project_escargot.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 233,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "IPython.OutputArea.prototype._should_scroll = function(lines) {\n",
       "    return false;\n",
       "}"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%javascript\n",
    "IPython.OutputArea.prototype._should_scroll = function(lines) {\n",
    "    return false;\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {},
   "outputs": [],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "from matplotlib import colors as colors\n",
    "import numpy as np\n",
    "\n",
    "import sys \n",
    "sys.path.append('../')\n",
    "import Utils.utils as util\n",
    "from importlib import reload\n",
    "reload(util)\n",
    "\n",
    "from sklearn.linear_model import Lasso\n",
    "from sklearn.multiclass import OneVsRestClassifier\n",
    "from sklearn.preprocessing import normalize\n",
    "\n",
    "np.set_printoptions(threshold=np.inf) #print full arrays\n",
    "from sklearn import linear_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "metadata": {},
   "outputs": [],
   "source": [
    "def split_patch(patch):\n",
    "    h = patch.shape[0]\n",
    "    #Convert one patch into a column vector for training\n",
    "    patch = patch.flatten()\n",
    "    pixel_central = patch[(h*(h//2)+(h//2))*3:(h*(h//2)+(h//2))*3+3]\n",
    "    patch_sans_central = np.delete(patch, range((h*(h//2)+(h//2))*3, (h*(h//2)+(h//2))*3+3))\n",
    "    return np.array((patch_sans_central, pixel_central))\n",
    "    \n",
    "def convert_patch_into_vector(patch):\n",
    "    #Convert one patch into a column vector for training\n",
    "    #patch = patch.flatten()\n",
    "    h = patch.shape[0]\n",
    "    patch = np.reshape(patch, h*h*3)\n",
    "    return np.array(patch)\n",
    "\n",
    "def convert_vector_into_patch(vector, h):\n",
    "    #Convert a one column vector into a patch of size h\n",
    "    return vector.reshape((h,h,7))\n",
    "    \n",
    "def convert_patches_into_training_set(patches):\n",
    "    #Convert a set of patches into a set of vectors\n",
    "    return np.array(list(map(convert_patch_into_vector, patches))).T\n",
    "\n",
    "def find_index_noisy_lines(label):\n",
    "    \"\"\"\n",
    "    indexes = []\n",
    "    for index, value in enumerate(label):\n",
    "        if(value == -1):\n",
    "            indexes.append(index)\n",
    "    \"\"\"\n",
    "    label = np.array(label)\n",
    "    indexes = np.where(label < 0)\n",
    "    return indexes\n",
    "\n",
    "def delete_all_indexes_from(dictionary, label):\n",
    "    indexes = find_index_noisy_lines(label)\n",
    "    dictionary = np.delete(dictionary, indexes , 0)\n",
    "    label = np.delete(label, indexes , 0)\n",
    "    \n",
    "    return (dictionary, label)    \n",
    "\n",
    "#test the alpha on a lot of cleared patches and returns an average value\n",
    "def estimate_alpha(noisy_image, dictionary_full, h):\n",
    "    lassocv = linear_model.LassoCV()\n",
    "    listalpha = []\n",
    "    width = noisy_image.shape[1]\n",
    "    height = noisy_image.shape[0]\n",
    "    #params\n",
    "    pas = h+10\n",
    "    start = h//2\n",
    "    endx = height-h//2\n",
    "    endy = width-h//2\n",
    "    #end params\n",
    "    i=0\n",
    "    total = int( (endx-pas)/pas * (endy-pas)/pas )\n",
    "    for x in range(start,endx,pas):\n",
    "        for y in range(start,endy,pas):\n",
    "            label = util.get_patch(x, y, h, noisy_image)\n",
    "            label = list(convert_patch_into_vector(label))\n",
    "            dictionary, label = delete_all_indexes_from(dictionary_full,label)\n",
    "            lassocv.fit(dictionary_full, label)\n",
    "            alpha_ = lassocv.alpha_\n",
    "            listalpha.append(alpha_)\n",
    "            i+=1\n",
    "            if(i%10==0): print(i, \"/\", total)\n",
    "    return np.mean(np.array(listalpha))\n",
    "    \n",
    "\n",
    "def inpaint_image(noisy_image, dictionary, coord_noisy_pix, h, clear_image):\n",
    "    noisy_image_orig = np.array(noisy_image)\n",
    "    dictionary_full = np.array(dictionary)\n",
    "    \n",
    "    #todo question: vérifier qu'il faut bien faire le lassocv pour trouver le alpha sur le clear?\n",
    "    lassocv = linear_model.LassoCV(max_iter=1000)  #1000 by default, 100 works OK\n",
    "    \"\"\"\n",
    "    y, x = coord_noisy_pix[0]\n",
    "    label = util.get_patch(x, y, h, noisy_image)\n",
    "    #label = util.get_patch(x, y, h, clear_image)\n",
    "    label = list(convert_patch_into_vector(label))\n",
    "    lassocv.fit(dictionary_full, label)\n",
    "    alpha_ = lassocv.alpha_\n",
    "    \"\"\"\n",
    "    #alpha_ = estimate_alpha(noisy_image, dictionary_full, h)\n",
    "    #autre idée: en prendre 10 aléatoirement\n",
    "    alpha_ = 92 #calculé pour 12.jpg\n",
    "    \n",
    "    #alpha_ = 0.01\n",
    "    \n",
    "    print(\"alpha = \", alpha_)\n",
    "       \n",
    "    i = 0\n",
    "    nbiter  = int(len(coord_noisy_pix))\n",
    "    while i < nbiter: # DONT USE FOR FOR THIS BECAUSE WE WILL MODIFY THE LIST\n",
    "        x, y  = coord_noisy_pix[0]\n",
    "        \n",
    "        if(i%10==0): print( i, \" / \",  nbiter)\n",
    "        \n",
    "        #print(\"Correction de : (\", x, \", \", y,\")\")\n",
    "        patch = util.get_patch(x, y, h, noisy_image)\n",
    "        labelfull = list(convert_patch_into_vector(patch))\n",
    "        \n",
    "        #print(\"labelfull : \", labelfull)\n",
    "        \n",
    "        dictionary = np.array(dictionary_full)\n",
    "        dictionary, label = delete_all_indexes_from(dictionary,labelfull)\n",
    "\n",
    "        #print(\"--- lasso regression ---\")\n",
    "        \n",
    "        #recalcule du alpha à chaque patch\n",
    "        if(i%1==0) :\n",
    "            lassocv.fit(dictionary, label)\n",
    "            alpha_ = lassocv.alpha_\n",
    "            print(\"alpha = \", alpha_)\n",
    "    \n",
    "        #fin recalcule du alpha\n",
    "        \n",
    "        model = Lasso(fit_intercept=True,alpha=alpha_,positive=True)\n",
    "        \n",
    "        model.fit(dictionary, label)\n",
    "        \n",
    "        res = model.predict(dictionary_full)\n",
    "        \n",
    "        #coef = model.coef_.reshape((1,-1)) #poids\n",
    "        #print(\"nb patch not at zero: \", np.count_nonzero(model.sparse_coef_.toarray()), \" / \", coef.shape[1])\n",
    "       \n",
    "        # corrige juste le pixel central\n",
    "        \"\"\"\n",
    "        newPixel = util.get_centered_pixel(res, h)\n",
    "        print(\"new pixel (lasso): \", newPixel)\n",
    "        noisy_image[x][y] = newPixel\n",
    "        \"\"\"\n",
    "        \n",
    "        def removearray(L,arr):\n",
    "            ind = 0\n",
    "            size = len(L)\n",
    "            while ind != size and not np.array_equal(L[ind],arr):\n",
    "                ind += 1\n",
    "            if ind != size:\n",
    "                L.pop(ind)\n",
    "            else:\n",
    "                raise ValueError('array not found in list.')\n",
    "        \n",
    "        # corrige tous les pixels noirs dans le patch\n",
    "        labelfull = np.array(labelfull)\n",
    "        list_indexes_noise = np.where(labelfull < 0)[0]     \n",
    "        #print(list_indexes_noise)   \n",
    "        indexeach3 = list(range(0,len(list_indexes_noise),3))\n",
    "        listindexes2 = list_indexes_noise[indexeach3] #on prend que 1 sur 3 car 3 valeurs par pixel\n",
    "        xpatch = x-h//2\n",
    "        ypatch = y-h//2\n",
    "        for val in listindexes2:\n",
    "            val = (val/3)\n",
    "            xlocal = int((val)//h)\n",
    "            ylocal = int(val-(xlocal)*h)\n",
    "            #print(\"xlocal = \",xlocal,\" ylocal=\",ylocal)\n",
    "            xglobal = xpatch + xlocal\n",
    "            yglobal = ypatch + ylocal\n",
    "            #print(\"xglobal=\",xglobal,\" yglobal=\",yglobal)\n",
    "            patch = res.reshape(h,h,3)\n",
    "            newpixel = patch[xlocal, ylocal]\n",
    "            noisy_image[xglobal][yglobal] = newpixel\n",
    "            #supprimer de la liste pour pas le refaire\n",
    "            coord_noisy_pix = list( coord_noisy_pix)\n",
    "            currentpixel = [xglobal, yglobal] #ok vérifié\n",
    "            #print(\"remove : \", currentpixel)\n",
    "            \n",
    "            removearray(coord_noisy_pix, np.array(currentpixel))\n",
    "            coord_noisy_pix = list( coord_noisy_pix)\n",
    "            #print(\"left to correct : \",len(coord_noisy_pix))\n",
    "            i = i + 1\n",
    "            if(i%10==0): print( i, \" / \",  nbiter)\n",
    "            \n",
    "            \n",
    "        \n",
    "        #Test a supprimer\n",
    "        #_,clear_patches = util.get_patches(noisy_image, h, width, height)\n",
    "        #dictionary_full = convert_patches_into_training_set(clear_patches)\n",
    "        #print(\"shape : \", dictionary_full.shape)\n",
    "\n",
    "    util.display_im(noisy_image_orig)\n",
    "    util.display_im(noisy_image)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "taille de l'image :  (120, 200, 3)\n",
      "nb pixels à corriger:  (100, 100, 2)\n",
      "200  ,  120\n",
      "nbclear: 68\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsMAAAGxCAYAAACHjrsVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nCCGEDJb/D1D+TF34eqHxAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe968151b00>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "img_matrice = util.read_im(\"Data/imgdataset/10small.jpg\")\n",
    "\n",
    "print(\"taille de l'image : \", img_matrice.shape)\n",
    "\n",
    "#util.display_im(img_matrice)\n",
    "\n",
    "#Taille des patchs\n",
    "h=15 #nombre impair\n",
    "\n",
    "noisy_image, deletepixels = util.delete_rect(img_matrice, 10, 10, 100, 100) #66, 55, 75, 15\n",
    "#10small 65, 50, 80, 15\n",
    "#12.jpg 85, 75, 30, 6\n",
    "\n",
    "print(\"nb pixels à corriger: \", np.array(deletepixels).shape)\n",
    "\n",
    "width = img_matrice.shape[1]\n",
    "height = img_matrice.shape[0]\n",
    "\n",
    "print(width,\" , \", height)\n",
    "\n",
    "util.display_im(noisy_image)\n",
    "\n",
    "_,clear_patches = util.get_patches(noisy_image, h, width, height)\n",
    "\n",
    "nbclear = np.shape(clear_patches)[0]\n",
    "print(\"nbclear:\", nbclear)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(675, 68)\n"
     ]
    }
   ],
   "source": [
    "training = convert_patches_into_training_set(clear_patches)\n",
    "print(training.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(9, 2)\n",
      "(3, 3, 2)\n",
      "center= ['5' 'a']\n",
      "(3, 1)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[0, 2, 6]])"
      ]
     },
     "execution_count": 238,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = np.array([[1, 'a'], [2, 'a'], [3, 'a'], [4, 'a'], [5, 'a'], [6, 'a'], [7, 'a'], [8, 'a'], [9, 'a']])\n",
    "print(a.shape)    \n",
    "a = a.reshape(3,3,2)\n",
    "print(a.shape)\n",
    "print(\"center=\",util.get_centered_pixel(a, 3))\n",
    "\n",
    "#coefsize = (1, 2510)   dicofullsize = (27, 2510)  ressize= (27, 2510)\n",
    "coef = np.array([0,1,2])\n",
    "coef = coef.reshape(-1,1)\n",
    "print(coef.shape)\n",
    "dico = [1,2,3]\n",
    "coef.T*dico"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1  2  3  4\n",
    "5  6  7  8\n",
    "9 10 11 12"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([['1', 'a'],\n",
       "       ['2', 'a'],\n",
       "       ['3', 'a'],\n",
       "       ['6', 'a'],\n",
       "       ['9', 'a'],\n",
       "       ['8', 'a'],\n",
       "       ['7', 'a'],\n",
       "       ['4', 'a'],\n",
       "       ['5', 'a']],\n",
       "      dtype='<U21')"
      ]
     },
     "execution_count": 239,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def spyral_transformation(a):\n",
    "    out = []\n",
    "    while (a.size):\n",
    "        out.append(a[0])\n",
    "        a = np.rot90(np.delete(a, 0, 0))\n",
    "    return np.concatenate(out)\n",
    "\n",
    "spyral_transformation(a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  92\n",
      "0  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  68.7763377047\n",
      "10  /  10000\n",
      "20  /  10000\n",
      "30  /  10000\n",
      "40  /  10000\n",
      "50  /  10000\n",
      "60  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  66.2403309216\n",
      "70  /  10000\n",
      "80  /  10000\n",
      "90  /  10000\n",
      "100  /  10000\n",
      "110  /  10000\n",
      "120  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  64.6396598662\n",
      "130  /  10000\n",
      "140  /  10000\n",
      "150  /  10000\n",
      "160  /  10000\n",
      "170  /  10000\n",
      "180  /  10000\n",
      "190  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  69.080349695\n",
      "200  /  10000\n",
      "210  /  10000\n",
      "220  /  10000\n",
      "230  /  10000\n",
      "240  /  10000\n",
      "250  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  262.334171234\n",
      "260  /  10000\n",
      "270  /  10000\n",
      "280  /  10000\n",
      "290  /  10000\n",
      "300  /  10000\n",
      "310  /  10000\n",
      "320  /  10000\n",
      "320  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  240.595894834\n",
      "330  /  10000\n",
      "340  /  10000\n",
      "350  /  10000\n",
      "360  /  10000\n",
      "370  /  10000\n",
      "380  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  234.250762206\n",
      "390  /  10000\n",
      "400  /  10000\n",
      "410  /  10000\n",
      "420  /  10000\n",
      "430  /  10000\n",
      "440  /  "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10000\n",
      "alpha =  235.362676184\n",
      "450  /  10000\n",
      "460  /  10000\n",
      "470  /  10000\n",
      "480  /  10000\n",
      "490  /  10000\n",
      "500  /  10000\n",
      "510  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  233.696015604\n",
      "520  /  10000\n",
      "530  /  10000\n",
      "540  /  10000\n",
      "550  /  10000\n",
      "560 "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " /  10000\n",
      "570  /  10000\n",
      "alpha =  71.0810105323\n",
      "580  /  10000\n",
      "590  /  10000\n",
      "600  /  10000\n",
      "610  /  10000\n",
      "620  /  10000\n",
      "630  /  10000\n",
      "640  /  10000\n",
      "640  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  248.668372052\n",
      "650  /  10000\n",
      "660  /  10000\n",
      "670  /  10000\n",
      "680  /  10000\n",
      "690  /  10000\n",
      "700  / "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 10000\n",
      "alpha =  223.557685418\n",
      "710  /  10000\n",
      "720  /  10000\n",
      "730  /  10000\n",
      "740  /  10000\n",
      "750  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "760  /  10000\n",
      "alpha =  103.46171309\n",
      "770  /  10000\n",
      "780  /  10000\n",
      "790  /  10000\n",
      "800  /  10000\n",
      "800  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  14.2984367179\n",
      "810  /  10000\n",
      "820  /  10000\n",
      "830  /  10000\n",
      "840  /  10000\n",
      "850  /  10000\n",
      "860  /  10000\n",
      "alpha =  10.6117851051\n",
      "870  /  10000\n",
      "880  /  10000\n",
      "890  /  10000\n",
      "900  /  10000\n",
      "910  /  10000\n",
      "920  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  0.682307819914\n",
      "930  /  10000\n",
      "940  /  10000\n",
      "950  /  10000\n",
      "960  /  10000\n",
      "970  /  10000\n",
      "980  /  10000\n",
      "990  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  53.2017668553\n",
      "1000  /  10000\n",
      "1010  /  10000\n",
      "1020  /  10000\n",
      "1030  /  10000\n",
      "1040  /  10000\n",
      "1050  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  22.0453243886\n",
      "1060  /  10000\n",
      "1070  /  10000\n",
      "1080  /  10000\n",
      "1090  /  10000\n",
      "1100  /  10000\n",
      "1110  /  10000\n",
      "1120  /  10000\n",
      "1120  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  1215.5059226\n",
      "1130  /  10000\n",
      "1140  /  10000\n",
      "1150  /  10000\n",
      "1160  /  10000\n",
      "1170  /  10000\n",
      "1180  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  56.324146535\n",
      "1190  /  10000\n",
      "1200  /  10000\n",
      "1210  /  10000\n",
      "1220  /  10000\n",
      "1230  /  10000\n",
      "1240  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  13.5233917153\n",
      "1250  /  10000\n",
      "1260  /  10000\n",
      "1270  /  10000\n",
      "1280  /  10000\n",
      "1290  /  10000\n",
      "1300  /  10000\n",
      "1310  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  5.95423278987\n",
      "1320  /  10000\n",
      "1330  /  10000\n",
      "1340  /  10000\n",
      "1350  /  10000\n",
      "1360  /  10000\n",
      "1370  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  1.23398425987\n",
      "1380  /  10000\n",
      "1390  /  10000\n",
      "1400  /  10000\n",
      "1410  /  10000\n",
      "1420  /  10000\n",
      "1430  /  10000\n",
      "1440  /  10000\n",
      "1440  /  10000\n",
      "alpha =  1.9229270062\n",
      "1450  /  10000\n",
      "1460  /  10000\n",
      "1470  /  10000\n",
      "1480  /  10000\n",
      "1490  /  10000\n",
      "1500  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  1.45314868408\n",
      "1510  /  10000\n",
      "1520  /  10000\n",
      "1530  /  10000\n",
      "alpha =  126.848661825\n",
      "1540  /  10000\n",
      "1550  /  10000\n",
      "1560  /  10000\n",
      "1570  /  10000\n",
      "1580  /  10000\n",
      "1590  /  10000\n",
      "1600  /  10000\n",
      "1600  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  176.559500142\n",
      "1610  /  10000\n",
      "1620  /  10000\n",
      "1630  /  10000\n",
      "1640  /  10000\n",
      "1650  /  10000\n",
      "1660  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  210.197857398\n",
      "1670  /  10000\n",
      "1680  /  10000\n",
      "1690  /  10000\n",
      "1700  /  10000\n",
      "1710  /  10000\n",
      "1720  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  238.182593858\n",
      "1730  /  10000\n",
      "1740  /  10000\n",
      "1750  /  10000\n",
      "1760  /  10000\n",
      "1770  /  10000\n",
      "1780  /  10000\n",
      "1790  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  225.970671942\n",
      "1800  /  10000\n",
      "1810  /  10000\n",
      "1820  /  10000\n",
      "1830  /  10000\n",
      "1840  /  10000\n",
      "1850  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  234.324121289\n",
      "1860  /  10000\n",
      "1870  /  10000\n",
      "1880  /  10000\n",
      "1890  /  10000\n",
      "1900  /  10000\n",
      "1910  /  10000\n",
      "1920  /  10000\n",
      "1920  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  231.070661657\n",
      "1930  /  10000\n",
      "1940  /  10000\n",
      "1950  /  10000\n",
      "1960  /  10000\n",
      "1970  /  10000\n",
      "1980  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  222.234538959\n",
      "1990  /  10000\n",
      "2000  /  10000\n",
      "2010  /  10000\n",
      "2020  /  10000\n",
      "2030  /  10000\n",
      "2040  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  240.176521317\n",
      "2050  /  10000\n",
      "2060  /  10000\n",
      "2070  /  10000\n",
      "2080  /  10000\n",
      "2090  /  10000\n",
      "2100  /  10000\n",
      "2110  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  242.860226513\n",
      "2120  /  10000\n",
      "2130  /  10000\n",
      "2140  /  10000\n",
      "2150  /  10000\n",
      "2160  /  10000\n",
      "2170  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  243.026043286\n",
      "2180  /  10000\n",
      "2190  /  10000\n",
      "2200  /  10000\n",
      "2210  /  10000\n",
      "2220  /  10000\n",
      "2230  /  10000\n",
      "2240  /  10000\n",
      "2240  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  835.04893102\n",
      "2250  /  10000\n",
      "2260  /  10000\n",
      "2270  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  83.4708681322\n",
      "2280  /  10000\n",
      "2290  /  10000\n",
      "2300  /  10000\n",
      "2310  /  10000\n",
      "2320  /  10000\n",
      "2330  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  91.5589790349\n",
      "2340  /  10000\n",
      "2350  /  10000\n",
      "2360  /  10000\n",
      "2370  /  10000\n",
      "2380  /  10000\n",
      "2390  /  10000\n",
      "2400  /  10000\n",
      "2400  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  33.5610674567\n",
      "2410  /  10000\n",
      "2420  /  10000\n",
      "2430  /  10000\n",
      "2440  /  10000\n",
      "2450  /  10000\n",
      "2460  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  64.6704573094\n",
      "2470  /  10000\n",
      "2480  /  10000\n",
      "2490  /  10000\n",
      "2500  /  10000\n",
      "2510  /  10000\n",
      "2520  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  115.838882928\n",
      "2530  /  10000\n",
      "2540  /  10000\n",
      "2550  /  10000\n",
      "2560  /  10000\n",
      "2570  /  10000\n",
      "2580  /  10000\n",
      "2590  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  18.9755823179\n",
      "2600  /  10000\n",
      "2610  /  10000\n",
      "2620  /  10000\n",
      "2630  /  10000\n",
      "2640  /  10000\n",
      "2650  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  59.3777759245\n",
      "2660  /  10000\n",
      "2670  /  10000\n",
      "2680  /  10000\n",
      "2690  /  10000\n",
      "2700  /  10000\n",
      "2710  /  10000\n",
      "2720  /  10000\n",
      "2720  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  43.4344781516\n",
      "2730  /  10000\n",
      "2740  /  10000\n",
      "2750  /  10000\n",
      "2760  /  10000\n",
      "2770  /  10000\n",
      "2780  /  10000\n",
      "alpha =  113.826784221\n",
      "2790  /  10000\n",
      "2800  /  10000\n",
      "2810  /  10000\n",
      "2820  /  10000\n",
      "2830  /  10000\n",
      "2840  /  10000\n",
      "alpha =  11.029371639\n",
      "2850  /  10000\n",
      "2860  /  10000\n",
      "2870  /  10000\n",
      "2880  /  10000\n",
      "2890  /  10000\n",
      "2900  /  10000\n",
      "2910  /  10000\n",
      "alpha =  125.731994262\n",
      "2920  /  10000\n",
      "2930  /  10000\n",
      "2940  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  58.4066175778\n",
      "2950  /  10000\n",
      "2960  /  10000\n",
      "2970  /  10000\n",
      "2980  /  10000\n",
      "2990  /  10000\n",
      "3000  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  51.4243876162\n",
      "3010  /  10000\n",
      "3020  /  10000\n",
      "3030  /  10000\n",
      "3040  /  10000\n",
      "3050  /  10000\n",
      "3060  /  10000\n",
      "3070  /  10000\n",
      "alpha =  49.5444255843\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3080  /  10000\n",
      "3090  /  10000\n",
      "3100  /  10000\n",
      "3110  / "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 10000\n",
      "3120  /  10000\n",
      "3130  /  10000\n",
      "alpha =  1.68184567051\n",
      "3140  /  10000\n",
      "3150  /  10000\n",
      "3160  /  10000\n",
      "3170  /  10000\n",
      "3180  /  10000\n",
      "3190  /  10000\n",
      "3200  /  10000\n",
      "3200  /  10000\n",
      "alpha =  1.83608224168\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3210  /  10000\n",
      "3220  /  10000\n",
      "3230  /  10000\n",
      "3240  /  10000\n",
      "3250  / "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 10000\n",
      "3260  /  10000\n",
      "alpha =  2.27483058064\n",
      "3270  /  10000\n",
      "3280  /  10000\n",
      "3290  /  10000\n",
      "3300  /  10000\n",
      "3310  /  10000"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "3320  /  10000\n",
      "alpha =  2.11007269375\n",
      "3330  /  10000\n",
      "3340  /  10000\n",
      "3350  /  10000\n",
      "3360  /  10000\n",
      "3370  /  10000\n",
      "3380  /  10000\n",
      "3390  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  2.41600889448\n",
      "3400  /  10000\n",
      "3410  /  10000\n",
      "3420  /  10000\n",
      "3430  /  10000\n",
      "3440  /  10000\n",
      "3450  /  10000\n",
      "alpha =  22.3777781409\n",
      "3460  /  10000\n",
      "3470  /  10000\n",
      "3480  /  10000\n",
      "3490  /  10000\n",
      "3500  /  10000\n",
      "3510  /  10000\n",
      "3520  /  10000\n",
      "3520  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  1.33578907707\n",
      "3530  /  10000\n",
      "3540  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3550  /  10000\n",
      "3560  /  10000\n",
      "3570  /  10000\n",
      "3580  /  10000\n",
      "alpha =  2.95769985685\n",
      "3590  /  10000\n",
      "3600  /  10000\n",
      "3610  /  10000\n",
      "alpha =  41.3509799116\n",
      "3620  /  10000\n",
      "3630  /  10000\n",
      "3640  /  10000\n",
      "3650  /  10000\n",
      "3660  /  10000\n",
      "3670  /  10000\n",
      "3680  /  10000\n",
      "3680  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  8.86474531665\n",
      "3690  /  10000\n",
      "3700  /  10000\n",
      "3710  /  10000\n",
      "3720  /  10000\n",
      "3730  /  10000\n",
      "3740  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  18.1812832062\n",
      "3750  /  10000\n",
      "3760  /  10000\n",
      "3770  /  10000\n",
      "3780  /  10000\n",
      "3790  /  10000\n",
      "3800  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  8.67697950913\n",
      "3810  /  10000\n",
      "3820  /  10000\n",
      "3830  /  10000\n",
      "3840  /  10000\n",
      "3850  /  10000\n",
      "3860  /  10000\n",
      "3870  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  447.96064826\n",
      "3880  /  10000\n",
      "3890  /  10000\n",
      "3900  /  10000\n",
      "3910  /  10000\n",
      "3920  /  10000\n",
      "3930  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  22.3929018513\n",
      "3940  /  10000\n",
      "3950  /  10000\n",
      "3960  /  10000\n",
      "3970  /  10000\n",
      "3980  /  10000\n",
      "3990  /  10000\n",
      "4000  /  10000\n",
      "4000  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  13.8101926839\n",
      "4010  /  10000\n",
      "4020  /  10000\n",
      "4030  /  10000\n",
      "4040  /  10000\n",
      "4050  /  10000\n",
      "4060  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  36.6519938277\n",
      "4070  /  10000\n",
      "4080  /  10000\n",
      "4090  /  10000\n",
      "4100  /  10000\n",
      "4110  /  10000\n",
      "4120  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  1.41592110644\n",
      "4130  /  10000\n",
      "4140  /  10000\n",
      "4150  /  10000\n",
      "4160  /  10000\n",
      "4170  /  10000\n",
      "4180  /  10000\n",
      "4190  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  1.66196732654\n",
      "4200  /  10000\n",
      "4210  /  10000\n",
      "4220  /  10000\n",
      "alpha =  71.5908485076\n",
      "4230  /  10000\n",
      "4240  /  10000\n",
      "4250  /  10000\n",
      "4260  /  10000\n",
      "4270  /  10000\n",
      "4280  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  60.2813751304\n",
      "4290  /  10000\n",
      "4300  /  10000\n",
      "4310  /  10000\n",
      "4320  /  10000\n",
      "4330  /  10000\n",
      "4340  /  10000\n",
      "4350  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  55.3431215184\n",
      "4360  /  10000\n",
      "4370  /  10000\n",
      "4380  /  10000\n",
      "4390  /  10000\n",
      "4400  /  10000\n",
      "4410  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  44.8408243002\n",
      "4420  /  10000\n",
      "4430  /  10000\n",
      "4440  /  10000\n",
      "4450  /  10000\n",
      "4460  /  10000\n",
      "4470  /  10000\n",
      "4480  /  10000\n",
      "4480  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  19.0436294938\n",
      "4490  /  10000\n",
      "4500  /  10000\n",
      "4510  /  10000\n",
      "4520  /  10000\n",
      "4530  /  10000\n",
      "4540  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  18.6157237298\n",
      "4550  /  10000\n",
      "4560  /  10000\n",
      "4570  /  10000\n",
      "4580  /  10000\n",
      "4590  /  10000\n",
      "4600  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  23.4372347113\n",
      "4610  /  10000\n",
      "4620  /  10000\n",
      "4630  /  10000\n",
      "4640  /  10000\n",
      "4650  /  10000\n",
      "4660  /  10000\n",
      "4670  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  27.8436190058\n",
      "4680  /  10000\n",
      "4690  /  10000\n",
      "4700  /  10000\n",
      "4710  /  10000\n",
      "4720  /  10000\n",
      "4730  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  26.534099513\n",
      "4740  /  10000\n",
      "4750  /  10000\n",
      "4760  /  10000\n",
      "4770  /  10000\n",
      "4780  /  10000\n",
      "4790  /  10000\n",
      "4800  /  10000\n",
      "4800  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  2.41499108655\n",
      "4810  /  10000\n",
      "4820  /  10000\n",
      "4830  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  4.84179252243\n",
      "4840  /  10000\n",
      "4850  /  10000\n",
      "4860  /  10000\n",
      "4870  /  10000\n",
      "4880  /  10000\n",
      "4890  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  14.0607641082\n",
      "4900  /  10000\n",
      "4910  /  10000\n",
      "4920  /  10000\n",
      "4930  /  10000\n",
      "4940  /  10000\n",
      "4950  /  10000\n",
      "4960  /  10000\n",
      "4960  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  38.6817270818\n",
      "4970  /  10000\n",
      "4980  /  10000\n",
      "4990  /  10000\n",
      "5000  /  10000\n",
      "5010  /  10000\n",
      "5020  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  48.44264321\n",
      "5030  /  10000\n",
      "5040  /  10000\n",
      "5050  /  10000\n",
      "5060  /  10000\n",
      "5070  /  10000\n",
      "5080  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  2.46256530589\n",
      "5090  /  10000\n",
      "5100  /  10000\n",
      "5110  /  10000\n",
      "5120  /  10000\n",
      "5130  /  10000\n",
      "5140  /  10000\n",
      "5150  /  10000\n",
      "alpha =  1132.69103836\n",
      "5160  /  10000\n",
      "5170  /  10000\n",
      "5180  /  10000\n",
      "5190  /  10000\n",
      "5200  /  10000\n",
      "5210  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  12.3586493988\n",
      "5220  /  10000\n",
      "5230  /  10000\n",
      "5240  /  10000\n",
      "5250  /  10000\n",
      "5260  /  10000\n",
      "5270  /  10000\n",
      "5280  /  10000\n",
      "5280  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  70.5723156507\n",
      "5290  /  10000\n",
      "5300  /  10000\n",
      "5310  /  10000\n",
      "5320  /  10000\n",
      "5330  /  10000\n",
      "5340  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  2.89690448085\n",
      "5350  /  10000\n",
      "5360  /  10000\n",
      "5370  /  10000\n",
      "alpha =  2.78074061198\n",
      "5380  /  10000\n",
      "5390  /  10000\n",
      "5400  /  10000\n",
      "5410  /  10000\n",
      "5420  /  10000\n",
      "5430  /  10000\n",
      "5440  /  10000\n",
      "5440  /  10000\n",
      "alpha =  3.57060925021\n",
      "5450  /  10000\n",
      "5460  /  10000\n",
      "5470  /  10000\n",
      "5480  /  10000\n",
      "5490  /  10000\n",
      "5500  /  10000\n",
      "alpha =  36.2554795041\n",
      "5510  /  10000\n",
      "5520  /  10000\n",
      "5530  /  10000\n",
      "5540  /  10000\n",
      "5550  /  10000\n",
      "5560  /  10000\n",
      "alpha =  16.7934159991\n",
      "5570  /  10000\n",
      "5580  /  10000\n",
      "5590  /  10000\n",
      "5600  /  10000\n",
      "5610  /  10000\n",
      "5620"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  /  10000\n",
      "5630  /  10000\n",
      "alpha =  11.2145637062\n",
      "5640  /  10000\n",
      "5650  /  10000\n",
      "5660  /  10000\n",
      "5670  /  10000\n",
      "5680  /  10000\n",
      "5690  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  9.92085946751\n",
      "5700  /  10000\n",
      "5710  /  10000\n",
      "5720  /  10000\n",
      "5730  /  10000\n",
      "5740  /  10000\n",
      "5750  /  10000\n",
      "5760  /  10000\n",
      "5760  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  10.1222201653\n",
      "5770  /  10000\n",
      "5780  /  10000\n",
      "5790  /  10000\n",
      "5800  /  10000\n",
      "5810  /  10000\n",
      "5820  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  6.35163504124\n",
      "5830  /  10000\n",
      "5840  /  10000\n",
      "5850  /  10000\n",
      "5860  /  10000\n",
      "5870  /  10000\n",
      "5880  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  19.1787311928\n",
      "5890  /  10000\n",
      "5900  /  10000\n",
      "5910  /  10000\n",
      "5920  /  10000\n",
      "5920  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  173.369138067\n",
      "5930  /  10000\n",
      "5940  /  10000\n",
      "5950  /  10000\n",
      "5960  /  10000\n",
      "5970  /  10000\n",
      "5980  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  228.822280289\n",
      "5990  /  10000\n",
      "6000  /  10000\n",
      "6010  /  10000\n",
      "6020  /  10000\n",
      "6030  /  10000\n",
      "6040  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n",
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  3.0920505453\n",
      "6050  /  10000\n",
      "6060  /  10000\n",
      "6070  /  10000\n",
      "6080  /  10000\n",
      "6090  /  10000\n",
      "6100  /  10000\n",
      "6110  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  510.338836408\n",
      "6120  /  10000\n",
      "6130  /  10000\n",
      "6140  /  10000\n",
      "6150  /  10000\n",
      "6160  /  10000\n",
      "6170  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  13.2711229411\n",
      "6180  /  10000\n",
      "6190  /  10000\n",
      "6200  /  10000\n",
      "6210  /  10000\n",
      "6220  /  10000\n",
      "6230  /  10000\n",
      "6240  /  10000\n",
      "6240  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  15.079331266\n",
      "6250  /  10000\n",
      "6260  /  10000\n",
      "6270  /  10000\n",
      "6280  /  10000\n",
      "6290  /  10000\n",
      "6300  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  83.8121368131\n",
      "6310  /  10000\n",
      "6320  /  10000\n",
      "6330  /  10000\n",
      "6340  /  10000\n",
      "6350  /  10000\n",
      "6360  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  13.1839199624\n",
      "6370  /  10000\n",
      "6380  /  10000\n",
      "6390  /  10000\n",
      "6400  /  10000\n",
      "6400  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  61.1059611072\n",
      "6410  /  10000\n",
      "6420  /  10000\n",
      "6430  /  10000\n",
      "6440  /  10000\n",
      "6450  /  10000\n",
      "6460  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  21.9920395828\n",
      "6470  /  10000\n",
      "6480  /  10000\n",
      "6490  /  10000\n",
      "6500  /  10000\n",
      "6510  /  10000\n",
      "6520  /  10000\n",
      "alpha =  9.34034043824\n",
      "6530  /  10000\n",
      "6540  /  10000\n",
      "6550  /  10000\n",
      "6560  /  10000\n",
      "6570  /  10000\n",
      "6580  /  10000\n",
      "6590  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  25.1346508908\n",
      "6600  /  10000\n",
      "6610  /  10000\n",
      "6620  /  10000\n",
      "6630  /  10000\n",
      "6640  /  10000\n",
      "6650  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  16.1738249221\n",
      "6660  /  10000\n",
      "6670  /  10000\n",
      "6680  /  10000\n",
      "6690  /  10000\n",
      "6700  /  10000\n",
      "6710  /  10000\n",
      "6720  /  10000\n",
      "6720  /  10000\n",
      "alpha =  18.1927588116\n",
      "6730  /  10000\n",
      "6740  /  10000\n",
      "6750  /  10000\n",
      "6760  /  10000\n",
      "6770  /  10000\n",
      "6780  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  9.05599563554\n",
      "6790  /  10000\n",
      "6800  /  10000\n",
      "6810  /  10000\n",
      "6820  /  10000\n",
      "6830  /  10000\n",
      "6840  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  3.15207622396\n",
      "6850  /  10000\n",
      "6860  /  10000\n",
      "6870  /  10000\n",
      "6880  /  10000\n",
      "6880  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  103.77718921\n",
      "6890  /  10000\n",
      "6900  /  10000\n",
      "6910  /  10000\n",
      "6920  /  10000\n",
      "6930  /  10000\n",
      "6940  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  57.8670248743\n",
      "6950  /  10000\n",
      "6960  /  10000\n",
      "6970  /  10000\n",
      "6980  /  10000\n",
      "6990  /  10000\n",
      "7000  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  99.8678179999\n",
      "7010  /  10000\n",
      "7020  /  10000\n",
      "7030  /  10000\n",
      "7040  /  10000\n",
      "7050  /  10000\n",
      "7060  /  10000\n",
      "7070  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  123.260627435\n",
      "7080  /  10000\n",
      "7090  /  10000\n",
      "7100  /  10000\n",
      "7110  /  10000\n",
      "7120  /  10000\n",
      "7130  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  625.581326166\n",
      "7140  /  10000\n",
      "7150  /  10000\n",
      "7160  /  10000\n",
      "7170  /  10000\n",
      "7180  /  10000\n",
      "7190  /  10000\n",
      "7200  /  10000\n",
      "7200  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  18.7112169735\n",
      "7210  /  10000\n",
      "7220  /  10000\n",
      "7230  /  10000\n",
      "7240  /  10000\n",
      "7250  /  10000\n",
      "7260  /  10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/users/Etu7/3262287/.local/lib/python3.4/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alpha =  29.1883990267\n",
      "7270  /  10000\n",
      "7280  /  10000\n",
      "7290  /  10000\n"
     ]
    }
   ],
   "source": [
    "deletepixels_transformed = spyral_transformation(np.array(deletepixels))\n",
    "\n",
    "inpaint_image(noisy_image, training, deletepixels_transformed, h, img_matrice)"
   ]
  },
  {
   "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": "3.4.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}