1_notmnist.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "5hIbr52I7Z7U"
   },
   "source": [
    "Deep Learning\n",
    "=============\n",
    "\n",
    "Assignment 1\n",
    "------------\n",
    "\n",
    "The objective of this assignment is to learn about simple data curation practices, and familiarize you with some of the data we'll be reusing later.\n",
    "\n",
    "This notebook uses the [notMNIST](http://yaroslavvb.blogspot.com/2011/09/notmnist-dataset.html) dataset to be used with python experiments. This dataset is designed to look like the classic [MNIST](http://yann.lecun.com/exdb/mnist/) dataset, while looking a little more like real data: it's a harder task, and the data is a lot less 'clean' than MNIST."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "cellView": "both",
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     }
    },
    "colab_type": "code",
    "collapsed": false,
    "id": "apJbCsBHl-2A"
   },
   "outputs": [],
   "source": [
    "# These are all the modules we'll be using later. Make sure you can import them\n",
    "# before proceeding further.\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import os\n",
    "import sys\n",
    "import tarfile\n",
    "from IPython.display import display, Image\n",
    "from scipy import ndimage\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from six.moves.urllib.request import urlretrieve\n",
    "from six.moves import cPickle as pickle"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "jNWGtZaXn-5j"
   },
   "source": [
    "First, we'll download the dataset to our local machine. The data consists of characters rendered in a variety of fonts on a 28x28 image. The labels are limited to 'A' through 'J' (10 classes). The training set has about 500k and the testset 19000 labelled examples. Given these sizes, it should be possible to train models quickly on any machine."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "cellView": "both",
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     },
     "output_extras": [
      {
       "item_id": 1
      }
     ]
    },
    "colab_type": "code",
    "collapsed": false,
    "executionInfo": {
     "elapsed": 186058,
     "status": "ok",
     "timestamp": 1444485672507,
     "user": {
      "color": "#1FA15D",
      "displayName": "Vincent Vanhoucke",
      "isAnonymous": false,
      "isMe": true,
      "permissionId": "05076109866853157986",
      "photoUrl": "//lh6.googleusercontent.com/-cCJa7dTDcgQ/AAAAAAAAAAI/AAAAAAAACgw/r2EZ_8oYer4/s50-c-k-no/photo.jpg",
      "sessionId": "2a0a5e044bb03b66",
      "userId": "102167687554210253930"
     },
     "user_tz": 420
    },
    "id": "EYRJ4ICW6-da",
    "outputId": "0d0f85df-155f-4a89-8e7e-ee32df36ec8d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found and verified notMNIST_large.tar.gz\n",
      "Found and verified notMNIST_small.tar.gz\n"
     ]
    }
   ],
   "source": [
    "url = 'http://yaroslavvb.com/upload/notMNIST/'\n",
    "\n",
    "def maybe_download(filename, expected_bytes):\n",
    "  \"\"\"Download a file if not present, and make sure it's the right size.\"\"\"\n",
    "  if not os.path.exists(filename):\n",
    "    filename, _ = urlretrieve(url + filename, filename)\n",
    "  statinfo = os.stat(filename)\n",
    "  if statinfo.st_size == expected_bytes:\n",
    "    print 'Found and verified', filename\n",
    "  else:\n",
    "    raise Exception(\n",
    "      'Failed to verify' + filename + '. Can you get to it with a browser?')\n",
    "  return filename\n",
    "\n",
    "train_filename = maybe_download('notMNIST_large.tar.gz', 247336696)\n",
    "test_filename = maybe_download('notMNIST_small.tar.gz', 8458043)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "cC3p0oEyF8QT"
   },
   "source": [
    "Extract the dataset from the compressed .tar.gz file.\n",
    "This should give you a set of directories, labelled A through J."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "cellView": "both",
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     },
     "output_extras": [
      {
       "item_id": 1
      }
     ]
    },
    "colab_type": "code",
    "collapsed": false,
    "executionInfo": {
     "elapsed": 186055,
     "status": "ok",
     "timestamp": 1444485672525,
     "user": {
      "color": "#1FA15D",
      "displayName": "Vincent Vanhoucke",
      "isAnonymous": false,
      "isMe": true,
      "permissionId": "05076109866853157986",
      "photoUrl": "//lh6.googleusercontent.com/-cCJa7dTDcgQ/AAAAAAAAAAI/AAAAAAAACgw/r2EZ_8oYer4/s50-c-k-no/photo.jpg",
      "sessionId": "2a0a5e044bb03b66",
      "userId": "102167687554210253930"
     },
     "user_tz": 420
    },
    "id": "H8CBE-WZ8nmj",
    "outputId": "ef6c790c-2513-4b09-962e-27c79390c762"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['notMNIST_large/A', 'notMNIST_large/B', 'notMNIST_large/C', 'notMNIST_large/D', 'notMNIST_large/E', 'notMNIST_large/F', 'notMNIST_large/G', 'notMNIST_large/H', 'notMNIST_large/I', 'notMNIST_large/J']\n",
      "['notMNIST_small/A', 'notMNIST_small/B', 'notMNIST_small/C', 'notMNIST_small/D', 'notMNIST_small/E', 'notMNIST_small/F', 'notMNIST_small/G', 'notMNIST_small/H', 'notMNIST_small/I', 'notMNIST_small/J']\n"
     ]
    }
   ],
   "source": [
    "num_classes = 10\n",
    "\n",
    "def extract(filename):\n",
    "  #tar = tarfile.open(filename)\n",
    "  root = os.path.splitext(os.path.splitext(filename)[0])[0]  # remove .tar.gz\n",
    "  #print('Extracting data for %s. This may take a while. Please wait.' % root)\n",
    "  #sys.stdout.flush()\n",
    "  #tar.extractall()\n",
    "  #tar.close()\n",
    "  data_folders = [\n",
    "    os.path.join(root, d) for d in sorted(os.listdir(root)) if d != '.DS_Store']\n",
    "  if len(data_folders) != num_classes:\n",
    "    raise Exception(\n",
    "      'Expected %d folders, one per class. Found %d instead.' % (\n",
    "        num_classes, len(data_folders)))\n",
    "  print(data_folders)\n",
    "  return data_folders\n",
    "  \n",
    "train_folders = extract(train_filename)\n",
    "test_folders = extract(test_filename)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "4riXK3IoHgx6"
   },
   "source": [
    "---\n",
    "Problem 1\n",
    "---------\n",
    "\n",
    "Let's take a peek at some of the data to make sure it looks sensible. Each exemplar should be an image of a character A through J rendered in a different font. Display a sample of the images that we just downloaded. Hint: you can use the package IPython.display.\n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "A(notMNIST_small/A/QnJlYWtiZWF0IEJUTiBPdXRsaW5lLnR0Zg==.png)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAABwAAAAcCAAAAABXZoBIAAABbUlEQVR4nIWRPUucQRSFn3tndguz\nChpMYaxsJJDKQgixswjxJ1gnP8BaRVJaiSmFkHRqkzKFhRA0XQgohMQPCCr4BX4QlTWrMzfFuzvv\nu+yGnGbm3Dtn7pkzgMLo0mG0hNuNN/04BcGFwffPfq7sH/0RAKyr78lY5+xMlIjjxf33YYoQyq/C\nWg8CT+OSot67BO+UgbPPJWB9pwNHM6TEiE3CSxvD04ISH24esnzgyJyIqqpqRhzPbZyTxXZChMrv\nd/7RZoP3ViKglxcAJte7/djrzI5j2aq1WtVmG3z1i0czXWT67b1g/oAIYKilecb2dtpmixTM1H1a\nTJVCMy+m422e0U4pzSObm3mxTfNB2ZQot9XWmY6F872trb3zqfyPkjIy9/FyPkx0/8htF0L4+i0e\nh1WN+fBiCC56cPwrBDNCIYv/hKDJnUYRnCSCp0YIDU8hFK69K3uGfnXWuV73xtFKg1w9PpXW0OoI\n7tNfJh+IMjkgDUEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "B(notMNIST_small/B/RmFnb05vVGYtTWVkaXVtQ2Fwcy5vdGY=.png)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAABwAAAAcCAAAAABXZoBIAAABx0lEQVR4nGWSu2tUURDGvzlzL7sa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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "C(notMNIST_small/C/Q29tc2F0LVN0YXRpb24ub3Rm.png)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAABwAAAAcCAAAAABXZoBIAAABqklEQVR4nLWSvWtUURDFz9w7EUJA\no2wsNRaCuwZEEITVtbGJVrERgzZa2QkWFv4TWgkWFgo2giCIglpYiEUKIRaBgIWoiKIBBeP63r1z\n51i8jZtdsfRXzplhznwACED3+tKXRGenRTJ/fXXjuCAAQMDcIyPJDZEk/flhBCDi9E+6WTLL7LSK\nmWUzZ76IAMwbje4knft3NnWFVngWMv2yUyJl/d6yId6tzhOxfWYHpcSPXVxglWu+mWsMAIBgdpl1\n/sUruE1z8gQmVFUBVdUJHKrphU+lt8skVA+NGCJxccoZP8uebQ6x1YLNyCD1GYvz23bIiKqqqlEz\nXKITo1iTIggydDrKP8L/WyQxbnZDFIiMDfkHLSiiZTwcAYAyM0nAP/mINlif9tpZQv9mvbmvxMtb\nDfoOt2gkTzUnk2ap6Bm98DHOsco13x8dmhIcWGWdK16SqRcHTSnpyUpBvLZ+FYh7T05SSnzbBY71\nmf96MM9eFoCI+TV6sZRSzXbLUkrJivPHIiIQsPtOn01lZ2bw3dX9NgIECOS+he7stHLLkbUVonz/\nsPTgNYLjN5W0Cp7rM5L8AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "D(notMNIST_small/D/QWxpeDIudHRm.png)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAABwAAAAcCAAAAABXZoBIAAABw0lEQVR4nGXSPWjTcRDG8e/d/dJi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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "E(notMNIST_small/E/RGlnaUdyb3Rlc2tORUYtQm9sZENvbmRlbnMub3Rm.png)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAABwAAAAcCAAAAABXZoBIAAAAmElEQVR4nI2SSw7CMBBD7TbdIy7Z\nM3IIxAEQR0Bd8imNWSCUZNJEk+2LR7ZnKDRfHNoMCNnHSsnO2Ez5vK0smaDf++hyhDWQlCTIFoQg\nY8AZhQEhtuD7hdWM/ecUl9NjLKGcJWizLGaGauOusXG4zvfJGErdng9gQwmEyZagwu1mGvKeSW9l\nqleWcu6cibPbndPMYG28G+ULIX9HSjhCA9sAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F(notMNIST_small/F/QWt6aWRlbnpHcm90ZXNrQkUtTWQub3Rm.png)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAABwAAAAcCAAAAABXZoBIAAAAj0lEQVR4nN2RsQ3CMBRE72JDwwKs\nkQFYkV0YBIkhUoSCGiQnPgqwlfgnVmj51Zffv/N9m8JKje7crDEA8KkZrDJ6M5WK8J8z8XkpYTxc\nIUmK6qp3koYpQy3stG0VlLaawgXbBN2xUOoRwK9AYY7c63RrkpL7OeSOG9Ma25DTivfWBEL+lbGv\n2P7+CNW0fwTf79MvKmeLXGoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "G(notMNIST_small/G/RGVhclRlYWNoZXItTm9ybWFsLnR0Zg==.png)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAABwAAAAcCAAAAABXZoBIAAAB/0lEQVR4nG2STUhVURSF197nXlN6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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "H(notMNIST_small/H/QmVybnN0ZWluLURlbWlCb2xkLm90Zg==.png)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAABwAAAAcCAAAAABXZoBIAAABTklEQVR4nMWSMS9EURCFv3vfDctD\nlggJjaCUiFqELZAodKJQ+Qvih6gkKo1CiJIIOgqJhAKNkCgotlnF7go7991RvLUsiWjEKe5M5syZ\nOzkZsz91XSg9E8ftQztzDr+wcZcvln1TnO1vc3th2gIQDg5RuNkeHwCAizMgVxYRKeWoovvSV/xr\nsgJEEcfqvR7hIgCTYUklhMogDTbB5lHlwZIAqDcPgHl+MmIxIQCkL4CmTYli+QF/RLpaZoypi5/J\noJImgv9GNra6tOh8y+fBWyr6BUGfspiqUs3tiU1tsKFvQuuUoqsfsyZV6pTQ7Gp/tn/f1r9v6ZNf\nmfBPpI0AXK3NRABRxmAjDb0YQ08gPTCnvYDGXdoAzLyIiBTH3pWdF77iX5M1YHm/6rXsLhJZRjYf\nq4XzNTc7elUol4lbOmYK6wYG5u9PiyVpirP9w29BN6oA2ywJYwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "I(notMNIST_small/I/SVRDIEdvdWR5IFNhbnMgTGlnaHQucGZi.png)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAABwAAAAcCAAAAABXZoBIAAAAiElEQVR4nGPsYVDUUGVlQAe/b16/\ny8CykMGWTQGL5L29hxlYnjK8+fYfQ47h//f3zxlYvjP8+oMpx8Dw9+d3BpbvDL/+YtP59/cPBiZs\numBgVHJUEgJYOBhYmRkxxRmZWdgZWDgZ2Jmx6WJm52RgkWEQ4camk1NIkoEljkFRCTPBM7AqOSsw\nAADTqCBgVmBgnwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "J(notMNIST_small/J/QW1wbGl0dWRlQ29tcC1MaWdodC50dGY=.png)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAABwAAAAcCAAAAABXZoBIAAAAqklEQVR4nO3KPwqCcBiH8e/7801D\nhP5N5WBEWxdoiK7REbpGtwk6RtQRlIKaHRyENC1F3waDBhE6gM/6eYBa20i+qTr+arHFP5CbSaQZ\niyRnEAEQQEAAkdYxCAAk8x8Mkeqk6hjOHQWUop7nK8Mc6ZTEWV4KVLdnrzYOl0oVgXc4MS13i/47\nTtMsJzatgQ4AcvP2x/BFPF3P7PHEskxDI2RxFAb+xXXvAOEDV1tAW1Ux6M0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import random\n",
    "#display(Image(url='http://www.google.fr/images/srpr/logo3w.png'))\n",
    "examples = []\n",
    "for letterdir in test_folders:    \n",
    "    letters = os.listdir(letterdir)\n",
    "    letter = os.path.join(letterdir, letters[int(len(letters)*random.random())])\n",
    "    examples.append(letter)\n",
    "    print(letterdir[len(letterdir)-1] + '(' + letter + ')')\n",
    "    display(Image(filename=letter))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "PBdkjESPK8tw"
   },
   "source": [
    "Now let's load the data in a more manageable format.\n",
    "\n",
    "We'll convert the entire dataset into a 3D array (image index, x, y) of floating point values, normalized to have approximately zero mean and standard deviation ~0.5 to make training easier down the road. The labels will be stored into a separate array of integers 0 through 9.\n",
    "\n",
    "A few images might not be readable, we'll just skip them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "cellView": "both",
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     },
     "output_extras": [
      {
       "item_id": 30
      }
     ]
    },
    "colab_type": "code",
    "collapsed": false,
    "executionInfo": {
     "elapsed": 399874,
     "status": "ok",
     "timestamp": 1444485886378,
     "user": {
      "color": "#1FA15D",
      "displayName": "Vincent Vanhoucke",
      "isAnonymous": false,
      "isMe": true,
      "permissionId": "05076109866853157986",
      "photoUrl": "//lh6.googleusercontent.com/-cCJa7dTDcgQ/AAAAAAAAAAI/AAAAAAAACgw/r2EZ_8oYer4/s50-c-k-no/photo.jpg",
      "sessionId": "2a0a5e044bb03b66",
      "userId": "102167687554210253930"
     },
     "user_tz": 420
    },
    "id": "h7q0XhG3MJdf",
    "outputId": "92c391bb-86ff-431d-9ada-315568a19e59"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "notMNIST_large/A(0)\n",
      "('Could not read:', 'notMNIST_large/A/Um9tYW5hIEJvbGQucGZi.png', ':', IOError('cannot identify image file',), \"- it's ok, skipping.\")\n",
      "('Could not read:', 'notMNIST_large/A/SG90IE11c3RhcmQgQlROIFBvc3Rlci50dGY=.png', ':', IOError('cannot identify image file',), \"- it's ok, skipping.\")\n",
      "('Could not read:', 'notMNIST_large/A/RnJlaWdodERpc3BCb29rSXRhbGljLnR0Zg==.png', ':', IOError('cannot identify image file',), \"- it's ok, skipping.\")\n",
      "notMNIST_large/B(52909)\n",
      "('Could not read:', 'notMNIST_large/B/TmlraXNFRi1TZW1pQm9sZEl0YWxpYy5vdGY=.png', ':', IOError('cannot identify image file',), \"- it's ok, skipping.\")\n",
      "notMNIST_large/C(105820)\n",
      "notMNIST_large/D(158732)\n",
      "('Could not read:', 'notMNIST_large/D/VHJhbnNpdCBCb2xkLnR0Zg==.png', ':', IOError('cannot identify image file',), \"- it's ok, skipping.\")\n",
      "notMNIST_large/E(211643)\n",
      "notMNIST_large/F(264555)\n",
      "notMNIST_large/G(317467)\n",
      "notMNIST_large/H(370379)\n",
      "notMNIST_large/I(423291)\n",
      "notMNIST_large/J(476203)\n",
      "('Full dataset tensor:', (529114, 28, 28))\n",
      "('Mean:', -0.020221893)\n",
      "('Standard deviation:', 0.14220417)\n",
      "('Labels:', (529114,))\n",
      "notMNIST_small/A(0)\n",
      "('Could not read:', 'notMNIST_small/A/RGVtb2NyYXRpY2FCb2xkT2xkc3R5bGUgQm9sZC50dGY=.png', ':', IOError('cannot identify image file',), \"- it's ok, skipping.\")\n",
      "notMNIST_small/B(1872)\n",
      "notMNIST_small/C(3745)\n",
      "notMNIST_small/D(5618)\n",
      "notMNIST_small/E(7491)\n",
      "notMNIST_small/F(9364)\n",
      "('Could not read:', 'notMNIST_small/F/Q3Jvc3NvdmVyIEJvbGRPYmxpcXVlLnR0Zg==.png', ':', IOError('cannot identify image file',), \"- it's ok, skipping.\")\n",
      "notMNIST_small/G(11236)\n",
      "notMNIST_small/H(13108)\n",
      "notMNIST_small/I(14980)\n",
      "notMNIST_small/J(16852)\n",
      "('Full dataset tensor:', (18724, 28, 28))\n",
      "('Mean:', -0.074780457)\n",
      "('Standard deviation:', 0.45423415)\n",
      "('Labels:', (18724,))\n"
     ]
    }
   ],
   "source": [
    "image_size = 28  # Pixel width and height.\n",
    "pixel_depth = 255.0  # Number of levels per pixel.\n",
    "\n",
    "def load(data_folders, min_num_images, max_num_images, min_examples_from_folder = 0):\n",
    "  dataset = np.ndarray(\n",
    "    shape=(max_num_images, image_size, image_size), dtype=np.float32)\n",
    "  labels = np.ndarray(shape=(max_num_images), dtype=np.int32)\n",
    "  label_index = 0\n",
    "  image_index = 0\n",
    "  for folder in data_folders:\n",
    "    print(folder + '(' + str(image_index) + ')')\n",
    "    file_list = os.listdir(folder)\n",
    "    if(min_examples_from_folder == 0):\n",
    "        min_examples_from_folder = len(file_list) \n",
    "    for image in file_list[:min_examples_from_folder]:\n",
    "      if image_index >= max_num_images:\n",
    "        raise Exception('More images than expected: %d >= %d' % (\n",
    "          image_index, max_num_images))\n",
    "      image_file = os.path.join(folder, image)\n",
    "      try:\n",
    "        image_data = (ndimage.imread(image_file).astype(float) -\n",
    "                      pixel_depth / 2) / pixel_depth\n",
    "        if image_data.shape != (image_size, image_size):\n",
    "          raise Exception('Unexpected image shape: %s' % str(image_data.shape))\n",
    "        dataset[image_index, :, :] = image_data\n",
    "        labels[image_index] = label_index\n",
    "        image_index += 1\n",
    "      except IOError as e:\n",
    "        print('Could not read:', image_file, ':', e, '- it\\'s ok, skipping.')\n",
    "    label_index += 1\n",
    "  num_images = image_index\n",
    "  dataset = dataset[0:num_images, :, :]\n",
    "  labels = labels[0:num_images]\n",
    "  #if num_images < min_num_images:\n",
    "  #  raise Exception('Many fewer images than expected: %d < %d' % (\n",
    "  #      num_images, min_num_images))\n",
    "  print('Full dataset tensor:', dataset.shape)\n",
    "  print('Mean:', np.mean(dataset))\n",
    "  print('Standard deviation:', np.std(dataset))\n",
    "  print('Labels:', labels.shape)\n",
    "  return dataset, labels\n",
    "#min_examples_from_folder == 0 means load all\n",
    "train_dataset, train_labels = load(train_folders, 450000, 550000, 0)\n",
    "test_dataset, test_labels = load(test_folders, 18000, 20000)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "vUdbskYE2d87"
   },
   "source": [
    "---\n",
    "Problem 2\n",
    "---------\n",
    "\n",
    "Let's verify that the data still looks good. Displaying a sample of the labels and images from the ndarray. Hint: you can use matplotlib.pyplot.\n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU0AAAESCAYAAAB5KIfmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ffa9df0fa10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU0AAAESCAYAAAB5KIfmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ffa9df0f990>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.cm as cm\n",
    "import matplotlib.gridspec as gridspec\n",
    "\n",
    "def display_sample(dataset, labels, title_label, model=None):    \n",
    "    gs = gridspec.GridSpec(4, 4)\n",
    "    gs.update(wspace=0.3, hspace=1)\n",
    "    fig = plt.figure()\n",
    "    fig.suptitle(title_label)\n",
    "    for d in range(0, 16):\n",
    "        a=fig.add_subplot(gs[d])\n",
    "        letter_idx = int(len(dataset)*random.random())\n",
    "        #print(letter_idx)\n",
    "        plt.imshow(dataset[letter_idx], cmap = cm.Greys_r)\n",
    "        l_title = chr(labels[letter_idx] + ord('A'))\n",
    "        if model is not None:\n",
    "            l_title = l_title + ' / ~' + chr(model.predict(dataset[letter_idx].reshape(1, -1)) + ord('A'))\n",
    "        a.set_title(l_title)\n",
    "        a.axis('off')    \n",
    "    plt.show()\n",
    "    \n",
    "display_sample(train_dataset, train_labels, 'train dataset')\n",
    "display_sample(test_dataset, test_labels, 'test dataset')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "GPTCnjIcyuKN"
   },
   "source": [
    "#### Next, we'll randomize the data. It's important to have the labels well shuffled for the training and test distributions to match."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "cellView": "both",
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     }
    },
    "colab_type": "code",
    "collapsed": false,
    "id": "6WZ2l2tN2zOL"
   },
   "outputs": [],
   "source": [
    "np.random.seed(133)\n",
    "def randomize(dataset, labels):\n",
    "  permutation = np.random.permutation(labels.shape[0])\n",
    "  shuffled_dataset = dataset[permutation,:,:]\n",
    "  shuffled_labels = labels[permutation]\n",
    "  return shuffled_dataset, shuffled_labels\n",
    "train_dataset, train_labels = randomize(train_dataset, train_labels)\n",
    "test_dataset, test_labels = randomize(test_dataset, test_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "puDUTe6t6USl"
   },
   "source": [
    "---\n",
    "Problem 3\n",
    "---------\n",
    "Convince yourself that the data is still good after shuffling!\n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU0AAAESCAYAAAB5KIfmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSQghElDQJIQQCShoEkKIBBQ0CSFEgv8ACo48953iWocAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ffaa17cf210>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU0AAAESCAYAAAB5KIfmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nwgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ffaa17cf1d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display_sample(train_dataset, train_labels, 'train dataset')\n",
    "display_sample(test_dataset, test_labels, 'test dataset')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "cYznx5jUwzoO"
   },
   "source": [
    "---\n",
    "Problem 4\n",
    "---------\n",
    "Another check: we expect the data to be balanced across classes. Verify that.\n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[  0.00000000e+00   5.29090000e+04]\n",
      " [  1.00000000e+00   5.29110000e+04]\n",
      " [  2.00000000e+00   5.29120000e+04]\n",
      " [  3.00000000e+00   5.29110000e+04]\n",
      " [  4.00000000e+00   5.29120000e+04]\n",
      " [  5.00000000e+00   5.29120000e+04]\n",
      " [  6.00000000e+00   5.29120000e+04]\n",
      " [  7.00000000e+00   5.29120000e+04]\n",
      " [  8.00000000e+00   5.29120000e+04]\n",
      " [  9.00000000e+00   5.29110000e+04]]\n",
      "[[  0.00000000e+00   1.87200000e+03]\n",
      " [  1.00000000e+00   1.87300000e+03]\n",
      " [  2.00000000e+00   1.87300000e+03]\n",
      " [  3.00000000e+00   1.87300000e+03]\n",
      " [  4.00000000e+00   1.87300000e+03]\n",
      " [  5.00000000e+00   1.87200000e+03]\n",
      " [  6.00000000e+00   1.87200000e+03]\n",
      " [  7.00000000e+00   1.87200000e+03]\n",
      " [  8.00000000e+00   1.87200000e+03]\n",
      " [  9.00000000e+00   1.87200000e+03]]\n"
     ]
    }
   ],
   "source": [
    "import scipy.stats as stats\n",
    "\n",
    "print stats.itemfreq(train_labels)\n",
    "print stats.itemfreq(test_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "LA7M7K22ynCt"
   },
   "source": [
    "Prune the training data as needed. Depending on your computer setup, you might not be able to fit it all in memory, and you can tune train_size as needed.\n",
    "\n",
    "Also create a validation dataset for hyperparameter tuning."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "cellView": "both",
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     },
     "output_extras": [
      {
       "item_id": 1
      }
     ]
    },
    "colab_type": "code",
    "collapsed": false,
    "executionInfo": {
     "elapsed": 411281,
     "status": "ok",
     "timestamp": 1444485897869,
     "user": {
      "color": "#1FA15D",
      "displayName": "Vincent Vanhoucke",
      "isAnonymous": false,
      "isMe": true,
      "permissionId": "05076109866853157986",
      "photoUrl": "//lh6.googleusercontent.com/-cCJa7dTDcgQ/AAAAAAAAAAI/AAAAAAAACgw/r2EZ_8oYer4/s50-c-k-no/photo.jpg",
      "sessionId": "2a0a5e044bb03b66",
      "userId": "102167687554210253930"
     },
     "user_tz": 420
    },
    "id": "s3mWgZLpyuzq",
    "outputId": "8af66da6-902d-4719-bedc-7c9fb7ae7948"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('Training', (469114, 28, 28), (469114,))\n",
      "('Validation', (60000, 28, 28), (60000,))\n"
     ]
    }
   ],
   "source": [
    "train_size = 469114\n",
    "valid_size = 60000\n",
    "\n",
    "valid_dataset = train_dataset[:valid_size,:,:]\n",
    "valid_labels = train_labels[:valid_size]\n",
    "train_dataset = train_dataset[valid_size:valid_size+train_size,:,:]\n",
    "train_labels = train_labels[valid_size:valid_size+train_size]\n",
    "print('Training', train_dataset.shape, train_labels.shape)\n",
    "print('Validation', valid_dataset.shape, valid_labels.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "tIQJaJuwg5Hw"
   },
   "source": [
    "Finally, let's save the data for later reuse:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "cellView": "both",
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     }
    },
    "colab_type": "code",
    "collapsed": true,
    "id": "QiR_rETzem6C"
   },
   "outputs": [],
   "source": [
    "def save_pickle(name, train_dataset, train_labels, valid_dataset, valid_labels, test_dataset, test_labels):\n",
    "    #will use global variables for datasets and labels. sorry\n",
    "    pickle_file = name + '.pickle'\n",
    "\n",
    "    try:\n",
    "      f = open(pickle_file, 'wb')\n",
    "      save = {\n",
    "        'train_dataset': train_dataset,\n",
    "        'train_labels': train_labels,\n",
    "        'valid_dataset': valid_dataset,\n",
    "        'valid_labels': valid_labels,\n",
    "        'test_dataset': test_dataset,\n",
    "        'test_labels': test_labels,\n",
    "        }\n",
    "      pickle.dump(save, f, pickle.HIGHEST_PROTOCOL)      \n",
    "      f.close()\n",
    "      statinfo = os.stat(pickle_file)\n",
    "      print('Compressed pickle size:', statinfo.st_size)\n",
    "    except Exception as e:\n",
    "        print('Unable to save data to', pickle_file, ':', e)\n",
    "        raise"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "cellView": "both",
    "colab": {
     "autoexec": {
      "startup": false,
      "wait_interval": 0
     },
     "output_extras": [
      {
       "item_id": 1
      }
     ]
    },
    "colab_type": "code",
    "collapsed": false,
    "executionInfo": {
     "elapsed": 413065,
     "status": "ok",
     "timestamp": 1444485899688,
     "user": {
      "color": "#1FA15D",
      "displayName": "Vincent Vanhoucke",
      "isAnonymous": false,
      "isMe": true,
      "permissionId": "05076109866853157986",
      "photoUrl": "//lh6.googleusercontent.com/-cCJa7dTDcgQ/AAAAAAAAAAI/AAAAAAAACgw/r2EZ_8oYer4/s50-c-k-no/photo.jpg",
      "sessionId": "2a0a5e044bb03b66",
      "userId": "102167687554210253930"
     },
     "user_tz": 420
    },
    "id": "hQbLjrW_iT39",
    "outputId": "b440efc6-5ee1-4cbc-d02d-93db44ebd956"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('Compressed pickle size:', 1720211761)\n"
     ]
    }
   ],
   "source": [
    "#save original pickle\n",
    "save_pickle('notMNISTfull', train_dataset, train_labels, valid_dataset, valid_labels, test_dataset, test_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "gE_cRAQB33lk"
   },
   "source": [
    "---\n",
    "Problem 5\n",
    "---------\n",
    "\n",
    "By construction, this dataset might contain a lot of overlapping samples, including training data that's also contained in the validation and test set! Overlap between training and test can skew the results if you expect to use your model in an environment where there is never an overlap, but are actually ok if you expect to see training samples recur when you use it.\n",
    "Measure how much overlap there is between training, validation and test samples.\n",
    "\n",
    "Optional questions:\n",
    "- What about near duplicates between datasets? (images that are almost identical)\n",
    "- Create a sanitized validation and test set, and compare your accuracy on those in subsequent assignments.\n",
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "enumerated elements:300000\n",
      "duplicates:25409 which is 8.46966666667%\n",
      "check validation dataset against training\n",
      "enumerated elements:60000\n",
      "duplicates:9278 which is 15.4633333333%\n",
      "now check test dataset against training\n",
      "enumerated elements:18724\n",
      "duplicates:3909 which is 20.8769493698%\n"
     ]
    }
   ],
   "source": [
    "#count duplicates by hashing\n",
    "def check_duplicates(dataset, hashes=None):\n",
    "    if hashes is None:\n",
    "        hashes = {}\n",
    "    total = 0\n",
    "    dups = 0\n",
    "    dataset.flags.writeable = False\n",
    "    for letter in dataset:\n",
    "        h = hash(letter.data)\n",
    "        total = total + 1\n",
    "        if(h in hashes):\n",
    "            dups = dups + 1\n",
    "        else:\n",
    "            hashes[h] = h\n",
    "    print('enumerated elements:' + str(total))\n",
    "    print('duplicates:' + str(dups) + ' which is ' + str(dups*100.0/total) + '%')\n",
    "    return hashes\n",
    "\n",
    "#check test dataset against training\n",
    "hashes = check_duplicates(train_dataset)\n",
    "print('check validation dataset against training')\n",
    "hashes = check_duplicates(valid_dataset, hashes)\n",
    "print('now check test dataset against training')\n",
    "hashes = check_duplicates(test_dataset, hashes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "enumerated elements:300000\n",
      "removed duplicates:25409 which is 8.46966666667%\n",
      "dedup sizes (set/labels): 274591/274591\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU0AAAESCAYAAAB5KIfmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f5c142bbd50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "enumerated elements:60000\n",
      "removed duplicates:9278 which is 15.4633333333%\n",
      "dedup sizes (set/labels): 50722/50722\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU0AAAESCAYAAAB5KIfmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEonEB6RoSiQSiQ9I0ZRIJBIf+P8I/A7ZxXnKOAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f5beaa0ffd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "enumerated elements:18724\n",
      "removed duplicates:3909 which is 20.8769493698%\n",
      "dedup sizes (set/labels): 14815/14815\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU0AAAESCAYAAAB5KIfmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n24CMAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f5b82504fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('Compressed pickle size:', 1068002361)\n"
     ]
    }
   ],
   "source": [
    "def deduplicate(dataset, labels, hashes=None):\n",
    "    if hashes is None:\n",
    "        hashes = {}\n",
    "    #create deduplicated empty arrays\n",
    "    #dedup_dataset = np.empty((0, dataset.shape[1], dataset.shape[2]))\n",
    "    #dedup_labels = np.empty((0))\n",
    "    unique_values = []\n",
    "    unique_labels = []\n",
    "    i = 0\n",
    "    dups = 0\n",
    "    dataset.flags.writeable = False\n",
    "    while i < dataset.shape[0]:\n",
    "        letter = dataset[i]    \n",
    "        h = hash(letter.data)        \n",
    "        if(h in hashes):\n",
    "            dups = dups + 1\n",
    "        else:\n",
    "            hashes[h] = h\n",
    "            unique_values.append(letter)\n",
    "            unique_labels.append(labels[i])\n",
    "            #np.append(dedup_dataset, [letter], axis=0)\n",
    "            #np.append(dedup_labels, labels[i], axis=0)\n",
    "        i = i + 1\n",
    "    dedup_dataset = np.array(unique_values)\n",
    "    dedup_labels = np.array(unique_labels)\n",
    "    print('enumerated elements:' + str(i))\n",
    "    print('removed duplicates:' + str(dups) + ' which is ' + str(dups*100.0/i) + '%')\n",
    "    print('dedup sizes (set/labels): ' + str(dedup_dataset.shape[0]) + '/' + str(dedup_labels.shape[0]))\n",
    "    \n",
    "    return dedup_dataset, dedup_labels, hashes\n",
    "    \n",
    "dedup_train_dataset, dedup_train_labels, hashes = deduplicate(train_dataset, train_labels)\n",
    "display_sample(dedup_train_dataset, dedup_train_labels, 'dedup train')\n",
    "dedup_valid_dataset, dedup_valid_labels, hashes = deduplicate(valid_dataset, valid_labels, hashes)\n",
    "display_sample(dedup_valid_dataset, dedup_valid_labels, 'dedup valid')\n",
    "dedup_test_dataset, dedup_test_labels, hashes = deduplicate(test_dataset, test_labels, hashes)\n",
    "display_sample(dedup_test_dataset, dedup_test_labels, 'dedup test')\n",
    "\n",
    "#save dedup pickle\n",
    "save_pickle('notMNISTdedup', dedup_train_dataset, dedup_train_labels, dedup_valid_dataset,\n",
    "            dedup_valid_labels, dedup_test_dataset, dedup_test_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "L8oww1s4JMQx"
   },
   "source": [
    "---\n",
    "Problem 6\n",
    "---------\n",
    "\n",
    "Let's get an idea of what an off-the-shelf classifier can give you on this data. It's always good to check that there is something to learn, and that it's a problem that is not so trivial that a canned solution solves it.\n",
    "\n",
    "Train a simple model on this data using 50, 100, 1000 and 5000 training samples. Hint: you can use the LogisticRegression model from sklearn.linear_model.\n",
    "\n",
    "Optional question: train an off-the-shelf model on all the data!\n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#train model\n",
    "\n",
    "def train_models(models, dataset, labels):\n",
    "    #reshape train dataset so 2d image is an 1d list of features\n",
    "    dataset_f = dataset.reshape((dataset.shape[0],-1))\n",
    "    #print(train_dataset[0])\n",
    "    #train_dataset.reshape((train_dataset_shape[0],-1))[0]\n",
    "    for model in models:\n",
    "        print('training ' + model[0])\n",
    "        model[1].fit(dataset_f, labels)\n",
    "    \n",
    "dedup_models = [\n",
    "        ('lbfgs multi', LogisticRegression(solver='lbfgs', multi_class='multinomial', max_iter=1000, verbose=1, n_jobs=1)),\n",
    "        #('newton-cg multi', LogisticRegression(solver='newton-cg', multi_class='multinomial', max_iter=300, verbose=1, n_jobs=1)),\n",
    "        ('liblinear l2', LogisticRegression(verbose=1, n_jobs=-1)),\n",
    "        ('liblinear l1', LogisticRegression(penalty='l1', verbose=1, n_jobs=-1)),                   \n",
    "        ('sag l2', LogisticRegression(solver='sag', max_iter=300, verbose=1, n_jobs=-1))\n",
    "         ]\n",
    "models = [\n",
    "        ('lbfgs multi', LogisticRegression(solver='lbfgs', multi_class='multinomial', max_iter=1000, verbose=1, n_jobs=1)),\n",
    "        #('newton-cg multi', LogisticRegression(solver='newton-cg', multi_class='multinomial', max_iter=300, verbose=1, n_jobs=1)),\n",
    "        ('liblinear l2', LogisticRegression(verbose=1, n_jobs=-1)),\n",
    "        ('liblinear l1', LogisticRegression(penalty='l1', verbose=1, n_jobs=-1)),                   \n",
    "        ('sag l2', LogisticRegression(solver='sag', max_iter=300, verbose=1, n_jobs=-1))\n",
    "         ]\n",
    "train_models(dedup_models, dedup_train_dataset, dedup_train_labels)\n",
    "train_models(models, train_dataset, train_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "scoring newton-cg multi\n",
      "0.8242\n"
     ]
    }
   ],
   "source": [
    "#check against test data\n",
    "def score_models(models, dataset, labels):\n",
    "    dataset_f = dataset.reshape((dataset.shape[0],-1))\n",
    "    for model in models:\n",
    "        print('scoring ' + model[0])\n",
    "        print(model[1].score(dataset_f, labels))\n",
    "\n",
    "#score_models(dedup_models, dedup_test_dataset, dedup_test_labels)\n",
    "#score_models(dedup_models, dedup_valid_dataset, dedup_valid_labels)\n",
    "\n",
    "score_models(models3, valid_dataset, valid_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU0AAAESCAYAAAB5KIfmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEkkI/H9b3yUPmKtgSgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdc6afea050>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU0AAAESCAYAAAB5KIfmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzsvXl4VMeV8P073a2WWvvOIoE2NiFAIHYDtsE22GCb2MRk\nHE8Sx/bE+SaL32TszExiv0m+yZc48eTNZCbzJsGZOHEcEpzYsbHBGPCCAbMLgxA7aGcRktDeavVS\n3x/SvbRWtK/1e55+1H1v3aq6R1WnTp3aRCmFRqPRaLqGZbAzoNFoNMMJrTQ1Go2mG2ilqdFoNN1A\nK02NRqPpBlppajQaTTfQSlOj0Wi6gVaamjaISE3z33Ei8upNwj4lIkF+v98WkfB+ytdtIvJW8/f7\nRORbzd9fEpEHuxp+oPDPl4h8RUTOiYhXRKIHMh+avkUrzVGCiHTnf60AlFKXlVLrbxL2fwHB5oNK\n3auUqu5BFrubt7eUUj/ph/A95iYy3gPcART0Zx40/Y9WmsMcEUkSkVMi8oqInBSRVw3LT0TyROR5\nETkMfFpEUkXkHRE5JCK7RGRKc7hkEflYRI6JyL+1ijun+btFRF4QkRwR+aTZcvoaMB74QETe80sz\nuvn7N5vDHxeRp/ziPCkiG0TkhIhsE5HA5ntfF5Hc5vg33uS9vyAi/+V36a7m9zotIms6C99sAf5c\nRPaKyHl/K1VEnhaRg815+K7f9b81x58jIk/4Xa8RkX8XkaPAoo7yq5Q6ppQqBKSz99IMfWyDnQFN\nnzAV+KJSar+I/A/wj8D/ab5XppSaByAiO4EnlVIXRGQB8EuarJ+fA/+tlPqjiPxjq7iNJWNPAknA\nLKWUEpFIpVSliHwDuF0pdd0/vIhkAV8A5gNW4ICIfAhUApOAzyilviQim4B1wEbgn4FkpZS7i118\n/+VsSUqp+SIyiSYlnnaT8GOVUktEJB3YDLwuIncBk5VSC0REgM0islQptadZvpXNDdIhEXmt+Z1D\ngH1Kqae7kF/NCEBbmiODQqXU/ubvrwBL/e5tAhCREOAW4C/NVtGvgTHNYZYAf27+/ocO0rgD+LVq\nXnerlKpsvi60bz0tBf6mlGpQStUBrwPLmu/lKaVymr8fAZKbvx8DNorII4C30zduy6vN+ToPXACm\n3ST8G83hTwHxzddW0mSxZgPZNDVGk5vv/S8R+QTYDyT6Xfc0v5tmlKAtzZGJv0VV1/zXAlxXSmV1\nEN54ZiC6jy6/717AGEhaA9wK3A98R0RmKKV8XYzT/52l1e+b5UH8/v5IKfWif0ARuQ1YASxUSrlE\n5AO/PDcYDUkP8qkZhmhLc2QwUUQWNn//LLC7dQClVA2QJyKfNq6JyKzmr3uBh5u/P9JBGjuAJ0XE\n2vxsVPP1asC/K20ooN3Ap0QkqNnKfcAvXx0p5olKqV3AvzTHGdpBuPZ4SJpIA1KAM9141sjPu8Bj\nzflFRMaLSBwQQVOD4xKRabT0XXa3kenIMtcME7TSHBmcAb4iIieBSOBXzddbWzWPAI83D3KcoMmi\ng6YR8K+IyDFgXAdp/AYoAo43d+8NJfsisM0YCOLGaPVR4HfAIWAfsEEpdayDfCEiNuCV5jwcAX7e\njVF4BRQCB4EtNPltG28Svs1vpdQOmnyr+0TkOPAXmhT3NiBARHKBHza/T0dxtZuOiHxNRIqABOCY\niGzoyotphh6it4Yb3ohIEvC2UmrmYOdFoxkNaEtzZKBbvgFCRD4Qkcc6uf8r/ylJPYj/CyLSxr2i\nGTr0q9IUkc82z22rEZESEdkiIktu8szp5mkjvUl3VfM8xGoRudpc0O/rTZxDERH5LPBXIEXLd8hw\nD7C1l3GMukZQmub3rmh17aYNiIgsEpG9vUg3SUR80o3FH/2mNEXkmzTNFfwBTVM6JgL/DXRYuUQk\nFbA0Txvpabqfpmn6ye+ABKXUGOB/A/f2NM6hiJbv0ENEZtI0YHRpsPMygrhZA7KGJj92f6bRgn6Z\nctQ8Mfn7wBeUUm/63dpK563wmpvc7wo/Bb6vlHrJuKCU2k07I8rDFS3fIctqel+BNd1jNfD4QCbY\nX5bmYiCQ5gnE3aBXhU5EptI08fi1nsYxTNDyHZp02CiJSESzv7NQmpaRPisiKSIyUUReFJEJA5zX\n4UCnU7NEZCwQr5T6ZIDyA/Sf0oyhafleVycmIyIOYB7wYQf3p0nTDjpXRGSPiDwmInEikiUir/il\nC3C5N5kfBmj5DjFEJIKmFUQfdxDkbpqmUqUCn6JJlntoUrJHlVJFA5HPIc4bIlJhfGhyN3XGapqm\ng7WLiHyxeXrdZRH5vYjcKiKRIvI5Efl6TzPZX0qzHIjtjnOVpmV6Hyul3B3c/yzwAk0bRPwrTQ73\nXOA/aZpDaKQLHc81HClo+Q49VgHvd7Q6SCm1SSn1olLKo5Q6q5T6hlIqQSk1Qyn1fwc4r0OVtUqp\naOND0x4KnbGazt1N99C0NHYyTQs4fkZTmV4B/LGnmeyvZZT7aFqm9im6vi63UwEopf633892fWhK\nqTPNE4jXcWPDipGIlu/QQ/sze0+XV0o1L4a4DXi0ozCttjXc0PzpNf1iaTav5Pgu8N8islZEHCJi\nE5G7ReT5Dh67h74pdP8EPNc8XSFMmlgqIr/ug7iHBFq+g0571uQqOukq9kH8mpYsBY4ppWr7IK5u\nLWvttw07lFL/R0QuA8/StPNODU0+nf+vdVgRyQBqlFLFfZDua9K08/izwH8BTppM8hd6G/dQQst3\n0AjnhpsCABGZDxQopcr6I/5RQncbir6YCdKjtIfEMkoReQaIUUr9y2DnZSSi5ds3NDc+B4Fp/gM3\nzUozWin1bi/jDwLepsk3+sNeZXaEI037AKxTSp0e6LSHytZweTRtBKvpH7R8e0mz2+MR4FutR7qV\nUof6IP4ZNPmR3wN+0dv4RjIiEgD8fjAUJgwRS1Oj0WiGC3rDDo1Go+kG/d49FxHjzBgAAgICiIuL\nY9asWcyfP5+MjAySk5OJi4sjPDyc0NBQAgIC6MgCVkrhdDqpqamhvLycoqIicnNz2bt3L0eOHKG0\ntJTGxkYzPaUUSqkRvemrIePWWCwW4/2x2WxMmjSJlStXsmLFCsaNG0dISAg+n4+qqiouX75MYWEh\n5eXl1NbW4na7CQgIICgoiODgYKKjo0lISGD8+PGEhYUhIlRWVnL+/Hn279/PL3/5yxEvY6NMTZo0\niSeffJInnniC8PBwlFKIiFnm+isL/Rn5YNNRGQYICgpi7Nix3HfffTzyyCMsWLCghcx9Ph8iwtWr\nV3nnnXd4/vnncTqdLFy4kK997WssW7bMSAPADH/27Fk2bdrEpk2bcLvdnD17tksyHhCfpoiglGLa\ntGmsXr2a5cuXM3bsWGJjY4mKiiI4OBir1WqG9f/bHqGhoYSGhhIfH09qaiqzZs3ijjvu4NKlS+zd\nu5etW7dy4sQJUzijEUPmFouFZcuWsWrVKrKysoiOjqa6upoPPviAEydOUF5eTl1dHbW1tVRXV+Ny\nuXC73fh8PiwWC1arFZvNRlBQECEhIYSFhRESEkJ8fDyLFi1i+fLlLFy48OYZGgEopQgMDOTuu+9m\nzZo1hIWFmZXXuN8eo7UM9hUul4v8/HwAoqPbHhlvyLe4uJht27Zx9uxZABYvXkxAQABAi/+TQUFB\nAYcPH+bkyZOm/ukKA2FpApCcnMynPvUpPv/5zzN58mSsVqtZyPwLndvtpq6ujqtXr1JVVYXH4zHj\nslgsjB8/ntjYWIKCgrBarQQHB+NwOBg/fjwzZ85k6tSpZGRksHPnTnbt2mUKe7Tgb9GPHz+e9evX\nc+uttxIdHU1eXh5btmwhLy+PM2fOUFRURENDg/lc60rfXmU3FHFwcDC1tbVkZWUxf/78/n+xQcaQ\nz8qVK7n//vtJTU3VynCAUErhcDhISEhgzJgxbe4BNDY2kpeXx/79+83/1fTp01uE99c3Pp+PvLw8\nzp071+JeVxgQSzMwMJA777yTBx54gGnT2h4SWFtbS1FREefOnePSpUuUlpZy5coVqqqqcLvdZuE0\nlObEiRNJTk4mJSWFxMREIiMjERGsVitpaWkkJCQwffp00tLS2Latr+YbDx8iIyPJzMxk7dq1rFmz\nhrNnz7J582Z27drF0aNHcbmazhRr3WgZ3Xn/e/4Y3SGLxUJdXR2lpaVUVFQM0FsNPpMnT+ahhx5i\nzpw5WK1WfL6bL/0fgG77iMZQgIZryOFwmNfhhgV57do1zpw5Q3Fx01TkwMBA0tLS2rVMAZxOJwUF\nBWb4IaU0lVLMmTOHtWvXkpmZaV4zXrqwsJDs7Gx27drFxx9/TF5eHuXlTXN7Wxc2o4WIiIhg8uTJ\nZGVlMW/ePObOncu0adMIDAwEmnwgc+bMYcKECe0q6ZGKiBAdHc3ixYv53Oc+x+23387777/Pr371\nKw4fPkx9fb0ZzsDwefpjuDX8fUCGhemfltFQjQaMiltXV8eBAwc6rWQigtfrJTg4mGnTpjF+/PgW\n9zRdx1CaU6ZMISEhAZvNZl6HG8ouLy+vhUsuLi6O8ePHExwc3Ca8oWQvX75MbW1ti3i6Qr8rzdjY\nWL70pS+xcOFC7HZ7C4VZVVXFn/70J/7whz9w+vTpFhm3WCxtWmnj2aqqKg4fPsyRI0d47bXXuPXW\nW3nyySdJT08nOjqawMBArFYrMTExPPDAA/39ikMCo8u8dOlSHn/8cbKysvjb3/7Gd7/7XUpLS81w\nrZWkoQgDAwOJiIgwB+KsVqvZCDU2NuLxePB6vdTX11NfX9+jwjbcOXr0KCdPngTo1F9utVpxOp2k\npqbyne98hwcffND0EWt6xqRJk4iPbzqe3r/MGQ3UqVOnOHr0KNBUpqdNm0ZsbKzp02ztfjp//jxX\nrlxpMZjUVfpdaT7xxBOsWLGC2NhY4EbmlVJs376djRs3cvr06TaZ7+wl/JXp9evXeeONN9i2bRuP\nP/449957L5MnTyYkJMQMFxcX19+vOagYlXHBggU8+uijzJ07lzfffJN//Md/NGXQnkVpWIphYWHM\nmDGD++67j6VLlxITE4PVam3xnMfjoaamhuPHj/PRRx+xf/9+goKCRpXlVFVV1aVwhvXtdrtN617T\nM4zGKSMjo4XFDjeMqOvXr3Pq1CnOnz9vDl7OmzePiIiIDhv1CxcucOXKlR7lqd+V5jPPPEN4eHiL\na0opamtr+eMf/0hhYaF5ratWS+uwIkJDQwMvvvgiv//97wkODjYrtFJqxA8G+Xw+EhISePTRR1m0\naBHvvPMOzz77bIv77aGUYs2aNTz22GMsXLiQ6upqsrOz+dvf/kZxcTHl5eWmO2TcuHGkpaWRlZXF\nihUrqKuro6amhpiYmHbjHs34NySjqVHpL8LDwxk7diwhISFAW5nm5uZy8eJFAFNpLliwgIiIiDZx\niQhut5vc3FyKi4t71FPqd6VpDNLADbPa5XJx/PhxTpw4gdPp7LZ53BojXrfbjcvlora2dtR1hR59\n9FEWLlzIxx9/zIsvvkhlZSUWi6VDudpsNr7//e+zevVqysvLef7559m5cyc1NTXU1dWZXXJospxa\nz9lcvHgxq1evJiMjYyBfc1jgXxFHk/uirzEGJmfPns3YsWPNrraBoVeOHDnC2bNnTas0PDyc5ORk\nHA5Hu41WTU0NBQUFXL9+Hej+/2jAphz5d/W8Xi/V1dU4nc5eKcvWGHEZaYwWkpOTWb58OV6vl717\n95KTkwN0bGFGRETwzDPPsG7dOvbu3csbb7zBkSNHWnRX2uvK+1vuxcXFHD16lHnz5rF27VpzArFG\n05cYXXPD+DK65P5d82PHjlFUVISIYLfbSU1NJT4+3hw0MjCeOX/+PBUVFaaSHXJKsyN0C9x3rFy5\nkpSUFPbt28e+ffs6tLRFhKioKO68804eeeQRsrOz2bRpE/v27aOurq5DC6m1T9Tj8XDp0iWuXr3K\nmTNnOHnypFaamj7HmLGRmZnZoX8yNzeXgoICnE6nORiamZlJcHBwG4PNICcnx7Qyh6TS1D6d/ufe\ne+8lODiY7Oxszpw5A7RvKQIkJiby+c9/ntDQUDZt2sThw4epq6tr9xmDjuZser1erly5wjvvvNOX\nr6PRmAQFBZGenm4u3W3dsH/00UeUlJSYPaHQ0FAWLVqE3W5vEc7A5/Nx7NixHnfNYehsDafpBXPn\nzsXtdlNSUmKO8LZnKQYFBZGcnMzixYs5evQoR44c6XHhaT0xXqPpS0SEgIAAJkyYwNixY83pb/7l\nzel0cvToUerq6ggJCcFisTBmzBiysrJaKE1/nE4nubm5VFZW9ri3q5XmCCA+Pp6TJ09SXV1tzgf0\n92caBS0mJobJkydjsVjYvn071dXVvU5bu1k0fY1hNTocDhYtWkREREQbd5NSisrKSqKiopg9e7a5\nUm3KlCkkJiaaU+b88Xq9XL16laKiIpxOZ4/zp5XmCMBms5mbbXSGw+EgKioKr9fLpUuXcLs7OphS\noxl8HA4HmZmZbRbFQJPSdLlcfPnLXzYb7oCAAGJiYoiMjDTD+fvjXS4Xx44do6GhoUe+TAOtNEcI\nDQ0NN50xoJQyLdDRNiVLM3wwlFxwcDC33HILQUFB5nVDcZaVlfHkk09y/Phx3G43SikmTZrEY489\nZirS1pZmQ0MD+/fvN6c5GnF2F600RwBKKSIiIky/T3v3lVJUVVVRUlKCzWYjLS2tw/DdRfs1NX2N\n1WolIiKC1NRUAgIC2liM+fn5ZGdnt/Dhh4SEMHPmzBbx+CvF2tpaDh48iNPp7JVbSZsbI4Dr16+b\n+5K2t7LKaHWrq6u5cOEC169fZ9WqVe2umOguvenmjGQGaGPiEYchs4iICKZPn96uP/P69evs3buX\n2tpas7xbLBaio6NJTk5uN16Px8P169fJycmhoaGhWysQW6OV5gjg2LFj5sbAhuJsr7K63W7y8vJ4\n6623mDp1KitXrmTcuHHAjQ1SuoJRsI0RztZ7HI52jMposVh6bc37fL6b+qpHEkYZjIqKIiMjw5yg\n7l+mS0tL2blzJ16v12y0w8PDSUxMNJf1ti7L9fX15OfnU1lZ2esFNf2uNHuj0TVdY+vWrTQ2NjJr\n1ixSU1OBtoXG+B9cuXKFP//5z+Tk5PDggw+yZs0aEhMTzXD+CrG9jz+RkZEsXryYr3zlK/38hsML\nQ9bBwcE93izGqDc+n8/cUWo0YMguKiqKGTNmtLgGTUuwi4qKOHLkSAvlN2bMGFJTU03/p398hmvq\nxIkTbbY47AmDZmnqbkvf8fbbb1NQUMD8+fNZsGBBh3PUlFLU19fzySef8ItfNJ0Su27dOv7u7/6O\nOXPmEBUV1WL+ZUefmJgY5syZwwMPPMBXvvIVnnnmmQF716GMf8MSGhpKSkpKh41YV2lsbDQ3oxgN\nGOUvJiaG9PT0NvevXbvGyZMnuXbtWgvX0IQJE5g8eXKLOOCG3K9fv84nn3zS5npPGJBNiP1fzn87\nMofDgcVi6bN14sb8RP8KPhqs3NOnT/PRRx/x8MMPs2zZMvbu3cuZM2fafX+lmg6me/XVV2lsbOSL\nX/wiTzzxBEuXLuUvf/kL2dnZuFwuXC4XHo/HlKfNZsPhcBAYGMiCBQtYtWoVCxYsYMyYMX02oDSc\n8VeYwcHBZGVlcffddzNhwgTzflcx/meGhbR9+/ZRdaSIw+EgPj7e3ArOX3b5+fns27evzbNJSUlM\nnTq13Xh9Ph9lZWWmpQm9m1/c70qzqqqKsLCwFpNNAwMDmTVrFhkZGZSXl1NTU9MrBWfEa7PZsNvt\nOByOFlvDjQZefPFFJk2axO233861a9d47rnnzFM521OcFouFN954g5ycHP7+7/+e9evX8x//8R84\nnU7z7JRr167hcrmw2WzExsaSnp5OUlISDoeDEydOsG/fPqZOncrcuXMH6a0Hh9buCv+GxW63c9tt\nt/GlL32Je+65p81zndG6Qns8HgoLC/nrX//Kd77znX54k6GFIZ/x48czbdo082gLA4/Hw4ULFzh4\n8KB5zefzERAQQGxsbIs9e/1xuVxcvXq1R+cBtUe/K80XXniBL3/5y2aLC00vFRYWxiOPPEJeXh4n\nT55sswlxZ7T2rxlLBB977DHWrFnTZhPikY7FYqG4uJhXXnmF8PBw7r//fsrKyvjJT36CzWbD5/O1\nkatR0fPz8/nJT37Chg0bmDp1Kvfccw9z585l/vz55tZabrfbPMdp27Zt7N69m7Nnz5KZmTmquub+\n0178K57NZiM1NZU77riD+++/n9mzZxMdHd2jo0D8BzzOnj3Lr371K3PXqpGOIV/DPwktl+uWlpZy\n4cIFrl692qJeT5gwgaSkJMLCwlrEZ8iyoqKCoqKiNr3entLvSvM3v/kNU6dOZfXq1cTGxraYirFy\n5UouXrzIK6+80uXjLozzaox4oqOjWbZsmXncRUxMjHncxWgZhDKWTu7evZugoCAeffRRvvjFLwKw\nYcMGqqqqTOe3v/I05NnQ0EBpaSmVlZWcPn3aPPLCeMbYaq+xsZG6ujrq6upoaGjA5XKNCvkaKKWw\n2+3ExcWRlpZGRkYGM2fOJC0tjXHjxhEVFUVkZKTpdjLorOFubV2KCB6Ph+zsbH7729/y5ptv9un2\niUMZo24nJCS029U+efIkOTk5bY4OSU5OZvz48eb+m61dcyUlJeTm5gL0iTuw35XmtWvX2LBhA7Gx\nsdx5550tlkRFRkby8MMPM2XKFD766CPzYLWKiop2FZ7/9IIpU6YwZ84c82C19PT0Fov6RYTy8nJ2\n7drFpz/96f5+zUHH5/NRU1PDrl27cLvdPPzww/zDP/wDKSkpbNiwgdzcXPNkT/9upSFjr9drKkLj\nTKH2HOrG99GkLKHpJMqHH36YyZMnEx0dTWRkJNHR0URHRxMWFobdbm+3Z9O60ff/21qmPp+PoqIi\n3n77bd566y2OHz/eq40lhhtKKaxWK2PHjm3TM/V6vRw/frxdq3v69Olt5mf6W5VXrlzps645DNDW\ncNnZ2bz55pvEx8czb968FoUlKSnJ9JctX76cy5cvc/XqVfMIX4/HY7681Wpl3LhxTJw4kZSUFFJS\nUpgwYUKL3eGhabnUmTNn2Lp1K++8886oUJrQVCDKy8vZvXs31dXVXL58mTvuuIOoqCj27NnD3r17\nOXnyJI2NjS2e62hKUesw/t/9p8SMBmpqarDb7SxcuJDk5OR2u97tya91499azj6fj7q6Ok6ePMnB\ngwfJycnhwIEDnDt3blRZ8v5d88TExDZH5JSVlXH+/HkuX77c4qhpESEpKanDucKNjY0UFxdz4cIF\noPOzx7rKgCyjdLlc7Nixg9jYWMLCwpg0aVKLQhcaGsr06dNJT0/H4/FQV1fXQmkaGOeex8XFERQU\n1KL7CE3WUkFBAQcPHmTHjh3s2rWLvLy8gXjFIUVlZSUff/wxFy9e5MqVKyxZsoQHHniAzMxMsrOz\nKSws5MKFC+ZuL/4WUEeV1L+QGq14UFBQGz/SSOXq1av87W9/w+VysWbNGjIyMjpdSNAehjXpdDop\nKyujqKiIoqIi8vPzycnJ4eDBg1y6dKnNufSjAaPbnJycTFJSUptd13Nycjh//jwul8tUmkopYmNj\nGT9+vFkO/deUizSdXHvp0iXzWPC+YMCmHOXn5/PGG2/Q2NjI8uXLGTt2LLGxsURGRhIcHGwq0YCA\nACIjI1vsVNJRvF6vF5fLRWVlJaWlpVy+fJm9e/eyZcsW0/cxGgaC/DEqmtvtprCwkBdeeIF9+/ax\natUq5syZQ3p6unmq5IkTJ7h27Rp1dXXU19ebvsrGxka8Xq/pOzLOCHI4HISGhhISEkJISAiLFi0i\nMjKSixcvmo77kYpSiiNHjnD+/Hny8vL4zGc+w9y5cxkzZkybzR+MhsXwA9fX11NZWUl1dTVlZWVc\nvXqV8+fPk5OTQ25uLvn5+Xg8nhYKeDQpTLih7FJTU5kwYUKLxhmaptVVVVUREhKCx+PB7Xbj9XqZ\nPHky48aN63DaW3FxsXmAWmdnZnWHAbE0jZc/ffo0Fy5c4M9//jOZmZnMnz+fGTNmkJSURFxcHOHh\n4YSEhBAQENCpxdPQ0EBNTQ3l5eUUFRWRm5vL3r17OXz4MKWlpeZUm9HoezMwColSit27d7N3716S\nkpJYvnw5d911F8uWLWP16tXmipPS0lJKSkrMKWBOpxOPx2POzwwODiY+Pp7ExERzuabL5eLEiRMc\nOXKEH//4x4P9yv2OSNP6/Y0bN3LmzBkeffRR1q5dS3x8PD6fD7fbTUNDAw0NDdTV1VFZWUlZWRkl\nJSWcOnWKM2fOcOrUKS5fvtxiaaSImBV6tJbX1NRU8+jdiRMntrjn8/kYM2YMCxcuZMyYMVy9epXi\n4mJCQ0NZsmSJ2XD5l3kDw8JMTEzk6tWrfaI0pb//SSKiWv1uM0qe3Lyb+B133MHSpUtJSkoydzaB\nlq1uTU0NR48eZefOnezYsYPjx4+32FDU33Q3UEqNaHOztYxb01omxpSvW265haVLlzJv3jxSU1OJ\niIggODjYtC6NiuzxeMyR82vXrpGTk2Mq4rNnz+L1ekeNjP0b44SEBB5++GGefvppnE4nFy9eJDc3\nlxMnTpCbm8vZs2cpKytrU1FbL77oSkUe6fI9dOiQio+PJzo6ukXP0x+fz0djYyO5ubls2rSJhx56\niNTUVMLDw83ufOuepdPppLS0lC1btvDMM890eg59V2Xc70pTo9FoRhJ6lyONRqPpBlppajQaTTfQ\nSlOj0Wi6gVaaGo1G0w200tRoNJpuoJWmRqPRdAOtNDUajaYbaKWp0Wg03UArTY1Go+kGWmlqNBpN\nN9BKU6PRaLqBVpoajUbTDbTS1Gg0mm6glaZGo9F0A600NRqNphtopanRaDTdQCtNjUaj6QZaaWo0\nGk030EpTo9EMKUTkAxF5rJP7vxKRJ3oR/xdEZHdPn+8XpSki+SJSLyJVIlIhIntE5Em5yXm6IhIg\nItdEJHig0x4JiEieiKy4SZhtInJnL9J4SUT+354+P9zRMh4S3ANs7WUcPT4crb8sTQWsUUpFAEnA\n88A/A/8knal4AAAgAElEQVRzk+duBY4qpTo+Mq7/0h7xNDdGc4Fdg52XkYqWcf8iIjOB60qpS4OV\nh/7snguAUqpGKfU28BngCyIyvZNnVtP7FqSnaY8G7gD2KqXcg52REYyWcf+yGtgymBkYMJ+mUuoQ\nUAws6yRYvwiki2mPBga9wI0CtIz7lzV0YFiJSESzv7NQRE6KyLMikiIiE0XkRRGZ0BcZGOiBoEtA\ndHs3RCQVsCqlznVwf7GIvC8iV0TkXRF5SESiReRWEflFb9IeRXRoyYvIeBH5k4iUiMgREXmq+dpU\nEXljgPM5nNEy7idEJAKYCnzcQZC7gSNAKvApIAbYQ9P/46hSqqgv8jHQSjMBqOjg3mrgnU6e/Tvg\nG81x/Bx4DDgNfAd4sZdpj3hEZAZQqZQq6SDIQ8CrQCLwBDALOApsBF4ZkEwOc7SM+51VwPtKqXYH\ncZRSm5RSLyqlPEqps0qpbyilEpRSM5RS/7evMmHrq4huhojMB8bTpPnbYzVNyrBdlFJP+f3cSjd8\nn11IezTQqb9YKeUv+6PA4/2eo5GHlnH/MiRcH/1uaYpImIjcC/wJ+INSKredMA5gPvDBQKc9ihgS\nBW6Eo2Xcd7RnTa4CtvVj/F2iPy3Nt0TEA/iAk8C/A7/uIOwKYJ9SqnEQ0h5JtFsQmn1B6XTsC+p1\nGqMILeP+Jxwo97/Q3FssUEqV9Uf83aFflKZSKqWbj/TVVKOepD1iUEqldnBrJfCBUsrXB2l0uFJj\nNKBl3L+ISAYwjSb3RWu+2wfxBwFrgfd7GsdQWUZ5FPjbYGdiBHMd+NlgZ2KEo2XcS0TkeZq6399q\nPdKtlDqklHq3l/HPAC4DlUBXZty0H08HA1EajUajaYd+Hz0XEdXqNwEBASQmJvJP//RPLF68mCtX\nrnDu3DmsVivTp08nLCwMh8OB3W7HarUiIogIPp8Pp9PJ9evXOXPmDHv27GHPnj2cP38eAIvFgs/X\ntneklBrR686nTJmivve977F8+XJiYmKw2WyICEoplFL4fD58Ph9erxev12t+9/l8GI2mxWLBYrFg\ntVqxWq3md4vFYsrfWL7v8/lwuVyUlpaya9cufvCDH3Du3LkRLePc3FwFTXKy2+2EhoYSGRmJ3W43\n5dJblFK4XC4qKyuprq7G6/Vy9epVnn/+ebZt2zai5dtaT3QSjqioKKZPn864ceP47Gc/y5IlS4iN\njTXv3wyjzLvdbo4dO8bGjRt5//33OXbsWJdkPGBTjtokbLNx2223MWXKFFJSUggICGD79u3s2rUL\npZRZSY3K63A4iI+PZ9KkScyePZsVK1YwZ84c7rrrLg4ePMjWrVspKCjA7XYz2qzntWvXsnDhQuLi\n4kwlB7SQoT9dkU/rwuf/22KxEBQURGJiIrfccgv33XdfH7zF0OZf/uVfgBuNS0hICPPmzeORRx4x\nKyx0rdL6YzRsIkJpaSl//OMfOXLkCHV1dSilcDqd5OTk9Om7DGeUUlRUVLBv3z5CQkIoKSnB5/Nx\n5513EhYWZsqyK5w8eZJXX32V1157jcuXL3c5DwOuNJVSeDweKioqzMobGhpKVVUV7777LsePHzdf\n2rgvIthsNqKiopg4cSIzZ84kKyuLRYsWcffdd5OVlcXkyZPZsmULBw8ebBH3aGD27NmEh4e3UI6t\n37831lB7slRKYbFYiIyMZN68eT2Oe7iwZcuNmURGeSwsLGTWrFksX768W5W1NcazOTk5vP766xw8\neBCv12veG01luat4vV5qa2v5+OOPWbhwIenp6UybNq3T/4O/HH0+H8eOHeO9996jqKgIq9Xa5bQH\nZSDI5/NRVVVldkEAamtrKStrmk1gVH7/LqHb7aa0tJTDhw/z0ksv8cILL7BhwwaOHDlCYmIiTz75\nJF/96le57bbbiI2N7bMu03CguLiYhoYGs1B0VMn8Lc+bffxlb8Tp/zG6/0Y3faTj76KwWCx4vV5K\nS0s5c+YM0DXr/WacOXOGa9eu4fP5WtQBTfsYMrp48SKXLnVv0yOPx0N+fj55eXkA7br1Oky3Wyn1\nIYbFqZTCZrMRExNDYmKiea+9Supf4YuKinj55Zf54Q9/SHZ2NgD33HOPqTiDg4PbdEtHKq+99hrH\njh3j+vXruN1uPB4PjY2N1NbWUlFRQX19PV6vt1tWi78/1Ol0UlFRQU1NDY2NjXg8HlwuF+Xl5Rw9\nepRNmzb18xsOPoZf2PgYsnG7+24zI4/H0yJu46NpH6Ms19XV0dDQ0K1nvV4vVVVVVFVVtYirKwya\nTxMwKzKA3W4nJCSkw7CtrSiLxYLb7ebjjz/m6aefZuvWrURGRnL77bdTXV3N9evXee+990aF4jx0\n6BDPPvssX/rSl1i+fDkOh4OKigry8vI4e/Ysc+bMISsri7i4OKCl26M1rbs3lZWVfPLJJxw6dIgJ\nEyYwdepUIiMjqaqq4v333+fll1/mxIkTA/OiIxzdDe8ZhqHQHYyGqScyH1Sl6W/5WK1WbLauZ8fn\n8yEieL1e8vLy+PnPf87XvvY14uLiuO222ygpKeHYsWOUl/d44v+w4tSpU3zve9/jxz/+sTnTwOv1\n4nK5SE5O5umnn+bBBx8kICDgpnH5jy6+++67/OxnPyMvLw+73Y7NZjNnKdTX11NTU9Pfr6bRdMpA\n+30HVWm2picjjz6fj+rqajZu3MgDDzxAZGQk4eHhLFiwgHvvvZff/e53/ZPZIYbb7TZ9woa1aFjZ\np06dori4mMbGxi4pTYPGxkYKCwvJzc2lsbFphavROmtfm2a0MiL6roZT9/jx41RVVSEipKSkcMcd\ndxAaGjrY2RsQ/P2/xm/DH+Z0OnE6neagW1fxer00NDTgdDoBWnRnbjbopNGMVEaE0jS6o0eOHDG7\n4xEREaSnpzNr1qxBzt3g0d6AWk+eb/3d/75GM9oY9krTvzLn5ORw/fp1lFJYrVbi4uJYtWrVIOdw\n+KKVokbTlmGvNOFG5T579qypNAHCw8O59dZbBzNrGo1mhDGilGZVVRV1dXV4vV5EhKCgIJKSkgY5\ndxqNZiQxIpSmgcvlor6+3pxwbLVaiYiIGORcaTSakcSIUpperxe3222OElssFgIDAwc5VxqNZiQx\nopQmtBy8EJFuLcTXaDSamzGilKbFYjH3koQmBdrduYkajUbTGSNKaVqtVnOpH2Au9dNoNCOX1jty\n9TcjSmna7XYCAwOxWq0t9u3UaDQjF/+NtwckvQFLqR8xBDZu3DgiIiJMP6bT6TSPwtBoNEOf7ig/\nY/wiKChoQAd8h73S9DfNk5OTCQ8PNzfIraqq4sMPPxzcDGo0mpti1GP/vXO7qkCjo6MJCwvr5xze\nYMjsctRdQbV+FmDGjBlERUUhIng8HkpLS3nvvff6OqujBr2TkWag+PKXv0xcXByHDh3i0KFD5h4S\n/sq0I5KTkxk7duxAZXVwlaZ/pQwICMDhcPQoHmOt+dy5c4mJiQHg+vXrnDhxgtOnT/dJXocjrY9o\n6K4S9G/E2iu8NyvMGk1XWb16NbNmzWL+/Pls376dffv2cfr0aerq6toNb5S9lJQUMjIyiI+PN693\nhd7swTloSlNEsNvtWCwWlFLExMQwffp0Dhw4QFlZWbdeymq1kpCQwIwZMwgPD0cpxcWLF9m5c6e5\nrdlIp/V5Pv4KLzQ0lODg4G5t8gw3Tp10OBw0Nja2mMrlr0y14tT0lpqaGoKDg1m1ahXp6enMnj2b\nHTt2UFhYSFVVFfX19bhcLnMKod1uJyIigvvuu485c+Z0eupDexhHWPeEQVGaFouFsLAwwsPDzYqc\nkpLCunXrqKysZOPGjVRXV9+0MhqKISQkhPvvv5+xY8dis9moqqri8OHDbNu2bdR0Me12O2FhYQQF\nBZk7qxs7t0+dOpWJEydit9u7FWdgYCATJkwgPT2d/Px88xx6I36n00ltba25QbFG01NOnTrF0qVL\niY6OJiUlheTkZB566CEKCwvJzs6muLiY8vJycwphREQEc+bM4fbbbzd7l3BzS9No8BsaGnp8vtOA\nKk1DyQUHB7N06VJiYmJMSzMoKIjMzEy++c1vsnXrVmprazttCfwtqaioKL7+9a8THR0NwO7du9m8\nebO5IfFoIDMzky9+8YssW7YMh8NBdXU1BQUFnDx5kvnz5zN79uwuW5qG9RgQEMCqVasYM2YMhw4d\nYvLkyaSkpBAZGUl1dTXvvfcev//97/UZQZpec+nSJbM3Yyi20NBQpk+fzrRp04C2WxW2PrGzO3W9\ntLTUPFStu72lAVeaALGxsXzjG98gMjISi8VCY2MjFy9e5L333uPtt9/m8uXLna7k8fevTZkyhaee\neorExEQCAgI4dOgQr776Kvv27TM3Jx7pzJ07l+eee46FCxeaswd8Ph9paWnccsstOBwOgoKCzPCd\nFa7W9yIjI1m0aBGZmZkEBgYSEBCA1WrF6/Uyfvx4UlNT+d73vtdfrzaqGC0NfHsUFRWZhpJhSEFL\nV1DrQxJb+9c7o/XJtseOHSM/P79LA02tGRClaQxEKKUYM2YM99xzD3PnziUwMJCCggLeeustduzY\nwblz58yzbFq/hL9laRy7kJGRwSOPPMLatWux2+0cOXKEX/7yl+zcubNDB/JI5NOf/jSZmZlER0e3\nKDz+J3x2d2aCEdZqtRIcHExwcHCL+8axy1lZWaxfv74P3mJo419hDfeEsQKtrwgICGj3zPnRQEFB\nATU1NXi93haDlt1p4NujvVNs8/LyePfddzl9+vTQPY3S0PCxsbEsX76cz33ucwQGBrJ79262bNnC\n9u3bOXv2LC6Xq9OXMO5FR0ezcOFC7r33XvMcoO3bt7Nx40a2b9/O1atXW4Qf6UyYMIGgoKB2C5q/\nDLp6rk97hbH1NeM00ODgYCZMmNDjvA8X/I899vl82Gw2xowZw9SpU83rvWXKlCnExcWRl5dn9rRG\nSxkuKSmhuroaj8dDQEBAp1ZkR0dP+39vL4zL5eLs2bP84Q9/YPfu3VRWVrZ5tisMiNKMjY0lJSWF\n+fPnc9ddd5GQkMDmzZv5y1/+wgcffMC1a9datK7+L2FYS2FhYURGRjJhwgQmT57MXXfdxdSpU6mq\nqmLTpk1s3ryZDz/8kLq6ulFT0AxycnJYvHhxC0vT3zfUW1pbPka8Simqq6vJycnpdRpDndWrVwM3\nek1hYWEsWLCAmTNnmtd7ivHsrFmzeOCBB5g4caK5mfZooba2ltLSUpxOZ4+nHraHx+Ohurqa4uJi\nTp06xa5du3jttdd6dbS39LeCiY2NVXfeeSd33303M2bMwOPxsHfvXjZs2EB+fr45gtVePkJCQkhL\nS2PatGlMmDCB5ORkFixYQGRkJPX19Zw6dYodO3awZcsWSktLO/RNKKVGdF8nPT1dfe973+P2228n\nOjoaq9VqysL4+Hw+8+P1es1r/t0WQyG094Ebldvn8+FyuSgrK2PXrl384Ac/4MyZMyNaxrm5uQqa\n5BQQEEB4eDgRERHY7fY+60orpXC5XFRWVlJTU4Pb7Tb/PxkZGSNaviKivvKVr7BmzRqSkpIIDg42\nffGGH90o1+014kZ5NvbUra+vp7a2lrKyMs6ePcvevXvZsWMHFy9ebNc4a/7dJRn3u9L81Kc+pb71\nrW8RExPDRx99xMaNG82ljVartUXFNTPVXOEzMjL48pe/zLp167BarXg8HkpKSti5cydbtmzh2LFj\n1NbWAjf8TO0x0pWmiKglS5bw1a9+lbvuuovIyEigqZX1er00Njaax/jW19dTV1dHQ0MDLpcLt9uN\nxWLBbrdjt9sJCgoiJCSkRaE1phoZhbahoYHCwkI2b97Mb37zG86dOzfiZQy0qSj+XfY+SaDz+Ea0\nfEVE2e12xo4dy7Rp05g5cybTpk0jNTWV+Ph4IiIiCAkJISgoCJvN1sIw8Pl8ZhmvrKzk2rVrHD9+\nnMOHD7N//34uXLhgztfuSGE2XxsaSlOj0WhGEsN+ww6NRqMZSLTS1Gg0mm6glaZGo9F0A600NRqN\nphtopanRaDTdQCtNjUaj6QZaaWo0Gk030EpTo9FouoFWmhqNRtMNtNLUaDSabqCVpkaj0XQDrTQ1\nGo2mG2ilqdFoNN1AK02NRqPpBlppajQaTTfQSlOj0Wi6gVaaGo1G0w200tRoNJpu0G9KU0TyRaRe\nRKpFpFxE3hKRhJs8EyAi10QkuLNwfZ3ucEXLeGAQkQ9FpEJEAroQttfy7Um6w5lW5amm+e9/3uSZ\nbSJyZy/TXSAiW0TkuoiUich+EXn0Zs/1p6WpgDVKqXBgHFAK/NdNnrkVOKqUqh/gdIcrWsb9jIgk\nAQtoesf7u/BIX8i3J+kOZ8zypJQKa/779Y4CNzdIc4FdPU1QRBYD7wEfAGlKqVjg/wFW3ezZ/u6e\nC4BSqhH4KzD9JuFXA1sHId3hjJZx//J5YAfwMvBoF8L3lXy7m+5wpzunbd4B7FVKuXuR3k+Al5RS\n/66UqgBQSh1VSj18swcHxKfZ3DJ8Bth3k6CrgS2DkO6wR8u43/g8sAn4C7BKROJuEr6v5NvddEcT\nvZKxiDiAxcBrPYrAOGi9rz9AHlANVACNQDGQ0Un4VOBsJ/cXA+8DV4B3gYeAaJq6Q7/oabrD+aNl\n3O/yXQrUA6HNv48CTw2AfLuV7nD/tCpP15v/Pt5J+AIgoYN744E/ASXAEeCp5mtTgTf8wviAKT3K\nbz8LYnnzdwEeAMqB+A7CfxX4eSfx/RzIBKw0tTTv0OTveRfI7Gm6w/mjZdzv8t1gVLTm398GsjsJ\n31fy7Va6w/3jX566EHYGcKyT+081l0cB5gD/A1xtVqCfbg7jADzAbT3Kbz8LYkWra6XAgx2E3wqs\nGuh0h/NHy7hfZRsEVNJkAV1u/pQDXmBmf8m3J+kO90975amTsN8CftQHae4G/rMnzw7YPE0RWQtE\nAqfauecA5tM0kjVg6Y40tIz7lAdoskbSabIOM5u/7wa+0DpwH8q3W+mOQvrKZ/wt4FER+ScRiQYQ\nkUwR+dNNn+zn1qOOphazCjgO/F0HYdcAmwc63eH+0TLuV9m+A/yknesPAZcAS3/It7vpjoRPq/Jk\nfF5rJ1wETV3tPpEBMI+m3sF1oIymwcy/v9lz0vzwoCIi/w3kKKV+Ndh5GaloGfcvWr79j4g8BKxT\nSv3dYObDNpiJ+3EUeGuwMzHC0TLuX7R8+5/rwM8GOxNDwtLUaDSa4YLesEOj0Wi6Qb93z1NSUlRE\nRATLli1j1aqmZZ0vvfQShw4dwmKxoJQiJCSEtWvXsn79eo4ePcqGDRsoKyvD5/MRGBhIeHg448eP\nZ9KkSZSUlHDhwgXi4+MJCgri/PnzfPLJJ53mQSnVnSVaw47Y2Fj105/+lBUrVhAZGdnkrJaOX9no\nXYgISil8Ph8ej4fGxkacTic1NTWUlpZy4cIF9u/fz549e7h48aIZZ3u9k5EuYxHpcpfMkNPYsWP5\nzGc+w/e//31+//vf86Mf/YgrV64A7cuwM0azfO12OykpKaSlpZGWlsakSZMYN24c0dHRhISE4PP5\nqK+vp76+Ho/Hg9frNcu08dvr9WKxWHA4HAQGBhIWFkZ0dDQNDQ0cPXqUXbt28de//rVLMu53pZmf\nn4+IcO3aNcrLy7nrrrtYt24deXl5nDx5EpfLxcMPP8yKFSu4dOkSv/jFLzh+/Dg+n8+Mw2q1cv78\nefLz88nMzCQgIICCggLGjRvHnDlzSEhIYOvWrd0uiCOF8vJyLl68yNKlSwkJCWlXaYoILpeLa9eu\nUVRUhNVqxeFwEBUVRXR0NOHh4aYSBWhsbCQzM5MFCxawcuVKDh48yF//+ldKS0vxer2jVtZdRSlF\nYmIiq1evJiQkhDvvvJOXX36ZsrIy3O7eLJkeHdjtdsaNG8ftt99OQkICEyZMICUlhcTERKKioggJ\nCUFEOHDgAGVlZeYzQUFBOBwOHA4Hdrsdu92OzWZDRGhsbKSqqorCwkJOnz5NQUEBxcXFFBcXm41Z\nVxiwgaBLly7x0UcfERMTw5o1a2hsbDQr3qxZs8jIyODdd9/l6NGjWK1W8zmlFB6Ph4qKCqqrqwkJ\nCWH8+PFUVFRw+vRpQkJCmDdvHm63m71791JXVzdQrzSkqKqqwuVytXvPkHN5eTm7du3irbfeIiAg\nALvdTkREBBMmTGD27NlkZWWZyjMwMJD4+Hji4uJIT09nzpw5jB8/no0bN3Lu3DmcTudAvt6wQilF\nREQEGRkZZGVlISKkpqaycOFCioqKKC0tNcNp2icxMZH77ruPjIwMamtrmTFjBtOnTycyMhKfz0dZ\nWRl//vOf+fDDDykvL8dqtRIcHNxCaQYGBhIYGEhAQAAWiwWXy0VlZSVFRUXk5+dz+fJl6urqWhgL\nXWFAlKaRocrKSk6fPs24ceM4efKked/ophstcHsvISJ4PB4OHTrEQw89RFJSEtnZ2Rw8eJD4+Hge\ne+wxrly5wpkzZzpUHiMZq9WKiJgWpsVyw13t9XqBJsWanZ3Nq6++it1ux+Px4PP5iIuL45ZbbmH9\n+vXcfffdhIeHt3g+KCiIadOmkZqaCsBvf/tbzp8/36I3oGnCkH9ycjK33HKLWcntdjt33303Bw4c\n4Nq1a1phdoKIEBwcTGRkJJ988gm1tbXMnDkTq9WKUopr166xefNmfvjDH1JRUWGW756m5f+3Kwzo\nQJCI4PV6qampQSmFxWJpkdmb+MxME/vcuXNERkaSnJzM5cuX2bFjB8nJydx6663ExcW1UB6alhiy\nMWRvuE7efPNNnnvuOT744AOz0THuG41YYGAgTzzxBPPnzyc4ONgMo2mJ1WplxowZzJ07l5KSElOe\nS5YsYdKkSQQFBWml2QlKKU6dOsULL7zAX/7yF6ZOncqUKVMIDw+nvr6eAwcO8G//9m+UlZWZeqS7\nH39d4zfZvUsM+Oh5e5nzr5hdeTYvLw+v10tERAQ+n4/q6mr279/PPffcw8SJE/sl3yMBY9DHcHkY\nvkmLxYLVaqW4uJhnn32Wq1evtvhfGErW5/MRHR3NLbfcQlpamm6cWmGU4ejoaKZNm0ZDQwP/8R//\nQWlpKT6fj6ioKG699VZSU1NNmWrax+v1EhwczBe+8AUefPBB4uPjATh+/Dh//OMfKSkpMctzTz69\nabSG5X+toqICl8tl+iwaGxvJzc0lPj7eHAjRLXnXMQqS1+ulpKSEEydOUFtb26EcExISiI6OHoSc\nDm0MC2bJkiXMnz+f/Px8tmzZws6dO/F4PAAsX76c9PR0s8Jr2icsLIx58+bxz//8z6SmpmK326mu\nrubAgQO8++672Gy2Dq1If4zGqaP7PWFYKk2fz0dwcLA5gqaUoqGhgdDQUOx2+2Bnb1jjdrspLS2l\nsbGxzb3WXRpNS3w+HzabjczMTBwOB/v27aOkpIQ33niDuro6fD4fSUlJzJkzh8TEREC7N/zxdwdN\nnjyZxx57jJiYGAICAhARtm3bxuuvv05NTY05rai9jz/tWaO9Zagso+wWoaGhhIWFUVdXh8fjITg4\nmLS0NNNRrOk5IkJISEiLGQytuXr1KlVVVQOYq6GP0Xgbo7wlJSV8+OGHOJ1OPvnkE06ePMncuXNx\nOBwsWLCAjz/+mOLi4m6P3I4Gxo4dy9KlS7n99ttbXK+pqcHhcDBnzhyzZ+TxeFq4j+rr66moqKC2\nthaHw0FERAQxMTGMGTOGhIQExo0bx1tvvcXFixdpaGjoUf6GldI0BJOWlkZYWBgVFRXm9I7bbruN\nsrIyamtrBzubwwajwvqPuMfGxpKSkkJQUFC7lmVlZSU5OTmmT0nTkuXLlzNu3Dh27NjBuXPnUEpR\nUVHBO++8Yw4CTZ8+nTlz5vDBBx9QX9+r89dGFMYClvnz53PPPfcQF3fjhA+lFLNnz8Zut+N2u1tM\nXvdXmnV1dZSWllJZWUloaCgxMTHExcUxduxYYmNj8Xg8fPTRR72y8Iek0mxvYMi4FhAQwNy5c7FY\nLFy+fJmoqCiWLl1KVlYWv/71r7l06VKno/CjGf+BG/+/RqGLjo5m5cqVpKWlERQUZD5nKE2Px8Ou\nXbvYt2+fnmvYCqUUUVFRLFq0iLq6Oo4fP05jYyNWqxW3283mzZtZt24dMTExxMfHM2PGDNLS0jh+\n/LhZ4Uczc+fOZdmyZdx5551kZWUxZswY4Eb5KiwsxGKxMGXKFLO82mw27HY7gYGBOBwOQkJCCAwM\nNMuqMU2xpqaGa9euUVhYyL59+8jLy+uxlQlDTGm2rsjtXR8/fjxTpkzh0KFDXLhwgSVLlvDtb3+b\noqIiXnnlFfLz83VF7oTW8zhFBLvdTmRkJLfccgvf/e53iYqKAm4UWJ/PR0NDA+fPn+dHP/oRubm5\ng5P5IYqh9JYtW0ZycjI7d+7k4MGDKKXMOYQnTpwgJyeH5ORkoqOjmTFjBsuXLycnJ2eQcz80+OlP\nf0paWhqBgYEtejmG8fTyyy+zefNmSktLsdlsBAcHExoaalqRaWlpTJ48mYkTJxIVFWUOHF24cIGD\nBw/y4YcfcujQoXZ99d1lSClN/0rq3/IqpbDZbIwbN47169fzzjvvUFRUxPr16/mHf/gHLl++zJNP\nPklBQYE5Sqlpi38321iLGxUVxYIFC1i3bh333nsvsbGxbbouNTU17Nq1i2eeeYbCwkK9DLADHnro\nIRwOB2fOnOHKlStmhTcWb2zdupXZs2cTHR1NSkoKCxcuJDQ0lJqamsHO+qCzbds2ysvLyc3NZc2a\nNXzjG9/A4XCY5fXAgQNkZ2e32DfB/2OMjNtsNhwOBzabDafTicvlarP+vLdW/ZBQml6vF4fDwfz5\n83nuuefMRfiGsoyIiCA+Pp7AwEBqamrIzMxk7Nix2O12du7cyW9/+1vy8/PxeDxmAdXW5g0MqzIh\nIYGHH36YWbNmERQURHBwsOnziYmJISIiwlSYTqeTgoIC9uzZw/bt28nOzqakpKRXqy9GIoa8kpKS\nyE2JzyEAACAASURBVMjI4PDhw+Tm5rYof8b3Dz74gIceeoipU6cSGBjI5MmTWbFiBW+++eao76K/\n9NJLuN1uIiIizLXiBk6nE7fb3Uam/v54Q184nU5zKbX/fX+DrLcMCaVprBSqqKjg+PHjxMTEtBCK\nMegTHR1t7nISFhZGVVUVbrebp556itOnT7Nlyxby8/NxuVx6VLIdQkNDmT59OikpKVitVnP9uc3W\nshg0NjZSWFjIoUOH2L9/P7m5uRQXF7ewMLV8b1g7drud9evXExUVxd69e80BIP8KKyKUl5dz6NAh\nMjMzza7k3XffzZtvvjnqZWn4yNetW8cdd9xBQECAubT6d7/7XYduN0O2/vfaU4x9Kd8hozQ9Hg+X\nLl1i9+7dhIeHt7BobDabOc1oypQppKenk5WVhcViobi4mKioKNavX8+0adPYt28fu3fv5sKFC6O+\nILZGRExlaVifhpyNtevQZJmGhoaak4pnzpzJlStXKCoq4ujRoxQUFOB0Oke94jS63qGhoaxdu5ai\noiJOnTpFdXV1u7Lx+Xzs2bOHpUuXkpqaSlRUFHPnzmX69OmcPn3ajHM0k5mZSXp6egufe1RUFAEB\nAR0+05nM+kOeQ0Zp+nw+nE6nuZtRe9uPiQj79u0jNjaWVatWMW3aNMrLyykuLua2225j9uzZTJky\nhYkTJ/Lmm29y/Phx3Z3kRmtcVVXFyZMnOXbsmNkFCggIICQkhMjISOLj482ttxISEkhISACaFGtt\nbS15eXl8+OGHHDx4kOzsbAoKClrsVjXaMHaDmjVrFunp6fziF78w5162h8Vi4cSJE+Tm5rJkyRKi\no6PN3XwuXLjQpgs6mjDe22azmXOERQSbzcaaNWv47W9/O2QGIIeE0jToaPS89f2ysjJef/11Vq5c\nybx58zh06BBf+9rXWLNmDU888QSPPPIIiYmJ/PjHP+bcuXMD+g5DEaNAXr58mddff52f/exn2Gw2\nvF4vdrud8PBwEhISmDVrFnfccQfz5s0jKSmJoKAgc116REQEs2fPZvbs2RQWFrJp0yb+9Kc/cerU\nqT4ZkRxuGGU0Ojqaz3zmMzQ2NvL++++b+zK2V4atViu1tbUcO3aMs2fPsnDhQiIjI1m7di0vvfQS\n5eXlg/EqQ4rq6mrq6uparOzz7xkNBYZOTmg7eu6/A4n/cigRwel0smfPHo4cOcKDDz6IzWbj7bff\n5pvf/Cbvvfceq1at4l//9V/b+OtGM+1N4zI2Jv7kk094+eWXefzxx3n66afJyclpMRPB//8wceJE\nvv71r/Ptb3+btLS0UWkdGbKIjY3l/vvv58CBA5SUlAA3Knl7O+vYbDbTheTxeMzJ7vPmzWsx1Wa0\nUlJSwtWrV1tcO3/+/JDaJ3dYahSfz4fFYqGyspJTp06Rlvb/t3fmwVEdaYL/ZVWpVCVRuu8Tge4L\nhMQhARJGmua0ABuH2+dut8fRbbdju6e3d6NjY3p6emIneqfd3TO9Xm9sbEe3J4jBjgavDwawwWBj\nZG7ZIAkQAp3oRqADSaWz6u0f4j1XSSWhEpKqhN4vokJSvZcvMz/l+zLzyy+/XM769eu5ePEiDQ0N\n/PnPf0ar1bJ+/XpeeuklVxfXbXC0+gj27hswtsobGBjIG2+8wZo1a5R7bNN6eHhQUFDAT3/6U374\nwx/OYy1cjyyL0NBQCgoK8PPzw2Aw8NxzzzEyMqJ4cEyG7KQ9OjqKTqfDYDDwwgsvUFFRgdlsXtS2\n4itXrlBZWUlycrLynbvJY0EqTfg2ontHRwcVFRUkJydTVlbG0NAQ169f58svvyQ7O5tXXnnF1UV1\nKyZbgbRd7R0dHeXEiRNs2rSJFStW4Onpqdxra0IJCAggJyeH4uLi+Sm8mxEbG8uTTz6JJElERERM\nWw6SNHYu1sDAAAaDQemAYmJiuHPnzqIMoi1TXl5OZWUlTz75pDJLjI2N5ZlnnqG3t9ctNgMsaKUJ\nKAsUqampigG5v7+fyspKrly5wp49e1xZzAWFrUJtb2+ntbWVvr4+ZWua7YqmJElotVqCg4MpKChw\nVZFdgiRJmEwmkpKSiIuL48CBA3z++efTHhHJHVRmZiavvvoqJpNJGbU2NDQo03x3Gl3NF93d3Vy7\ndo3r16+TmZkJjLnKFRcX09DQQHt7u+Ke5CoWrNKUGRkZobu7G0BxS5AkidbWVsrKynjuuedcWbwF\niyRJDA8P2/lmOnIW9vT0VMKcLQbk+sfFxZGbm8vg4CB//OMfOX/+/LSigNuO1HNzc1m3bh3r169H\np9OxZcsWTp06RUtLy6JUmDDWri5fvsyxY8dITExUYiDExcWxdetWqqurOXz4sEvL6FYLQTPBarUq\ngRHkuHsAZrOZO3fuqNsqZ4jsimQbuMMRVqt10U0nbY+zuHHjBhcuXLCLBj5+i994m7H88/bt23zy\nySeMjo4iSRJZWVmkpqbi4+Nj10EtNm7evMmhQ4coKSlhcHBQkeuaNWvYsGGDy2PmLnilKa9IyrE0\nZQHLx3m606rbQkCegkdFRREdHc2SJUsAexca2xfabDbbHZL3OCNPv319fUlOTsZoNHLs2DElos54\nb4/JPrIXSFtbGydPnuTOnTvKlD8/P5/4+PhFqzBhrCMuLS3lF7/4BdXV1cpsx8/Pj+joaCUCkqtY\n8ErT09OTsLAwfH19GRoaUl7oiIgIVqxYMcF9QcUxsj+mzDPPPEN2drady9b4KaMkSbS1tfHxxx/P\nWzldiew2tG7dOtasWUNDQwMfffQR4PyeZnnrcFNTE5988omiGAoKCkhOTl708ROGhoaora1l3759\ndkevBAcHk5iYOOPnzsaxFy5Rmo5C0js7HZFHPn5+fmRmZiq7KeTvY2NjycjI4NChQ7Nd/AWL7WFe\ntjsvACUStk6n47XXXuP5559n6dKldmnl0ZQ8FW1oaOD48ePcvHlzvqviEuRdatnZ2YSHh1NeXk5n\nZ+eMXkBZjp2dnbz77rv09fVhtVoJDw8nJyeH+Ph4YHEfh9Hd3c17771HeXm54oq1Zs0a3nzzTd5+\n+20iIyOVnW3TVYazceyFS5TmZA1hJj1rSEgIOTk5nD59WjmHZdOmTezatYuenh727dv3qMV9rLDd\nQCD/bjAYWL58Oc8++yxvvfUWr7/+OgkJCXa2I1tlKYTgzp07fPrpp/zbv/3borBpyh1MVlYWmZmZ\ntLW1cerUqUe2mQ8NDXHt2jXKy8sZHBzEw8OD3NxcVq9erSiDxao4R0dHaW1t5e/+7u84ePAgLS0t\nmEwmUlNT2b17N7/73e9ITU3Fw8NjgjIcHzJOo9Hg7e3Nnj17ePHFFykoKJjxNH/OV89t7WDh4eEk\nJCQoDVC2Q8oNT150sI1BOBmSJLF06VKysrKUc1iGh4fZsGEDu3btQqfTsX///kWzjXKqY0nlBuTr\n68uqVavYu3cvXl5e6PV6vL29CQgIIDIykqSkJNLT0zGZTMqocvwuIqvVSk1NDUeOHOHgwYOLRr6y\nLIqKikhJSeH06dNcvnxZufYoz+3p6eHYsWOkpKRgNBpJTEwkLy+Ps2fP0tjYuGiVJoy16zNnzjA6\nOsqtW7fIz88nKyuL0NBQdu7cydDQEOXl5dTV1XH79m1aWlro6OhQdIrt/8bPz4833ngDo9FIS0sL\nf/7znzl69KjTZZo3l6PIyEhyc3NZunQply5dsrt269YtGhoaCA0NZdmyZdTX1z/0yIrw8HBWr15N\nREQEp0+fxmAwkJSURFFREVarlUOHDnHkyJE5r5e7IAdedYQsy8DAQPLz81m6dCleXl4YjUZMJhM+\nPj54eXkpnZmj6DzydsvKykq+/PJLjh49SkVFxaKxu+n1etLT09m8eTORkZHKyZNGo3HGRyfIUacM\nBgNtbW2KXdPX15e8vDyKi4v5+OOP6ezsnM2qLDgkSeL8+fPU1dVRXl5OQUEB6enpLFu2jJ07d7Jp\n0ybq6uqoqamhoaGBxsZGenp66Onpob+/n5GREby8vMjIyKCgoIDe3l6Gh4dnvAo/50ozIiICf39/\nNm3axOrVq2lsbOTKlStERUUpo8mqqiq++eYbCgsLeeGFF3j//ffp7u5mcHBQcceQR0seHh74+fmx\ndu1ali1bpoTjWrVqFVlZWZjNZg4fPszXX3+9aIK6BgYGEhcXp7iqyFPp8ej1eqKioib1q5RHq/Ip\nf4ODg5jNZrq6umhubqasrIwPPviAysrKReeVsHHjRl588UWysrLw8vIiPT2dvXv3UlZWxoULF2Zk\novDw8CA8PJzs7Gzy8/NZsmSJ0tbT09P53ve+h8Fg4OrVq3NQo4WFEIL29naOHDnCqVOnSEpKorCw\nkLVr1xIbG0t0dDSpqakYDAasVis9PT00NjbS3t6O2WzGZDKxfPlybt68SXl5Oe+//z7nzp2bWVnm\neqTwi1/8Qtq2bRvJyclUVlbyxRdfKEdxyg1kcHCQhIQEtmzZgr+/P0eOHOHLL7+kvr5eOeNYDsMV\nEhLCd77zHXx9fTlz5gylpaXK8b0lJSU0NDQoDViumyRJj/X85vnnn5d+/vOfk5ycPKVZY7IVWXkq\nbrFYGBkZobe3l7t371JbW8uVK1c4d+4cly5dUo7tdRQd/3GXcWdnp+Tj42O3eDY6OkppaSl79uxR\n3IamgzzyDwoKYs+ePbz11lvKgsb4XUUWi4Wenh4CAwMfa/kKIR4qPFv52G6uSEpKIi8vj7y8PDIy\nMoiJiWHJkiWK3/bw8DCtra2cP3+e9957j5MnTzo8tXa6bXjOlaaKiorK48SC99NUUVFRmU9Upami\noqLiBKrSVFFRUXECVWmqqKioOIGqNFVUVFScQFWaKioqKk6gKk0VFRUVJ1CVpoqKiooTqEpTRUVF\nxQlUpamioqLiBKrSVFFRUXECVWmqqKioOIGqNFVUVFScQFWaKioqKk6gKk0VFRUVJ1CVpoqKiooT\nqEpTRUVFxQlUpamioqLiBHOqNIUQ9UIIsxDivhCi98HP//mQNJ8KIYoeIc93hBBDD/KS87080+ct\nBIQQp4QQnUIIj2nc6yGE6BBCeM1nvguRce33nhDi34UQkQ9J88jynUm+CxkhxHeFEOeFEH1CiDYh\nxDkhxGsPSTMr7VgI8bwQ4tIDPdEshDgihFg/VZq5HmlKwA5JknwkSTI9+PmfJrv5gQCygS8fMd9/\nepCXnG/WIz7PbRFCxAJrgDtA8TSS5AOXJUkyz3O+CxGl/QLhjNX1rYekmQ35ziTfBYkQ4j8D/wz8\nExAqSVIY8EMg7yGd8SPLWQjxU+D3wH8HQoAY4G3gyanSzcf03JlT9AqBM5IkjcxVYR5DXgY+A/YB\n/3Ea928Hjrog34WKAJAkaRh4H0h9yP2zJV9n811wCCF8gF8Br0mS9KEkSf0AkiSVSZL00kP0wCPJ\n2Sbv1yVJ+liSpAFJkiySJB2VJOnnU6V1N5vmduCIqwuxwHgZ+AtwENgihAh+yP2zJWNn813QPJgF\nPQs87LDsWW3DTuS7EMkF9MChGaR9VDnnAp7AR84mnA+l+dEDu1fXg5+vTHHvpL2HECJCCPHeA7vD\n10KIHz/4LkkIMb7i/2Vcnu/MWm3cCCHEBiASOCRJ0i3gGvD8FPcvA7QP7nV0PVcI8fkDu9IxIcQz\nQogAIUS+EOJ/zTTfBc5HQohOoBsoAn472Y2zJV9n813ABAF3JUmyyl8IIc48eG/ND9rZBGZJzoHj\n85428sHrc/EB6oAnpnlvOlA2xfUfA3sYm7ZkAX8C2oGvgb02970D/MNc1stdPsD/BT6y+fu/Ad9M\ncf8bwB+muP4HYAWgZawD+4Qxe9oxYMVM812oH9v2+6Dd7QHuASFzLF+n8l2oH2ArMAxoHFy7DeTP\nlZyBLZPl/dByz0Oj2zzNe/8r8OtZyHNRKE3AwNgo5D7Q+uBzD7AAGZOkOQpsme98F+rHUft98PI9\nNVfynUm+C/UD+AK9wB4H1xqnUJqz0Y59HuTttEzdyaap2jOdYw8wCqQw1quuePB7CfAfxt8shDAC\nq4Ev5jPfxwkhxC7AD6h0cG225OtUvgsZSZJ6gH8A/rcQ4mkhxBIxxkrAoSvRbMlZkqT7wC+Bt4UQ\nu4QQRiGETgixVQjxPx6WeC57kjqgn7FRifz5f5P0OO3MYKjs4FnvAIM2+fUCd1zdq86BbD8BfuPg\n+2eAlvGyBHYwZoOc13wX8mdc++0ByoHvTnLvrMjX2Xwfhw/wHHAB6HugB84BrwC6uZSzTd6XHuiJ\nFuDfgXVTpREPEroUIcQzwNOSJH3X1WV5XBFCvA1USJL0f1xdlscRVb7zgzvIWeeqjMfRxZiDq8rc\ncZmxXlRlblDlOz+4XM5uMdJUUVFRWSi400KQioqKitsz59NzIcS0hrIxMTEUFRXxV3/1V3h7e2My\nmQgPD8fPzw+tVqsYYQcGBgDw8PBAp9Oh1WrRaCbqfiEEXV1dnDhxgldffdWZrZwLjunKeDwajQZJ\nknjiiSd44403KC4uRgiBEBPF1dLSwo9+9CNOnjxJf38/ALazFEmSFqWMNRoNGo2GjRs38tvf/pak\npCQMBoPSJsfL0rYdDwwMoNPpMJlMCCGwWCwMDg7S09NDfX09n3zyCX/6059ob29flPKVZSdJEvn5\n+RQXF5Oens7IyAg1NTVUVVVRXV1NX18fo6OjSnvUaDQEBASwfPlykpKSCAsLY2BggAsXLvCv//qv\nDAwM4GiGPV0Zu4tNk7t373Lq1CmqqqrQ6/V4enry6quvUlhYiI+PD1arla6uLv7lX/6Fe/fuodVq\n0ev1eHh44OnpiV6vx2QykZ6eTlZWFpGRkdy/f5/m5mZXV80tEUIgSRJ+fn6sXbuWtWvXKkpUvg7f\nvuQGg4HMzEzOnDmD2Wx22OgWI1br2IaS6upqfvWrX7F8+XJ2795Nbm4uOp1ugpysVisXLlzg008/\n5ebNm5hMJjZu3MjevXsxGo14e3vj7e1NUFAQAQEBeHp68tvfPo6bgaZGCIFWqyUkJIRXX32V1NRU\n7t27x5EjR6itraWlpYW2tjY6OzsZHh4eW9V+0KaFEBiNRq5fv05paSkREREkJSVRUFBAamoq7777\nLuXl5fT19c2obG6jNM1mM3V1ddTX12O1WomIiGBgYACtVgvA4OAg165d48CBA7S1tSGEQKPR2ClP\nk8lEQkICzz77LDt27KC/v59vvvnGxTVzT4QQWK1WVq9eTV5eHqGhocr3jvD09GTlypUYDIb5LOaC\nwGq10tTURFNTE/7+/oSGhpKWlkZAQMCEl7m2tpYPP/yQv/zlLzQ1NWE0GqmqqiItLY20tDT0ej0A\ner2e+Ph4nn76aa5fv+7iGs4vQghMJhNpaWkUFxcTHx/PrVu3OHXqFN988w337t0DmNAhyX9LkoTZ\nbKaxsZHGxkZ0Oh0VFRX09PSQnZ3Nli1b6OrqoqqqSun0nMFtlKaMRqPBarWyZs0aEhIS8PLyQpIk\n7t+/zxdffEFvby8w1lAtFgsjIyMMDg4C0N3dTUNDA76+viQkJGA0Grl586Yrq+PW+Pv7s2XLFlat\nWqWMMh2ZOmDMHJKYmIjRaJznUi4MZKXY1dVFZ2cnQ0NDDu8pKSnh5MmTNDU1odFoGBwc5Ouvv+bM\nmTMsXboUvV6vKFmdTkd0dDTf+973XFAj1yArzKysLJ566ikSExN59913OX78OHfu3JnRM0dHR6mp\nqaGxsZEzZ86wZcsWZTA2E9xKaSrOo0KwYcMGwsPDlRHR3bt3OXr0qKIgx2Nri2tvb6ehoYGIiAha\nW1vnswoLAlmma9asYe3atYSFhQFMqjCFEOh0OiIiIvDx8UGj0TA6OjqfRXZ7ZEVnMBgIDAzEZDIp\n1+Q2PTAwwNmzZ6mtrbVLZ7FYqKqqsmvbclv28vIiJydn/iriQuR2lp6ezt69e0lISOBv/uZvqKmp\nwWKxKO12utiORIUQjIyMUF5eTkVFxSOZl9xu9dxiseDp6Ul2djYhISEADAwMUFdXx6VLlxgZGcFq\ntTocmsvfyy91d3c39+/fd0U13B4hBC+99BIpKSnAxKnOeDQaDX5+fkRERGA0GiddMFrsREVFER0d\nzZIlSyZca2tro66uTmmTtgpgvDKQ5avVavHx8ZnbQrsBcqcTFhbGU089xbJly/jJT35CZWUlVqtV\n+cwUWT/MRpt1G6Up2ygNBgObNm0iMjJSMaS3trZy9uxZALthtZzG9m8hBP7+/nh7e9PZ2Tnv9XBn\nZPl4enpSVFTEqlWr8PHxmbYCFEKQmpqKr6/vPJR2YSG3w8zMTGJiYoBvR5gyFy9epKOjQ/lbnllp\nNBrCwsLw8JgYqFx+2RcLf/3Xf01MTAz79++ntrYWnU6HxWKZtYVHm+2TM8ZtlKaMwWBg586d+Pv7\nK73P7du3OX36NDDBzcVhg/Lz80Oj0VBfXz9fxV4QyLLz9vbm+9//PuHh4ZO6xsjYrqIDpKSk4Ofn\nN2WaxYjcDjMyMoiKirK7JsvJdhHDFq1Wy4oVK+zsxfLLPTo6Snt7+xyW3D2QJIkdO3aQnp7OjRs3\nOHnyJBaLZVYV5mzhNjZNSZLQarUEBASwfv16lixZghCC3t5eamtruX79+gSbRnh4OP7+/jQ0NNDf\n369Mz69du8bw8LCd7Uhl7OX18fEhNzeXjRs3KjKWGW8DckRCQoIyXXS3xuxKJEnC29ub+Ph4AgMD\nJ1zv6+vjypUrdHV12bl1yW41KSkpGAyGCf+Pvr4+Tp06xcsvvzxvdXEVO3bsoLW1lc8//5z29nY7\nFzh3wi1GmvL00Nvbm/T0dJYuXapMVRobGxV3AdtppL+/Pxs3bmT79u14enoC377EV69e5ciRI1y5\ncsU1FXJDZNlFRkby7LPPEhISopg6xr+oU01hoqOjlSm9OzZoVyCP1uPj44mMjFRGjLKMLBYLtbW1\nNDQ0KIs98v9jyZIlPPHEE4SFhSn/D1n+w8PD1NTUsH//ftdUbB7JzMwkLi6O8vJyrly54vSiz3zi\nNiNNgMDAQIqKivDy8lKEduPGDcXXUm5oFouFFStWsHXrVjQaDcPDw3bPkXdbqIwhv7w+Pj5kZmay\ndetWO9uwrPwsFgv3799Hq9Xi7e09wX4MEBQURFBQEAaDAbP5kQ60fGyQZbN69WrCwsImeCFYLBYu\nXbpEf3+/XcdvMBhISkrixRdfxMtrLHyk/L8YHR3l9u3bHD58mOPHj89jbVzDiy++SG9vLzU1NfT0\n9Li6OFPiFiNNGGt4oaGhbN68WdkaOTAwQGVlpTI1lxWmyWTiySefJDY2lnPnztHX1zepu4zKGEII\nUlJSKC4udjh9lCSJ3t5evvrqK2praxkcHHQ44vT09CQmJkZ5hmrX/FbR5eTkEBgYaCc3IQTDw8Oc\nPXvWTml6eHiwfPlynnnmGfLz8+06qJGREZqbm/n444/5539eHMG/tm3bxvXr12lqanL7GYzLNY08\nCjKZTCxfvpz4+Hjlu6qqKiorK+np6VHciLRaLT/72c8oLi6mvr6ew4cPu7oKbo8kSej1ejIzM/nO\nd77j8B4hBC0tLfzjP/4jN27ccOgPK68GL1u2jJCQkAneC4sVq9WKp6cnycnJ+Pr62nUk8vbf48eP\nc//+fSwWCxqNhpycHH72s5/xgx/8QFGkQghGR0e5ePEiv/zlL/nVr36lbOZ43AkLC6O8vFzxq3Zn\nxeny6bncwGJjY8nNzcXDw0N5OXU6HStXrsRoNOLn50doaCgZGRlkZGQwOjrKrVu3uHPnjlvbP1yN\nvMMqNzeXoqIifH197Vxh5N+7u7spLy+nqqqKurq6CY7Wto04JiaG4OBgJf1iRg7YkZ2dbec2JMtX\nXnz7wx/+wPDwMBqNBpPJRGRkJFFRUWg0GlpbW2lra6O8vJzTp09z5coV6uvrF5WJSaPR0N7ergSD\ncWdcrjTlqczSpUvJzc21m4YHBgZSWFhIfn4+3t7e+Pr6EhgYiMFg4IsvvuD69euMjIy47SqbOyDb\nMjdu3Mi6deuUiFEyskKsqanh0KFDmM1mqqurp7RXxsbGEhwcrMr8ATqdjnXr1jl0QhdC4OXlxebN\nm5W/PTw80Ov1aDQahoaGlC3Chw4dUmZWtlF7FgPywtdCGPy4hdL08/MjPj6eZcuWKd/X1dVx/Phx\nbt++jU6nU1bWCwsLFbeiqqoqF5bc/ZEV4po1a8jNzSU8PFz5HuxHmZcvX+b06dNYrVZqa2sVpelo\nlTw4OJjAwED0ev2ERbjFhrxHPC8vD29vb+U7Wcb37t3j0KFDtLS0oNPp0Ov1xMTEsGLFCuLi4vDw\n8MDf35/U1FQGBgbw8PCgrKxsUW7MsLUDu3OH4VKlKU8dly9fTnp6OiaTSWlwZ86c4Y9//CMVFRXA\n2B7cdevWkZaWRlBQELW1tTQ1NQHqFHEyZJeW7du3k5mZqYwyx7sYVVVVcfr0aVpaWhBC0NDQwP37\n97FarQ5X0L29vQkNDcXf35/29vZFvRik0Wjw9/cnMzNzQjCT0dFRGhoaePPNN6mursbDwwOj0UhS\nUhLFxcXs2bOHxMREgoOD2bp1K5s2bSI9PZ19+/Zx+vTpRac4x2+kcFdcPtLUaDSkp6eTmZmp2CZH\nR0c5ffo0zc3Nyjaz/v5+zp49S2NjIxaLhTt37jgMhqsyhtxbr1ixgnXr1ilBOcaPMs1mM+fOnePL\nL79U0ra3t9PV1cXIyIjDaDAeHh7KHmvZCXkxIoTAYDCQmppKcHAwWq1WkbscnqympoYbN24AY0p0\naGiI8+fPKzFhf/KTnyidmcFgoLi4GA8PD8xmMydOnHAYZ+FxRAihmCwsFourizMlLmvtsoLU6/Wk\npKSQkJAAjL3MLS0tXLp0ibt37ypbJeUXc3BwkOrqajo7O9XV2ymQbWff//737TwSxnPz5k1KS0tp\nampSXvqBgQG6u7uVxSBZwdo+IyQkRJnuL0ZsndPz8vIcxhnt6uri6tWrwJjd03ZxqLq6mlOnCyu/\nNAAACvtJREFUTnHv3j072QohKCws5Mknn8Tf33/CzOBxxWq1Ehoaqpg43BmXKk0Y2wmQnJyMt7c3\nkiQxODjIZ599Rnd3t11jkg3Eer2e+vp6h3t45ecuhkY2FXJQjpycHDZs2KDs47eVi9xpHT58mPPn\nz0/wyWxqaprSyTg6Opq4uLg5rcdCwGQysXnzZnS6bydtsqw7Ojq4cOEC8O2Cp+1CR29vr3KygG2w\nYr1eT0ZGBhs2bFCe97jT2trKypUriYiIcPt32OXT89zcXOLj45UV8L6+Po4ePTrhhZUblcFgoKWl\nRbk+WfTmxU5gYCA/+tGPHO5QkWVZX1/PtWvX6OvrIygoSDGF9Pb20tHRMeVxAIGBgUq094Ww4jnb\nyAtAISEhJCYmTpDx8PAwzc3Nik3ekYwsFovDYMVCCKKjo0lLS+Ojjz5yawUyWxw9epT09HSio6Op\nqqqacjFIq9ViNBoxGAyP/L7LgwdnXJ1cpjQlScJoNJKZmUlERAQw1tCampooLS11ePiR1Wrl2rVr\nyuqiI4GlpaVhsVhoamqa8RkgCxX55fL19WX16tUUFhYqsS/l67bTPZPJxAsvvEBRUZHdItHQ0BBJ\nSUlTHoFhMpkIDg7Gx8dn0cUsleXh4+NDUlKSMpKHbzuk3t5ebt++zd27dycoAPl3rVY76fEhXl5e\n+Pv7z3FN3If9+/fz+9//nrS0NMrLy6eM0h4UFMTWrVvJzs5WXLMepWPRaDRK/Irp4BKlKVcwMTGR\n5cuXK1Gue3p6uHTpEh0dHQ575pGREQ4cOEBlZaXDF9VkMrF9+3ba2tq4e/cufX19i6KXHk9MTAy7\nd+8mJCTEzo3D9ieMjRa3b98+6XM0Gs0E+cl/6/V6goKCiIiIWHRKUyYoKIicnBwl7qutrFpbW7lx\n44ZDP2J5RC8foOYIOdDHYqGiooJbt26RlZXFzZs3OXr0qOJdMx6LxYLZbKa7u3vGoePk/5Wfnx+J\niYmsXLly2mldqjSfeOIJIiMjFeG0t7dz/Phx5SiF8T306OgoJ06ccPg8SZJYuXIlGzZsoKSkZFH6\nD8rbUafaLmnLTBbRZFlrNBqCgoKIi4tTVocXE0IIgoODWbVq1YRrVquV+vp6h1G2bLcNR0VFKUpz\nfOfU29vL3bt356bwbsqhQ4d444032Lx5M1euXFHc2cYrxbt373Lw4EEOHjw447zkReRVq1axd+9e\nWlpaeOWVV6aVdt4XgmxHKoWFhYSEhCj7ypubmykpKVFC2zvqQcaPfuTK63Q6nn32WbRaLfX19fT2\n9i6qnUKyXFJSUigqKrKbWttOHWdDHvIzAgICiI+Pf+TnLTTk2K/BwcGK14et6WNkZISGhgblUL/x\nO7CEEMTGxrJ69WqMRqNDu3xbWxu3bt2akP5xRQjBp59+SmlpKYmJiRQXF+Pp6al4dMxmPvL/KjAw\nkFWrVhEZGcmvf/3raT/DJavnWq2WiIgIEhMTFReDe/fuUVVVpbhgTIatMpUFoNFoiI2NpaCgQAny\nsdiQ3bJWr17N1q1bnU4/viOaDgEBASQmJjqd10JGlk1QUBCJiYkO7Y7d3d00NTUpbnGOlN5kswH5\n/rq6OsrKyma/Am7O22+/TWVlJS+//DIbN25kdHTUzl1rNpCftXfvXvLz8/n444+pqamZdvp5nZ7L\nSk4O7SY3OHnrXklJybSeYzutlP273nrrLUJDQ/n666+pra1dFL2zjGzeWL9+PQUFBXbnbU/FZPZK\nR9cc3SfvDLJ1t3nckWcv8lbI8YtrALdu3VJGibKztjwTsFqt5Ofns2vXLmUBVE4rd3zV1dVcvHiR\n5ubmSe16jxuyDDs7Ozlw4AA6nY6//du/xWKxcPbsWWWw9CiykP93VquVH//4x2RkZPDFF19w7Ngx\np0xV89Laxy9CmEwmdu7cibe3txJEuLq6mtLSUrvRo6z4JnMr8vT0JC0tjVdeeYW8vDyamppobW1l\ncHDQ7fevziZWqxWDwcDmzZtZu3bthIUJWQ5dXV1cvHiRjz76CC8vr0kboFarpa+vj4yMDDZv3kxK\nSopix7TFYDAQEhKimAIWA1arFZ1OR2Ji4qRH6967d2/CFkjZNLJ27VplFGV7kJrtqvvhw4f57LPP\nGB4eXlQLmbJCu3HjBgcOHGBoaIi///u/58MPP+TgwYM0NjYq8pjOziFHg4LAwEB+8IMfEBUVxVdf\nfcWJEyfo7e11SlfMi9K0ddw1Go3ExcWxYsUK9Ho9AM3NzVy/fp22tjalN7ZNKyO7aAQHB7N06VLS\n09PJy8tj8+bN+Pj4cPXqVaWxLialCWNRw9euXWsXlGP8im1lZSX79u3j6NGjeHp6Tmrj1Ol0DAwM\nUF9fT1RUFCkpKQ7lqdPplGATiwEhxo5kyc3NZceOHURHRyvf2xIWFkZkZCR6vZ6RkRF0Oh2hoaFs\n2LCBnTt3UlBQoESJspVrb28vH3zwAR988AF1dXXA4rBn2iJvP62oqMBsNjMyMkJ+fj5RUVGcOXOG\nixcv0tjYOEEuU+1b12g0hIaGkpubq5zAWlJSQklJibK5wBnmXGnq9Xo8PT0VBRkWFkZ+fr5d9PCB\ngQHlqAvbEaYQQoniLkeI8fHxITIykuTkZDIyMoiLi1NGVl9//fWkO4UeZ7y8vNi5cycpKSmKLGQ5\nyi/mvXv3OH/+PCdPnlRchCZ7IeUpYU1NDbdv354w/bSdkvr6+pKdnT0v9XQl2dnZxMbGEhYWRmFh\nIRs3blR8+8a/sAkJCcoe8r6+PmWgsG3bNtLS0liyZIndLGp4eJj6+npKSkp45513uHz5MiMjI66p\nqBsgSRIDAwNcu3aN+vp6XnvtNXJycti1axcJCQlcvXqVlpYWmpub6erqsvOUsR2cBQYGEh4eTmho\nKLGxsaSmpmK1Wjly5AgXL16kq6tLyc8Z5lxprlq1irCwMEwmE0IIYmJiKC4ututhIyMjefrpp9m2\nbZtdgAidToeHh4fyMRgMGI1GPDw87F5ieftlWVmZ3Wl/i4WsrCyeeOIJOxvZ+FFmWVkZX331FR0d\nHYpSfNjUr729ndbWVoaHhx06/0rS2AmMjtxuHjdef/11ioqKCA4OVgJLgP0oU+6oAgICeOqpp9i6\ndStmsxlPT09MJpOygcBisTAyMoLZbKarq4uGhgaOHTvG/v37aWtrW3Tt1xGyDPr6+vjNb35DTk4O\nu3fvpqCggI0bN1JbW0tZWRl1dXXcv3/fzk1RnoYvW7aM5ORkwsPDMZvNlJaWsm/fPvr6+iY1/U2H\nOVeab775ptK7ajQah6HG/Pz8lLO0p8P4aaXFYqG9vZ26urpFGfnotddeY+nSpcqCjK3SFGLsjJqS\nkhK++uqraTlNy2n7+/vp6uqiv7/fodIUQiihzh53tm3bRlBQkDKSd2TjhW9lr9VqMZlMdoGJZWU5\nODhIR0cHFRUVHD9+nA8//JCOjg679CrfmvU0Gg2lpaWUlpYSExPD1q1b+e53v8vu3bvR6/XK+ehg\nv37S1dXFpUuX2L9/P59//jm9vb12LkczRaj/IBUVFZXpo8ZVU1FRUXECVWmqqKioOIGqNFVUVFSc\nQFWaKioqKk6gKk0VFRUVJ1CVpoqKiooTqEpTRUVFxQlUpamioqLiBKrSVFFRUXECVWmqqKioOIGq\nNFVUVFScQFWaKioqKk6gKk0VFRUVJ1CVpoqKiooTqEpTRUVFxQlUpamioqLiBKrSVFFRUXECVWmq\nqKioOIGqNFVUVFScQFWaKioqKk7w/wGan4dexZPD+gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdcce9e7dd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "ename": "AttributeError",
     "evalue": "'list' object has no attribute 'shape'",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-15-496470c2152a>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m      1\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mmodel\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mmodels\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 2\u001b[1;33m     \u001b[0mdisplay_sample\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtest_dataset\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtest_labels\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'predictions '\u001b[0m \u001b[1;33m+\u001b[0m \u001b[0mmodel\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mmodel\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;32m<ipython-input-6-33a7164a244b>\u001b[0m in \u001b[0;36mdisplay_sample\u001b[1;34m(dataset, labels, title_label, model)\u001b[0m\n\u001b[0;32m     15\u001b[0m         \u001b[0ml_title\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mchr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlabels\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mletter_idx\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m+\u001b[0m \u001b[0mord\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'A'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     16\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mmodel\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[0mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 17\u001b[1;33m             \u001b[0ml_title\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0ml_title\u001b[0m \u001b[1;33m+\u001b[0m \u001b[1;34m' / ~'\u001b[0m \u001b[1;33m+\u001b[0m \u001b[0mchr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpredict\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdataset\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mletter_idx\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mreshape\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m-\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m+\u001b[0m \u001b[0mord\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'A'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     18\u001b[0m         \u001b[0ma\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mset_title\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0ml_title\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     19\u001b[0m         \u001b[0ma\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0maxis\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'off'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m/usr/local/lib/python2.7/dist-packages/sklearn/linear_model/base.pyc\u001b[0m in \u001b[0;36mpredict\u001b[1;34m(self, X)\u001b[0m\n\u001b[0;32m    266\u001b[0m             \u001b[0mPredicted\u001b[0m \u001b[1;32mclass\u001b[0m \u001b[0mlabel\u001b[0m \u001b[0mper\u001b[0m \u001b[0msample\u001b[0m\u001b[1;33m.\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    267\u001b[0m         \"\"\"\n\u001b[1;32m--> 268\u001b[1;33m         \u001b[0mscores\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdecision_function\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    269\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mscores\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    270\u001b[0m             \u001b[0mindices\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0mscores\u001b[0m \u001b[1;33m>\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mastype\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mint\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m/usr/local/lib/python2.7/dist-packages/sklearn/linear_model/base.pyc\u001b[0m in \u001b[0;36mdecision_function\u001b[1;34m(self, X)\u001b[0m\n\u001b[0;32m    244\u001b[0m         \u001b[0mX\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcheck_array\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0maccept_sparse\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m'csr'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    245\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 246\u001b[1;33m         \u001b[0mn_features\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcoef_\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    247\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mX\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m!=\u001b[0m \u001b[0mn_features\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    248\u001b[0m             raise ValueError(\"X has %d features per sample; expecting %d\"\n",
      "\u001b[1;31mAttributeError\u001b[0m: 'list' object has no attribute 'shape'"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPQAAABTCAYAAACoNiqoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYS+zpQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdcce967e50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for model in models:\n",
    "    display_sample(test_dataset, test_labels, 'predictions ' + model[0], model[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Data set of sizes ('Training', (89998, 28, 28), (89998,)) ('Validation', (10000, 28, 28), (10000,))\n",
    "--------\n",
    "\n",
    "\n",
    "#### Results for original dataset:\n",
    "on test dataset:\n",
    "scoring newton-cg multi\n",
    "0.887951292459\n",
    "scoring liblinear l2\n",
    "0.884693441572\n",
    "scoring liblinear l1\n",
    "0.885921811579\n",
    "scoring sag l2\n",
    "0.884640034181\n",
    "scoring lbfgs multi ((our winner: time vs accuracy))\n",
    "0.887897885067\n",
    "\n",
    "on validation dataset:\n",
    "scoring liblinear l2\n",
    "0.8178\n",
    "scoring liblinear l1\n",
    "0.8186\n",
    "scoring sag l2\n",
    "0.8179\n",
    "scoring newton-cg multi\n",
    "0.8242\n",
    "\n",
    "\n",
    "#### Results for deduped dataset\n",
    "test dataset:\n",
    "scoring liblinear l2\n",
    "0.882522602474\n",
    "scoring liblinear l1\n",
    "0.883964723501\n",
    "scoring newton-cg multi\n",
    "0.885628709302\n",
    "scoring lbfgs multi\n",
    "0.885739641688\n",
    "scoring sag l2\n",
    "0.882578068667\n",
    "\n",
    "\n",
    "validation dataset:\n",
    "scoring liblinear l2\n",
    "0.796029536376\n",
    "scoring liblinear l1\n",
    "0.798692652221\n",
    "scoring newton-cg multi\n",
    "0.802203123109\n",
    "scoring lbfgs multi\n",
    "0.802566275269\n",
    "scoring sag l2\n",
    "0.796029536376\n"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "default_view": {},
   "name": "1_notmnist.ipynb",
   "provenance": [],
   "version": "0.3.2",
   "views": {}
  },
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}