printing_the_results.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "nb_epochs 60000\n",
      "learning_rate 0.0001\n",
      "size_of_lrs [200, 300, 400]\n",
      "batch_size 10\n",
      "gamma=0.2\n",
      "\n",
      "alpha=0.0\n",
      "\n",
      "The autoreload extension is already loaded. To reload it, use:\n",
      "  %reload_ext autoreload\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import scipy.linalg\n",
    "from scipy import stats\n",
    "from itertools import product as prod\n",
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "from constants_of_experiments import *\n",
    "from noise_models_and_integration import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "dictionary = np.load(\"results/eff_fid_lstm/dim_2x1/PLG_50_100tys/statistic_aSxbSy_id_spinChain_dim_2x1.npz\")[\"arr_0\"][()]\n",
    "dictionary['gamma0.8_alpha0.0'] = dictionary['gamma0.8_alpha0.']\n",
    "dictionary['gamma0.6_alpha0.0'] = dictionary['gamma0.6_alpha0.']\n",
    "del dictionary['gamma0.8_alpha0.']\n",
    "del dictionary['gamma0.6_alpha0.']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'gamma0.8_alpha0.0': [0.96, 0.97, 0.96], 'gamma0.6_alpha0.0': [0.94, 0.96, 0.94], 'gamma0.6_alpha0.2': [0.98, 0.98, 0.96], 'gamma0.8_alpha0.5': [0.9, 0.99, 0.99], 'gamma0.6_alpha0.5': [0.99, 0.99, 0.99], 'gamma0.6_alpha0.8': [0.99, 0.98, 0.98], 'gamma0.8_alpha0.8': [0.99, 0.99, 0.99], 'gamma0.8_alpha0.2': [0.98, 0.98, 0.98]}\n",
      "gamma  0.6\n",
      "alpha  0.0\n",
      "0.9619728\n",
      "gamma  0.6\n",
      "alpha  0.2\n",
      "0.96604335\n",
      "gamma  0.6\n",
      "alpha  0.5\n",
      "0.9885367\n",
      "gamma  0.6\n",
      "alpha  0.8\n",
      "0.9888214\n",
      "gamma  0.8\n",
      "alpha  0.0\n",
      "0.9717829\n",
      "gamma  0.8\n",
      "alpha  0.2\n",
      "0.97772264\n",
      "gamma  0.8\n",
      "alpha  0.5\n",
      "0.99072057\n",
      "gamma  0.8\n",
      "alpha  0.8\n",
      "0.9915163\n"
     ]
    }
   ],
   "source": [
    "\n",
    "print(dictionary)\n",
    "iterat = 0\n",
    "for gamma in [0.6,0.8]:\n",
    "    for alpha in [0.,0.2,0.5,0.8]:\n",
    "        print(\"gamma \",gamma)\n",
    "        print(\"alpha \", alpha)\n",
    "        data0 = np.load(\"results/eff_fid_lstm/dim_{}/PLG_50_100tys/statistic_{}_gam_{}_alpha_{}_{}.npz\".format(model_dim,\n",
    "                                                                                           noise_name,\n",
    "                                                                                           gamma,\n",
    "                                                                                           alpha, int(iterat)))[\"arr_0\"][()]\n",
    "        data100 = np.load(\"results/eff_fid_lstm/dim_{}/PLG_50_100tys/statistic_{}_gam_{}_alpha_{}_{}.npz\".format(model_dim,\n",
    "                                                                                           noise_name,\n",
    "                                                                                           gamma,\n",
    "                                                                                           alpha, int(iterat+100)))[\"arr_0\"][()]\n",
    "        data200 = np.load(\"results/eff_fid_lstm/dim_{}/PLG_50_100tys/statistic_{}_gam_{}_alpha_{}_{}.npz\".format(model_dim,\n",
    "                                                                                           noise_name,\n",
    "                                                                                           gamma,\n",
    "                                                                                           alpha, iterat+200))[\"arr_0\"][()]\n",
    "        dictionary[\"gamma{}_alpha{}\".format(gamma,alpha)].append(data0[0][0])\n",
    "        dictionary[\"gamma{}_alpha{}\".format(gamma,alpha)].append(data100[0][0])\n",
    "        dictionary[\"gamma{}_alpha{}\".format(gamma,alpha)].append(data200[0][0])\n",
    "        \n",
    "        iterat+=1\n",
    "        print(data0[0][0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gamma  0.6\n",
      "alpha  0.0\n",
      "0.9509780351320902\n",
      "gamma  0.6\n",
      "alpha  0.2\n",
      "0.9734388065338134\n",
      "gamma  0.6\n",
      "alpha  0.5\n",
      "0.9892206839720408\n",
      "gamma  0.6\n",
      "alpha  0.8\n",
      "0.9865614394346873\n",
      "gamma  0.8\n",
      "alpha  0.0\n",
      "0.9693977161248525\n",
      "gamma  0.8\n",
      "alpha  0.2\n",
      "0.9709230665365854\n",
      "gamma  0.8\n",
      "alpha  0.5\n",
      "0.9752432413895925\n",
      "gamma  0.8\n",
      "alpha  0.8\n",
      "0.9909477384885151\n"
     ]
    }
   ],
   "source": [
    "for gamma in [0.6,0.8]:\n",
    "    for alpha in [0.,0.2,0.5,0.8]:\n",
    "        print(\"gamma \",gamma)\n",
    "        print(\"alpha \", alpha)\n",
    "        print(np.mean(dictionary[\"gamma{}_alpha{}\".format(gamma,alpha)]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6\n",
      "0.0\n",
      "1.0\n",
      "aSxbSy_id_spinChain_dim_2x1\n",
      "0.97886056\n",
      "0.6\n",
      "0.1\n",
      "0.8\n",
      "aSxbSy_id_spinChain_dim_2x1\n",
      "0.97887844\n",
      "0.6\n",
      "0.1\n",
      "0.6\n",
      "aSxbSy_id_spinChain_dim_2x1\n",
      "0.9812311\n",
      "0.6\n",
      "0.2\n",
      "0.6\n",
      "aSxbSy_id_spinChain_dim_2x1\n",
      "0.9557622\n",
      "0.6\n",
      "0.3\n",
      "0.6\n",
      "aSxbSy_id_spinChain_dim_2x1\n",
      "0.97750616\n",
      "0.8\n",
      "0.0\n",
      "1.0\n",
      "aSxbSy_id_spinChain_dim_2x1\n",
      "0.90486413\n",
      "0.8\n",
      "0.1\n",
      "0.8\n",
      "aSxbSy_id_spinChain_dim_2x1\n",
      "0.9841345\n",
      "0.8\n",
      "0.1\n",
      "0.6\n",
      "aSxbSy_id_spinChain_dim_2x1\n",
      "0.97732115\n",
      "0.8\n",
      "0.2\n",
      "0.6\n",
      "aSxbSy_id_spinChain_dim_2x1\n",
      "0.97354627\n",
      "0.8\n",
      "0.3\n",
      "0.6\n",
      "aSxbSy_id_spinChain_dim_2x1\n",
      "0.97896427\n"
     ]
    }
   ],
   "source": [
    "params_list = [(0.6,0.,1.),(0.6,0.1,0.8),(0.6,0.1,0.6),(0.6,0.2,0.6),(0.6,0.3,0.6),\n",
    "                   (0.8,0.,1.),(0.8,0.1,0.8),(0.8,0.1,0.6),(0.8,0.2,0.6),(0.8,0.3,0.6)]\n",
    "for params in params_list:\n",
    "    gamma,alpha,beta = params\n",
    "    print(gamma)\n",
    "    print(alpha)\n",
    "    print(beta)\n",
    "    print(noise_name)\n",
    "#         data = np.load(\"results/eff_fid_lstm/dim_{}/statistic_{}_gam_{}_alpha_{}.npz\".format(model_dim,\n",
    "#                                                                                              noise_name,\n",
    "#                                                                                              gamma,\n",
    "#                                                                                              alpha))['arr_0']\n",
    "#     data = np.load(\"results/eff_fid_lstm/dim_{}/PLG_tr8k_ts4k_60tys_iter/statistic_{}_gam_{}_alpha_{}_beta_{}.npz\".format(model_dim,\n",
    "#                                                                                              noise_name,\n",
    "#                                                                                              gamma,\n",
    "#                                                                                              alpha,beta))['arr_0']\n",
    "    data = np.load(\n",
    "        \"results/eff_fid_lstm/dim_{}/PLG_final_batch_10_50tys_8k_to_4k/statistic_spinChainDrift_spinChain_dim_2x1_gam_{}_alpha_{}_beta_{}.npz\".format(model_dim,\n",
    "                                                                                         gamma,\n",
    "                                                                                         alpha,\n",
    "                                                                                         beta))['arr_0']\n",
    "    data2 = np.load(\n",
    "        \"results/eff_fid_lstm/dim_{}/PLG_final_batch_10_50tys_8k_to_4k/statistic2_spinChainDrift_spinChain_dim_2x1_gam_{}_alpha_{}_beta_{}.npz\".format(model_dim,\n",
    "                                                                                         gamma,\n",
    "                                                                                         alpha,\n",
    "                                                                                         beta))['arr_0']\n",
    "\n",
    "    data = data[()]\n",
    "    a = []\n",
    "    \n",
    "    for i in range(len(data)):\n",
    "#         print(list(data[i])[0])\n",
    "        a.append(list(data[i])[0])\n",
    "    data2 = data2[()]\n",
    "#     print(data2)\n",
    "    for i in range(len(data2)):\n",
    "#         print(list(data2[i])[0])\n",
    "        a.append(list(data2[i])[0])\n",
    "    print(np.mean(a))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6\n",
      "0.0\n",
      "aSxbSy_id_spinChain_dim_2x1\n",
      "0.929525\n",
      "0.95366395\n",
      "0.95160675\n",
      "0.93236196\n",
      "0.95506155\n",
      "0.96455604\n",
      "0.96338636\n",
      "0.9500231\n",
      "0.6\n",
      "0.2\n",
      "aSxbSy_id_spinChain_dim_2x1\n",
      "0.96862584\n",
      "0.9710865\n",
      "0.9670774\n",
      "0.96892995\n"
     ]
    }
   ],
   "source": [
    "params_list = [(0.6,0.),(0.6,0.2)]\n",
    "for params in params_list:\n",
    "    gamma,alpha = params\n",
    "    print(gamma)\n",
    "    print(alpha)\n",
    "    print(noise_name)\n",
    "    data = np.load(\"results/eff_fid_lstm/dim_{}/stretch_batch50_100tys/statistic_{}_gam_{}_alpha_{}.npz\".format(model_dim,\n",
    "                                                                                             noise_name,\n",
    "                                                                                             gamma,\n",
    "                                                                                             alpha))['arr_0']\n",
    "\n",
    "\n",
    "    data = data[()]\n",
    "    a = []\n",
    "#     print(data)\n",
    "    for i in range(len(data)):\n",
    "        print(list(data[i])[0])\n",
    "        a.append(list(data[i])[0])\n",
    "    print(np.mean,(a))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(32, 2, 2500)"
      ]
     },
     "execution_count": 172,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# dane = np.load(\"results/NN_as_approx/dim_{}/stretch/{}_gam_{}_alpha_{}_epsilon_1e-{}.npz\".format(model_dim,\n",
    "#                                                                                         noise_name,\n",
    "#                                                                                         gamma,\n",
    "#                                                                                         alpha,\n",
    "#                                                                                         eps_order))[\"arr_0\"]\n",
    "\n",
    "# dane_rev = np.load(\"results/NN_as_approx/dim_{}/stretch/{}_reversed_gam_{}_alpha_{}_epsilon_1e-{}.npz\".format(model_dim,\n",
    "#                                                                                         noise_name,\n",
    "#                                                                                         gamma,\n",
    "#                                                                                         alpha,\n",
    "#                                                                                         eps_order))[\"arr_0\"]\n",
    "\n",
    "dane_rev = np.load(\"results/NN_as_approx/qutip/dim_{}/{}_gam_{}_alpha_{}_epsilon_1e-{}.npz\".format(model_dim,\n",
    "                                                                                        noise_name,\n",
    "                                                                                        gamma,\n",
    "                                                                                        alpha,\n",
    "                                                                                        eps_order))[\"arr_0\"]\n",
    "\n",
    "dane = np.load(\"results/NN_as_approx/qutip/stretch/{}_reversed_gam_{}_alpha_{}_epsilon_1e-{}.npz\".format(\n",
    "                                                                                        noise_name,\n",
    "                                                                                        gamma,\n",
    "                                                                                        alpha,\n",
    "                                                                                        eps_order))[\"arr_0\"]\n",
    "# print(dane)\n",
    "# print(dane_rev)\n",
    "dane = np.concatenate((dane,dane_rev),axis=2)\n",
    "np.shape(dane)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "456\n"
     ]
    }
   ],
   "source": [
    "def sieve_data(data):   \n",
    "    tmpdata = []\n",
    "    for i in range(np.shape(data)[2]):\n",
    "        if not((data[:,:,i] <0).any()) and (data[:,:,i] >0).any():\n",
    "            tmpdata.append(data[:,:,i])\n",
    "\n",
    "    print(len(tmpdata))\n",
    "    data = np.zeros((np.shape(data)[0],np.shape(data)[1],len(tmpdata)))\n",
    "    for i in range(len(tmpdata)):\n",
    "        data[:,:,i] = tmpdata[i]\n",
    "    return data\n",
    "\n",
    "data = sieve_data(dane)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-1.15037126e-01, -3.15511528e-02,  3.42568188e-02, -2.80410067e-01,\n",
       "        3.79908734e-01, -2.21027141e-01, -8.07595270e-02, -2.42670588e-01,\n",
       "       -2.68328365e-01, -3.60111971e-01,  3.95289445e-03, -2.05431049e-01,\n",
       "       -1.28801888e-01, -2.04321686e-01, -2.05162373e-01, -4.63367687e-01,\n",
       "       -1.76647786e-01, -2.48655965e-01, -2.03828155e-01, -3.05482203e-01,\n",
       "       -4.88976863e-01, -1.72312436e-01, -2.18385147e-01,  4.75421276e-01,\n",
       "       -1.82389603e-01, -3.00313385e-01, -2.68756208e-01, -7.84793301e-02,\n",
       "       -2.19955908e-01, -1.70026452e-01, -2.88503770e-01, -1.27462619e-01,\n",
       "        3.65021122e-01,  2.14035112e-01,  2.23634124e-02, -5.10903242e-01,\n",
       "       -1.62645223e-01, -3.59152669e-01, -3.73340879e-02, -5.11113925e-01,\n",
       "       -3.74656475e-01, -1.37152916e-01, -2.28452541e-01, -8.86728634e-02,\n",
       "       -3.17148888e-01, -3.58037442e-01, -1.51475297e-01, -1.69708616e-01,\n",
       "       -1.14246364e+00, -3.98036757e-03, -8.45933092e-02,  6.04443478e-02,\n",
       "       -2.77797817e-01, -2.02346144e-01, -2.43186955e-01, -7.42057094e-01,\n",
       "       -7.44742457e-01, -2.28914978e-01, -2.71970135e-01, -2.82298367e-01,\n",
       "       -2.32308732e-01, -6.85779300e-01, -1.86279130e-01, -2.07721766e-01,\n",
       "       -1.55617854e-01, -1.40930790e-01, -2.02180562e-01, -1.72237822e-01,\n",
       "       -6.42929266e-01, -1.66663158e+00, -1.74206815e-01, -2.47248116e-01,\n",
       "        8.64619816e-01, -9.24824606e-01, -1.57856630e-01, -7.61380108e-01,\n",
       "       -1.85858678e-01, -3.30423400e-01,  4.90791122e-01, -4.68859991e-01,\n",
       "       -2.68438985e-01, -2.04425512e-01, -3.00646972e-01, -2.25291044e-01,\n",
       "       -3.29646715e-01, -1.65716674e-01, -1.71201100e-01, -9.52215971e-02,\n",
       "       -1.19997283e-04, -1.17746529e-01, -2.44486971e-01, -1.32276695e-01,\n",
       "       -2.23668860e-01,  4.77160117e-01, -2.81959627e-01, -2.73237619e-01,\n",
       "       -2.60700077e-01, -4.51796806e-01, -2.04996728e-01, -3.83310747e-01,\n",
       "       -2.14282504e-01, -2.96559101e-01, -2.23624434e-01, -2.05339102e-01,\n",
       "       -9.68212808e-02, -2.99426672e-01, -1.79310425e-01, -3.57118881e-02,\n",
       "       -9.14064221e-02, -1.53609999e-01, -1.15071403e+00, -5.82695958e-01,\n",
       "       -2.00500729e-01, -2.42713908e-01, -1.26103189e-01, -1.13811763e-01,\n",
       "       -3.20367227e-01, -6.34765679e-02, -3.62890941e-01, -2.51308490e-01,\n",
       "       -2.25892483e-01, -2.14970761e-01,  2.49269663e-01, -1.17952761e-01,\n",
       "       -4.78752756e-01, -1.05362665e+00, -3.20971538e-02, -2.35664821e-01,\n",
       "       -1.79380412e-01,  1.43248756e-01,  1.36356096e-01, -3.50407673e-01,\n",
       "       -2.59568336e-01, -1.69481568e-01, -6.23144513e-01, -2.44383663e-01,\n",
       "       -2.84920451e-01, -2.70815657e-01, -2.53048025e-01, -4.27586139e-01,\n",
       "       -2.55237339e-01, -4.00567343e-01, -5.28859124e-02, -2.00256756e-01,\n",
       "       -2.85622232e-01, -4.18355897e-01, -1.59083428e-01, -1.04441557e-01,\n",
       "       -1.25028610e-01, -8.12426905e-02, -2.25891281e-01, -2.35609328e-01,\n",
       "       -1.58316049e-01, -2.23367159e-01, -4.77707982e-01, -1.34119269e-01,\n",
       "       -2.47339339e-01, -2.14168101e-01, -1.79971308e-01,  2.83838290e-02,\n",
       "       -1.35818002e-01, -2.06247199e-01, -1.94569832e-01, -1.19261819e-01,\n",
       "       -2.58864326e-01, -6.98484356e-02, -4.47681077e-01, -2.48546101e-01,\n",
       "       -4.94338547e-01, -2.69712466e-01, -1.73019567e-01, -4.14424999e-01,\n",
       "       -1.41323868e-01, -2.84849150e-01, -1.91627260e-01, -1.69133275e-01,\n",
       "       -6.33910165e-01, -7.37244776e-01, -9.62664385e-02, -1.91370691e-01,\n",
       "       -5.66310670e-01, -2.18899248e-01, -1.73148490e-01, -2.45341863e-01,\n",
       "       -4.21238360e-01, -1.60100177e-01, -1.71049912e+00, -3.08319957e-02,\n",
       "       -1.94839410e-01, -3.03646746e-01,  3.67085298e-01, -1.89736901e-01,\n",
       "       -4.56766145e+00, -1.81380817e-01, -2.26830344e-01, -1.02848470e-01,\n",
       "       -1.89678809e-01, -1.30585388e-01, -1.50097253e-01, -2.59323293e-01,\n",
       "        1.81149281e-02, -5.86702629e-01, -2.16860988e-01, -2.63749332e-01,\n",
       "       -2.23567590e-01,  7.54636125e-02, -1.96767800e-01, -1.89282828e-01,\n",
       "       -1.87615919e-01, -2.40769382e-01, -2.32177573e-01, -2.42896712e-01,\n",
       "       -2.88539673e-01, -3.19820473e-01, -2.30754293e-01, -1.57941199e-01,\n",
       "       -3.49479762e-01, -1.50302297e-01, -5.45184582e-01, -2.01867240e-01,\n",
       "       -5.82816818e-01, -2.92316146e-01, -2.05060904e-01,  4.92454749e-01,\n",
       "       -9.25747338e-02, -1.89290596e-01, -1.13493225e-01, -3.39261780e-01,\n",
       "       -2.05622138e-01, -8.45368780e-02, -2.04462982e-01, -3.70075547e-01,\n",
       "       -1.44130063e-01, -7.77342210e-01, -3.31065818e-01, -3.82533908e-01,\n",
       "       -8.50575573e-02, -4.16567797e-01, -5.51015345e-01, -1.48495122e+00,\n",
       "       -4.76085381e-01, -2.64930390e-01, -1.61383021e-01, -2.23757806e-01,\n",
       "        1.03759024e+00, -1.75873423e-01, -1.43501019e+00, -1.64520534e-01,\n",
       "        2.55909318e-02, -2.71662173e-01, -7.92607173e-01, -1.60767270e-01,\n",
       "       -2.26092633e-01, -9.31703579e-01,  2.34535867e-01, -2.11516663e-01,\n",
       "       -2.58138287e-01, -2.79567093e-01, -1.79604719e-01, -2.10764786e-01,\n",
       "       -1.97247643e-01, -2.58469717e-01, -2.24742731e-01, -1.54121733e-01,\n",
       "       -3.53251191e-01, -1.05526059e-01, -1.37001569e-01, -3.97809921e-01,\n",
       "       -2.80556838e-01, -2.25393087e-01, -4.13162024e-02, -1.61969599e-01,\n",
       "       -2.92687116e-01,  6.61997575e-03, -1.88658288e-01,  8.29823361e-04,\n",
       "       -2.83545096e-01, -1.61054861e-01, -2.72436066e-01, -1.58972559e-01,\n",
       "        3.56114148e-01, -4.11662368e-01, -2.62989715e-01, -1.69842407e-01,\n",
       "       -5.15510011e-02, -1.16938748e-01, -2.23140930e-01,  1.06239231e-02,\n",
       "       -1.40731424e-01, -2.22557529e-01, -2.72152998e-01, -1.55796463e-01,\n",
       "       -1.32859948e-01, -2.24659243e-01, -4.08926609e-02, -3.11908780e-01,\n",
       "       -2.88556546e-01, -6.55058845e-01, -3.82052909e-01, -7.36766854e-01,\n",
       "       -1.40112269e-01, -1.18984582e-01,  4.40470673e-02, -2.02529295e-01,\n",
       "       -6.70489042e-02, -2.99327510e-01, -2.28778367e-01, -6.81378634e-01,\n",
       "       -5.28972975e-01, -3.67441009e-01, -3.28699035e-01,  1.38100331e+00,\n",
       "       -2.15275565e-01, -1.52697377e-01, -2.26209703e-01, -2.45290882e-01,\n",
       "       -2.90434228e-01, -2.38487668e-01, -3.81617282e-01, -2.39211007e-01,\n",
       "       -2.43559061e-01, -2.73400888e-01, -2.09593901e-01, -1.84744467e-01,\n",
       "       -2.50467965e-01, -2.15468349e-01, -8.62207159e-03, -3.06957808e-01,\n",
       "       -1.35272137e-01, -4.11742004e-01, -1.42893512e-01, -2.06514444e-01,\n",
       "        3.71589896e-01, -6.17853024e-02, -1.58219971e-01, -9.42778766e-01,\n",
       "       -2.32666948e-01, -2.80778214e-01, -9.34463728e-01, -3.85967118e-01,\n",
       "       -3.45842133e-01, -1.44987717e-01, -1.01692307e-01, -1.36981417e-01,\n",
       "       -1.00753883e-01, -2.24308333e-01, -1.35448673e-01, -2.64359267e-01,\n",
       "       -6.24633115e-01,  2.28004207e-01, -2.79990028e-01, -2.86164744e-01,\n",
       "       -1.63815058e-01, -1.36959795e-01, -1.45352890e-01, -1.61582367e-01,\n",
       "       -3.24429167e-01, -2.25783833e-01, -1.78524044e-01, -7.62525978e-01,\n",
       "        3.95556725e-02, -1.46127720e-01, -2.73674146e-01, -1.29658044e-01,\n",
       "       -1.62270518e-01,  1.89042837e-01, -2.76009028e-01, -4.24658385e-01,\n",
       "        1.02326625e+00, -2.40946063e-01, -6.78830262e-02, -2.30675308e-01,\n",
       "       -2.63380585e-01, -3.22361568e-01, -5.75596245e-02, -2.00337631e-01,\n",
       "       -8.45302748e-02, -2.37695177e-01, -3.21797508e-01, -4.17031223e-01,\n",
       "       -1.19604255e-01, -2.92892734e-01, -1.64927978e-01, -2.43956733e-01,\n",
       "       -3.23168192e-01, -1.81187673e-01, -3.29973548e-01, -2.78125868e+00,\n",
       "       -1.97689332e-01, -1.99197125e-01, -2.93141258e-01, -2.75376394e-01,\n",
       "       -1.49722870e-01, -2.87778308e-01, -2.44512718e-01, -2.58883341e-01,\n",
       "       -1.56027089e-01,  1.63574142e+00, -7.31678386e-02, -5.19495979e-01,\n",
       "       -1.81208085e-01, -3.24476121e-01, -4.92776613e-01, -3.69143289e-01,\n",
       "       -1.50467541e-01, -2.18510049e-01, -1.62998852e-01, -3.06351925e-01,\n",
       "       -4.89835708e-01, -3.95547516e-02, -1.28798333e-01, -1.66525211e+00,\n",
       "       -5.86422669e-02, -1.39476162e-01, -2.56572588e-01, -3.03147249e-01,\n",
       "       -1.89724423e-01, -2.46130265e-01, -7.56502921e-01, -5.19100533e-01,\n",
       "       -2.23120679e-01, -2.79736586e-01, -5.51395671e-01, -2.29116090e-01,\n",
       "       -1.65172034e-01, -2.27454397e-01, -1.04440467e+00, -2.08029895e-01,\n",
       "       -1.68506814e-01, -1.16263582e-01, -2.21015085e-01, -7.08645562e-01,\n",
       "       -1.01731928e+00, -1.89912473e-01, -2.49359670e-01, -1.03963683e+00,\n",
       "       -8.52167912e-02, -7.00887591e-01, -2.10262046e-01, -3.53053386e-01,\n",
       "       -3.39816019e-01, -2.77850819e-01, -1.80717172e-01, -3.75320722e-01,\n",
       "       -2.24607968e-01, -3.24241639e-01, -1.83759587e-01, -1.49170132e-01,\n",
       "       -1.47162280e-01,  2.49452684e-03, -1.78910120e-01, -2.30294665e-01,\n",
       "       -1.35181752e-01, -1.72647204e-01, -2.09894638e-01, -2.56115420e-01])"
      ]
     },
     "execution_count": 174,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[2,0]-data[31,1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXkAAAD7CAYAAACPDORaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3Xlc1HX+B/DXhxtUGG4FQQTxxgMGrzJNx8wu01AzM2t/\nhVttdm12bLvV1la61dZ2uNHutmWZV1Ze2ybe5ZGIBwqKgiKCAoIDKPfw+f3BFx1xZBAGvnO8no8H\nj5nvNby/wrz4+Pl+5/MRUkoQEZF9clK7ACIiaj8MeSIiO8aQJyKyYwx5IiI7xpAnIrJjDHkiIjvG\nkCcismMMeSIiO8aQJyKyYwx5IiI75qJ2AQEBATIiIkLtMoiIbMrevXvPSSkDze2neshHREQgJSVF\n7TKIiGyKECKnJfuxu4aIyI4x5ImI7BhDnojIjjHkiYjsGEOeiMiOMeSJiOwYQ56I2t2+U+dRVlWr\ndhkOiSFPRO1GSomPNx/HlE924PGvU8E5pTseQ56I2kV9vcTrazPw1/8dRZ/gLth+7Bx+2J+vdlkO\nhyFPRBZXa6jHsysO4N+/nMBDN0Rg3bwbMSRMg9fXpkNfUaN2eQ6FIU9EFlVZY8DcxXvx3b48/P6W\n3vjTHf3h4uyEt6bGQF9ZizfXZ6hdokNhyBORxZRW1OL+f+3GlqOFeHNKDH43LhpCCABAv27eeHh0\nTyxPOY1d2cUqV+o4GPJEZBEFZVWY/ulOpJ0uxcf3xeK+4eFX7fPU+N4I8/PES9+loarWoEKVjoch\nT0RtduLcRUz9ZAdOn6/Afx6Kx6SYbib383Rzxht3xyC76CI+2ZLVwVU6JoY8EbXJobxSJCzagcpa\nA75JHIFRvQKa3X9M70BMHhKCRVuO43hheQdV6bjMhrwQIkEIoRNCzL/G9lhlnwSjdQuUx0TLlUpE\n1mZnVjHuTdoFD1dnrPztSAzqrmnRcX+8oz+83Fzw0qpDqK/nvfPtqdmQF0LEAoCUMhmAvnG5iRel\nlCsBRBptTxRCZAHItmi1RGQ1Si7W4NGv96Krjwe+fXQUIgM7t/jYgM7ueOm2vvj1ZAmWp+S2Y5Vk\nriU/A4BeeZ4NQGe8UWm97wEAKeVCKWWqsukRKWWU8seBiOzQwh+P4EJVHT6ZFYuuPh7Xffx0bRiG\n9fTDm+szUFRe3eLjzl2oxuajhfz0bAuZC3kNgBKjZf8m2+MB+CtdNsbdOZHNdfEQkW3bm3MeS/fk\n4v9u7InewV1a9RpCCLw5JQZVtfX489p0s/vnllTgTz8cwg1vb8JDn+/B6gP89GxLWOLCa3FjC76x\nX15p1Sej4Q+ArukBQohEIUSKECKlqKjIAiUQUUepM9Tj5e8PoZuPB+aNj27Ta/UK6ozHbo7CmgP5\n2HK00OQ+R8+W4+ll+zH2nS345tdTmDwkBP27eeOt9UdQUVPXpu/vCMyFvB6An/JcA6DpJxiKcbnf\nXQ8gXgnwBKPtkU1fVEqZJKXUSim1gYFmJxsnIivy5c4cZJwpwyt39kcnd5c2v96jY6MQFdgJL39/\n6IrQ3ptzHg9/sQcT39+G/x0+iwdHRWDb/JuxMGEwXr97AM6WVeGTzbwN0xxzP6FlALTK80gAyQAg\nhNBIKfUAVgJoDHQNGvrns3E5+KMAfGrJgolIPQVlVXhvQybG9gnExAFdLfKa7i7OeHNKDGYk7cL7\nyccwKsofn2zJwq8nSqDxcsVTumjMGRkB305ul46J6+GHKUNDkbQ9G9O1YQj397JILfao2Za8UTeM\nDoDe6MLqRmV7NhruukkA4C+lXKnsM11Zl2V0DBHZuL+sy0CNoR6v3TXg0nAFljA80h8ztGFI2paN\nBz/fg9ySCvzxjv7Y8cI4PKXrfUXAN3phUl+4OAm8sc58f74jM/t/LSllkol1cSa2r2zuGCKybb8c\nP4fVB/LxlC4aPfw7Wfz1X7qtH2oN9RgR5Y+7h4TCzaX53uRgbw88fnMv/PV/R/HzsXO4Mbr5D2E5\nKqH2bUharVampKSoWgMRNa+6zoBJH2yHoV7if0/dBA9XZ7VLAgBU1Row8f1tcHN2wvonR8PV2XE+\nxC+E2Cul1Jrbz3H+RYio1f65/QSyiy7itbsGWE3AA4CHqzNevr0/jhVewOKdOWqXY5UY8kTUrNyS\nCny46Rhui+mKsX2C1C7nKrp+QRgdHYC/JWei+ELLP1TlKBjyRNSs19akw0kI/PGO/mqXYpIQAq/c\n2R+VNQa881Om2uVYHYY8EV3ThvQCJGcU4CldNLr5eKpdzjX1CuqCOaMisHTPKRzKK1W7HKvCkCey\nAwdy9fhw4zHszi622KiOlTUGvLr6MHoHd8ZDN/S0yGu2p3njo+Hn5YZXVx/muDZG2v5xNSJSRVWt\nAWsO5GPxrhwcPH259Rqq8cTkISGYMjQU0a0cVwYAPtp8DHn6SiyfO9Im7lrx8XTFcxP74IVVaVh9\nIB+Th4SqXZJVYMgT2ZiT5y7i6905WJ5yGqWVtYgO6ow/Tx6AWwd0xc7sYny3Lw+fbsvGJ1uyMCDE\nG1OGhuKuwSEI8m7ZSJGVNQak5ZUiaVs27ontjmE9/cwfZCWmacPw1e4cvLX+CCb0D4aXm+UirrCs\nCt6erlZ1d1FL8D55IhtgqJfYdKQQi3flYFtmEVycBCYO7IrZI3pgeE+/qz59WlRejbUH8/H9vjwc\nOF0KJwHc0CsAdw8JxejoABRfrMHp85XIO1+BPH1lw3PlseRiDYCGlvHGZ8cgoLO7GqfcaiknS5Dw\nj514YlwvPHtLnza/npQSS349hdfWpCM6qDOWPDICPp6uFqi0bVp6nzxDnshKSSlxtKAcPx0uwLI9\nucjTVyLY2x33DeuBe4eFIbiFLfOsogv4YV8evtufh9ySyqu2u7s4obuvJ0J9vRCq8UR334YvbYQf\nQjXWe7G1OU8t3Yf1h84i+ekxbRrXpryqFi+uSsPag2cQG65BWl4pBnfXYPH/DYenm7oteoY8kQ26\nWF2HHVnF2Hy0EFuOFCK/tAoAcEMvf8we0QPj+wW3un9cSonUU3ocyNUj2NtDCXZP+Hdys+g4NNbg\nbGkVxr27BdoIP3wwY4jJsW/MOZRXit8tScWpkgo8e0sfPDomCv89dBZPfJOKG6MD8dkDcXB3US/o\nGfJENuLkuYvYdKQQm48WYnd2CWoM9ejk5owbowNwc58gjO0T1KqZlxzdf345gVfXpMPLzRn3DQvH\nw6MjW/TvKKXEV7ty8PraDPh1csPfZw694rrE8j25mP/tQUwa2BUfzhwKF5UuSjPkiaxYTV09Ptly\nHD/sz8eJcxcBAJGBnTCuTxBu7huE+Ag/swN0kXmZBeVYtCULqw/kw1kI3BMXirk3RSEiwPQAa2VV\ntXjh24NYn3YWY/sE4r3pQ+Bn4n8B//r5BF5fm46EuO5YeM8gODl1/P+EGPJEViq3pAK/+2YfDuTq\nMTo6AOP7NgR7e4zsSA1ySyrw6bYsLE85jTpDPW4fFILHxkahXzfvS/uknS7F40tSkaevxHMT+yBx\ndGSz4f1+cibeTz6GB0dF4JU7+3d4l1dLQ563UJLNqa+XWHMwH/ERfgixsQuDPx46g+dWHgQALJoV\ni0kx3VSuyDGE+XnhjbtjMG98NP718wl8tTMHaw7kY1zfIDw2NgqH88vwl3UZ8O/shmWJI6CNMH/b\n6JPjo1FeVYd//XwC3h4ueMYCd/K0B7bkyaacu1CNp5ftx/Zj56Dt4YsVvx3Z5haUlBJpeaUoKKtG\naWXtpa+yylroK2quWNfZ3QUPjIzAXUNCrusCaFWtAW+tz8AXO3MwOEyDj2YORZgfZzNSS2lFLb7c\neRL//uUEzlfUAgDG9w3CO9MGX9dFWiklXvg2DctScvGH2/rhkZuumu203bC7huzOruxizPtmH/SV\ntZjQPxjrDp7B32cOxV2DQ9r0up9uzcJb/z1y1fouHi7QeLnCx/PyV1bhRRwtKEeIjwd+c2NPzBwW\nbnae0xPnLuJ3S1JxOL8Mj4zuiecm9mV/u5WoqKnD8j25cHF2wqzh4a1qMBjqJeYt3Yd1B8/grakx\nmDksvB0qvZrFumuUafz0AGKllAtNbI+FMlm3lHJlS44huh719RKfbDmO9zZkood/J3z+UDz6dvXG\niaKLeHt9Bib0C271Pcu5JRX4W3ImxvUNwlO66Eth3sXDFc4m+mOllNhytAiLtmbhjXUZ+HDTcTww\nsgfmjIow+aGhH/bn4aVVaXB1ccI/H9BC1z+4VXVS+/Byc8GDbRyXx9lJ4G/Th+BidR1e+i4Nnd1d\ncGcbGx6W1GxzQglwSCmT0TCXa6yJ3V5Uwj1SCBHbwmOIWuTchWrM+fxXvPNTJu4YFII1T9yIASE+\ncHZqGF42v7QKSduyzb+QCVJK/OmHQ3AWAn+ZMhCDumvQw78TNF5uJgMeaBjW9ua+QVg+dyRWPTYK\nIyL98NHm47jh7U14+fs0nCquANAwNMCLqw7iyaX70a+bN9bPG82At2NuLk5YNCsO8RF+eHrZfiz8\n8QhySyrULguA+Zb8DAAblOfZAHQALk3MrbTY9wBAY4tdCLGguWOIWmpnVjGeXLoPpZW1eGtqDO6N\nD7viv9PDI/1xW0xXLNp6HNO03a/7Iux/D53F5qNF+OMd/Vs1jG5suC8+na1FVtEFJG3NxvI9p7Fk\n9yncFtMNxwou4GhBOR4bG4VnJvRW7V5q6jiebs741xwt5q88iH9szcKirVm4sVcA7hsWDl3/1n+I\nra3MhbwGQInRsn+T7fHApRa/Tgl6c8cQNctQL/Hx5uN4PzkTEf6d8MVvhl1xq5uxFyf1Q3JGIRb8\neAQf3Du0xd+jvKoWr605jAEh3pgzskeb6o0K7IwFCYPwzC298e9fTmDJrlNwc3HCF78ZhjG9A9v0\n2mRbuni4YtH9cThTWolle3KxbE8uHv06FQGd3TFN2x0z48PbNMxCa1jiFspiKWWqEEKntOzNEkIk\nAkgEgPDwjrlIQbahqLzh7pmfj5/DlKGheOPugc1e2Azz80Li6Eh8tPk4HhgZgbgevi36Pu/+lInC\n8mokzdZarJUd7O2BFyf1w9O63nASghdXHVg3H088peuNJ8ZFY2tmIZbszsWnW7OwaEsWRkcHYOaw\ncOj6BXfI74i576AH0HjDqAZAcZPtxWjokmncN74Fx0BKmSSl1EoptYGBbOlQg+OF5bjro5+x52QJ\nFt4zCO9NH2z2zhUAeHRsFIK6uOPPaw63aMKMg6f1+GLnSTwwogcGh2ksUPmVPFydGfAEoOGi7Li+\nwfjnHC1+eWEcnpnQG1mFF/DY16kY9fZG7M6+Kh4tztxv4jIod84oj8kAIIRofGesNNquQUP/vMlj\niJqz79R5JPxjJ2oNEt8+OgrTm/S/N6eTuwuev7UvDpwuxXf78prdt85Qj5e+S0NgZ3c8O9E6P7xC\n9qmbjyfmjY/G9ufH4fMH4xHXw7dNk7q0VLMhL6VMBQAhhA6AvnEZwEZlezYa7qBJAOAvpVzZzDFE\nJm3NLMJ9n+2Gj6crVj06CgNDfa77NaYMDcXgMA0W/HgEF6vrrrnflztzcCivDK/cOQDeHuqPCU6O\nx9mp4Q6tT2drTY6LY2n8MBSp6of9efj9igPoFdQFX/wmHkFdWj/a4t6c87hn0Q48fnMUnpvY96rt\nZ0oroXt3K+J7+uHzB+Ptbnhdciwt/TAUOw5JNf/55QSeWrYfQ8N9sWzuiDYFPADE9fDF3UNC8Nn2\nEybvUX5tdToMUuL1yQMZ8OQwGPLU4aSUeO+no3h1TTom9AvGl78ZZrGuk+cn9YWzEHjrvxlXrN+Y\nUYAfD5/FvPHRHDOGHApDnjqUoV7i5e8P4e+bjmOGNgyfzIq16MTI3Xw88dsxUVifdha7lDsXKmrq\n8KcfDiM6qDMevrHjBpAisgYMeeow1XUGPPFNKr7efQqPjo3C2/fEtMsnQRNvikSIjwf+vCYdhnqJ\nD5KPIU9fiTenxvDWRnI4/I2nDnGhug4Pfb4H69PO4uXb++H5W/u2W7+4p5szXrytH9LPlOGNden4\n588nMEMbhvgWjBFOZG8Y8tQhXl+Tjt0nSvDe9MF4eHT7d5ncMagb4iN88fkvJ+Hj6YoXJl19tw2R\nI2DIU7urrjNgfdoZTB0aiqmx3Tvkewoh8MqdA+Dj6YpX7xpwXRNBENkTTv9H7W575jmUV9fh9kEd\nO9XdwFAf7H1ZxxEgyaHxt5/a3bq0M/DxdMUNvQI6/Hsz4MnR8R1A7aqq1oDk9AJMHKDeeNpEjozv\nOmpX2481dtVYz3RoRI6EIU/tat3BfGi8XDEqinPHEKmBIU/tpqrWgOSMQkzs35VdNUQq4TuP2s22\nzCJcUOGuGiK6jCFP7WZd2hn4erliJLtqiFTDkKd2cfmuGnbVEKnJ7LtPCJGgTNI9/xrbFyiPic2t\nI8eyNbMIF2sM7KohUlmzIS+EiAUAKWUyGqb5izWxW6IQIguXJ/S+1jpyIOsOKl01keyqIVKTuZb8\nDAB65Xk2AJ2JfR6RUkYpfwiaW0cOouGumgLcOrArP3FKpDJz70ANgBKjZVPNskgT3Tmm1pGD2HK0\nCBU1Btweww9AEamtzc0sKeVCpcXuL4TQXWsdOY51aWfg18kNIyI5fjuR2syFvB5A4ztVA6DYeKMQ\nIlEIkaAsFqOhBX/VuqYvquyTIoRIKSoqan31ZHWqag3YmNFwVw27aojUZ+5duAyXQzoSQDIACCE0\nyrqUxnUAopRlU+uuIKVMklJqpZTawMDA1ldPVmfL0UJU1BhwB++qIbIKzY4nL6VMFUJolS4XvZQy\nVdm0EUCcsj1RCFECIKtxu6l15BjWHjwD/05uGN6TXTVE1sDspCFSyiQT6+LMbL9qHdm/yhoDNmYU\nYkpsKLtqiKwE34lkMVuOFqKy1oA7YthVQ2QtOP2fncgtqcC8pftwRl+FYG93BHt7KF/uCFKed1WW\nfTxdIYSweA1r084goLMbhrGrhshqMOTtwN6cEiR+uRe1hnpM6N8VRReqkVNcgV9PlkBfUXvV/u4u\nTujbzRtDwzQYGq5BbLgvuvt6tin4K2sM2JRRiKnsqiGyKgx5G/f9vjzMX3kQIRoP/PvBeEQGdr5i\ne1WtAUXl1ThbVoWCsioUlFUjX1+JtLxSLN1zCv/ZcRIAENDZHUPDL4f+oO4+8HJr+a/HZqWrhmPV\nEFkXhryNqq+X+FtyJj7cdBwjIv2waFYcfDu5XbWfh6szwvy8EObnddW2WkM9jp4tx75cPfadOo99\np/TYkF4AAHB2EujfzRtzRkXg7iEhZlvn6w42dNUM78mxaoisCUPeBlXWGPD7FQewLu0MZmjD8Prd\nA+Hmcv1dJK7OThgY6oOBoT6YPaIHAOD8xRrsV0J/Q0Yhfr/iAD7efBzzxvfCXYND4ex0dZdORU0d\nNh4pQEJcd5PbiUg97Dy1MYVlVbg3aSfWHzqDP9zWD2/fE9OqgL8W305uuLlvEJ65pQ/Wz7sRSbPj\n4OHqjKeXHcAtf9uK1QfyYaiXVxyz+UgRqmrrOVYNkRViyNuQw/mlmPzxLzhWeAFJs7V45KbIdrlL\nppEQArcM6Ip1T9yIRbNi4eLkhHnf7MOt72/DuoNnUK+E/bq0fAR0duddNURWiN01NmJDegGeXLoP\nPp6uWPHbkRgQ4tNh39vJSWBSTDdMHNAV6w+dwQfJx/D4klT07doFj46NwqYjhZgWF8auGiIrxJC3\nAUt2n8Ifvk/DoFAffPaAFkHeHqrU4eQkcMegEEwa2A1rD+bjg43H8OTS/QDAu2qIrBRD3sqdKq7A\na2sOY3R0ID69Pw6ebs5qlwRnJ4HJQ0Jxx6AQrD6Qh6NnLyA+gl01RNaIIW/FpJR4dc1huDgJLLxn\nkFUEvDFnJ4EpQ7urXQYRNYMXXq3YhvQCbDpSiKcn9EZXH3W6aIjItjHkrVRFTR1eW5OOPsFdMGdU\nhNrlEJGNYneNlfpo03Hk6SuxfO5IuHIsGCJqJaaHFTpeeAGfbc/G1NhQ3ntORG3CkLcyUkq8svoQ\nPFyd8eKkfmqXQ0Q2zmzICyEShBA6IcT8a2xfoDwmtvQYura1B8/gl+PFeG5iHwR2cVe7HCKycc2G\nvBAiFgCklMkA9I3LTSQKIbIAZF/HMWTCheo6vLEuHQNDvTFreA+1yyEiO2CuJT8DgF55ng1AZ2Kf\nR6SUUUqot/QYMuH9DZkoLK/GG3fHcIgAIrIIcyGvAVBitGxqsPDIJl0zLTmGmjhytgyf7ziJe+PD\nMSRMo3Y5RGQn2nzhVUq5UGnF+wshWtRqF0IkCiFShBApRUVFbS3B5kkp8cfvD8HbwwXzJ/ZRuxwi\nsiPmQl4PoPEePg2AYuONSlgnKIvFACLNHQMAUsokKaVWSqkNDAxsbe1249vUPOw5eR4vTOprcnYn\nIqLWMhfyy9AQ3FAekwFACNHYn5DSuA5AlLJs8hgyrbSiFm+tz0BsuAbT4sLULoeI7EyzIS+lTAUA\npRtG37gMYKPR9ulKaz5LSpnazDFkwjs/HcX5ihq8fvdAOPFiKxFZmNlhDaSUSSbWxZnZftU6ulra\n6VJ8tTsHc0ZGdOgkIETkOPiJVxV9uOkYfL3c8MwtvdUuhYjsFENeJfn6SiRnFODe+DB4e7iqXQ4R\n2SmGvEq++fUUJICZw8LVLoWI7BhDXgU1dfVYuicX4/oEIczPS+1yiMiOMeRV8FP6WRSVV+P+kRyf\nhojaF0NeBYt35iDMzxNjovlBMCJqXwz5DnasoBy7T5Rg1vAevC+eiNodQ76DfbUrB24uTpiu5adb\niaj9MeQ70MXqOnybmofbY7rBj2PUEFEHYMh3oO/35+FCdR3uH8ELrkTUMRjyHURKicU7c9C/mzdi\nwzlePBF1DIZ8B0k9dR5HzpZj9sgeEIIXXImoYzDkO8jinTno4u6CyUNC1C6FiBwIQ74DFF+oxvq0\ns7gnrju83MwO/ElEZDEM+Q6wPOU0agz1uH8Ex6khoo7FkG9nhnqJr3fnYESkH3oFdVG7HCJyMGZD\nXgiRIITQCSHmm9lvvtHzBcpjYttLtG3bMotw+nwlZo+IULsUInJAzYa8ECIWAKSUyQD0jcsm9tMB\nmGC0KlEIkQUg21KF2qrFu3IQ2MUdtwwIVrsUInJA5lryMwDolefZAHQtfN1HpJRRyh8Hh5VbUoHN\nRwsxMz4Mrs7sGSOijmcueTQASoyW/ZvuIISINRHmkS3p4rF3S349BSchMHM4L7gSkTos0bz0a7pC\nSrlQCX5/pSvH4VTXGbBsTy50/YLQzcdT7XKIyEGZC3k9Loe4BkCx8UZTrXghRKIQIkFZLAYQ2fRF\nlX1ShBApRUVFravcyv037SxKLtZwnBoiUpW5kF+GyyEdCSAZAIQQjYOvRCp33yQC8FMuzKY07gcg\nSlm+gpQySUqplVJqAwPtc+KMr3bloGdAJ9wQFaB2KUTkwJoNeSllKnDp7hl94zKAjcr2lVLKlco6\njdEx05XWfJbRMQ7jUF4pUnLOY9bwcE4MQkSqMvsZeyllkol1cSb2SWruGEdRa6jHi6vS4Ovlimlx\nnBiEiNTFgVQs7MNNx5GWV4pFs2Lh4+WqdjlE5OB487YF7c/V4+PNxzF1aCgmxXRTuxwiIoa8pVTW\nGPDMsv0I7uKOV+4aoHY5REQA2F1jMQt+PILscxfx9cPD4ePJbhoisg5syVvAz8fO4T87TuKhGyJw\nQy/eMklE1oMh30allbV4buUBRAV2wvO39lW7HCKiK7C7po1eXX0YheXV+O6xUfBwdVa7HCKiK7Al\n3wbr087gu315eGJcLwzqrjF/ABFRB2PIt1JhWRX+8F0aBnf3weM391K7HCIikxjyrSClxAur0lBR\nY8C704dwrHgislpMp1ZYuicXm44U4sVJfdErqLPa5RARXRND/jqdKq7A62vTcUMvfzwwMkLtcoiI\nmsWQvw6GeolnV+yHs5PAXxMGc4RJIrJ6vIXyOizdcwp7Tp7Hu9MGI0TD2Z6IyPqxJd9C+ooavPO/\noxgR6YepsaFql0NE1CIM+RZ696dMlFXV4dW7BkAIdtMQkW1gyLfA4fxSfL07B7NH9EDfrt5ql0NE\n1GJmQ16Zw1UnhJhvZr/513uMLZBS4tXVh6HxcsPTut5ql0NEdF2aDXllYm5IKZMB6BuXTeynAzDh\neo6xFT/sz8eek+cxf2IfzvRERDbHXEt+BgC98jwbgK4Fr9maY6zSheo6vLk+A4O6+2C6lvO1EpHt\nMRfyGgAlRsv+TXcQQsQqrfYWH2MrPtx0DIXl1XjtrgG8J56IbJIlLrz6WeA1rE5W0QX8++cTSIjr\njqHhvmqXQ0TUKuZCXo/LIa4BUGy80UQr3uwxynGJQogUIURKUVHR9VfdzqSU+POadHi4OHMiECKy\naeZCfhmASOV5JIBkABBCNA6eHqncSZMIwE+5yGryGGNSyiQppVZKqQ0MDGzrOVhcckYhtmYW4Uld\nNAK7uKtdDhFRqzUb8lLKVODS3TP6xmUAG5XtK6WUK5V1GjPH2ISqWgNeX5uO6KDOmDMqQu1yiIja\nxOzYNVLKJBPr4kzsk9TcMbbis23ZOFVSga8fHs5x4onI5jHFjOTpK/HxluOYNLArbugVoHY5RERt\nxpA38ua6DADAH27vp3IlRESWwZBX7Dh+DuvSzuDRMb3Q3ddL7XKIiCyCIQ+gvl7i1TWH0d3XE3PH\nRJo/gIjIRjDkAWw9VoTMggv4/S194OHqrHY5REQWw5AH8OWOkwjs4o7bYrqpXQoRkUU5fMifPHcR\nWzKLMHNYONxcHP6fg4jsjMOn2le7cuAsBGYND1e7FCIii3PokK+oqcPylFxMHNgVwd4eapdDRGRx\nDh3yP+zPR1lVHeaMjFC7FCKiduGwIS+lxBc7TqJv1y6Ij+BQwkRknxw25FNyzuPI2XLMGRUBITgh\nCBHZJ4cN+S92nIS3hwsmDwlRuxQionbjkCFfUFaFHw+dxXRtGLzczA7ESURksxwy5JfsPgWDlLh/\nRA+1SyEialcOF/I1dfVY8uspjO0diIiATmqXQ0TUrhwu5H88fBZF5dV4gLM+EZEDMBvyyhyuOiHE\n/Gts1ym/hJLEAAAHpElEQVRfC4zWLVAeEy1XqmV8ueMkevh7YUy09c0tS0Rkac2GvDIxN6SUyQD0\njctNtk9QtscabU8UQmQByG6HmlvtUF4pUnLOY/aIHnBy4m2TRGT/zN1aMgPABuV5NgAdgEsTcyuT\ndDcuRxpN2j1NCX6rsnhnDjxcnTAtLkztUoiIOoS5kNcAKDFa9je1k9KVM9doVazyAaNYKeXCNlVo\nIfqKGny/Pw9TY0Ph4+WqdjlERB3CIhdelSCfK4TQNC4rLXl/IYTOEt+jrZan5KK6rh6zR0SoXQoR\nUYcx15LXA/BTnmsAFBtvNOqzT0VDd06iECJbWbdS2f+q+fSUC7KJABAe3v5D/BrqJRbvysGwCD/0\nD/Fu9+9HRGQtzLXkl+FySEcCSAaAxhY7Gvrojf8IZCtfjf3xUQBSmr6olDJJSqmVUmoDA9v/Lpet\nmYXILanEA6P44ScicizNhnzjhVSly0VvdGF1o/KYBCBSCJGg7L9S2We6si7L6BjVfLEjB0Fd3DFx\nQFe1SyEi6lBmB26RUiaZWBenPOrREPQAsLK5Y9Ry4txFbM0swtO63nB1drjPfhGRg7P71Fu8Mweu\nzgIzh/O2SSJyPHYd8lW1Bqzcm4uJA7oiqAun9yMix2PXIb8hvQBlVXWYEc9WPBE5JrsO+RV7TyPE\nxwOjogLULoWISBV2G/L5+kpsP1aEhLjucOY4NUTkoOw25FelnoaUQALHqSEiB2aXIS+lxIq9pzG8\npx/C/b3ULoeISDV2GfK/nihBTnEFpmvZiicix2aXIb9i72l0dnfBpBh+wpWIHJvdhfyF6jqsTzuD\n22O6wcvN7Ad6iYjsmt2F/PqDZ1BRY8D0+O5ql0JEpDq7C/kVe3MRGdgJseG+apdCRKQ6uwr57KIL\n2HPyPKbFhUGZmYqIyKHZVciv3HsaTgKYGhuqdilERFbBbkLeUC+xKjUPY3oHItibg5EREQF2FPLb\njxXhbFkV740nIjJiNyG/IuU0fL1cMb5fsNqlEBFZDbMhL4RIEELohBDzr7Fdp3wtaOkxlqavqMGG\n9AJMHhIKNxe7+btFRNRmzSaiECIWAKSUyQD0jctNtk9QtscKIWLNHdMeftifjxpDPbtqiIiaMNfs\nnQFArzzPBqAz3iilTJVSPq8sRiqTdjd7THtYnpKLASHe6B/i3d7fiojIppgLeQ2AEqNlf1M7Kd0y\nc6/nGEtJzy/D4fwyTIvjJ1yJiJqySAe2lHIhgLlCCI0lXu96rNibCzdnJ0wewnvjiYiaMhfyegB+\nynMNgGLjjcZ98Gjomkk0d4xyXKIQIkUIkVJUVNTa2lFTV4/v9+VhQv9g+HZya/XrEBHZK3MhvwxA\npPI8EkAyABi12HW4MtCzr3WMMSllkpRSK6XUBgYGtrr4jRkFOF9RiwQtu2qIiExpNuSVC6kQQugA\n6BuXAWxUHpMARAohEpT9VzZzjMWt2HsaXb09cFN06/9QEBHZM7MDrkspk0ysi1Me9WgIegBY2dwx\nllZQVoUtRwvx2zFRnKibiOgabPaTQ6tS81AvgQTeVUNEdE02G/KVtQbc1DsQkYGd1S6FiMhq2ez8\neM9M6A0ppdplEBFZNZttyQPgxCBERGbYdMgTEVHzGPJERHaMIU9EZMcY8kREdowhT0RkxxjyRER2\njCFPRGTHhNofKBJCFAHIaeXhAQDOWbAcNfAcrAPPwXrYw3l0xDn0kFKaHZ1R9ZBvCyFEipRSq3Yd\nbcFzsA48B+thD+dhTefA7hoiIjvGkCcismO2HvLtPm59B+A5WAeeg/Wwh/OwmnOw6T55Uo8QItZ4\n1i9ldjA9gFhlYnerZ+IcFkgpnxdCJHbExDdEHcFmW/JCiAQhhE4IMV/tWlpLCLFAeUxUu5broUzt\nuMJoORYApJTJAPRGk7tbLeUcPmuyOlEIkYWGuYptghAiUflaYLTOpt4b1zgHm3pvKP/eOmv8Odhk\nyNtiqFyDzYUKcOnf3bjmGWhoxUNZr+vwoq6Tcg4lTVZPk1JGKdusnvKHKln5X0ekEig29d4wdQ7K\nJpt5byg1T1P+zWOFELHW9HOwyZCHDYbKNdhUqDRDgysD01+tQtoo1hpaXtchEpd/97OVZVt7b5g6\nBwB4xFbeG1LKZCnlXGUxUukCtJqfg63ODGVPoQLYUD+2PWv8GQghJgghdNYeME2uG8QCWAYgDjb0\n3rjGOQCXW/U2895QGgeNYW81GWWrLXm7IKVcqASJv9F/U22RHoCf8lwDoFjFWlpF6T9NUBaLcblF\nafWUroBU44vItqbpOdjie0P5YzRXCKFRuxZjthryDBXrsgyX648EYNUt4GvIxuW6owCkqFjL9dJJ\nKZ9Xntvqe+PSOSgXYW3mvWHcB4+G36NEWNHPwVZDnqGiIuUNqG18Iza2vpQWl94WWpTXOIfpynKW\nLZwD0BCIRt1MOtjge8PEOaTAtt4bOlwZ6Nmwop+Dzd4nr9xalY2GCx02eU+zcg4laDgHm+h3JOth\ndCtrCRpCZpqUMtmW3htmzsEm3htK98x0ZTGu8SKstfwcbDbkiYjIPFvtriEiohZgyBMR2TGGPBGR\nHWPIExHZMYY8EZEdY8gTEdkxhjwRkR1jyBMR2bH/B0DuMkyQi6ECAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f07362084e0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot([np.mean(data[i,1]) for i in range(32)])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD7CAYAAAB68m/qAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3Xl4VNX9BvD3TFaykxUIBAhZQAhLEnZFwABqRQTBDbW2\nliBobVGrxVqtdavYWncRrfpDRFSUTWuBgAoKAoEASVgChCU72Sb7Mpk5vz8ygRBCJstk7tw77+d5\neGa5NzPfS3LfnJx75hwhpQQREamfTukCiIjIOhjoREQawUAnItIIBjoRkUYw0ImINIKBTkSkEQx0\nIiKNYKATEWkEA52ISCMY6EREGuFsyzcLDAyUAwYMsOVbEhGp3v79+4uklEGW9rNpoA8YMADJycm2\nfEsiItUTQpxtz37sciEi0ggGOhGRRjDQiYg0goFORKQRDHQiIo1goBMRaQQDnYisav/ZUhRX1ild\nhkNioBORVUgp8e+tGbj13V2Y9fbPOFNUpXRJDoeBTkRdVtdgxJLPD+L1bSdww7BeqKprwNzlu3E0\nr1zp0hwKA52IuqS0qh73fLAX6w/m4k8zovHO/Fh8+cB4uDgJ3P7ebuw/W6J0iQ6DgU5EnZZZWInZ\n7/yMg9l6vHnnKDw4JQJCCEQEe+PLB8YjwMsN8z/Ygx+On1e6VIfAQCeiTtmTWYw57+5CeW0DPlsw\nFjNH9Llke9+eHvjygfEID/TCgpXJ2HQoV6FKHQcDnYg6bF1KNu7+zx74e7pi3eIJiOvv3+p+gV5u\nWLNwHEb164mH16Tg0z3tmmOKOomBTkTt1jSSZcnnhxDf3x/rFk1E/wDPNr/Gx90FK+8fg6nRwfjL\nujS8/f1JSCltVLFjYaATUbs0H8kyN64v/u+3Y+Dr4dKur3V3ccLye+Jwy8g+eGXzcbz436MM9W5g\n0/nQiUi9fr86BVuOFOBPM6KxePIgCCE69PUuTjq8ettI+PZwwfs7T0NfbcBLc2Lg7MR2pbUw0InI\not2nii+E+YNTIjr9OjqdwN9uHgpfD1e8se0Eiirr8OZdsfByYxRZA381ElGbpJRYtvkYevm44/6r\nB3b59YQQeGRaFF6YPQw7ThRh3vLdyCursUKlxEAnojYlHT2PlHN6/CEhEu4uTlZ73flj++PD+0Yj\nq6Qas976GWk5ZVZ7bUfFQCeiKzKZJP65+TgGBnpiblxfq7/+tVFB+GrRBLg46TBv+W5sPVJg9fdw\nJAx0IrqijYdycbygAkumRcGlmy5eRvfyxroHJyAqxAuJnyTjPz+d5giYTmKgE1Gr6htMeHVrBq7q\n7YObYnp363sFe7tjTeJ4XD+0F5775gie3pCOBqOpW99TixjoRNSqz5OzcK6kGn+aEQ2drmNDFDuj\nh6sT3r4rFguvDccnv5zF71Ymo6LW0O3vqyUMdCK6TE29EW9uO4HRA3picnSQzd5XpxNYesMQvDQn\nBjvNI2By9BwB014c/ElEl/l41xmcr6jD2/NjO/wBImu4c0wY+vX0wKJV+3HTGztxTWQQhoX6YFgf\nXwwN9YVvj/Z9QtXRMNCJ6BJlNQYs//EUpkQHYfSA1ifdsoWrIwPx9eIJ+NeWDOw/W4qNzWZrDPP3\naAz4UF8M6+OLYaG+8Pd0VaxWe8FAJ6JLrNhxCmU1Bjw2I1rpUhAZ4o3l98QBAIor65CeW4603DKk\n5ZQhLacc/03Nv7BvTKgvVv52DHo6cLAz0InogvMVtfjwpzOYOaIPhvbxVbqcSwR4uWFSVBAmRV3s\n0y+rNiA9twwHs/V4besJLP70AFbeP6bbhljaO8c8aiJq1dvbT6LeaMIj06KULqVdfD1cMCEiEIsn\nR+ClOTHYnVmM5785onRZimELnYgAAFkl1Vi99xxui++LgYFtz3Fuj26N64vjBRVYsSMT0b18cNfY\nMKVLsjm20IkIAPDvpAwIIfDwdZFKl9JpT1w/GNdGBeHpDWnYk1msdDk2x0AnImQUVGBdSg5+Pb4/\nevv2ULqcTnPSCbxx5yiEBXhg0acHkFVSrXRJNsVAJyL8a8txeLo6Y9Hkzs91bi98e7jgg3vjYTCa\nsGBlMqrqGpQuyWYY6EQ2kFdWY1fBUtdgRH5ZLY7klmNdSjY2pxdgwTXhmhnLHR7khbfuikVGQQUe\n+eIgTCbHmOyLF0WJulFmYSXe3H4SGw7mYPygAKy6f6xNPnlZ32DCTycLsftUMYqr6lFSVY/SqnqU\nVNejtMqAyha/XIK93XD/NV1fvMKeXBsVhCdvHILnvz2K17edwBKVjNzpCgY6UTc4U1SFN7afwPqU\nHLg663B1ZBB2ZBTim8N5mDmiT7e8p9EksSezGJsO5+K7tHzoqw1wc9Yh0MsN/p6u6OnpioGBnvD3\ndIO/pwt6errC36Px+cG9vDW5DNz9Vw/EsfwKvL7tBKJ7eePGbp41Umna+w4SteF0URVW7j6DJdOi\n4ONu/flAzhZX4c3tJ7EuJQfOOoHfThyIhdcOgr+nK25+6ye88O1RTB0cDE8rhaeUEgfO6bHpUC6+\nTc1DYUUdPFydMP2qEMwc0QfXRAbB1dlxe1aFEHhh9jBkFlbi0S8OoX+Ah919YMqahKWJ5IUQcwHo\nAcRKKZe1sj0WQDgASCnXtvVa8fHxMjk5ufPVEnVBWY0Bs9/+GZlFVbh7XBievyXGaq+dVVKNN7ef\nwFcHGoN8/tj+eGByOIK93S/ss/9sKW59dxcWXhuOpTcM6dL7Hcsvx/qUXGw6lIscfQ1cnXWYGh2M\nmSP6YOrgYPRwtd5ScVpwvqIWs976GTohsOGhiQj0clO6pA4RQuyXUsZb2q/NZoI5rCGlTBJChAsh\nYqWUB1rstlRKOU8I8fgVthMpzmiS+MOaFGSVVmNydBBW/XIOt4wMRXwXJ5/KL6vFa0kZWLs/Gzqd\nwD3j+mPx5EEI9nG/bN+4/j0xN64v/rPzNObF9UNEsFen3vPrA9l45ItDcNIJXBMZiEemRWHa0JBu\n+YtDK4K93fH+vfGYu3wXJi37HgFervBxd4GPuwt8e7jAp4dzs/uNt5Ojg+Dnoa6LxG220IUQLwPY\nag70BLRopZtb7+GttdxbwxY6KeUf3x3D8h9P4cXZMZg1sg+m/3sHPFyd8O3D13S6S6Ki1oCZb/6E\nXH0t7hzTD4smR6CX7+VB3lxRZR2m/PMHjOjrh0/uH9PhC6Sp2WWYu3wXRoX54Z35cZoZlWIrezKL\n8d/UPJTXNqC8xoCyGgPKaw0or2lAWY0BNQbjhX2H9vHBhgcnwtkO5oWxSgsdgB+AkmaPA1psH21+\ns1gACe0NdiJb2nAwB8t/PIW7x4Vd+Dj4c7cMxW8/TsaKHafw0NSOfzJSSomlX6ciq7QGny0YhzED\n29fSD/Ryw2PTo/HMxnR8l5bfoYt0xZV1eGDVfgR4uuLtu2IZ5p0wNjwAY8NbxthF9Q0mVNQasP3Y\nefxp7WG8v/M0Fk0eZMMKu8Yav3qKm7pZzC32SwghEoUQyUKI5MLCQiu8HVH7pWaX4fG1hzFmoD+e\nvmnoheenDg7Br2J6443tJ5FZWNnh11299xy+OZyHR6dHtTvMm8wfG4YhvX3w/DdHUF3fvrHpDUYT\nHlqdgsLKOrx3TzwCVNYHrBauzjoEeLlhXnw/XD+0F15LyujUz4dSLAW6HkDTT6sfgJaTIxQDyGy2\n7+iWLyClXCGljJdSxgcF2W4pK6LCijokfpKMQC83vDM/9rKulWdmXgU3Zx2eXJfaoVXmj+SW49lN\nR3BtVBAemNTx1puzkw7PzRqK3LJavLX9ZLu+5h/fHcPuzGK8NDsGMX21O0rDnvx91lC4Oevw569T\nu/TBJCklko4UdOhnrLMsBfrnMI9gMd8mAYAQws/83Npm2/0A7LN2gUSdUd9gwuJP96O0uh7v3RPX\n6qiGYB93LL1hCH7JLMGX+7Pb9bqVdQ14cPUB9PRwwau3jej04snxA/wxJzYU7+/MxCkLLcANB3Pw\nwU+ncd+EAbg1rm+n3o86LtjHHU/ddBX2ni7B6r3nOv06b20/id+tTMb/0vIt79xFbQZ6s66UBAD6\nZiNYtpm3ZwLQm7taAiwNWySylb9tSse+M6V4Ze4IDAu9cov2jtH9MHpAT7z436Moqqxr8zWllPjL\nulScLa7CG3eM6nK3x9IbhsDd2Ql/25h+xdZbem4ZnvjqMMYM8MdfftW1oY7UcfPi+uLqiED847tj\nyO3EYtVfJmfhX1szMGdUKK4f1qsbKryUxT50c5dJkpRyRbPn4lpsXyulfKK7iiTqiE9+OYvVe85h\n8eRBFj+VqdMJvDQnBlV1DRYXRvh8XxY2HMzFI9Oi2ryw1l5B3m5YMi0KO08UYXP65a230qp6LPxk\nP/x6uOLt+bEOuwqPkoQQeHF2DIwmiafWp3Wo2+THjEIs/Tq18RfCrcNtMuUDf0JIU37JLMazG9Mx\ndXAwHp3evjUxI4K9sWhyBNYfzMWPGa1fuD+aV45nNqbjmsjG1XGs5d7x/TG4lzee++YoauovDplr\nMJrw+89ScL68DsvviUOQNy+CKiUswAOPzYjG9mPnL1moui1pOWVYtGo/okK88e7dl1+/6S4MdLJ7\nW9LzsXrPOWxOz8f+syU4U1SFilrDZa2l7NJqLP70AMICPPDaHSPh1IH+7cWTByE8yBNPrU+9JFgB\noMrcb+7TwwWv3jay0/3mrXF20uHvs4YhR1+Dt7+/eIH0lS3H8dPJIjx/yzCM7OfXxiuQLdw3YQBG\n9vPDs5uOoNhC11xWSTXu+2gfenq44qPfjIa3DT/wxblcyK4dzSvHwlX70dpfuq7OOgR6uiLAyw0B\nXq44U1QFg9GE9++N7/CnJt1dnPDi7BjcseIXvLYt48JH86Vs/FP7TFEVVv1ubLe0lMcM9MfsUaFY\nsSMTt8b1RXpuGd77MRPzx4bhttH9rP5+1HFOOoFlc4fjV2/sxN+/OYLX7xjV6n6lVfX49Ud7YTCa\nsCZxLEJa+cRwd2Kgk13715YMeLk5Y93iCag1mFBUWYfiynoUVzXeFjW77+Kkw9t3xWJQUOc+Uj8u\nPAC3x/fDBztP42bzqvdfJmdjXUoOliREYcKgQCsf3UVLbxiMrUcKsOTzgzieX4G4/j3xzMyhlr+Q\nbCYqxBsPTonAa0kncPOIPrhuSMgl22sNRtz/f/uQXVqDT383FhHB3javkYFOdivlXCmSjhbgselR\nNjs5lt44GNuOFWDp16l4+dbheHpjGiYMCsBDU7t3JZ9gH3f8MSESz397FMHebni3lXHzpLzFkyPw\nXWo+nlqfhjED/S90pxhNEg9/loKULD3euSsWo7s4R1Bn8SeG7NY/txxHgKcrfjPRdgsv+Hm44umZ\nQ3E4uwxz390FLzeXDvfHd9avJwzA4smD8OF9o1ud3IuU5+qsw8tzh6OgvBYv/+8YgMZuuWc3pWPL\nkQI8fdNVuEHBOdcZ6GSXdp0sws8ni7F4SoTV5g5vr5nDe2NydBCqDUa8fsfIS6bA7U4uTjo8fv3g\nNsfNk/JG9vPDbyYOxKpfzmFPZjGW/5iJlbvPInFSuE0bH62xOB+6NXG2RWoPKSXmvLsL+WW1+P6x\nyXB3sf3c3lV1DThdVMVwpVZV1zdgxms7UF1nRHFVPWaO6IPXb7fuCKjm2jvbIlvoZHe2HzuPlHN6\nPHxdpCJhDgCebs4Mc7oiD1dn/GPOcBRX1WN8eAD+OW94t4V5R/CiKNkVk0nilc3H0T/AA3M5bwnZ\nsYkRgfj24asRHugFN2f7WCGKLXSyK9+m5uFYfgUemRbFj7qT3Rvax9eulvvjGUN2o8Fowr+3ZiA6\nxBszh7c9BwsRXY6BTnbj6wM5yCyqwiPTo+yiP5JIbRjoZBfqGox4fdsJjOjri+lXhVj+AiK6DAOd\n7MJne84hR1+Dx2ZE22SaUSItYqCT1Z0trsKR3PJ2719d34C3vj+FceH+uDqi++ZLIdI6Dlskqyiq\nrMM3h3Kx/mAuDmbpATTOIvjQlAhcExnYZqv7411nUFRZh/fuiWXrnKgLGOjUaZV1DdiSno8NB3Px\n08kiGE0SQ3r7YOkNg+HspMMHOzNx74d7MbyvLxZPjsD0q0Iuu9hZVmPAez9mYurgYMT1V2ZCIyKt\nYKBThxiMJuzIKMT6g7nYeiQftQYTQv16YOGkcNwyKhRRIRdnRbxnXH+sS8nGuz+cwgOr9iMi2AuL\nJw/CzSP6wNk8xvyDnZkoqzHg0elRSh0SkWZwLhdqt+3HCvDoF4dQWm1ATw8X/Gp4b8waGYq4sJ5t\nDjNsMJrwbWoe3vn+FI4XVKCffw8snDQIUwcHI+HVHzElOhhvz4+14ZEQqUt753JhC53aJaOgAr9f\nnYJ+/h7457wRuCYyqN3zdTs76TBrZChmDu+DbcfO463vT+Kp9WlwcRIwmiSWTGPrnMgaGOhkkb66\nHgtWJqOHqzM++s1o9Pbt0anX0ekEpl0VgoQhwdh1qhgrdmQiJtQXEcGdW2GIiC7FQKc2Na0+n6uv\nwZrEcZ0O8+aEEJgYEYiJHKJIZFUMdGrTss3HsfNEEV6aE8NRKER2jh8soital5KNFTsyce/4/rhz\nTJjS5RCRBQx0atXhbD2e+CoVYwf64683XaV0OUTUDgx0usz5ilos/GQ/grzc8M78WM5LTqQS7EOn\nS9Q3mLB41QGUVtfjq0UTEODlpnRJRNRODHS6QEqJZzamIflsKd68cxSG9uGamkRqwr+l6YJVe87h\ns71ZWDx5EGaO4IpBRGrDQCcAwJ7MYjy7MR1TBwfj0enRSpdDRJ3ALhcNk1Jic3oBcvQ1qDUYUddg\nQl2DEXWG5reN95PPliIswAOv3TESTlz+jUiVGOga1WA04an1aVizL+uS512ddXB31sHNxQluzjq4\nm29jQn3x7M1D4ePuolDFRNRVDHQNqq5vwEOrU7D92Hk8NCUCCyaFw81ZB1cnHRdfJtIwBrrGFFXW\n4f6P9yE1pwwvzB6G+WP7K10SEdkIA11DzhRV4dcf7UVBeS3euyce064KUbokIrIhi6NchBBzhRAJ\nQojHr7D9ZfNtorWLo/Y7mKXHre/uQnmNAasXjGOYEzmgNgNdCBELAFLKJAD6psctJAohTgHI7Ib6\nqB22HyvAnSt+gYebE75aNAGxYT2VLomIFGCphX47AL35fiaAhFb2WSClHGQOfbKxNXvPYcHKxvU6\nv140EeFBXCyCyFFZCnQ/ACXNHge0sk94W10y1D2klHh1awb+/HUqro4IxJrEcQjy5rwrRI6syxdF\npZTLAEAIMU0IkdCypW7uW08EgLAwzqltSa3BiENZelTVN6Cqzohq821VXQOq6i8+zi6txp7TJZgb\n1xcvzYnhjIhEZDHQ9QCalqnxA1DcfKM5rEuklGvN28JbvoCUcgWAFQAQHx8vu1qwlkkpce+He7H3\ndEmr212ddfB0dYKHqzM83ZzwpxnRWDx5EITg2HIishzonwOIN98PB5AEAEIIPymlHkAyLl4MHQTg\nve4o0lH8kFGIvadL8PB1kZg6OBhebubwdnVGD1cnuDqzFU5EV9ZmoEspDwgh4oUQCQD0UsoD5k3b\nAMSZtycKIUoAnGq2nTpISol/bTmOfv498NCUCIY3EXWYxT50c5dJy+fi2tpOHbc5vQBpOeV4Ze5w\nhjkRdQqTww6YTBL/3pqB8EBPzB4VqnQ5RKRSDHQ78E1qHo4XVOAPCZFw5mgVIuokpofCGowmvJaU\ngegQb8wczlWCiKjzGOgKW38wF5mFVVgyLYpT2xJRlzDQFWQwmvD6tgwMC/XBjKGcTIuIuoaBrqAv\nk7ORVVKDR6dF88NBRNRlDHSF1BqMeHP7CYwK88Pk6CClyyEiDWCgK2TN3nPIK6vFY9PZOici62Cg\nK6Cm3oi3vj+FsQP9MWFQaxNYEhF1HANdAZ/8cgZFlXV4lK1zIrIiBrqNVdY14N0fTmFSVBDGDPS3\n/AVERO3EQLexj346jdJqAx6ZFqV0KUSkMQx0GyqrNmDFzkwkDAnByH5+SpdDRBrDQLehD37KREVt\nA1vnRNQtGOg2UlJVjw9/Oo1fxfTGVX18lC6HiDSIgW4jb24/gWqDEX9MiFS6FCLSKAa6DRzNK8fK\n3Wdx55gwRIZ4K10OEWkUA72bSSnx1/Vp8HF3xuMzopUuh4g0jIHezb4+kIPks6X48w2D4efhqnQ5\nRKRhDPRuVFZjwEvfHcWoMD/Mi+undDlEpHEWF4mmznt1y3GUVNXj49+M4eIVRNTt2ELvJmk5Zfjk\nl7O4e1x/DAv1VbocInIADPRuYDJJ/HVDGnp6uOLR6bwQSkS2wUDvBmv3ZyPlnB5LbxwC3x4uSpdD\nRA6CgW5l+up6/ON/xxDfvyfmjApVuhwiciAMdCt7ZfNxlNUY8Nwtw3ghlIhsioFuRYez9Vi99xzu\nHd8fQ3pzvhYisi0GupWYTI2fCA30csMSzqZIRApgoFvJmn1ZOJRdhr/cOAQ+7rwQSkS2x0C3gpKq\neizbfAxjBvpj1sg+SpdDRA6KgW4Fr2w+horaBjw3axgXfSYixTDQu0BKiS+Ts7BmXxZ+O3EAontx\nalwiUg7ncumkoso6PPl1KrYcKcDoAT3xhwReCCUiZTHQO+F/aXl4cl0aKmsb8OSNg3H/1eFw4phz\nIlIYA70DyqoNeGZjGtYfzMWwUB+8ettIRHEFIiKyExYDXQgxF4AeQKyUclkb+z3e1na1++H4eTzx\n1WEUV9bjjwmReHBKBFyceAmCiOxHm4kkhIgFACllEgB90+NW9ksAMM365Smvsq4BT65LxX0f7YOP\nuwvWLZ6IPyZEMcyJyO5YaqHfDmCr+X4mgAQAB7q1IjuyJ7MYj609hOzSGiROCscj06Lg7uKkdFlE\nRK2y1Mz0A1DS7HFAyx2EELHmFrymnC6qwl0f7IGAwBcLx+PJG4cwzInIrlnjoqi/FV7D7mw4mAOT\nlPhi4Xj08nVXuhwiIosstdD1uBjYfgCKm2/UautcSolNh3IxbmAAw5yIVMNSoH8OINx8PxxAEgAI\nIfyanhNCzBVCJALwb+2iqRAiUQiRLIRILiwstFbd3epoXgVOFVZh5gjOy0JE6tFmoEspDwAXRrHo\nmx4D2GbevlZKudb8nF8rLwEp5QopZbyUMj4oKMhKZXevjYdy4awTuH5YL6VLISJqN4t96FLKFa08\nF9fKPpftp0ZN3S1XRwbC39NV6XKIiNqNg6lbSMnSI0dfg5nD2d1CROrCQG9h06FcuDrrMG1oiNKl\nEBF1CAO9GaNJ4tvDeZgSHcRVh4hIdRjozew9XYLzFXUc3UJEqsRAb2bT4Vx4uDph6uBgpUshIuow\nBrqZwWjCd6l5SBgSAg9XzipMROrDQDf7+WQRSqsN7G4hItVioJttOpQHH3dnTIoKVLoUIqJOYaAD\nqDUYsSU9H9cP6wU3Z86oSETqxEAH8MPxQlTUNbC7hYhUjYGOxtEtAZ6uGB9+2XTvRESq4fCBXlXX\ngG1HC3BjTG84c1k5IlIxh0+wpKMFqDWY2N1CRKrn8IG+6VAeevm4I75/T6VLISLqEocO9LJqA37M\nOI+bhveGTieULoeIqEscOtA3H8mHwSjZ3UJEmuDQgb7pUC7C/D0wvK+v0qUQEXWZwwZ6UWUddp0q\nxswRvSEEu1uISP0cNtC/S8uH0cTuFiLSDocN9E2HchEZ7IXoEG+lSyEisgqHDPS8shrsO1OCm0f0\nYXcLEWmGQwb6t4fzICVwE7tbiEhDHC7QpZTYeCgXMaG+GBjoqXQ5RERW43CBvjk9H4ezyzAvvq/S\npRARWZVDBXplXQP+tvEIBvfyxp1jwpQuh4jIqhxq8cxXt2SgoKIW79wdCxfOrEhEGuMwqZaWU4aP\nd53GXWPCEBvGibiISHscItCNJokn16XC39MNj18/WOlyiIi6hUME+qpfzuJwdhn+etMQ+PZwUboc\nIqJuoflALyivxSubj+OayEDczHHnRKRhmg/0v39zBPVGE56bNYyfCiUiTdN0oP9w/Dy+PZyHh6ZE\nYAA/REREGqfZQK+pN+KvG9IQHuSJhdeGK10OEVG30+w49De3n0BWSQ0+WzAObs5OSpdDRNTtNNlC\nzyiowIodmbg1ti/GDwpQuhwiIpvQXKCbTBJ/WZcKL3dnPHkjx5wTkePQXKCv3Z+NfWdKsfSGwQjw\nclO6HCIim9FUoBdX1uHF745i9ICemBfXT+lyiIhsyuJFUSHEXAB6ALFSymWtbE8w350mpXzCyvV1\nyCubj6OytgEvzI6BTscx50TkWNpsoQshYgFASpkEQN/0uNn2BADzzNtjW263pVqDERsO5mJefF9E\ncZ1QInJAlrpcbkdj6xwAMgEkNN8opUySUi40PwyXUh6wcn3ttvNEEWoMRtwY01upEoiIFGUp0P0A\nlDR73OoYQCHE4wAWtrbNVjan58Pb3RnjwjlMkYgck1Uuipr71hcKIfxabhNCJAohkoUQyYWFhdZ4\nu8s0GE3YdrQA1w0O5sIVROSwLKWfHoC/+b4fgOLmG4UQzfvNMwEktnwBKeUKKWW8lDI+KCioq/W2\nau+ZEpRWGzBjaK9ueX0iIjWwFOifA2iaCCUcQBIANGuJJ+DSwM+0doHtsSW9AG7OOlwb3T2/MIiI\n1KDNQG+6yGkezaJvdtFzm/l2BYBwIUSief+13VVoGzViS3o+rokMgoerZqemISKyyGICSilXtPJc\nnPlWj8ZQV0xaTjlyy2qxZFqUkmUQESlO9VcQN6fnw0knkDAkROlSiIgUpYlAHzPAHz09XZUuhYhI\nUaoO9MzCSpw4X4npQ9k6JyJSdaBvTi8AAEzncEUiIrUHej5iQn0R6tdD6VKIiBSn2kAvKK/FwSw9\nZrC7hYgIgIoDfcuRxu4WfjqUiKiRegM9PR/hgZ6ICPZSuhQiIrugykAvqzZg96liTB/aC0JwIQsi\nIkClgb79eAEaTJL950REzagy0DenFSDY2w0j+l42Wy8RkcNSXaDXGoz4MaMQ04eGcN1QIqJmVBfo\nTUvNcXQLEdGlVBfom9Pz4cOl5oiILqOqQL+w1NyQEC41R0TUgqpS8eJScxzdQkTUkqoCvWmpuUlR\nXGqOiKiQ2FchAAADXklEQVQl1QQ6l5ojImqbagI9NacMuWW17G4hIroC1QT6lvQCLjVHRNQG1QQ6\nl5ojImqbKgK9aak5drcQEV2ZKgKdS80REVmmikA3GE24OiIQfbjUHBHRFali/N/D10UqXQIRkd1T\nRQudiIgsY6ATEWkEA52ISCMY6EREGsFAJyLSCAY6EZFGMNCJiDSCgU5EpBFCSmm7NxOiEMDZTn55\nIIAiK5ajBB6DfdDCMQDaOA4eQ/v0l1JaXNnHpoHeFUKIZCllvNJ1dAWPwT5o4RgAbRwHj8G62OVC\nRKQRDHQiIo1QU6CvULoAK+Ax2ActHAOgjePgMViRavrQSTlCiFgp5YFmj+cC0AOIlVIuU66y9mvl\nGF6WUj4hhEiUUtrNCUnUFapooQsh5gohEoQQjytdS2cJIV423yYqXUtHCCESAHzZ7HEsAEgpkwDo\nmx7bM/MxvN/i6UQhxCkAmQqU1ClCiETzv5ebPaeqc+MKx6Cqc8P8/51gj98Huw90NQbIFaguQIAL\n/+/Na74dja1zmJ9PsHlRHWQ+hpIWT8+TUg4yb7N75l9KSea/JsLN4aGqc6O1YzBvUs25Ya55nvn/\nPFYIEWtP3we7D3SoMECuQFUB0gY/XBqOAUoV0kWx9tCi6oBwXPzZzzQ/Vtu50doxAMACtZwbUsok\nKeVC88Nwczee3Xwf1LBikZYCBFBRv7OWNX0PhBDThBAJ9h4mLfr5YwF8DiAOKjo3rnAMwMXWumrO\nDXNDoCnY7Saj1NBC1wQp5TJzaAQ0+1NTjfQA/M33/QAUK1hLp5j7O+eaHxbjYkvR7pn/nD/Q/AKv\n2rQ8BjWeG+ZfPAuFEH5K19KcGgKdAWJfPsfF+sMB2HXL9goycbHuQQCSFayloxKklE+Y76v13Lhw\nDOYLpKo5N5r3maPx5ygRdvR9UEOgM0AUZD7Z4ptOuqZWlbklpVdDS/EKx3Cb+fEpNRwD0Bh+zbqK\nEqDCc6OVY0iGus6NBFwa3pmwo++DKsahm4czZaLxIoQqxwybj6EEjcegin5Csh/Nho+WoDFQ5kkp\nk9R0blg4BlWcG+YultvMD+OaLpDay/dBFYFORESWqaHLhYiI2oGBTkSkEQx0IiKNYKATEWkEA52I\nSCMY6EREGsFAJyLSCAY6EZFG/D8deCSjTgFScAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f0736208da0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot([np.mean(data[i,0]) for i in range(32)])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlUAAAJaCAYAAAAGULJnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nUAUA0IFQBQDQgVAFANDB/w+GVnNHDyMvGwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f07360c5080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rc(\"text\", usetex=True)\n",
    "my_font_size = 15\n",
    "#PyPlot.rc(\"font\", family=\"serif\")\n",
    "\n",
    "plt.rc(\"text.latex\", unicode=True)\n",
    "fig = plt.figure()\n",
    "fig.set_size_inches(10, 10)\n",
    "\n",
    "\n",
    "\n",
    "plt.subplot(2,2,1)\n",
    "plt.hist(data[0,0], bins='auto')  \n",
    "plt.tick_params(labelsize = my_font_size)\n",
    "plt.xlabel(r'$l_2(\\textrm{nnDCP}-\\textrm{nnDCP}^{0,x})$', fontsize=my_font_size)\n",
    "\n",
    "plt.subplot(2,2,2)\n",
    "plt.hist(data[0,1], bins='auto')  \n",
    "plt.tick_params(labelsize = my_font_size)\n",
    "plt.xlabel(r'$l_2(\\textrm{nnDCP}-\\textrm{nnDCP}^{0,z})$', fontsize=my_font_size)\n",
    "\n",
    "plt.subplot(2,2,3)\n",
    "plt.hist(data[31,0], bins='auto')  \n",
    "plt.tick_params(labelsize = my_font_size)\n",
    "plt.xlabel(r'$l_2(\\textrm{nnDCP}-\\textrm{nnDCP}^{31,x})$', fontsize=my_font_size)\n",
    "\n",
    "plt.subplot(2,2,4)\n",
    "plt.hist(data[31,1], bins='auto')  \n",
    "plt.tick_params(labelsize = my_font_size)\n",
    "plt.xlabel(r'$l_2(\\textrm{nnDCP}-\\textrm{nnDCP}^{31,z})$', fontsize=my_font_size)\n",
    "# plt.savefig(\"../txt/LSTM_approx/histograms_{}_eps_e-{}.pdf\".format(noise_name,eps_order))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [],
   "source": [
    "kruskal = np.zeros((64,64))\n",
    "for i,j in prod(range(n_ts),range(n_ts)):\n",
    "    for k,l in prod([0,1],[0,1]):\n",
    "        kruskal[i+k*32,j+l*32] = stats.kruskal(data[i,k],data[j,l])[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAccAAAFvCAYAAAA/swAxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAAXEQAAFxEByibzPwAAIABJREFUeJzt3X2MHeV1P/DvIWCDeVuveQ0x4F1wGpsm8dpWCCKtEq9t\nUCmUX3dBQlREIl7T5A9U/VIvRklTRU3MugWlSiB4DVIaqRJmNylKiJrihShqFbUxu1AodsWvu+al\nIGOwvYZgMG/n98fM7M5z7tyZe+/M3Llz7/cjrexz577M7s7ec5955pxHVBVEREQ074Sid4CIiKjV\nMDkSEREZTI5EREQGkyMREZHB5EhERGQwORIRERlMjkRERAaTIxERkcHkSEREZDA5EhERGUyORERE\nBpMjERGRweRIRERknFj0DhARETVCREYAHAIwC6AXwC5VncrkublkFRERFUVEBgDsBLBMVWfreNwk\ngG2qOu7HXQAmAaxX1Zm0+8WRIxERNZWfyHYCmAHQA6CrzscPAegJEiMAqOqsiIwDGAOwOvU+tvPI\nMc8hNxERpecnuh0AFtc6chSR3QBmVXXQ3D4ALznW/FzVtO0FOf6Qe4+qblfVUQDbAIyJSE/Bu0ZE\nROn0V7l9NmF7zdoyOVYbcgMIhtxERFRC/ilZADgcc7fUg6C2TI4ABgFMRNy+B0Bf6IdLRETl0u3/\nG3XaNC5h1qVdL8jphzdKtGYTticSkQMAFgF4ubFdozawFMAxVT2v3gfy+CGkO35+Bu/6iVawFFWO\nY1Vd2YTXjxvkLEn75G2XHJsw5F4E4HQAK4IbVqxYUf3e1Hamp6dx/PjxRh++aOHChaf39vbyoOlQ\nKY+fXoTeewrW8DeRUtx7ezCqPJT2RdouOSKjIbeIPFdl0wJ7w3PPVbsrtaOVK1di7969jY78Xu7t\n7V3BY6ZzpTx+Wsl0k0aIDr9kA5h/rw8LBkesc4yR65CbiKgofnIoTAuUAE4g+j2+O7Q9lXZMjpkM\nuat9IvJHlK1yWoOIOhCTI8YAjETc3gtgKm2NI9CGV6uGfii5DblXrFgBVZ37EhHni4goT/Y9p9lf\nGat6lk9EukRE/br1OX7t+mG/6H/uvgCGAGzKYqfaceQINGHITURUlBNOKHZc8+GHH6Z+DhHZAe89\nOSjYf1xEngQw6Se/YH5xBtEDmtUAtoYau6wFsC6rLmjtmhxzH3ITERWhFc5QiUjqU6uqurnG+0WW\nrvjv48OpdiJG251WBZoz5CYiKkqbnVZtSe06cgRyHnKH2U9Q9uBpgclramEHDx504nPOOaegPaGy\n6JQEVaS2TY55D7mJiIrC5Ji/tk2ORETtiskxf0yOREQlIiKFX62axQU5rY7JMQecg6R62DnGpOOD\nowbiMZA/JkciopJhcswfkyMRUckwOeaPyZGIqGSYHPPH5NgEnIOketjj44MPPnDiE0/kn22nY3LM\nH//KiIhKpFWuVm13TI5ERCXTCcmpaEyOREQlw+SYPybHAnAOkuph5xiPHTvmxCeffLITF33Kjagd\nNJQcReSzALpV9YmM94eIiBJw5Ji/RkeOdwJYBeBSABCRVQBuALBNVd/MaN+IiCgCk2P+Gj3/skdV\nLw0CVX1KVbcC2JrNbhERUZSi13LslDUdGx05TojIHgD3A3hcVV/wbz+cyV51mKQ5yI8++ih2O3WW\nRYsWOfE777zjxKecckozd4cKwHnl/DX6E94MYD+APwcwIyIf+smyO7M9IyKiSBw15q/RkeO0qt4W\nBCLSB2/OcXcme0VERFV1SoIqUqPJcYmInBFcfKOqUwCmROQrAHgFKxFRjpgc89dQclTVO0TkL0Vk\nt6o+DQAi8hGAMQAPZLmDncjOMdr5hffff9+J2Wuzs9k5xiVLljjxq6++6sQLFy7MfZ8oP61warPo\n12+Ght9VVfVvzU294AU5RES564TkVLTMhhyquj+r5yIioup4tWr+MkuOInIGgH4/nAqVdxARUYY4\ncsxflpNV4wBWq+oSEflTEelXVc4/NsAe+HaO8aSTTnLif/qnf3LiP/mTP8lnx6gUDhw44MRXXnml\nE3/zm9904quuusqJOYfd+pgc85flX8FmAF0AoKo/yfB5iYjIxwtymoNzjkREJdMJyalobXv+RERG\nABwCMAvvStpdfj0mEVGp8YKc/NWUHEXkhwBGwhfZhJsA+Kty3Ajg/mZciCMiAwB2AlimqrMR2yfh\nrRAy7sddACZFZL2qzuS9f1mzc0B2jvH66693YtZBdjY7J33vvfc68fr165143759Tnzeeefls2OU\nGY4c81frx48Z+POJACAiDwOYFZH/JyL/F147uTvgzTvmQkS6RGTMHxHeGN4fc78hAD1BYgQAP4GO\nw2tSQERUauytmr+akqOq/m3QCce3R1VPgLdE1UYAR/zG4z057GOwD7OqOqiqw4jv4ToIYCLi9j0A\n+vxRJBFRKYkITjjhhEK/OiFBNnriekpEtgF4TFU3AFgC4C4AQ5ntWeP6q9w+m7CdiKgUOHLMX6O9\nVR8H8LiILAPwpn/asvDyjdCoMK6NXW6j22axdYxJdZCTk5NOvHz5cic+7bTTMtw7ajVr1qxx4tde\ne82JH330USe2c46f//znnbhT3hyps6W6UiMo3/BHkQPwRmf3q+qDGexbI4L1JCsu0kGdfV9F5Lkq\nm3rr2iMioozxatX8pb6M0Z9rXB26aVREVqvqV9M+dwpx84pLYrYREbU8jt7zlyo5isgmeCOpzQCe\n9G9eD+AOEdleQH/VuNFhMKo8VMsTqerKqNv9EeWKOveLiCgTrTDvV/TrN0PakeMAvH6q4e44T4nI\nFLyEuTXl89dFVWf9X1p3xOZgNFm6Osckto7RzjGuXr3aiR980D3r/eUvf9mJecqmvS1YsMCJr776\naic+//zzndj2aj355JPz2TGqGf9G85c2OUpU2zhVnfDrDYswgejTqt2h7UREpdUJI7eipU2OURe+\nBIoaoY0BGIm4vRfeUlpx+0xE1PI4csxf2uQYN8cXObcnIl9X1b9L+bpVL7hR1VERGRaRAdM+bgjA\nupSvS0RUqHabc0zTBzv03g54F1t2ARhT1dRnCNMmx8Ux27TK7WsbfTER2QHv9GhQyP+4iDwJYFJV\nR0N3XQ1gq4gENY1rAazrlMbjto7RzjHeeuutTnzTTTc5sZ2T4qfU9rZw4UInvv/++534pZdecuLe\nXrea6WMf+1g+O0ZVtcvfZAZ9sEdU1Wlb6rcZ7Qq3EG1E2uQ4KCIKIOpjRJ+IrI+4veEONfaHEHO/\nWQDDjb4OEVErK3rkmIVqfbBFJOiDvbrqgzG3AMVkxKZNAB6H10+7YVks13BDzLaogvlqI0oiIqpB\nOyRHxPfB3uKP/uKuEVkLoA/AaMS21D200ybHGXjfYK0XuQiAx1K+JhFRxwoajxe9DxnoR/TobjZh\ne2A3vCS6G8BgKJFuRfRFmXVJmxynVPWpeh4gIiylyJntlWrrGO0c4ymnnOLEf/3Xf+3EAwMDTnzO\nOedUvGZXl/tBzc5DqWrsdiqOfaO98cYbndjOWX/ve99z4u9+97tOfOzYMSe2x+Opp55asQ88HupT\n9pFjFn2w/ZLBcXj19kdEZDO8s5V70s43AumTYyPzeqkzOhFRJyt65OjrrdaDulqHsZBM+mCr6qB/\ntesWADsATAHYVuvj46T6CUc1AMjjMURENK+NlqxK1QfbvyinC17b0hl4c5D7RST10oRpe6s+DGCZ\nuXlYVZ/wt58J7/zvGxnUNhIREVrmtOp0DSPEalL3wfavdl2vqoP+Tb0isgXe2ckxEVmWpulL2tOq\nQwAeBrAG3inWh1X1aLDR//8dInKmiNwF4CFVfTrla1Kd7CkYW8do5xht/NFHHzmxXd8PAC6//HIn\ntv03P/zwQye285z2j71F/vg7kv3Z9/e7H8K/853vOLGdw7Z1kbbudsWKyr79Sf1aeXzMa4cLcjLq\ngz0CMzhT1e3+dS2TSL6gJ1ba9RxnRWQG3pVCR2PuFyTJu0RkRlXfTPO6RESdrE0+HDTcB9u/oCey\n1ENVp/wLdaISb81SffwQka/D61BQNTEa29DklTqIiNrNCSecUOhXRsbgnXW0Evtg+9tmQ13QrC6k\nXGQi7Xe5tp4LbPwk2hYfeYiIqHF+y8/D/kU1AJxeqZvCt4mI+q3mwgbhzS06o09/LnJ3je3nqsqi\nQ069zizgNSnEfvKzdYx2jvHb3/62E19zzTUVz3nGGWc48UUXXeTEb731lhNffPHFTmzr3OwalW1y\nGqmUzj33XCe+8847nfiBBx5w4qkpt4XxF7/4RSf+6le/WvEa9nixv+96j4ei5+Ty1kZ/D4l9sEPT\nd06y8+scNwHYKSKH4ZWFtEzj8UZa9MQWdhIRUXXttCpHrX2wVTWqFSn8JDoYtS2t1Isdi8jFqvpC\njXdeBp5WJSJKpd1Hxq0gbXIch9eVYGON978fXukHERE1qOiRYydIW8oxKiJbROQhAJtU9a2o+4nI\nGQB2wlue5IGo+1BxbK9UW8do5xgfffTRiue47rrrnNjWrb366qtOfP755zvxSSed5MR2DrLZbwa2\nF2wnvxnZ301fX58Tb9myxYkXLVrkxL/85S+dOGrO2h4PdmRkf/5JvXtbbWRl9y+tVvv+2lEWF+Rs\nAPA/8NZ2HIN3+WzQ/aAbXluf4IqPyPPGRERUm3aac2xlqZOjqs6IyCXwlg+5AZWTowLvKqP1tc5N\nEhFRdRw55i+TUg6/nqTXr1e5AfNXpM4A2KWqP8nidYiIqDNGbkXLtM7RX0Mr9TpaRERUHZNj/opo\nAkAtxi5UbJuI2wJ/e/ENAGzatMmJr732Wifet2+fE993331O/PGPf9yJly1zF3uxF/gkvTnYRgb2\n/vYCCRt/8MEHTmyL0DuJvfjlk5/8pBMPD7tlarbR+G9+8xsnvv322yte45577nHiM890e4V84hOf\niN3+/vvvO/HixYudOOk0ZNIFM/Z4suzz2+cLN97P4uIcJsf8ZfoXLyIXw7v4JrjwZhbeaVWuxEFE\nlAFekNMcmSTHUKlGcFVq8JNTAFtEZDeAzar6YhavR0TUyTohORUtdXIUkVXw1s4CvPnGGcwvUtkL\nr5xjAMCMiPSp6n+mfU0iok7G5Ji/VMlRRM4E8DiAUVW9LeF+dwJ4wl+dmes5thA7p2Tn92xT6KiF\nae0c489+9jMntvNQ//Ef/+HEv/d7v+fEdg7SLtCcVASeNOdoF1+2j3/vvfeqPj7rgu5WZ392Cxcu\ndOIrrrjCic8++2wntvO14+OV1+zt2bPHie2coX0OO8d37NgxJ7bz5EmLJdd7vNTLzommxeSYv7Qj\nxzsAPByXGIG5paqG/V/oVnBNRyKihjE55i9tchxQ1UtrvbOqDovIHjA5EhE1REQKbwLQCck5bXI8\n2sBjjjT6Yv6iliN+GDQaGA6v/RW67wi8uc9ZeHOfu6LuR0RUNp2QnIqWNjk2stLybCMv5CfGnao6\nGLptBMCkiKwPL27prxi9zW9KEDw2uF+q1aHbUVxNFlC5ULFtIg5U1jHaOcbnn3/eiZcuXerEF1xw\ngRPbZtdJbwZ2e9In66Q5y7g5rk5/Y7LHxyuvvOLE9nf3wgsvOPHBgwcrntP+/u2c85IlS5z4lFNO\nceKk33+9x0+9x1PS82ddJ9vpx2AzpB2bH06+SyaPAbwRo1NprqrD8JLtWHCbiAzBW/1jPHS/WXhX\n0o6BiIgoQdrkuDj5Lpk8BvB6tk5G3D4BoEtEgtOsg/5t1h4Aff4okoiotIJGAEV9dYK0Y/1BEVHM\nF/3XYiD5LpFm4Cc3fyRoBUmvH9H9XWcTthMRlUKnJKgiZXEi/IY6799QkZiqrq6yKVh5dSY0Kow7\nddsTs22OiDxXZVPbrUlp59/sfM7FF1/sxHZhWqCyV6qtY7RzjLfccosTb9y40Yn/5m/+xokvu+wy\nJ7a1drZO7Z133nFiO+dj6xhtL1U7Lxb+npP6bLY7+7O3vztbk2qPH7sdAG67za0GO+uss5z4+uuv\nd+J169Y58Ysvus23br31Vic+7bTTnNgmF3s82L+JN990S7Pt4+0Cz/YYOXJk/jrEtDWTUa9P2cvi\ngpxB1H6RzWIAu1K+5hwR6YeX7Lar6mzo1GrU/jQ610lE1DJa4dRm0a/fDGmT45SqPlXH/feLSD33\nT7IDwLh/YU5Y3Lzikphtc1R1ZdTt/ohyRW27R0SUvU5ITkVLmxxtUsrrMRVEZAeACVXdHLo5bnTY\n7f97KOY+REQtj8kxf6mSo6rub8ZjLBHZ4j9XODHCP7UKzCfCsGA0yTrHBPYPz86/2Do2oLIuzfZK\ntXVsdp7qX/7lX5z4z/7sz5zY1k3aeSs7Z/i73/3OiW0/2KNH3f4Vx48fd+KZGfcwCc8pZTFn1E7s\nHKQ9Xux83CWXXFLxHKtWrXJiO4do14S0vVPtHLGtzbX7YPfR/v7t8RWeMwQq6x7t89lj5PDhw1W3\nNYLJMX9NX8FVRHap6o0pHj8AoDecGIO5Rr/AfwLRp1WDhBlV5kFEVBpMjvnLJTn6ix73IzpJ9ad4\n3j4Aa+2IEV55SFCeMYb5FnNhvfDmSBvq0ENE1CqYHPOXeXIUkb9EdHIKNFTK4Y8OxwCM+23jAC/5\ndgPoU9XtAKCqoyIyLCIDpn3cEIB1EU9NRFQavFq1OfIYOa6Fd9ozcm5RRB5r8HnH4JVtbInYZucR\nVwPYGirtWAtgHRuP18Ye+LZG0M6vAMCyZcuc2M5B2nlKW8do5xhvvvlmJ/7FL37hxH/wB3/gxK+9\n9poT2zUDe3rc8tZnn33WiQ8cOODEP/3pT51406b5zoV2/qnT2OMjqc+p/d1/7nOfq3jOu+++24nt\nz3hqyv3Tfewx922kt9ctP/7Rj37kxJs3uyebbN3jo48+Gvt8P/7xj53Y/k1s2LDBie3xFN7f119/\nHWl1QnIqWh7JcU/CRTc7GnnSmCYAUfedRUZXxRIRtRomx/w1/YIcANMFvCYRUdtgcsxfHitmTojI\n12O2c6FjIqIU2Hg8f3mMHIcA9IjIVnhzgbYwv+GrVakYSWvdAZV1hLZOzD7G9kq1dYx2jvGP/uiP\nnHhszF197B//8R+d+N1333Xie+65x4ltnaVdL/D73/++Ez/00ENz/7c1cZ2u3rUzTz311Ir72DpH\nWwu4Zs0aJ77uuuuc2NY92l6tdh/sHOGWLe6lDLY36rXXXuvEtnev3R87ZxnuDRtVJ1yvTklQRcoj\nOd4ILyEGy0vxt0hElJFWGL0V/frNkEdynFHVNdU2isjDObwmEVHH6ITkVLQ8kuOmhO3bcnhNIqKO\nweSYv8yTY9IqHXWu4kElkdSP1bL9OO0cpa1jtHOMg4ODTjw87FbufPnLX3bicG9LoLIO086Z2nmx\nvr6+uf9fc801Fb07qXZRc5S2VtKyvVEXL14c+5y2btLOQdo6WBvb3rpXXnll7P7ddNNNsfsX7i38\nyCOP4NChdOsfMDnmL4+rVWOJSGbrORIREeWhVL1ViYiII8dmKE1vVSIi8jA55q9MvVWpRFTdz0D2\nj9nWkdn1GG2vVFvHaOcYR0bcz2PnnXeeE+/du9eJr7jiCie2c0S2Ti1cd/nmm2+CGmePDaCyrtHe\n5/3333diu16nnbO0vVJPP/10Jw73ygWABx980ImffvppJ7Zzmnb/HnnkESdeunSpE/f3z58wszW4\n9Wq3Ug5/IYlDAGbhrZ60q94+2KHnAIAljTyHVZreqkRE5Ck6OWZFRCYBbDMrKE2KyHp/fd6kx3cB\neBzAsKpO+LftALAT3gIUDWNvVSKikknqSlQGIjIEoCdIjIC3aISIjMNbhamW5DYG4MkgMfp6UNmZ\nrW7srUpEVDJt0lt1EMBExO17APT5o8K4n0E/vAs8nTkVVV2vquvT7hx7q1Im7Byije0nXdub0s4h\n2bozO09j6xjtHONf/MVfOPHll1/uxLZu8uyzz3biq6++2onvu+++uf/b+VGqjz02gMrfv/0Z2/Ud\nn3nmGSe2c3zf+ta3nHj37t1OfOmllzrxD37wAydevdodtHzjG99wYns82t6wP//5z534V7/61dz/\nDx48iLTa5LRqP4DxiNtnE7YHhgHM1nL6tRHsrUpEVCLtcEFOaFQYd/qzJ2Yb4CXPKRHpg5d3DsG7\noGfMnGZtCHurEhGVTNHJ0dcrIs9FbVDVlQmP7fb/nY3YljhfaE65rlHV4dC2aREZDs9lNoK9VYmI\nSqZFkmMW4uYVl8RsC/Sp6qi5bQLAThGZUNWo5FsT9lalTNSy5mPYiSe6h57tbdrT455Rsesx2l6p\nto7RzjH++7//uxNfc801TmzXc1y50v3g+6lPfWru/88//zzXdEwh6tiwx4O9j61jPP/8853Y1qna\nOcX9+93qMvv4Z5991ontvGi4ty5QOYc+MeGexbNz6uFewDMzM5nUOraA6RpGiNXEjQ6DUWXVBrT+\nVa2Ad12LNQnv2pc1iL7gpyZFlHIQEVEKLZIcGxZKbt0Rm4PRZNKFNjOIPi0bSJqzjMXkSERUMmVP\njr4JRJ9W7Q5tjzOF+OqHVFexlr+SlIiogxRd45jh1bJj8E59Wr0ApmqYL9wGoCuiHnI1vBKPVFes\ncuRImbC9Jm2vTLu+43vvvefER48edWI7B7Rx40Yntusx2l6pto7RzjHa3ptnnnmmE9tenuH19+z3\nRvHssRFV5/j22287sT0+Xn/9dSe2x8fy5ctjny88ZwwA3d3u2bx169Y58WWXXebEP/zhD53Yzjnf\ncsstTrxv3z4nDtc52n1rRDuMHFV1VESGRWTAtI8bAjD3C/FvOwIvYa4OPX7K76azFV7NY3DfG5B8\nYWgiJkciopJph/ZxvtUAtopIMD+4FsC6cNNwf35yBhGnSVV1UERG/H6qs/DmGZ3HN4rJkYioZNph\n5Ah4iQ/+qC/hfr0x2xIf34jSJUf/E8bm0E19AHZEFXxmsRQKEVGraZfk2MpKlRz9xDisqptDtw0A\nGBORzeFi0LRLoVB97LxSUmx7Z9o5nAMHDjixrUO0dZG2zs32SrWPt3OMds7TrtkYros7fPgw+6um\nELWeo51jtLGdp7P9SW1vXXs8LFq0yIlPOukkJ7bHhz1+7D7bx5922mlOfNFFFznx5OTk3P85Z10O\nZTtxPQBgyE+IAIDQiDHcPihyKRR4TWzdKzWIiEqk6KtUW6G3azOUauQIr65lFtGFn+Hb4pZC2SIi\nXWnaChERFakTklPRSjVyVNUJVV0crl8JjSJ3hO5arTB0NmE7EVHL46gxf2UbOTr8ecQRANuD+UbJ\nZikUihE1Z2Tn4Oycke2daeeMZmbcaeCf/vSnTvz973/fiU899VQn7u11L2az6zHaXqm2jtHOMdo6\nyNtvv33u//v370/dG7OdRB0Pcdvtzx4AXn31VSeenXVP7ITn7ADg17/+dez9N2zY4MR2DnHhwoVO\n/IUvfMGJbd3kVVdd5cS2lML28n3llVeceHp6eu7/zzzzTEXv1Xp1SoIqUimTY+iK1QF4p0/DK32k\nWgol9BqRS7HAu+qViKgwTI75K9Vp1YCqzqjqsF/7Mg1gv4jYU6Vpl0IhImpJPK2av1KOHMNUdbuI\nbIVXzrEMKZdCCT1v5FIs/ohyRd07SkSUgVZIUEW/fjOUKjn6p1MRUaf4JLyLbPpVddz/xaVZCoVi\nRP1hJK3HZ+do7Hp6tg5t0ya3NeJDDz3kxHZ9vV/84hdOfN999zmx7a0Z7pUatT/hOUYA+Pu//3tQ\ntHrfKG2NIABceOGFTnzuuec6sa1bPOecc5z4kksuceLrrrvOie0cpD0eRkfd9XI//elPO7Gd87Tf\ns51jf/rpp534ggsuQJbaqH1cyypVcoR3ChUisjihFCPtUihERC2rE0ZuRStbcpwF8GREYgyuPg2S\n3hi8q1itWpdCISJqWUyO+StbcqxoMOvXOfYA2BwkvVqXQmnQ0unp6YrSAHLZy/ftH7Ndtsi21Dpy\n5IgT2/ZydgkqW4phS0uef/752Nc7fNidqt6/fz8SLE26Q7XHdfrxE1X6YX8fSe0GbSnNggULnNi2\nm7v11lud2JZyvPTSS04cLr2Iej7rjTfeiN2/cOmSv63R4wcAk2MzlCo5+kmvX7zlSQI9ANZHLGyZ\nuBRKg44dP34ce/fufTnl81AKb731Vl33t8nVSnrzNZYCOFbXDszj8VOAF198sa7711vHGlW7GbM9\nzfHDC3KapFTJEfC65KCGOcNal0Jp4PXPS74XUTQeP5SFTkhORStdciQi6nRMjvljciQiKhkmx/wx\nORIRlQyTY/6YHImISobJMX9MjkREJSIihXfI6YTkzB5ERERERluPHEVkBF6T8Vl43XF2ZVDnSERU\nqE4YuRWtbZOjiEwC2GY65EyKyPqIxuVERKXB5Ji/tkyOIjIEoCdIjIDXFEBExuH1XV1d2M4REaXE\n5Ji/tkyOAAYR3UVnD4AtItLVaPNxETkAYBEAtv/qXEsBHGuk2w2PH0KK4wfgBTnN0q7JsR/AeMTt\nswnba7Fo4cKFp/f29nLB4w41PT2d2Ks1xiIApyO0YPaKFTyUOknK4wdAZySnorVdcvTnFgHgcMzd\nemK2Bc/zXJVNC3p7e/Hcc9U2U7tbuXJlmsbhLyOUGAHwWOowKY8fAEyOzdB2yRHzCxpHnTaNS5hE\nRKXA5Ji/dkyOga6YbUuSHqyqkQvu+SNKngcjosIwOeavHZNj3OgwGFUeynMHwgubAsA555yT58tR\nyaxYscI5lWrf6KIWAyYK8IKc5mi75OiXbADziTAsGE2yzpGISqsTklPR2i45+iYQfVq1O7SdiKiU\nmBzz167JcQzASMTtvQCmGq1xJCJqBUyO+WvL5KiqoyIyLCIDpn3cEIB1eb++nWO0c0g8sCks6fjg\nHCRZRc85doK2TI6+1QC2ikhQ07gWwDo2HieiMhORwj9gF/36zdC2ydE/dTpc9H4QEWWtE5JT0do2\nORIRtSsmx/wxORbggw8+cOITT+SvgeZxDpKSMDnmj+/KREQl004X5GS5KL2I9AHYrKqb0+4XkyMR\nERUih0XDtkWcAAAgAElEQVTpxwBkctElkyMRUYm0y9WqWS9KLyJbEN0ZrSFMjk1gDyQ7x3js2DEn\nXrRoUe77ROXBOUiyik6OGclsUXq/ZG8WGbYGbZ8T10REHSIYPRb1lZH+KrfPJmyPsllVR1Puj4Mj\nRyKikin7BTlZLUrvP9cQgB2pd8pgciQiKpk2OK2ayaL0fpLtavDinVhMji3g5JNPduJ33nnHiU85\n5ZRm7g61OM5BdrYWuiCn11/8vUK1xeIjpFqUHsBWVc2lE1pDyVFEPquqT4fiM1T1zex2i4iIqik6\nOWYg9aL0IjIAYFdme2Q0OnI8KiJfV9W/8+NeEbkLwCCTJBFRvlokOU7XMUJ0pF2U3j+dujZcBpK1\nRpPjIQBnicj/UdWfqupTodqUjdntHhERWWW/IMeXZlH6fgA9ImIvxOkL3T6tqtsb3blGk+Pj8K4O\nelxEvqSqT8AbJq9tdEc6mT3Q7RzjkiXuqfcDBw448UknnZTPjlEpJM1B2jlsO8dN5dMiI8e0Gl6U\n3h8xVowaReQGABNZtI9r9OPHYgCHVfUovFY/qwCsQRMWEiYi6mRF1zhmdUGQX5d42J87DL63YFH6\nTeHbRET9VnO1iLvIp2aNjhw3ANgC4KeqelREZgH0qupTWewUERFV1yYjR6CGRen9+ckZxM9BjsA7\npdoFoF9ExgDsTtMYoKHk6NeU3BaK94vIsIh8RVUfaHRniIgoWZvMOda8KL2q9iZsz7ycI7M6R1Xd\nD4CJMQevvvqqE1955ZVOfO+99zrxmjVrct8nal1JdbJPPeWe4Pn93/99J/7Yxz6Wz45RZtpo5Niy\n2uPjBxERUYYyGzmKyBmYbxQ7paovZPXcREQ0jyPH/GXZPm4cwGpVXSIifyoi/Zx/JCLKVgu1j2tr\nWSbHzfAvoVXVn2T4vB1v4cKFTvzNb37TidevX+/Er732mhMvWLAgnx2jlmTrGO0c46pVq5z4xRdf\ndOKlS5c6cSe8EZYNfyf5y/qCHCIiyhmTY/54QQ4REZFR08hRRH4IYCR8kU14JQ6/Q86NAO5vlQtx\n/KLQQ/DWC+sFsCtcWEpEVFYcOeav1tOqMwi15BGRhwEMiMg0gPsB7FTVO0RkG4Ct2e+my283tBPA\nsqj+e36boW1Bx3a/JdGkiKzPY1HMZrvqqquceN++fU786KOPOvHVV1/txHYOs10KiimarWO0c4wX\nXXSRE7/11ltOfNppp+WzY9QwJsf81ZQcVfVvzU17VPUGP0kNAdguIlOIae+Tlp/gdvqv0YMq/fNE\nZAhAT3gpE7/9ULBqyOq89pGIqBmYHPPX6JBhyh8lPqaqG+Ct2HwXvESZC1WdVdVBv03Q7pi7DiJ6\nqZM9APr8JEtERFRVQ8lRVR9X1a3wkmKQuH7ir9JRtP4qt88mbCcianlFr8bRCnWWzZCqlCMo3/BH\nkQPwEtD9qvpgBvtWt9Co8HDM3XpitpXCiSe6v7bzzjsvNj7//POd+P7773fiG2+80Yk74cDvJLZX\nqq1jtHOMp59+uhO/++67TmznrKn5+Deav9R1jiKyB+483qiIrFbVr6Z97gYEK0hHLZIZlzCJiEqD\nyTF/qZKjiGyCVyaxGcCT/s3rAdwhItsLLOuIm1dcUssTiMhzVTbFLp1CRJQ3Jsf8pR05DsDrpxru\njvOUf+XqZjShrMOIGx0Go8pDzdgRIqK8MDnmL21ylKi2cao64ZdUNJVfsgHMJ8KwYDRZU7mJqq6M\nut0fUa5oaAeb5POf/7wTHzhwwIlfeuklJ37wQXeKuL/fvWbp3HPPdeKTTjqp4jXtvBb/eFuX/d3Y\nOkY7x2h7tT7//PNOvGzZMidWVSe2dbRcLzI9/n3lL21yjJrbCxRVbD+B6NOq3aHtRESl1ApXixb9\n+s2QtjVK3GnMyNOXIvL1lK+ZZAzAmojbe+GtMxmX0ImIiFInx8Ux27TK7WtTviYQc8GNqo4COOx3\n7wEwV+IxBGBTBq9NRFQo1jjmL+1p1UERUQBRP60+EVkfcXvDRfgisgPe6dHgOR4XkScBTPpJMbAa\nwFYRCWoa1wJY1ymNx+3Ba+eMenvdC26/973vOfF3vvMdJ77zzjuduK+vr+I1P/nJTzqxrYX78MMP\nY7dbnfIH2Irs78bOMS5fvtyJ/+u//suJDx1yTxp9/OMfd+JPfOITia9p8Xhw8eeRvyzWc7whZltU\n2UO1EWUiVd1c4/1mAQw3+jpERK2MyTF/aZPjDLxeprXO4wmAx1K+JhFRR2NyzF/a5Dilqk/V8wAR\n4dWiREQpMDnmL21ybOTU5UjK16SUbJ3Zd7/7XSe+6aabnPiBBx5w4i1btlQ85/CweyhcccUVTvzK\nK6848caNG53YzjmdcsopTtzsNSdtrV4nvxnZOkY7x3jZZZc58Wc/+1knfvXVV5343nvvrXiNL33p\nS05s+wfb48Mew/b3FVWLWyS7f2m0wkUxRb9+M6R6x4lqAJDHY4iIiJopbW/VhwEsMzcPq+oT/vYz\n4bWQe0NV/y7NaxERkacTRm5FS3tadQjAw/CK7ocBPBxe09H//x0icqaI3AXgIVV9OuVrEhF1NCbH\n/KVdz3FWRGYADMYtdBxKkneJyIyqvpnmdSlbx44dc2Lbe3Vqyi0PXbRoUcVz2Nq3s88+24ntHNCC\nBQucmL1ZW5edL7N1jHaO8emn3c+/dj7RricJVM4p2t+/nXO2cZZzenng8Vw+qeYc/VZwI3GJ0diG\n5q/UQUTUVtghJ39pT6uurWcuUVWPSqf8ZImIcsK30fxl0SGnXmcW8JpERG2DyTF/aZNj1QbgMXqS\n70LNZNfzs/OHX/ziF534l7/8ZcVz/OY3v3FiW6f2wgsvOPHFF1/sxHYe085RZr0GYNIcVavPYTWT\nnd+zvVJtHaOdY3ziiSeceHJysuI1LrzwQie2v/9TTz01dh8/+ugjJ7b7nDaZJB0PSc/POsfySb3Y\nsYhcrKov1HjnZYhuUk5ERDXqhORUtLTJcRzADgAbk+7oux9e6QcRETWIyTF/aTvkjALoFZGHROT0\navcTkTNEZBeAHlV9oNr9iIiIWkEWF+RsAPA/8NZ2HAMwAeCwv60bwHoAwcLDUUtYUcHsfM6KFSuc\n+Ktf/aoTX3PNNRXPcfvttzvx+Pi4Ex88eNCJbZ3jJZdc4sSf+9znYvcxqc7NzkHZT9p2u338+++/\n78ThObBOm4+08712PUbbK9XWMdo5xq997WsVr2GPhzPPdK/bW7lypROfddZZTmxrde0+2jlwK+l4\n+eCDD2Ifn9TrNXw8ZXH8cOSYv9TJUVVnROQSALvhre04aO4i8Ja2Wl/r3CQREVXH5Ji/TEo5VHUG\n3unVAXgJMrgidQbALlX9SRavQ0RE1AyZ1jmq6ji8i3SIiCgHLOVojiKaABARUQrtlJxEZATAIQCz\n8K5L2aWqU/GPmntsF+bXCA7OWA7X+vg4mSZHEbkY3sU3wYU3s/C+Ua7E0cLsxQQnn3yyE1900UVO\nfP7551c8xz333OPEe/bsceILLrjAiW+77TYnXrVqlRPffffdsdvtYsgffvihE//ud79zYntBxttv\nv+3E7733nhPbwvZwkbp9rU5jm4Tbon+73Rb424tvAGDTpk1OfM455zjxH//xHzuxXUz7f//3f53Y\nXvTT1eX2K7EXdNkLeuzxYput2+R0xhlnOLE9RsKLfdtjrRHtkhxFZBLANv+sY5DsJkVkvT9dF/fY\nLgA7VXUwdNtI6PETafYtk+QoImcA2In5q1KD35wC2CIiuwFsVtUXs3g9IiIqNxEZglfeNzcV56/0\nNA5gDMDqhKcYAeB8qlLVYf95xwAsTrN/qZOjiKwCEFyrPQ7vIpzgY1YvvHKOAQAzItKnqv+Z9jWJ\niDpZm4wcB+GV/ll74A2qulR1NubxNwDoR2WJ4ASAARHpSRp9xkmVHEXkTACPAxhV1dsS7ncngCdE\nZBnXcyQi6nj9iL6AczZhe2AGQF9MEm2k9/ectCPHOwA8HJcYgbnFjof9TztbwTUdSyVp4Vmgsmh7\n8WL3jIZtVm2LuF980T3jfuTIESdOmuezhdW2aNt+D3bex8azs+7f2rnnnlv1tTqdnZ+zP2vbRNwe\nK0DlHKNtGvH666878eHDh534jTfecGL7+0z6ndnjy87Dv/POO05s/wZs8377fOE5TdtwoBFlHzn6\n84XAfMOYKLGLVKhqtdOuff6/DY8agfTJcUBVL631zv754D1gciQiKrteEXkuaoOqroy6PaTb/zdq\nxBeXMGOJSD+8pLo94ZRsorTJ8WgDjzmSfBciIqqm7CPHkLhTn0saeL4dAMZVdbjB/ZmTNjk2Mmxt\nOJvXU9OSpnaGiKhVtVATgOkaRojVxI0Og1HloZj7RO3TDgATqrq5wX1ypE2OjQx/Gxoy11PTkqZ2\nhio/lSbNKQHJjZ6XLHE/BF5//fVObBdLnppyP8esWbPGie3iyLZRuJ2ztHWRdg7L1j3aZtnhOdSk\nJtTtzv7+bV2jnY+zTeNtE3Ggso7R/n5+9rOfObH9/e/bt8+Jh4aGnLi7u9uJ7ffw1ltvObGdQ7R1\nlPb4trWbx48fd+LwYt/2WG1E0ckxLb9kA5hPhGHBaLLm92oR2eI/byaJEUi5ZBUaqyNptPYksqYF\n3shwLLitWu0MvKuexkBERK1gAtGnVbtD2xP5Pb17w4lRRHpEJPaCniRpR46DIqKYL/qvxUDyXSLV\nWtOStnaGiKillX3k6BvD/DRZWC+AqVrep0WkD8DaiBHjAFL2+c6iQ84Ndd6/0evga61pSVs7Q0TU\n0tohOarqqIgMi8iAmQIbArAuuJ9/2xF4CXN16PYeeAl23J9iA7w80A2gT1W3p9m/LC7IGUTtF9ks\nBrCrkReqpaYli9oZqvzDs7GtAQMqa9fsvJOd81u3bp0T296Ujz32mBNfd911TmzrKG0v1WeeecaJ\nbT/YZ5991oltXd2vf/1rJw7X4b377rugefZ4iKqDDbM1rkBlr1Rbx2jnGB966CEnXrZsmRP/93//\ntxPb/sB2H+2cop0jt3PQtlbRzrsePepeyP/b3/527v+2ZrIR7ZAcfasBbA2dAl0LYF344kl/fnIG\nlXOQY/Dez7dEPG/qa0vSJscpVX2qjvvvF5F67h/L1rSEfsCpa2eq1e+g8rQuERE1wD8LmFh2oaoV\n77sxA6ZMpE2OjdSSpK4/CalW05J17QwRUctoo5Fjy0qVHFV1fzMeE6VKTUtmtTPV6nf8EeWKmnaS\niChjLVTn2NaavtixiOxS1RtTPkdkTUvWtTMULapPpa3dsuvj2T8m20vVzvn19rpnUeycZNKc5tKl\nS53YzlEuX77cic877zwntr1VL7nkkrn/R61H2Mns8WBjOz9njw2gcs7P9kq1dYx2jnH/fvczt/19\n2v6u9vixx4c93i691O2Saesc7Tyqre0MH288fsohl+ToL3rcj+jTm/0pnzuypgUA/FKOTGpniIha\nVSeM3IqWeXIUkb9EdO1KoOElDWqsaUldO0NE1MqYHPOXx8hxLbyRXeTcoog8FnV7klprWmqtnSEi\nKismx/zlkRz3JFx0s6PB562npiWxdoZqZ+dnourYkuZs7GNuvfVWJ7a9LX/0ox858W23uUuG3n33\n3U786KOPOvG3vvUtJ7ZzRraXqt3/DRs2OHG4ztI+ttPZ+TzL/u5tH14A+NrXvubEdj1G2yvV1jHa\nOcYvfOELTvzP//zPTmz7u1599dVO/Id/+IdO/K//+q9ObGs7be/Wl19+2YnDc+B2PptaU9MvyAEw\n3ciD6qlpqbV2hoiojDhyzF/axuNRJkTk6zHbudAxEVEKQTlHUV+dII+R4xCAHhHZCu90p609THW1\nKhERUd7ySI43wkuIQTPCzviY0eHsvFJSf1a7Xp7tnbl5s3tBsp3jsXOQp59+uhPv3r3biW0d3Kc+\n9anY1z/77LOdODwHeeutt1bUaVJ1SeuDAkBXl1t9ZWsl7Zye7ZVq5z3tHKOdU/zzP/9zJ96yxb2U\nIbz+IgCMj7trFdjaTXu8XXjhhU78mc98Zu7/1157LaanG5pdoibKIznOqOqaahtF5OEcXpOIqCO0\nwqnNol+/GfJIjpsStm/L4TWJiDpGJySnomV+QU7SKh11ruJBRETUdKXsrUqtL+mTbdIakXZO0tYd\n2jmgTZvcExa2rtGu52jnsOyclV2fLzxHabdReklrQNrjJan21tYx2jnGH/7wh05se7fa3q52DtHa\nuHGjE9s63/D6kEl1obXgyDF/peutSkRElLdS9VYlIiKOHJshjyYAQW/VE6K+ADyew2sSERFlpky9\nVamFJa3hZz/p2t6Zx48fd2LbK9XWodk5xwcffNCJf/CDHzjxs88+68Tr1rk96MNzQkBlb87R0dG5\n/7/00kugxtljA6hc4/HDDz90Ytt7167/aHvj2rpGe/zYOUZb92p//3bO2h7v//Zv/+bER44cceKb\nb7557v8HDhxAWhw55i+PkWMSVr8SEVFLY29VIqKSYW/V/LG3KhERkcHeqpSLpN6qtq5xwYIFTtzb\n2+vEdp5qZsZdwvPpp5924tWr3RXO7OMvu+wyJ7a9VJcvX+7En/70p+f+Pz09jXfffRfUmKiRh+23\nmlT3aueIbV2hXY/R9kq1dYx2jtGu3/hXf/VXTmy/hyeffNKJL7/8cicOH09p62RbYfRW9Os3A3ur\nEhERGeytSkRUMp0wcisae6sSEREZTe+tSu3JzunZOjXb+/LNN990YlsX9uMf/9iJr732Wie+8sor\nY5//G9/4hhP39fU5se2taevWrrrqKieenJyc+//bb78Nqp392X7wwQcV9zl06JATv/POO05s6xrD\nvw+gspeunTO0dbG2V6qtY7RzjN/+9red2B5vX/nKV5zYrh961113zf3/6NGjSIsjx/wVUedIRETU\n0pgciYiIDCZHIiIig3OOVIik9fls3Zudg7LsvJatQ7TPb3u52jX27P05x5Oveo8HO8edVCcZ1c81\nzB4/Sftjny/p+LH7lxaPx/xx5EhERGRw5EhEVDIcOeavdMlRRHoAbA7d1Adgh6qOR9x3BMAhALMA\negHsUtWppuwoERGVVqmSo58Yh1V1c+i2AQBjIrJZVUdDt08C2BYkTRHpAjApIutVdcY+N6Vj51gs\n+0l30aJFTmzniDZs2ODE1113nRPfdNNNTvzII4848apVq5x4YmLCiW+55RYntr07bW/M8PqTb7zx\nBt5//31QbZL66gKVvVHt78P23rX9Sc866ywn7u7uduLdu3c78caNG53Yrsdoe6XaOkY7x3jvvfc6\n8UUXXVT18aOjoxW9XevFkWP+yjbnOABgyE+IAIDQiHE4uE1EhgD0hEeTqjoLYBzAWJP2lYiISqpU\nI0cAU/BOkc5GbAvfNghgIuI+ewBsEZEuP1kSEZUKV+VojlKNHFV1QlUXq+pc4guNIneE7lptzcjZ\nhO1ERESlGzk6/HnEEQDbg/lG/zagcpHlsJ68963T2DqxJEm9WA8cOODEdn3HxYsXO/HSpUud+Oc/\n/7kT2zrJffv2ObGdI3rllVecOLxeJNdyTCfqWLG/fxvbulTbn/TUU0914pdfftmJbS9VO8dpe/va\nOWfbK9XOsdvj58UXX3Ti8PGUxXx1J4zcilbK5Bi6YnUA3unT8DJYwUx81GnTuIRpX+O5Kpt6q9xO\nRER1SltVkFdVQimTo3+16TCAYRHZAmC/iAyGT7cC6Ip+NABgScw2IqKW1i4jx7RVBXlWJZRqzjGK\nqm73/zvm/2DiRofBqPJQzH2C510Z9QVgOuUuExF1vLRVBXlXJZRq5OifTkXEJ4In4V1k06+q4/6n\nqm5UCkaTrHNMoZY5IzuvYntL2jmew4fdzzSPPfaYE19//fVOfMEFFzhxf797jdWvfvUrJx4cHIzd\nbtcHnJ52PwOFX+/gwYOscwxJ6ktqt0f97Owc77Fjx5z4hRdecOLf/va3Trx8+XIntnPQn/nMZ5x4\nyRL35NHNN98c+3zh9RiByuPZ1kHa7+cf/uEfkKU2GTmmrSrItSqhbCPHaQDToYtuqplA9GnV7tB2\nIiIqTtqqglyrEko1coT3TT8Z8WkguPo0SHpj8K5itXoBTLHGkYjKrOwjx7RVBc2oSihbchy2N/h1\njj0ANgdJT1VHRWRYRAbMRO0QgHUp92Hp9PQ0Vq5cmfJp2ku9pRxJl+6//vrrTmzbddl2cba84uDB\ng048M+OeSX/77bdjX/+ZZ56J2u3way2teod4HX/8RB0r4fZ8QGWpjz0Va0tzbHu52Vn38++1117r\nxPZ4sqVDtj2dLR2xRkdHnbiG0+6NHj/Yu3dv4cfP3r17AaC32lX9/vUZcdJWFWRSlRCnVMnRT3r9\nIhIu+O8BsN5cqQoAqwFsDeYpAawFsC6DS3yPHT9+HHv37g3+Glv5Ap2g7KRV97Hh/Tt0KPGaKke9\ntYn+m2+1/VsK4BgaU5bjp9WPHaCOfbRzyFmr0is1j+NnGphLTkVrOMGHpK0qyK0qoVTJEfC65KCG\nOUN/FFkx0szg9c8D5usga/iEVJhW38dO3L+yHD+tvn9A6+9jTsfPtcn3KoW0VQWZVCXEKdsFOURE\nVHKh6z4aqipI+/haMDkSEVER0lYV5FqVwORIRERFGAOwJuL2WqsK0j4+FpMjERE1nb9YxOHw+ryh\nqoJN4dtERP1WcXU/vlGluyCHiIjaRmJVgarOisgMoucQ86pKgNRbn0ZERNTueFqViIjIYHIkIiIy\nmByJiIgMJkciIiKDyZGIiMhgciQiIjJY59gBRGQEXhPeWXjdI3ZlUQdEnYHHD3Ui1jk2oJXeLPzu\nEDsBLItql+R3ldhm1rWchLfMV6rGvAn71YX5BaeDAt3hqJ9TkT9Pv3h4c+imPgA7gp9XHvvJ46em\n/Wr546eIY4eaSFX5VccXvDeGgVDcBW+NtZ4m7kMXvL6CI/6/CqAr4n5DAI5E3D4CYDLv/Yt4TQXQ\n3yo/T3hvujvMbQP+fg7lsZ88ftrj+Cni2OFXc78K34EyfRX1ZlHDPlV7c9tt32T82weqPSajfdpR\nZX+OhH9+Rf88AWzxfw4D5nYFMJ31fhb9/fL4Ke+xw6/mf/GCnPoMInoZlD0A+vxTQa2kv8rtswnb\n07oB3qdlawJAV6gPYtE/zyl4P4uo7v3h27Laz6K/33rx+Kmu2ccONRmTY32KerOoW+iPLm7F7J6Y\nbWnMAOiJ+cMPbi/056mqE6q6WFXn3rxCHf53hO6a1X7y+KlNyx8/BRw71GS8WrVGBb9ZNCJY8DPq\nk23c95Caqq6usqnP/3emFX+eoYtAtqu3HE5mv/dW/H4T8PipQ57HDhWDI8faFfZmkVLcaZslzdoJ\nEemH90awXb2rIlvm5ykiPf7VhJPwToFtC23Oaj9b5vutE4+f+P1qxrFDBWByrF9LvFnUIO6PL/ij\nPdSMHfHtADCuqsPm9sJ/nqo6o6rDqtoL7yrC/f6bcVhW+1n491sjHj81aPKxQ03E5Fi7VnuziKXz\nNWvdEZuDP9bc6tTCRGQHgAlVHQzd3JI/T1Xd7v93zD8tltV+tuT3Ww2Pn/rleOxQAZgca9RKbxZ1\nmED0p9bu0PZcicgWAFDVcLF0S/w8/VNiUXM+T/r70J/VfrbC99sAHj/V96tpxw4Vg8mxPoW/WdRp\nDMCaiNt7AUxpREeULPlX7/WG39jMm0rRP89pANM1XE6f1X4W/f3Wi8dPdc0+dqjJmBzrU+ibRRVV\n/zj9q+YOhy4xD66gGwKwKc+dEpE+AGvtJ354BeSBon+es/BO19nXCb/5AtntZ9HfbxQeP41p9rFD\nTcbeqnUSkWl4PR7DvSb3A1inTeyV6M/DdMOrk+qCV5T8JLyuG6Oh+3UB2Ir5uY218Hpl5rav/if7\n3QDCPSa7/P3t8y9eCO5b2M9TRIaAuSQQ3DYA7w1ts7k9k/3k8VPTvrX88VPEsUPNxeRYpyLeLMrG\nb1bdV2XzjHlzK/Tn6V9ZGL7QowfASLi4O8v9LPr7LYOyHD/NPnaouZgciYiIDM45EhERGUyORERE\nBpMjERGRweRIRERkMDkSEREZTI7U8kRki1+XR1Q3Hj/UCJZyUMvzu6XMqir7UFLdePxQI5gciYiI\nDJ5WJSIiMk4segeIqvHbcwUNqAdruP8IvBZdvfAWx+0HW3V1rAaOn6Cn6yyAdfB6uXbx2OlMTI7U\nytar6rCIqIh0xa1g4K/7t01VZ0VkDMBOVV3tN32OWk+P2l/Nx49v2P/qht8LNbSAMXUYzjl2CP9T\n8Qi8T9LBsjrj8FYr7/Zv6/Jv21bLUjr+ygTBJ/Lg/ntUdbvfbHmnqg7W8NrBskljwWoG/mMOwxv9\njYSbTVfZl77gE77fuHpHeGUESqfdjx+zX8FixeOJd6b2par86qAveG8wCu+NJGr7DgBH4C0NVO05\neuAt9robQE/Eth3+tslaX9t/3JGIx+wGsKXO71HtfvGLx0+N31+QGAv/WfOr2C9ekNN5Yj/Rq7e4\n7MMAJqNWOfc/kU/DW6h1vZrL4/14GNELvFZ97dDj+vy5w7lP8ABG/cvxEwX3C/ar1sdRzdr2+PH3\nrUf9JadEpC/qe6DOwORIFXR+9fWdEZt3+/9WXQlevVNqww289JP+v8Fq7/2YXy39xuBOItIjIruD\nNy4/DuYHboS3cC9CMTVR2Y6f4DZ4p25vFJEB/5TviNZwepjaE5MjVTMOYMB/0wAwt9J5D4DxGt40\nHk7x2sFzTwCY8UcC4Q4nXfBGFv1+fBjAuP+GthvAhN8VZQTAthT7QY0r0/EDeKdSBwHsgpfUB+Eu\nZEwdhlertgj/tM9OeG8eD2O+FAHwyhH2aOjKuXrv34DgdNcAgOB5gk/fuyvv7lLvqtF6P/0H+78j\neA5EvEGpd+HNYv/NNup+E/Yx7Y7HD4AGjx//ttHwtjpfl9oQk2OL8P8og9KDHnifZLcD8/M0IjIe\nzA1XOdgAAAHQSURBVNHUe/8GHPL/DV/lF4wCDtf4PdWcpPyatK0ARrWGq0z9NzbWn/l4/PD4oWwx\nObaeGQBrVHV9cIOqzogI4F9ckPL+raDPr0sEgCXw3jRnAKyu4824O8Ubdzvj8VMbHj8Ui8mxNVX7\no6125Vy9969F8Ik/PDc0A+9y+pqK6mMKr2dSnrJDLaODDsbjJwGPH0rCC3JaU02nnVLcvxbBG9iu\n0G3B/9ejNluz2x2qA48fopSYHKmCf4n7AIAJDfWVVK9jyAy8qxBjRxX+9kNx96H2xOOH2gGTI0UJ\nPrFvjtgWfOqPqmFzniPtqS8qLR4/VHpMjp0n6RP7CIAhVLm4wb+tF95FEbvDdWz+47v859hlH0tt\ngccPdQQ2Hm8Rft3ZVsx39xiHV8DeA+8TeD+8ixsm1GvGXO/9a2ne3A2vRrCRxtEzmL/4wnm8/9rD\n/j6FXzttLR35ePwQZYvJkYiIyOBpVSIiIoPJkYiIyGByJCIiMpgciYiIDCZHIiIig8mRiIjIYHIk\nIiIymByJiIgMJkciIiKDyZGIiMhgciQiIjKYHImIiAwmRyIiIoPJkYiIyGByJCIiMpgciYiIDCZH\nIiIig8mRiIjIYHIkIiIy/j/tLnLJohCBVAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f073e87d198>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rc(\"text\", usetex=True)\n",
    "my_font_size =10\n",
    "#PyPlot.rc(\"font\", family=\"serif\")\n",
    "\n",
    "my_cmap = plt.cm.Greys\n",
    "\n",
    "plt.rc(\"text.latex\", unicode=True)\n",
    "fig = plt.figure(dpi=150)\n",
    "fig.set_size_inches(3 ,3)\n",
    "plt.subplot(2,2,1)\n",
    "plt.imshow(kruskal[:32,:32], interpolation='nearest', cmap=my_cmap, vmin=0, vmax=1)\n",
    "plt.ylabel(r'$ \\textrm{nnDCP}^{i,x}$', fontsize=my_font_size)\n",
    "plt.tick_params(labelsize = my_font_size)\n",
    "plt.xticks([])\n",
    "plt.yticks([0,10,20,30])\n",
    "plt.subplot(2,2,2)\n",
    "plt.imshow(kruskal[:32,32:], interpolation='nearest', cmap=my_cmap, vmin=0, vmax=1)\n",
    "plt.xticks([])\n",
    "plt.yticks([])\n",
    "plt.subplot(2,2,3)\n",
    "plt.imshow(kruskal[32:,:32], interpolation='nearest', cmap=my_cmap, vmin=0, vmax=1)\n",
    "plt.ylabel(r'$ \\textrm{nnDCP}^{i,z}$', fontsize=my_font_size)\n",
    "plt.xlabel(r'$ \\textrm{nnDCP}^{i,x}$', fontsize=my_font_size)\n",
    "plt.tick_params(labelsize = my_font_size)\n",
    "plt.yticks([0,10,20,30])\n",
    "plt.xticks([0,10,20,30])\n",
    "plt.subplot(2,2,4)\n",
    "plt.imshow(kruskal[32:,32:], interpolation='nearest', cmap=my_cmap, vmin=0, vmax=1)\n",
    "plt.subplots_adjust(hspace=0.1)\n",
    "plt.subplots_adjust(wspace=-0.1)\n",
    "plt.xlabel(r'$\\textrm{nnDCP}^{i,z}$', fontsize=my_font_size)\n",
    "plt.tick_params(labelsize = my_font_size)\n",
    "plt.yticks([])\n",
    "plt.xticks([0,10,20,30])\n",
    "plt.subplots_adjust(bottom=0.2, right=0.8, top=0.8)\n",
    "cax = plt.axes([0.8, 0.2, 0.075, 0.6])\n",
    "plt.tick_params(labelsize = my_font_size)\n",
    "c = plt.colorbar(cax=cax)\n",
    "# c.set_label(r'$p$--\\textrm{value}', fontsize=my_font_size,rotation=270)\n",
    "\n",
    "# plt.savefig(\"../txt/LSTM_approx/kruskal_mtx_plot_{}_eps_e-{}.pdf\".format(noise_name,eps_order))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.5.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}