdfzxJUl99bix6KfBm4JtJDnXH3gXsAm5Jci3w\nKHD1ZCLqTPMGImltGlroVfUlIAOe3jreOJKklfJOUUlqhIUuSY2w0CWpERa6JDXCQpekRljoktQI\nC12SGmGhS1IjLHRJaoSFLkmNsNAlqREWuiQ1wkKXpEZY6JLUiD7roUvSyE63rv4ju646g0nWD0fo\nktQIC12SGmGhS1IjLHRJaoSFLkmNsNAlqREWuiQ1YmihJ/lYkhNJ7l1y7Pwk+5M82D2eN9mYkqRh\n+txY9M/A3wMfX3JsJ3CgqnYl2dntXz/+eJJmxaAbhbxJaHYMHaFX1ReB7z3l8DZgT7e9B9g+5lyS\npBGtdA59Y1Ud67YfAzaOKY8kaYVWvZZLVVWSGvR8kgVgAWDz5s2r/XLCX30lLW+lI/TjSTYBdI8n\nBp1YVburar6q5ufm5lb45SRJw6y00PcBO7rtHcDe8cSRJK3U0CmXJJ8AXgFckOQIcAOwC7glybXA\no8DVkwypyTjd8qaS1p6hhV5V1wx4auuYs0iSVsE7RSWpEf7FIkmr4tTd7HCELkmNsNAlqRFOucyA\ncd0o5K++0vrmCF2SGmGhS1IjnHKRdMa5HtFkOEKXpEZY6JLUCKdcJsCrViRNgyN0SWqEhS5JjXDK\nZRVGnRJxCkUaL6+W+XmO0CWpERa6JDXCQpekRjiHLmlmOCe+Oo7QJakRFrokNcIpF0kzb61c8jvt\nKSNH6JLUiFUVepLXJPlWkoeS7BxXKEnS6FY85ZJkA/APwJXAEeCrSfZV1f3jCrfUtH6VWSu/6kka\nbtT/z4P6ZVZ7YTUj9CuAh6rq4ar6MXAzsG08sSRJo1pNoV8IfHvJ/pHumCRpCiZ+lUuSBWCh230i\nybfG+vnfN87P9v8uAL4zkc88GWsp71rKCuadtInkHVcvLPN5VpR3DHl+pc9Jqyn0o8DFS/Yv6o79\nnKraDexexdc545IcrKr5aefoay3lXUtZwbyTZt7xWs2Uy1eBS5NckuRs4I3AvvHEkiSNasUj9Ko6\nleRtwL8DG4CPVdV9Y0smSRrJqubQq+ozwGfGlGWWrKkpItZW3rWUFcw7aeYdo1TVtDNIksbAW/8l\nqREW+gBJ/jrJN5IcSvL5JL887UyDJPnbJA90eW9Pcu60M51Okj9Icl+SnyaZ2SsG1tLSFkk+luRE\nknunnWWYJBcnuSPJ/d2/g+umnel0kjwryVeSfL3L+55pZxrEKZcBkjyvqn7Qbf8p8MKqesuUYy0r\nyauBL3RvVL8PoKqun3KsgZL8GvBT4CPAn1fVwSlHeppuaYv/YsnSFsA1k1raYrWSvBx4Avh4Vf36\ntPOcTpJNwKaquifJc4G7ge0z/L0NcE5VPZHkLOBLwHVV9eUpR3saR+gDPFnmnXOAmf3JV1Wfr6pT\n3e6XWbwnYGZV1eGqGusNZhOwppa2qKovAt+bdo4+qupYVd3Tbf8QOMwM32Vei57ods/qPmayDyz0\n00jy3iTfBt4E/OW08/T0x8Bnpx2iAS5tcQYk2QJcDtw13SSnl2RDkkPACWB/Vc1k3nVd6En+I8m9\ny3xsA6iqd1fVxcBNwNtmOWt3zruBUyzmnao+ebW+JXkOcBvw9qf8RjxzquonVfViFn/7vSLJTE5r\nreu/WFRVr+p56k0sXm9/wwTjnNawrEn+EHgdsLVm4I2REb63s6rX0hZamW4u+jbgpqr61LTz9FVV\njye5A3gNMHNvQK/rEfrpJLl0ye424IFpZRkmyWuAdwKvr6r/mXaeRri0xYR0bzLeCByuqvdPO88w\nSeaevHIsybNZfKN8JvvAq1wGSHIbcBmLV2M8CrylqmZyhJbkIeAXgO92h748q1fkACT5feDvgDng\nceBQVf3udFM9XZLXAh/kZ0tbvHfKkQZK8gngFSyuBngcuKGqbpxqqAGSvAz4T+CbLP7/AnhXd+f5\nzEnyG8AeFv8dPAO4par+arqplmehS1IjnHKRpEZY6JLUCAtdkhphoUtSIyx0SWqEhS5JjbDQJakR\nFrokNeL/APEXBBh17xKwAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x123e3ff60>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(S[:,0],bins=50)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1.05778373, -0.05881601,  0.00954084, -0.02052903, -0.0627378 ,\n",
       "        0.01765268, -0.01276595, -0.03752322, -0.00955641])"
      ]
     },
     "execution_count": 216,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "C[1,1:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "matrix([[ 0.97093395,  0.00483192, -0.00124857, -0.00134383, -0.04884857,\n",
       "         -0.01170613, -0.03493499, -0.01353647,  0.00439961]])"
      ]
     },
     "execution_count": 190,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "C2[1,1:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ -3.23330264e-02,  -5.42963534e-02,   2.91393589e-02, ...,\n",
       "         -9.08881444e-05,   3.17666129e-03,   9.54992063e-03],\n",
       "       [ -3.81267195e-03,  -1.09199626e-02,   2.85962102e-03, ...,\n",
       "          2.61187388e-02,   6.81834340e-03,  -2.59857166e-02],\n",
       "       [ -1.13351998e-02,   1.66869346e-02,  -8.59725941e-03, ...,\n",
       "          4.17269658e-03,  -3.23213605e-02,  -6.76938947e-02],\n",
       "       ..., \n",
       "       [ -3.22049552e-02,  -1.86264279e-03,  -1.72076189e-02, ...,\n",
       "          3.40683941e-02,   1.97703868e-02,   8.70328792e-03],\n",
       "       [  3.09943079e-04,   5.13518582e-02,  -1.01703687e-02, ...,\n",
       "          4.05890881e-04,   5.91430163e-02,  -7.23158034e-02],\n",
       "       [ -8.44677541e-03,  -2.39635493e-02,   1.15079190e-02, ...,\n",
       "          2.13802364e-02,  -3.31688662e-02,   2.29927130e-02]])"
      ]
     },
     "execution_count": 191,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "A[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0021214258020596241"
      ]
     },
     "execution_count": 192,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.var(C[0,:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAARoAAAEECAYAAAD3b/tRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1242bfe10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.matshow(C2)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niakAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x124215630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(np.arange(C.shape[0]), C[:,3])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 325,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x123e21eb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from scipy.stats import gaussian_kde\n",
    "density = gaussian_kde(np.real(A[0]))\n",
    "x = np.arange(0., 8, .1)\n",
    "plt.plot(x, density(x))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 280,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x123fc5390>"
      ]
     },
     "execution_count": 280,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAD8CAYAAACGsIhGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1243ee0f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "df = pd.DataFrame(np.real(A[0]))\n",
    "df.plot(kind='density')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'mean' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-172-f31bb038b40e>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mmean\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mreal\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mA\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m: name 'mean' is not defined"
     ]
    }
   ],
   "source": [
    "mean(np.real(A))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "import csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('/Users/jul/git/pipe-generate-dataset/dataset.csv', 'r') as csvfile:\n",
    "    reader = csv.reader(csvfile, delimiter=',')\n",
    "    dataset = np.array(list(reader)).astype(\"float\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1000, 768)"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('/Users/jul/git/pipe-generate-dataset/classMatrix.csv', 'r') as csvfile:\n",
    "    reader = csv.reader(csvfile, delimiter=',')\n",
    "    dataClass = np.array(list(reader)).astype(\"float\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 1.,  0.],\n",
       "       [ 1.,  0.],\n",
       "       [ 1.,  0.],\n",
       "       ..., \n",
       "       [ 0.,  1.],\n",
       "       [ 0.,  1.],\n",
       "       [ 0.,  1.]])"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataClass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "from matplotlib.mlab import PCA\n",
    "model = PCA(dataset, standardize=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([  1.13209968e+09,   1.39246954e+09,   1.33643648e+09,\n",
       "         1.45046945e+09,   1.61032191e+09,   1.27310507e+09,\n",
       "         1.81382487e+09,   1.21403919e+09,   1.63738196e+09,\n",
       "         1.67249709e+09,   1.84033102e+09,   1.53927386e+09,\n",
       "         2.05963286e+09,   1.29978829e+09,   1.16866316e+09,\n",
       "         1.61257163e+09,   1.02352342e+09,   1.02295858e+09,\n",
       "         1.45688241e+09,   1.63637982e+09,   1.26973874e+09,\n",
       "         1.15310514e+09,   1.63498985e+09,   1.34311338e+09,\n",
       "         1.39670652e+09,   9.82014429e+08,   6.32647906e+08,\n",
       "         1.37874014e+09,   1.20267404e+09,   1.46346257e+09,\n",
       "         9.75242272e+08,   9.84675272e+08,   2.09757293e+09,\n",
       "         1.30760963e+09,   9.73991287e+08,   1.43533068e+09,\n",
       "         1.19612206e+09,   1.71973123e+09,   1.01400450e+09,\n",
       "         1.37629225e+09,   1.70546931e+09,   8.23144220e+08,\n",
       "         1.89368197e+09,   1.82890075e+09,   9.99906753e+08,\n",
       "         1.22853656e+09,   1.17308223e+09,   1.25100311e+09,\n",
       "         1.33984538e+09,   1.00280925e+09,   1.88341651e+09,\n",
       "         1.45876370e+09,   1.50331363e+09,   1.58479288e+09,\n",
       "         1.98497341e+09,   1.11122085e+09,   1.26426268e+09,\n",
       "         1.11699715e+09,   1.02217917e+09,   1.40297270e+09,\n",
       "         1.77799144e+09,   1.77938504e+09,   1.31059447e+09,\n",
       "         1.17682795e+09,   1.59280858e+09,   1.19418377e+09,\n",
       "         1.34404853e+09,   2.13103050e+09,   2.17816274e+09,\n",
       "         1.55256761e+09,   1.33106046e+09,   1.62281100e+09,\n",
       "         1.72892284e+09,   1.91434998e+09,   1.73565580e+09,\n",
       "         1.03276466e+09,   1.55452353e+09,   1.32289163e+09,\n",
       "         1.35060969e+09,   1.49522536e+09,   1.91967435e+09,\n",
       "         1.30304602e+09,   1.27747967e+09,   2.25998970e+09,\n",
       "         1.45198177e+09,   1.33076704e+09,   1.62114294e+09,\n",
       "         1.14085822e+09,   1.24233841e+09,   1.63874276e+09,\n",
       "         1.66234434e+09,   1.23122576e+09,   1.62476861e+09,\n",
       "         1.63900856e+09,   1.23113677e+09,   1.57441533e+09,\n",
       "         1.43496665e+09,   1.76888319e+09,   1.56147840e+09,\n",
       "         1.21336210e+09,   1.35763988e+09,   1.05538907e+09,\n",
       "         1.60168344e+09,   8.84453751e+08,   1.68521119e+09,\n",
       "         1.17359274e+09,   1.62030883e+09,   1.68040874e+09,\n",
       "         1.40270796e+09,   2.01672186e+09,   1.61039081e+09,\n",
       "         1.78302656e+09,   8.99268230e+08,   1.44899356e+09,\n",
       "         1.36391255e+09,   1.14973451e+09,   9.36325409e+08,\n",
       "         2.07151963e+09,   1.56580114e+09,   1.28758447e+09,\n",
       "         1.19624543e+09,   1.92450754e+09,   1.97280008e+09,\n",
       "         1.73416768e+09,   1.35873908e+09,   1.17540526e+09,\n",
       "         8.70352272e+08,   1.10319745e+09,   1.13366349e+09,\n",
       "         1.58328672e+09,   1.75431637e+09,   1.45437489e+09,\n",
       "         1.84098645e+09,   8.22931165e+08,   7.73136464e+08,\n",
       "         1.61604986e+09,   1.99551502e+09,   1.37090855e+09,\n",
       "         8.07342566e+08,   8.85316142e+08,   1.06768958e+09,\n",
       "         1.02589752e+09,   1.67546607e+09,   1.54623181e+09,\n",
       "         1.66530401e+09,   2.04271963e+09,   1.81140407e+09,\n",
       "         9.97341420e+08,   1.59306693e+09,   1.15897449e+09,\n",
       "         1.48521930e+09,   1.58135691e+09,   1.78349352e+09,\n",
       "         1.56689587e+09,   1.36312500e+09,   1.64719882e+09,\n",
       "         1.38585979e+09,   8.67998442e+08,   1.05300715e+09,\n",
       "         1.56386087e+09,   1.81525311e+09,   8.69217501e+08,\n",
       "         1.89427003e+09,   2.00795288e+09,   1.00225215e+09,\n",
       "         1.20415151e+09,   1.65374585e+09,   1.07146246e+09,\n",
       "         1.53919957e+09,   1.45826865e+09,   1.08548254e+09,\n",
       "         1.39453665e+09,   1.90456476e+09,   1.66539886e+09,\n",
       "         1.58282169e+09,   1.16713199e+09,   1.74205274e+09,\n",
       "         1.37952242e+09,   4.94630704e+08,   1.64059202e+09,\n",
       "         1.27640803e+09,   1.67971500e+09,   1.13713965e+09,\n",
       "         1.78222848e+09,   1.62959325e+09,   2.15414935e+09,\n",
       "         1.23882762e+09,   1.18065494e+09,   2.02654215e+09,\n",
       "         1.22546970e+09,   1.38074410e+09,   1.35677577e+09,\n",
       "         1.56678786e+09,   1.54035773e+09,   1.02673343e+09,\n",
       "         1.36225632e+09,   1.58350607e+09,   1.72686830e+09,\n",
       "         1.75514108e+09,   2.04433945e+09,   8.41663142e+08,\n",
       "         1.79266704e+09,   1.63144391e+09,   1.04761149e+09,\n",
       "         9.06325873e+08,   1.61408363e+09,   1.35337508e+09,\n",
       "         1.53589100e+09,   1.38990068e+09,   9.21266037e+08,\n",
       "         1.40691395e+09,   1.53032158e+09,   1.30073743e+09,\n",
       "         8.48338233e+08,   1.17503523e+09,   1.87387640e+09,\n",
       "         1.10975184e+09,   1.45738152e+09,   1.68481684e+09,\n",
       "         1.11273877e+09,   1.23226249e+09,   1.43693579e+09,\n",
       "         9.55538016e+08,   1.94286916e+09,   1.14051525e+09,\n",
       "         1.59852268e+09,   1.63056920e+09,   1.47940118e+09,\n",
       "         1.87941001e+09,   1.82316038e+09,   1.13076228e+09,\n",
       "         1.76198880e+09,   1.94130433e+09,   1.30308118e+09,\n",
       "         1.57493147e+09,   1.97065581e+09,   1.51300656e+09,\n",
       "         1.73005115e+09,   1.00016132e+09,   8.76527889e+08,\n",
       "         1.47481544e+09,   1.96814518e+09,   1.42305529e+09,\n",
       "         1.17605551e+09,   1.58706728e+09,   2.09154079e+09,\n",
       "         9.44124306e+08,   1.96373140e+09,   1.71432310e+09,\n",
       "         1.45415806e+09,   1.99972785e+09,   1.67848303e+09,\n",
       "         9.11836780e+08,   1.56676707e+09,   1.58892597e+09,\n",
       "         1.73041995e+09,   1.88300700e+09,   1.40585342e+09,\n",
       "         1.87540976e+09,   1.24546753e+09,   1.40529601e+09,\n",
       "         1.11791069e+09,   1.25993810e+09,   1.12000728e+09,\n",
       "         1.04460491e+09,   1.64621678e+09,   1.28061022e+09,\n",
       "         1.23961002e+09,   1.60711551e+09,   1.36106928e+09,\n",
       "         1.24468219e+09,   1.23369453e+09,   1.18334108e+09,\n",
       "         8.88234059e+08,   1.55737761e+09,   1.84290641e+09,\n",
       "         1.35943655e+09,   1.69549355e+09,   1.76541443e+09,\n",
       "         7.48935715e+08,   1.93203545e+09,   9.97569490e+08,\n",
       "         1.43849110e+09,   1.12544746e+09,   1.00973109e+09,\n",
       "         1.59333329e+09,   1.80737129e+09,   2.09097459e+09,\n",
       "         7.97463135e+08,   1.15831760e+09,   1.14371857e+09,\n",
       "         1.72833028e+09,   1.34433858e+09,   1.21320744e+09,\n",
       "         1.59185388e+09,   1.53868854e+09,   1.89948838e+09,\n",
       "         1.27082285e+09,   1.30007750e+09,   1.21828006e+09,\n",
       "         1.69076861e+09,   1.22101130e+09,   1.50025354e+09,\n",
       "         1.00657819e+09,   1.25550959e+09,   1.26720153e+09,\n",
       "         1.08878571e+09,   1.58469053e+09,   1.33496509e+09,\n",
       "         1.18605998e+09,   1.05375402e+09,   1.66837491e+09,\n",
       "         5.64289062e+08,   1.46602605e+09,   1.40855814e+09,\n",
       "         2.01457830e+09,   1.56125624e+09,   1.31904838e+09,\n",
       "         1.50991268e+09,   1.41871197e+09,   1.63592782e+09,\n",
       "         1.38722447e+09,   1.07777693e+09,   1.85486763e+09,\n",
       "         1.18806965e+09,   1.94526326e+09,   1.47321975e+09,\n",
       "         7.50289217e+08,   1.36370745e+09,   1.10098103e+09,\n",
       "         1.33636606e+09,   1.77219114e+09,   1.05154344e+09,\n",
       "         1.18443387e+09,   1.02708113e+09,   1.31613236e+09,\n",
       "         2.03306298e+09,   8.76035292e+08,   1.47534575e+09,\n",
       "         5.35230311e+08,   7.41085310e+08,   1.43743627e+09,\n",
       "         6.82678108e+08,   1.55947742e+09,   1.41235626e+09,\n",
       "         1.74979574e+09,   1.06197994e+09,   1.44247860e+09,\n",
       "         1.94635121e+09,   1.18335039e+09,   1.99984822e+09,\n",
       "         1.36743062e+09,   1.59935501e+09,   1.23452261e+09,\n",
       "         1.37535087e+09,   1.84495538e+09,   1.15517095e+09,\n",
       "         1.48638332e+09,   9.87497365e+08,   1.44905482e+09,\n",
       "         1.07076437e+09,   1.08311674e+09,   1.13260537e+09,\n",
       "         1.37618937e+09,   1.29977549e+09,   1.26318323e+09,\n",
       "         1.80942899e+09,   1.48652889e+09,   1.08144831e+09,\n",
       "         2.16181523e+09,   1.77512132e+09,   1.05160098e+09,\n",
       "         1.14207305e+09,   1.86761310e+09,   9.91736111e+08,\n",
       "         1.43464045e+09,   1.38386147e+09,   1.51276331e+09,\n",
       "         1.45760906e+09,   1.63269764e+09,   1.03717972e+09,\n",
       "         1.63806023e+09,   9.43992124e+08,   1.13479076e+09,\n",
       "         1.73317450e+09,   8.51992315e+08,   1.33901569e+09,\n",
       "         1.83983544e+09,   9.71867799e+08,   1.53890743e+09,\n",
       "         9.03109237e+08,   1.57661986e+09,   1.04087096e+09,\n",
       "         1.22342393e+09,   1.51898113e+09,   1.54967693e+09,\n",
       "         1.20362563e+09,   1.33412587e+09,   1.56931332e+09,\n",
       "         1.60020890e+09,   1.31596686e+09,   1.33895801e+09,\n",
       "         2.00990037e+09,   2.13910140e+09,   1.10786212e+09,\n",
       "         1.30556548e+09,   9.95206252e+08,   1.26372851e+09,\n",
       "         1.34640196e+09,   1.83055613e+09,   1.15071722e+09,\n",
       "         1.51470191e+09,   1.28594125e+09,   1.86013571e+09,\n",
       "         1.72826221e+09,   1.69647478e+09,   1.34691301e+09,\n",
       "         1.77491682e+09,   1.18464144e+09,   1.54747897e+09,\n",
       "         1.80677826e+09,   1.24543597e+09,   7.87162845e+08,\n",
       "         1.48756414e+09,   1.28567051e+09,   1.64052506e+09,\n",
       "         1.42509760e+09,   1.37327373e+09,   1.11971402e+09,\n",
       "         1.21035294e+09,   1.39702326e+09,   1.61804459e+09,\n",
       "         1.72250186e+09,   1.11941047e+09,   1.78467947e+09,\n",
       "         1.71200286e+09,   1.14302742e+09,   1.23278045e+09,\n",
       "         1.29569402e+09,   1.28996875e+09,   1.10053997e+09,\n",
       "         1.07861328e+09,   1.55059263e+09,   1.68802361e+09,\n",
       "         1.39926241e+09,   1.42218289e+09,   1.64955723e+09,\n",
       "         1.45943877e+09,   1.49117450e+09,   1.62070088e+09,\n",
       "         1.06215868e+09,   1.26483010e+09,   1.39661172e+09,\n",
       "         2.14453968e+09,   1.55026714e+09,   1.01326750e+09,\n",
       "         1.66720924e+09,   1.90661259e+09,   1.58121028e+09,\n",
       "         1.05917386e+09,   1.58645181e+09,   1.44210476e+09,\n",
       "         1.62711674e+09,   1.43640463e+09,   1.03727426e+09,\n",
       "         1.20096472e+09,   9.40617128e+08,   1.29612118e+09,\n",
       "         1.16324330e+09,   1.97083694e+09,   2.23648391e+09,\n",
       "         1.40080790e+09,   1.06222674e+09,   1.58602599e+09,\n",
       "         1.45409425e+09,   1.47422639e+09,   1.53995451e+09,\n",
       "         1.68005928e+09,   1.44933185e+09,   1.80091600e+09,\n",
       "         1.98197511e+09,   1.42185862e+09,   1.49737098e+09,\n",
       "         1.91620307e+09,   1.72844362e+09,   1.57232355e+09,\n",
       "         1.54319673e+09,   1.61131261e+09,   1.24549300e+09,\n",
       "         1.70568688e+09,   1.50139263e+09,   1.30965520e+09,\n",
       "         1.15581088e+09,   1.51637496e+09,   1.35859969e+09,\n",
       "         1.72381067e+09,   1.03070048e+09,   1.11198213e+09,\n",
       "         1.43437196e+09,   1.57194240e+09,  -1.89485701e+09,\n",
       "        -1.11327531e+09,  -9.06758100e+08,  -5.06301275e+08,\n",
       "        -1.25244503e+09,  -1.98433560e+09,  -1.26606297e+09,\n",
       "        -1.49207430e+09,  -1.00710629e+09,  -7.70171579e+08,\n",
       "        -1.03807210e+09,  -1.05649812e+09,  -1.60280148e+09,\n",
       "        -1.30516719e+09,  -1.93590029e+09,  -1.11657604e+09,\n",
       "        -1.85430135e+09,  -1.22931195e+09,  -1.32972782e+09,\n",
       "        -1.29465502e+09,  -4.31436603e+08,  -1.40350262e+09,\n",
       "        -1.27415161e+09,  -2.06950387e+09,  -2.03355533e+09,\n",
       "        -1.57188494e+09,  -1.23498605e+09,  -2.27339543e+09,\n",
       "        -1.55874542e+09,  -1.14541462e+09,  -8.95677914e+08,\n",
       "        -1.32830829e+09,  -1.47476938e+09,  -2.04702161e+09,\n",
       "        -2.03867754e+09,  -2.37603775e+09,  -1.65809689e+09,\n",
       "        -1.27475036e+09,  -1.29010049e+09,  -1.85990944e+09,\n",
       "        -1.20021931e+09,  -1.56836056e+09,  -1.20370421e+09,\n",
       "        -1.36374635e+09,  -1.22433418e+09,  -1.42506191e+09,\n",
       "        -9.56303062e+08,  -1.25852180e+09,  -1.44347172e+09,\n",
       "        -1.48659365e+09,  -1.61826243e+09,  -7.34210167e+08,\n",
       "        -2.08825804e+09,  -8.30539433e+08,  -1.43870641e+09,\n",
       "        -9.40966253e+08,  -1.62797364e+09,  -1.43594026e+09,\n",
       "        -1.29704412e+09,   4.96051463e+07,  -1.10716562e+09,\n",
       "        -1.68141574e+09,  -7.06016173e+08,  -2.16621896e+09,\n",
       "        -1.87563724e+09,  -1.87723748e+08,  -1.54762073e+09,\n",
       "        -1.79972552e+09,  -1.03999425e+09,  -8.71794208e+08,\n",
       "        -1.18820235e+09,  -2.05825774e+09,  -1.61137507e+09,\n",
       "        -1.68921799e+09,  -1.09647685e+09,  -1.36975565e+09,\n",
       "        -1.43676097e+09,  -1.65133133e+09,  -2.16740733e+09,\n",
       "        -1.70351610e+09,  -1.10090510e+09,  -1.14008119e+09,\n",
       "        -1.65204773e+09,  -1.54109665e+09,  -1.90240301e+09,\n",
       "        -1.49244381e+09,  -2.27981378e+09,  -1.81461130e+09,\n",
       "        -2.16881519e+09,  -8.52715780e+08,  -1.38105016e+09,\n",
       "        -1.36006047e+09,  -1.59856853e+09,  -1.48847491e+09,\n",
       "        -1.33979536e+09,  -2.09185681e+09,  -2.19516582e+09,\n",
       "        -9.03407621e+08,  -1.52311415e+09,  -1.86082554e+09,\n",
       "        -1.49630543e+09,  -1.59672947e+09,  -1.78463975e+09,\n",
       "        -2.27424814e+09,  -1.90857204e+09,  -8.54415987e+08,\n",
       "        -1.13109263e+09,  -1.64656080e+09,  -1.46715069e+09,\n",
       "        -2.46089441e+09,  -1.05481628e+09,  -1.76680371e+09,\n",
       "        -1.50045255e+09,  -1.17567042e+09,  -1.99042024e+09,\n",
       "        -1.58209539e+09,  -1.91319589e+09,  -1.41003861e+09,\n",
       "        -1.72694349e+09,  -9.02638872e+08,  -1.51189910e+09,\n",
       "        -1.62850189e+09,  -1.39712561e+09,  -1.71452754e+09,\n",
       "        -7.51624185e+08,  -1.63627015e+09,  -7.88857053e+08,\n",
       "        -2.00074327e+09,  -6.81111486e+08,  -8.88891249e+08,\n",
       "        -1.10160079e+09,  -1.61299767e+09,  -1.61397110e+09,\n",
       "        -1.07528149e+09,  -1.14596717e+09,  -2.79850077e+09,\n",
       "        -1.97846047e+09,  -1.22891397e+09,  -1.19735602e+09,\n",
       "        -1.22231418e+09,  -8.01710641e+08,  -6.44998200e+08,\n",
       "        -8.67206841e+08,  -5.05267268e+08,  -9.15699984e+08,\n",
       "        -2.10869347e+09,  -1.09398920e+09,  -1.22124142e+09,\n",
       "        -1.40206193e+09,  -1.52316960e+09,  -7.53272680e+08,\n",
       "        -2.14440206e+09,  -2.00170873e+09,  -6.70130968e+08,\n",
       "        -1.30356384e+09,  -1.80811625e+09,  -1.40733976e+09,\n",
       "        -1.69965510e+09,  -2.21713514e+09,  -4.27428481e+08,\n",
       "        -8.79444519e+08,  -2.32495713e+09,  -1.45212247e+09,\n",
       "        -1.38004875e+09,  -7.77054228e+08,  -1.05065352e+09,\n",
       "        -1.95992742e+09,  -9.25548436e+08,  -2.04340435e+09,\n",
       "        -1.19601537e+09,  -1.71735384e+09,  -1.40824726e+09,\n",
       "        -8.00599640e+08,  -1.47662492e+09,  -1.06722367e+09,\n",
       "        -1.06608254e+09,  -8.79689340e+08,  -1.61350818e+09,\n",
       "        -2.39492843e+09,  -1.59254099e+09,  -1.37165502e+09,\n",
       "        -2.21838964e+09,  -7.34027245e+08,  -1.64943496e+09,\n",
       "        -1.27789089e+09,  -1.14941799e+09,  -1.86012446e+09,\n",
       "        -5.95957166e+08,  -1.23507719e+09,  -1.08897465e+09,\n",
       "        -2.87452316e+09,  -7.53972045e+08,  -1.18518297e+09,\n",
       "        -2.15785328e+09,  -9.36200785e+08,  -1.28474014e+09,\n",
       "        -9.04866703e+08,  -2.29671804e+09,  -1.96634892e+09,\n",
       "        -2.23841433e+09,  -1.34106545e+09,  -1.83788668e+09,\n",
       "        -1.00246316e+09,  -1.55852239e+09,  -1.14370491e+09,\n",
       "        -2.23107789e+09,  -1.02839750e+09,  -1.33075452e+09,\n",
       "        -4.66480472e+08,  -1.90574655e+09,  -1.28575432e+09,\n",
       "        -1.86127826e+09,  -1.38329700e+09,  -9.98205749e+08,\n",
       "        -1.42327574e+09,  -1.78605635e+09,  -1.97609144e+09,\n",
       "        -8.97339914e+08,  -1.40562003e+09,  -1.12791091e+09,\n",
       "        -1.33225697e+09,  -1.83954894e+09,  -1.46043813e+09,\n",
       "        -2.07353512e+09,  -1.28254998e+09,  -7.21566833e+08,\n",
       "        -1.46988435e+09,  -1.94868400e+09,  -1.14986335e+09,\n",
       "        -1.09696617e+09,  -7.58159986e+08,  -1.40628525e+09,\n",
       "        -1.33465444e+09,  -1.42192630e+09,  -1.79344396e+09,\n",
       "        -1.64152918e+09,  -1.55146688e+09,  -1.60165648e+09,\n",
       "        -1.36743218e+09,  -1.33900831e+09,  -1.00517981e+09,\n",
       "        -8.28271019e+08,  -1.29397055e+09,  -9.65617486e+08,\n",
       "        -1.08523605e+09,  -7.39388494e+08,  -1.73259794e+09,\n",
       "        -2.37280937e+09,  -5.45903911e+08,  -1.87353713e+09,\n",
       "        -2.14354601e+09,  -1.87257995e+09,  -1.73880041e+09,\n",
       "        -1.22876813e+09,  -1.98268252e+09,  -1.70942907e+09,\n",
       "        -1.49630826e+09,  -1.78839092e+09,  -1.48473605e+09,\n",
       "        -1.44836892e+09,  -2.20393121e+09,  -5.28432513e+08,\n",
       "        -1.06821722e+09,  -1.77469842e+09,  -1.78736279e+09,\n",
       "        -1.88854320e+09,  -9.90215653e+08,  -1.68322086e+09,\n",
       "        -6.91703133e+08,  -6.42446939e+08,  -9.18404639e+08,\n",
       "        -8.55293685e+08,  -1.08700000e+09,  -1.07451122e+09,\n",
       "        -1.88428590e+09,  -1.58639881e+09,  -1.31350105e+09,\n",
       "        -1.32163364e+09,  -1.32966637e+09,  -1.25881181e+09,\n",
       "        -2.62435486e+09,  -1.15234346e+09,  -1.03550693e+09,\n",
       "        -1.59532960e+09,  -1.05181167e+09,  -1.55830397e+09,\n",
       "        -1.23864891e+09,  -1.55797803e+09,  -1.90462298e+09,\n",
       "        -1.68006211e+09,  -1.14284775e+09,  -6.65324643e+08,\n",
       "        -1.43791249e+09,  -1.65920947e+09,  -7.07640165e+08,\n",
       "        -1.08568065e+09,  -1.46726455e+09,  -1.45330177e+09,\n",
       "        -2.09801647e+08,  -1.29738981e+09,  -2.07957978e+09,\n",
       "        -1.57202200e+09,  -1.50584098e+09,  -1.69523826e+09,\n",
       "        -7.65928883e+08,  -1.22726909e+09,  -1.50247014e+09,\n",
       "        -1.16851250e+09,  -1.44760466e+09,  -1.49071483e+09,\n",
       "        -1.25499684e+09,  -1.41433561e+09,  -7.00909829e+08,\n",
       "        -1.02774417e+09,  -3.25013373e+08,  -1.21214709e+09,\n",
       "        -1.76341526e+09,  -1.03672386e+09,  -1.38163840e+09,\n",
       "        -1.11217595e+09,  -1.52861525e+09,  -5.70733092e+08,\n",
       "        -9.40421267e+08,  -1.63176465e+09,  -1.39018327e+09,\n",
       "        -1.31748539e+09,  -1.15596226e+09,  -2.22393918e+09,\n",
       "        -1.61090569e+09,  -1.29039845e+09,  -1.43982143e+09,\n",
       "        -1.23747300e+09,  -1.11150379e+09,  -2.30443898e+09,\n",
       "        -1.65055139e+09,  -1.62531672e+09,  -9.28861558e+08,\n",
       "        -1.88117695e+09,  -2.36348021e+09,  -1.73495990e+09,\n",
       "        -1.84647401e+09,  -1.74772615e+09,  -1.85771440e+09,\n",
       "        -9.86635644e+08,  -8.83077208e+08,  -2.13343564e+09,\n",
       "        -1.45688779e+09,  -2.16999992e+09,  -1.02203078e+09,\n",
       "        -1.81146886e+09,  -9.41042400e+08,  -1.71953149e+09,\n",
       "        -1.62074273e+09,  -2.28231237e+09,  -2.16932499e+09,\n",
       "        -1.42835478e+09,  -1.93959980e+09,  -2.07333713e+09,\n",
       "        -1.85128447e+09,  -8.03355969e+08,  -9.27193233e+08,\n",
       "        -1.21674610e+09,  -1.03843169e+09,  -1.43936129e+09,\n",
       "        -1.23074905e+09,  -1.44664126e+09,  -1.41561874e+09,\n",
       "        -1.33819718e+09,  -1.13402632e+09,  -1.06479236e+09,\n",
       "        -1.22591503e+09,  -1.88655557e+09,  -1.85963642e+09,\n",
       "        -7.97544538e+08,  -1.21422439e+09,  -1.74785433e+09,\n",
       "        -1.91117716e+09,  -9.99450262e+08,  -1.57956705e+09,\n",
       "        -1.86784003e+09,  -1.21964481e+09,  -1.32039415e+09,\n",
       "        -1.29274498e+09,  -1.52662091e+09,  -2.01030977e+09,\n",
       "        -5.49253726e+08,  -1.16122990e+09,  -9.81414760e+08,\n",
       "        -8.87630673e+08,  -1.23038422e+09,  -1.24895963e+09,\n",
       "        -2.18391416e+09,  -1.75066509e+09,  -1.50594869e+09,\n",
       "        -1.83308036e+09,  -1.31370321e+09,  -1.81206885e+09,\n",
       "        -1.42770419e+09,  -1.72912472e+09,  -1.10670660e+09,\n",
       "        -1.08641691e+09,  -1.66850816e+09,  -1.81552646e+09,\n",
       "        -1.13513424e+09,  -5.60542563e+08,  -5.43998985e+08,\n",
       "        -1.38863058e+09,  -1.36932519e+09,  -1.64910061e+09,\n",
       "        -2.27767186e+09,  -6.38076771e+08,  -1.47203888e+09,\n",
       "        -7.45273212e+08,  -1.20429417e+09,  -1.71927855e+09,\n",
       "        -2.24306897e+09,  -1.41486011e+09,  -1.36574446e+09,\n",
       "        -1.95689693e+09,  -2.35526966e+09,  -2.08895491e+09,\n",
       "        -1.90827929e+09,  -1.06211744e+09,  -4.29054945e+08,\n",
       "        -7.35219372e+08,  -2.19726959e+09,  -5.22052077e+08,\n",
       "        -1.68481670e+09,  -2.08541776e+09,  -1.83066836e+09,\n",
       "        -1.53386054e+09,  -1.11406024e+09,  -6.83268966e+08,\n",
       "        -1.36162818e+09,  -1.50855011e+09,  -1.52670100e+09,\n",
       "        -2.51817095e+09,  -1.05888183e+09,  -9.30050717e+08,\n",
       "        -1.55970649e+09,  -1.85663534e+09,  -1.53680594e+09,\n",
       "        -1.50671990e+09,  -1.63593105e+09,  -1.24915540e+09,\n",
       "        -1.78079304e+09,  -1.68292801e+09,  -1.33405370e+09,\n",
       "        -1.70551141e+09,  -8.53198035e+08,  -9.70803795e+08,\n",
       "        -1.80468588e+09,  -7.47349693e+08,  -1.07502495e+09,\n",
       "        -1.65311474e+09,  -2.02417995e+09,  -7.47152972e+08,\n",
       "        -1.18479565e+09,  -1.76963881e+09,  -9.14535335e+08,\n",
       "        -1.44772568e+09,  -1.56081682e+09,  -1.36503438e+09,\n",
       "        -1.66113169e+09,  -1.90830045e+09,  -1.97164048e+09,\n",
       "        -2.13905340e+09,  -1.81643053e+09,  -1.33966197e+09,\n",
       "        -1.99377172e+09,  -1.46080823e+09,  -2.21257683e+09,\n",
       "        -4.68982342e+08,  -1.40019080e+09,  -9.26865257e+08,\n",
       "        -1.45130617e+09,  -8.43959867e+08,  -8.87996722e+08,\n",
       "        -1.13858298e+09,  -1.77424822e+09,  -2.60584077e+09,\n",
       "        -2.55493788e+09,  -1.52997958e+09,  -1.22731864e+09,\n",
       "        -2.01090134e+09,  -1.93407933e+09,  -1.50857751e+09,\n",
       "        -1.94152335e+09,  -1.06478111e+09,  -1.35836802e+09,\n",
       "        -8.59855439e+08,  -7.58141748e+08,  -7.60587565e+08,\n",
       "        -1.54659961e+09,  -1.10676936e+09,  -1.55131027e+09,\n",
       "        -1.01412277e+09,  -8.29397325e+08,  -1.64519070e+09,\n",
       "        -1.71262617e+09])"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.Y[:,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x111894550>"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAEQCAYAAABY5nToAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11229eba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(model.Y[:,0], model.Y[:,1])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x1141956a0>"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYMAAAEDCAYAAADX1GjKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJAn4/3erQvVT7XBpAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1140b2be0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(np.arange(dataset.shape[1]), model.a[0,:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x11054f898>"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAEDCAYAAADUT6SnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZoMclgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10da0eac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(np.arange(dataset.shape[1]), np.std(model.a[:,:], 0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x1123f6b70>"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYMAAAEDCAYAAADX1GjKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAFdtJREFUeJzt3X+QXWd93/H3B/0wwgbLxh4hyzaC4CgxBSzYGjsOJAOi\nMk4TC4KDPaQxHRh3StoJbVFHHjP5VRiUKCEkE5JUNbTmRw0BbFnFToUtu+M0ExxWkY1sK8IGHCNZ\ntjYQQQnKRJa//eOeNavVXUm79+zeu7vv18ydPee5zz3PV7p372fPOc89N1WFJGl+e06/C5Ak9Z9h\nIEkyDCRJhoEkCcNAkoRhIEliFoRBko8nOZDkwZPo++Ik25N8Ncn/SXLuTNQoSbPdwIcB8D+Ay0+y\n7+8An6iqVwK/CXxouoqSpLlk4MOgqu4FvjO2LcmPJPnfSXYk+fMkP9bcdSFwd7N8D3DlDJYqSbPW\nwIfBBDYD/76qXgO8D/ijpv0B4K3N8luA5yd5YR/qk6RZZWG/C5isJKcBPwF8Lslo8ynNz/cBf5jk\nncC9wD7gyEzXKEmzzawLAzp7Mwer6qLxd1TVEzR7Bk1o/HxVHZzh+iRp1pl1h4mq6nvAN5NcBZCO\nVzXLZyUZ/TddD3y8T2VK0qwy8GGQ5GbgL4FVSfYmeRfwDuBdSR4AHuKHJ4p/GtiT5GvAMuCDfShZ\nkmadeAlrSdLA7xlIkqbfQJ9APuuss2rlypX9LkOSZo0dO3b8XVWdPdnHDXQYrFy5kuHh4X6XIUmz\nRpK/ncrjPEwkSTIMJEmGgSQJw0CShGEgSaKl2URJLgd+H1gA3FhVG8fdfwrwCeA1wLeBt1fVY22M\nPd77t+ziU19+fDo2LUkz6hcvOZ8PrHvFjIzV855BkgXAR4E30/k+gWuSXDiu27uAv6+qlwG/B/xW\nr+N2YxBImks+9eXHef+WXTMyVhuHiS4GHq2qb1TVPwGf4dgvlbkSuKlZ/jzwxoy5/nRbbr7vW21v\nUpL6aqbe19oIgxXA2Gr3Nm1d+1TV08B3ga5fOpPkuiTDSYZHRkYmVcgRr7MkaY6Zqfe1gTuBXFWb\nq2qoqobOPntyn6he0P7OhiT11Uy9r7URBvuA88asn9u0de2TZCFwOp0Tya265rXnnbiTJM0iM/W+\n1kYYfAW4IMlLkiwGrga2juuzFbi2WX4bcHdNw7WzP7DuFfziJee3vVlJ6ouZnE3UyvcZJLkC+Aid\nqaUfr6oPJvlNYLiqtiZ5LvBJYDXwHeDqqvrGibY7NDRUXqhO6p/3b9nFzfd9iyNVLEi45rXntfbm\ntGXnPtZ/7gEOP/PD96BFzwmbrnoV61Yffdrxso13s+/goQm3tWLpEv5iwxsAWLnh9lbqG7vN2STJ\njqoamuzjWjlnUFV3VNWPVtWPVNUHm7ZfraqtzfI/VtVVVfWyqrr4ZIJAUn+NTtUePYF5pKrVqY6/\nvvWho4IA4PAzxa9vfeiYvuvXrprwzWrBc8L6tauATsC0YcmiBc9uc74YuBPIkgbDRFMa25rqePDQ\n4ZNuX7d6BR9++0UsWXT0W9apixfwu82exJad+7j+lt6DakHCz79mxTF7J3PdQH+fgaT+mWhKY7+m\ncK9bffw36E3b9nDo8JGexzlSxRd27GPoxWfOq0Bwz0BSVxNNaWxrquMZz1s0qfYTeeI45xQm69Dh\nI2zatqe17c0GhoGkriaa0tjWVMdf+9mXs2jB0cGyaEH4tZ99+ZS2d87SJV3bxx9aGnWiSGszXGYD\nw0BSV6NTtUf3BBYkrU51XLd6BZve9ipWLF1C6Mze2fS2Y2cSnaz1a1exZNGCo9qWLFrAh976yq7/\njm9u/Bke2/gzrJggRCYKl7mqlaml08WppZImY8vOfWzatocnDh7inKVLWL921QnDZfTE89jzDZ0Q\necWsPGcw1amlnkCWNGec6CTzRI8BJh0ic41hIGlOmcrewVRCZK4xDCTNGeMP+ew7eOjZzx7M9zf7\nE/EEsqQ5o9tnDebjNNGpMAwkzRkTTQedb9NEp8IwkDRnTDQddL5NE50Kw0DSnLF+7aquH2Sbbxed\nmwrDQNLcMv6jU4P7UaqBYhhImjM2bdvT9bLYnkA+McNA0pzhCeSpMwwkzRmeQJ46w0DSnDHRxeo8\ngXxifgJZ0pzhdYamzjCQNKd4naGp8TCRJMkwkCT1eJgoyZnAZ4GVwGPAL1TV34/rcxHwx8ALgCPA\nB6vqs72MK2numcqlp3t5nI7W657BBmB7VV0AbG/Wx/sB8EtV9XLgcuAjSZb2OK6kOWT00tP7Dh6i\n+OGlp7fs3Dctj9Oxeg2DK4GbmuWbgHXjO1TV16rqkWb5CeAAcHaP40qaQ6Z66em2L1m9Zec+Ltt4\nNy/ZcDuXbbx7XoVKr7OJllXV/mb5SWDZ8TonuRhYDHz9OH2uA64DOP/883ssT9JsMNVPDrf5ieP3\nb9nFp7/8+LOXMppvX4xzwj2DJHclebDL7cqx/aqqOM4loZIsBz4J/OuqemaiflW1uaqGqmro7LPd\ngZDmg4k+IXz6kkXHfdzS53W/f7KfON6yc99RQTBqPn0xzgn3DKpqzUT3JXkqyfKq2t+82R+YoN8L\ngNuBG6rqy1OuVnOCJ/w03vq1q1j/uQeOucjcP/zT02zZua/r62PLzn18/x+fPqZ9Kpes3rRtz4R/\nyc6X6xr1es5gK3Bts3wtcNv4DkkWA7cCn6iqz/c4nmY5T/ipm3WrV3Dac4/92/TwkYmvONrtCqUA\npy5eOOk/Lo73hj9frmvUaxhsBN6U5BFgTbNOkqEkNzZ9fgF4PfDOJPc3t4t6HFezlN9Rq4kc/MHh\nru2TPS/w3UPdt3M8E73hB+bNdY16CoOq+nZVvbGqLqiqNVX1naZ9uKre3Sx/qqoWVdVFY273t1G8\nZh8vMayJTPaKo21eobTbBe4CvOOS8+fNIUw/gawZ5SWGNZHJXnG0zSuUrlu9gg+99RWsWLqEACuW\nLuH33n4RH1j3iklva7byQnWaUevXruL6W3YddajISwwLJn/F0bavUDrfL3CXzozQwTQ0NFTDw8P9\nLkMtczaRNH2S7KiqoUk/zjBQvxgKUvumGgYeJlJfjE4xHT1cNN8+7SkNGk8gqy+cYioNFsNAfeEU\nU2mwGAbqC6eYSoPFMFBftDlHXFLvPIGsvmh7jrik3hgG6pv5/iEfaZB4mEiSZBhIkgwDSRKGgSQJ\nw0CShGEgScIwkCRhGEiSMAwkSRgGkiQMA0kSLYRBkjOT3JnkkebnGcfp+4Ike5P8Ya/jSpLa08ae\nwQZge1VdAGxv1ifyX4B7WxhTktSiNsLgSuCmZvkmYF23TkleAywDvtTCmJKkFrURBsuqan+z/CSd\nN/yjJHkO8LvA+060sSTXJRlOMjwyMtJCeZKkEzmp7zNIchfwoi533TB2paoqSXXp9x7gjqram+S4\nY1XVZmAzwNDQULdtSZJadlJhUFVrJrovyVNJllfV/iTLgQNdul0KvC7Je4DTgMVJvl9Vxzu/IEma\nIW1809lW4FpgY/PztvEdquodo8tJ3gkMGQSSNDjaOGewEXhTkkeANc06SYaS3NjC9iVJ0yxVg3tY\nfmhoqIaHh/tdhiTNGkl2VNXQZB/nJ5AlSYaBJMkwkCRhGEiSMAwkSRgGkiQMA0kShoEkCcNAkoRh\nIEnCMJAkYRhIkjAMJEkYBpIkDANJEoaBJAnDQJKEYSBJwjCQJGEYSJIwDCRJ9BgGSc5McmeSR5qf\nZ0zQ7/wkX0qyO8nDSVb2Mq4kqV297hlsALZX1QXA9ma9m08Am6rqx4GLgQM9jitJalGvYXAlcFOz\nfBOwbnyHJBcCC6vqToCq+n5V/aDHcSVJLeo1DJZV1f5m+UlgWZc+PwocTHJLkp1JNiVZMNEGk1yX\nZDjJ8MjISI/lSZJOxsITdUhyF/CiLnfdMHalqipJTTDG64DVwOPAZ4F3Ah/rNl5VbQY2AwwNDXXb\nniSpZScMg6paM9F9SZ5Ksryq9idZTvdzAXuB+6vqG81jtgCXMEEYSJJmXq+HibYC1zbL1wK3denz\nFWBpkrOb9TcAD/c4riSpRb2GwUbgTUkeAdY06yQZSnIjQFUdAd4HbE+yCwjw33ocV5LUohMeJjqe\nqvo28MYu7cPAu8es3wm8spexJEnTx08gS5IMA0mSYSBJwjCQJGEYSJIwDCRJGAaSJAwDSRKGgSQJ\nw0CShGEgScIwkCRhGEiSMAwkSRgGkiQMA0kShoEkCcNAkoRhIEnCMJAkYRhIkjAMJEm0EAZJzkxy\nZ5JHmp9nTNDvt5M8lGR3kj9Ikl7HliS1o409gw3A9qq6ANjerB8lyU8AlwGvBP4Z8M+Bn2phbElS\nC9oIgyuBm5rlm4B1XfoU8FxgMXAKsAh4qoWxJUktaCMMllXV/mb5SWDZ+A5V9ZfAPcD+5ratqnZ3\n21iS65IMJxkeGRlpoTxJ0oksPJlOSe4CXtTlrhvGrlRVJakuj38Z8OPAuU3TnUleV1V/Pr5vVW0G\nNgMMDQ0dsy1JUvtOKgyqas1E9yV5KsnyqtqfZDlwoEu3twBfrqrvN4/5M+BS4JgwkCTNvDYOE20F\nrm2WrwVu69LnceCnkixMsojOyeOuh4kkSTOvjTDYCLwpySPAmmadJENJbmz6fB74OrALeAB4oKr+\nVwtjS5JacFKHiY6nqr4NvLFL+zDw7mb5CPBveh1LkjQ9/ASyJMkwkCQZBpIkDANJEoaBJAnDQJKE\nYSBJwjCQJGEYSJIwDCRJGAaSJAwDSRKGgSQJw0CShGEgScIwkCRhGEiSMAwkSRgGkiQMA0kShoEk\nCcNAkkSPYZDkqiQPJXkmydBx+l2eZE+SR5Ns6GVMSVL7et0zeBB4K3DvRB2SLAA+CrwZuBC4JsmF\nPY4rSWrRwl4eXFW7AZIcr9vFwKNV9Y2m72eAK4GHexlbktSemThnsAL41pj1vU1bV0muSzKcZHhk\nZGTai5MkncSeQZK7gBd1ueuGqrqt7YKqajOwGWBoaKja3r4k6VgnDIOqWtPjGPuA88asn9u0SZIG\nxEwcJvoKcEGSlyRZDFwNbJ2BcSVJJ6nXqaVvSbIXuBS4Pcm2pv2cJHcAVNXTwL8DtgG7gT+tqod6\nK1uS1KZeZxPdCtzapf0J4Iox63cAd/QyliRp+vgJZEmSYSBJMgwkSRgGkiQMA0kShoEkCcNAkoRh\nIEnCMJAkYRhIkjAMJEkYBpIkDANJEoaBJAnDQJKEYSBJwjCQJGEYSJIwDCRJGAaSJAwDSRKGgSSJ\nHsMgyVVJHkryTJKhCfqcl+SeJA83fX+llzElSe3rdc/gQeCtwL3H6fM08J+q6kLgEuCXk1zY47iS\npBYt7OXBVbUbIMnx+uwH9jfL/y/JbmAF8HAvY0uS2jOj5wySrARWA/cdp891SYaTDI+MjMxUaZI0\nr51wzyDJXcCLutx1Q1XddrIDJTkN+ALw3qr63kT9qmozsBlgaGioTnb7kqSpO2EYVNWaXgdJsohO\nEHy6qm7pdXuSpHZN+2GidE4ofAzYXVUfnu7xJEmT1+vU0rck2QtcCtyeZFvTfk6SO5pulwH/CnhD\nkvub2xU9VS1JalWvs4luBW7t0v4EcEWz/H+BiacbSZL6zk8gS5IMA0mSYSBJwjCQJNHjCWS1Y8vO\nfWzatocnDh7inKVLWL92FetWr+h3WZLmEcOgz7bs3Mf1t+zi0OEjAOw7eIjrb9kFYCBImjEeJuqz\nTdv2PBsEow4dPsKmbXv6VJGk+cgw6LMnDh6aVLskTQfDoM/OWbpkUu2SNB0Mgz5bv3YVSxYtOKpt\nyaIFrF+7qk8VSZqPPIHcZ6MniZ1NJKmfDIMBsG71Ct/8JfWVh4kkSYaBJMkwkCRhGEiSMAwkSUCq\nqt81TCjJCPC3U3z4WcDftVhOmwa5NrC+Xg1yfYNcG1hfL0Zre3FVnT3ZBw90GPQiyXBVDfW7jm4G\nuTawvl4Ncn2DXBtYXy96rc3DRJIkw0CSNLfDYHO/CziOQa4NrK9Xg1zfINcG1teLnmqbs+cMJEkn\nby7vGUiSTpJhIEmae2GQ5PIke5I8mmRDn2r4eJIDSR4c03ZmkjuTPNL8PKNpT5I/aOr9apJXT3Nt\n5yW5J8nDSR5K8isDVt9zk/xVkgea+n6jaX9JkvuaOj6bZHHTfkqz/mhz/8rprG9MnQuS7EzyxUGr\nL8ljSXYluT/JcNM2KM/v0iSfT/I3SXYnuXSAalvV/J+N3r6X5L0DVN9/aH4nHkxyc/O70t7rrqrm\nzA1YAHwdeCmwGHgAuLAPdbweeDXw4Ji23wY2NMsbgN9qlq8A/gwIcAlw3zTXthx4dbP8fOBrwIUD\nVF+A05rlRcB9zbh/ClzdtP8J8G+b5fcAf9IsXw18doae4/8I/E/gi836wNQHPAacNa5tUJ7fm4B3\nN8uLgaWDUtu4OhcATwIvHoT6gBXAN4ElY15v72zzdTcj/7Ez+AReCmwbs349cH2falnJ0WGwB1je\nLC8H9jTL/xW4plu/GarzNuBNg1gf8Dzgr4HX0vlk5cLxzzOwDbi0WV7Y9Ms013UusB14A/DF5s1g\nkOp7jGPDoO/PL3B684aWQautS63/AviLQamPThh8CzizeR19EVjb5uturh0mGv0PG7W3aRsEy6pq\nf7P8JLCsWe5bzc2u42o6f30PTH3NIZj7gQPAnXT29g5W1dNdani2vub+7wIvnM76gI8A/xl4pll/\n4YDVV8CXkuxIcl3TNgjP70uAEeC/N4fYbkxy6oDUNt7VwM3Nct/rq6p9wO8AjwP76byOdtDi626u\nhcGsUJ247uuc3iSnAV8A3ltV3xt7X7/rq6ojVXURnb/ALwZ+rF+1jJfkXwIHqmpHv2s5jp+sqlcD\nbwZ+Ocnrx97Zx+d3IZ3Dp39cVauBf6Bz2GUQantWc9z954DPjb+vX/U15ymupBOo5wCnApe3OcZc\nC4N9wHlj1s9t2gbBU0mWAzQ/DzTtM15zkkV0guDTVXXLoNU3qqoOAvfQ2f1dmmT0a1rH1vBsfc39\npwPfnsayLgN+LsljwGfoHCr6/QGqb/SvSKrqAHArnUAdhOd3L7C3qu5r1j9PJxwGobax3gz8dVU9\n1awPQn1rgG9W1UhVHQZuofNabO11N9fC4CvABc0Z9sV0dvW29rmmUVuBa5vla+kcqx9t/6VmZsIl\nwHfH7JK2LkmAjwG7q+rDA1jf2UmWNstL6JzP2E0nFN42QX2jdb8NuLv5621aVNX1VXVuVa2k8/q6\nu6reMSj1JTk1yfNHl+kc+36QAXh+q+pJ4FtJVjVNbwQeHoTaxrmGHx4iGq2j3/U9DlyS5HnN7/Do\n/117r7uZOBkzkzc6Z/i/Ruc48w19quFmOsf1DtP5a+hddI7XbQceAe4Czmz6BvhoU+8uYGiaa/tJ\nOru5XwXub25XDFB9rwR2NvU9CPxq0/5S4K+AR+nsvp/StD+3WX+0uf+lM/g8/zQ/nE00EPU1dTzQ\n3B4a/R0YoOf3ImC4eX63AGcMSm3NmKfS+Qv69DFtA1Ef8BvA3zS/F58ETmnzdeflKCRJc+4wkSRp\nCgwDSZJhIEkyDCRJGAaSJAwDSRKGgSQJ+P9cPKuyXZ9HOAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1122696a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(np.arange(dataset.shape[1]), model.a[1,:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x1117def98>"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAEDCAYAAADUT6SnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZoMclgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1123134e0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(np.arange(dataset.shape[1]), np.std(model.a[:,:], 0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}