analysis_ellipsoids.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Principal components analysis for order parameter construction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from analysis_tools.read_gsd import ReadGSD\n",
    "from analysis_tools.feature_creation import FrameToFeatures, TrajectoryToFeatures\n",
    "from analysis_tools.feature_creation import FrameToFeaturesComposition, TrajectoryToFeaturesComposition"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from numpy import array, arange"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Hard disks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "particle_inc = 100 #skip this many particles in generating features; #wp: because it selects index 0 be default, rounds up to integer if not a mulptiple of Ntot\n",
    "N_nn = 400 #number of nearest nighbors for pca analysis #wp:Should not be bigger than Ntotal\n",
    "nn_inc = 20 #reduces the number of nearest neighbors to include as features\n",
    "split = 1 #chunks the data up so it can be processed by the pca tool if really large\n",
    "\n",
    "method = ['distance']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#specify what data to read in and process\n",
    "#wp:there are 55 eta points given range below\n",
    "#file_data = [(arange(0.550, 0.690001, 0.005), 'ellipsoids'), (arange(0.695, 0.800001, 0.005),'ellipsoids')]  \n",
    "file_data = [(arange(0.550, 0.800001, 0.01),'ellipsoids/kappa_3.0')]  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Density range"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "#wp: the `_' variable is a dummy return variable like `~' in matlab\n",
    "#wp: Here, the for loop returns two arguments but only keeps the first that contains all the numbers\n",
    "etas = []\n",
    "[etas.extend(etas_) for etas_, _ in file_data]\n",
    "etas = array(etas)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Perfom PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from numpy import array_split\n",
    "from sklearn.decomposition import IncrementalPCA"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train on ideal gas structure first to identify trivial variance contributions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_components_ig = N_nn/nn_inc  #for this the more the better, however, 50 is probably enough"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "eta=0.69\n",
      "Frame partials:  1\n",
      "eta=0.8\n",
      "Frame partials:  1\n"
     ]
    }
   ],
   "source": [
    "###Note that this ideal gas variance comes from the fact that particles get sorted and creates an artificial correlation\n",
    "randomize = True\n",
    "#wp: incremental pca structure for ideal gas\n",
    "incpca_ig = IncrementalPCA(n_components=n_components_ig, whiten=False)\n",
    "\n",
    "frames = None\n",
    "features = None\n",
    "split_features = None\n",
    "\n",
    "#wp: return two variables from file data in for loop: the etas list and the file directory\n",
    "for etas_, file_base in file_data:\n",
    "    #wp: for each eta in the list\n",
    "    #for eta in etas_:\n",
    "    for eta in etas_[-1:]:\n",
    "        print 'eta={}'.format(eta)\n",
    "        #wp:access GSD files from directory given in 'file_base' variable\n",
    "        filename = \"{}/trajectory_{:.4f}.gsd\".format(file_base, eta)\n",
    "        #wp:Returns frames as dictionary [coords,diameters,types,L,D]. Size is (trajectory prod steps)/(trajectory collect period)\n",
    "        frames = ReadGSD(filename, shuffle_data=True, randomize=randomize)\n",
    "        \n",
    "        #wp: If array_split(array,1) then array is not split, although it does get transformed into a single list it seems\n",
    "        frames_split = array_split(frames, split)\n",
    "       \n",
    "        #wp: resets frame\n",
    "        frames = None\n",
    "        count = 0\n",
    "        \n",
    "        for frames_partial in frames_split:\n",
    "            #wp:All trajectories done at eta_ get passed down at once\n",
    "            features_partial = TrajectoryToFeatures(frames_partial, \n",
    "                                                    N_nn=N_nn, \n",
    "                                                    method=method, \n",
    "                                                    particle_inc=particle_inc, \n",
    "                                                    nn_inc=nn_inc)\n",
    "            \n",
    "            incpca_ig.partial_fit(features_partial)\n",
    "            count = count + 1\n",
    "            print \"Frame partials: \", count\n",
    "\n",
    "        frames_split = None\n",
    "        features_partial_ig=features_partial"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train on actual data with trivial data variance removed by the ideal gas model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 287,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_components = N_nn/nn_inc #this number will not affect the quality of the result just the computation time and amount of info returned"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 288,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "eta=0.55\n",
      "Frame split:  1\n",
      "eta=0.555\n",
      "Frame split:  1\n",
      "eta=0.56\n",
      "Frame split:  1\n",
      "eta=0.565\n",
      "Frame split:  1\n",
      "eta=0.57\n",
      "Frame split:  1\n",
      "eta=0.575\n",
      "Frame split:  1\n",
      "eta=0.58\n",
      "Frame split:  1\n",
      "eta=0.585\n",
      "Frame split:  1\n",
      "eta=0.59\n",
      "Frame split:  1\n",
      "eta=0.595\n",
      "Frame split:  1\n",
      "eta=0.6\n",
      "Frame split:  1\n",
      "eta=0.605\n",
      "Frame split:  1\n",
      "eta=0.61\n",
      "Frame split:  1\n",
      "eta=0.615\n",
      "Frame split:  1\n",
      "eta=0.62\n",
      "Frame split:  1\n",
      "eta=0.625\n",
      "Frame split:  1\n",
      "eta=0.63\n",
      "Frame split:  1\n",
      "eta=0.635\n",
      "Frame split:  1\n",
      "eta=0.64\n",
      "Frame split:  1\n",
      "eta=0.645\n",
      "Frame split:  1\n",
      "eta=0.65\n",
      "Frame split:  1\n",
      "eta=0.655\n",
      "Frame split:  1\n",
      "eta=0.66\n",
      "Frame split:  1\n",
      "eta=0.665\n",
      "Frame split:  1\n",
      "eta=0.67\n",
      "Frame split:  1\n",
      "eta=0.675\n",
      "Frame split:  1\n",
      "eta=0.68\n",
      "Frame split:  1\n",
      "eta=0.685\n",
      "Frame split:  1\n",
      "eta=0.69\n",
      "Frame split:  1\n",
      "eta=0.695\n",
      "Frame split:  1\n",
      "eta=0.7\n",
      "Frame split:  1\n",
      "eta=0.705\n",
      "Frame split:  1\n",
      "eta=0.71\n",
      "Frame split:  1\n",
      "eta=0.715\n",
      "Frame split:  1\n",
      "eta=0.72\n",
      "Frame split:  1\n",
      "eta=0.725\n",
      "Frame split:  1\n",
      "eta=0.73\n",
      "Frame split:  1\n",
      "eta=0.735\n",
      "Frame split:  1\n",
      "eta=0.74\n",
      "Frame split:  1\n",
      "eta=0.745\n",
      "Frame split:  1\n",
      "eta=0.75\n",
      "Frame split:  1\n",
      "eta=0.755\n",
      "Frame split:  1\n",
      "eta=0.76\n",
      "Frame split:  1\n",
      "eta=0.765\n",
      "Frame split:  1\n",
      "eta=0.77\n",
      "Frame split:  1\n",
      "eta=0.775\n",
      "Frame split:  1\n",
      "eta=0.78\n",
      "Frame split:  1\n",
      "eta=0.785\n",
      "Frame split:  1\n",
      "eta=0.79\n",
      "Frame split:  1\n",
      "eta=0.795\n",
      "Frame split:  1\n",
      "eta=0.8\n",
      "Frame split:  1\n"
     ]
    }
   ],
   "source": [
    "randomize = False\n",
    "incpca = IncrementalPCA(n_components=n_components , whiten=False)\n",
    "\n",
    "frames = None\n",
    "features = None\n",
    "split_features = None\n",
    "\n",
    "for etas_, file_base in file_data:\n",
    "    for eta in etas_:\n",
    "        print 'eta={}'.format(eta)\n",
    "        filename = \"{}/trajectory_{:.4f}.gsd\".format(file_base, eta)\n",
    "        frames = ReadGSD(filename, shuffle_data=True, randomize=randomize)\n",
    "        \n",
    "        frames_split = array_split(frames, split)\n",
    "        frames = None\n",
    "        count = 0\n",
    "        for frames_partial in frames_split:\n",
    "            features_partial = TrajectoryToFeatures(frames_partial, \n",
    "                                                    N_nn=N_nn, \n",
    "                                                    method=method, \n",
    "                                                    particle_inc=particle_inc, \n",
    "                                                    nn_inc=nn_inc)\n",
    "            #wp:project the *hard disk* feature vector unto a lower dimensionalized representation of the ideal gas space that captures most of the ideal gas variance\n",
    "            #wp:This then creates a 'hard disk' feature vector that strongly contains the sorting variance seen in the ideal gas\n",
    "            features_partial_ig_projection = incpca_ig.inverse_transform(incpca_ig.transform(features_partial))\n",
    "            #wp:remove this partial reconstructed hard disk ideal gas projection from real data features\n",
    "            incpca.partial_fit(features_partial - features_partial_ig_projection)\n",
    "            count = count + 1\n",
    "            print \"Frame split: \", count\n",
    "        #wp:resets frame split\n",
    "        frames_split = None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "eta=0.55\n",
      "Frame split:  1\n",
      "eta=0.555\n",
      "Frame split:  1\n",
      "eta=0.56\n",
      "Frame split:  1\n",
      "eta=0.565\n",
      "Frame split:  1\n",
      "eta=0.57\n",
      "Frame split:  1\n",
      "eta=0.575\n",
      "Frame split:  1\n",
      "eta=0.58\n",
      "Frame split:  1\n",
      "eta=0.585\n",
      "Frame split:  1\n",
      "eta=0.59\n",
      "Frame split:  1\n",
      "eta=0.595\n",
      "Frame split:  1\n",
      "eta=0.6\n",
      "Frame split:  1\n",
      "eta=0.605\n",
      "Frame split:  1\n",
      "eta=0.61\n",
      "Frame split:  1\n",
      "eta=0.615\n",
      "Frame split:  1\n",
      "eta=0.62\n",
      "Frame split:  1\n",
      "eta=0.625\n",
      "Frame split:  1\n",
      "eta=0.63\n",
      "Frame split:  1\n",
      "eta=0.635\n",
      "Frame split:  1\n",
      "eta=0.64\n",
      "Frame split:  1\n",
      "eta=0.645\n",
      "Frame split:  1\n",
      "eta=0.65\n",
      "Frame split:  1\n",
      "eta=0.655\n",
      "Frame split:  1\n",
      "eta=0.66\n",
      "Frame split:  1\n",
      "eta=0.665\n",
      "Frame split:  1\n",
      "eta=0.67\n",
      "Frame split:  1\n",
      "eta=0.675\n",
      "Frame split:  1\n",
      "eta=0.68\n",
      "Frame split:  1\n",
      "eta=0.685\n",
      "Frame split:  1\n",
      "eta=0.69\n",
      "Frame split:  1\n",
      "eta=0.695\n",
      "Frame split:  1\n",
      "eta=0.7\n",
      "Frame split:  1\n",
      "eta=0.705\n",
      "Frame split:  1\n",
      "eta=0.71\n",
      "Frame split:  1\n",
      "eta=0.715\n",
      "Frame split:  1\n",
      "eta=0.72\n",
      "Frame split:  1\n",
      "eta=0.725\n",
      "Frame split:  1\n",
      "eta=0.73\n",
      "Frame split:  1\n",
      "eta=0.735\n",
      "Frame split:  1\n",
      "eta=0.74\n",
      "Frame split:  1\n",
      "eta=0.745\n",
      "Frame split:  1\n",
      "eta=0.75\n",
      "Frame split:  1\n",
      "eta=0.755\n",
      "Frame split:  1\n",
      "eta=0.76\n",
      "Frame split:  1\n",
      "eta=0.765\n",
      "Frame split:  1\n",
      "eta=0.77\n",
      "Frame split:  1\n",
      "eta=0.775\n",
      "Frame split:  1\n",
      "eta=0.78\n",
      "Frame split:  1\n",
      "eta=0.785\n",
      "Frame split:  1\n",
      "eta=0.79\n",
      "Frame split:  1\n",
      "eta=0.795\n",
      "Frame split:  1\n",
      "eta=0.8\n",
      "Frame split:  1\n"
     ]
    }
   ],
   "source": [
    "########wp: System fit *without ideal gas*\n",
    "randomize = False\n",
    "incpca = IncrementalPCA(n_components=n_components , whiten=False)\n",
    "\n",
    "frames = None\n",
    "features = None\n",
    "split_features = None\n",
    "\n",
    "split=1\n",
    "for etas_, file_base in file_data:\n",
    "    for eta in etas_:\n",
    "        print 'eta={}'.format(eta)\n",
    "        filename = \"{}/trajectory_{:.4f}.gsd\".format(file_base, eta)\n",
    "        frames = ReadGSD(filename, shuffle_data=True, randomize=randomize)\n",
    "        \n",
    "        frames_split = array_split(frames, split)\n",
    "        frames = None\n",
    "        count = 0\n",
    "        for frames_partial in frames_split:\n",
    "            features_partial = TrajectoryToFeatures(frames_partial, \n",
    "                                                    N_nn=N_nn, \n",
    "                                                    method=method, \n",
    "                                                    particle_inc=particle_inc, \n",
    "                                                    nn_inc=nn_inc)\n",
    "\n",
    "            #wp: fits pure system features (no idea gas reduction)\n",
    "            incpca.partial_fit(features_partial )\n",
    "            count = count + 1\n",
    "            print \"Frame split: \", count\n",
    "        #wp:resets frame split\n",
    "        frames_split = None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "###Plots the feature matrix for the last eta\n",
    "fpartial_ig=np.asarray(features_partial_ig);\n",
    "fpartial=np.asarray(features_partial);\n",
    "sample_num=3000\n",
    "#wp:ideal gas\n",
    "plt.plot(fpartial_ig[sample_num,:],'g-^')\n",
    "#wp:system\n",
    "plt.plot(fpartial[sample_num,:],'r-o')\n",
    "plt.xlim(0,n_components+2);plt.ylim(0,fpartial[sample_num,-1]+2);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Compute the order parameters based on two nested models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 289,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "eta=0.55\n",
      "1\n",
      "eta=0.555\n",
      "1\n",
      "eta=0.56\n",
      "1\n",
      "eta=0.565\n",
      "1\n",
      "eta=0.57\n",
      "1\n",
      "eta=0.575\n",
      "1\n",
      "eta=0.58\n",
      "1\n",
      "eta=0.585\n",
      "1\n",
      "eta=0.59\n",
      "1\n",
      "eta=0.595\n",
      "1\n",
      "eta=0.6\n",
      "1\n",
      "eta=0.605\n",
      "1\n",
      "eta=0.61\n",
      "1\n",
      "eta=0.615\n",
      "1\n",
      "eta=0.62\n",
      "1\n",
      "eta=0.625\n",
      "1\n",
      "eta=0.63\n",
      "1\n",
      "eta=0.635\n",
      "1\n",
      "eta=0.64\n",
      "1\n",
      "eta=0.645\n",
      "1\n",
      "eta=0.65\n",
      "1\n",
      "eta=0.655\n",
      "1\n",
      "eta=0.66\n",
      "1\n",
      "eta=0.665\n",
      "1\n",
      "eta=0.67\n",
      "1\n",
      "eta=0.675\n",
      "1\n",
      "eta=0.68\n",
      "1\n",
      "eta=0.685\n",
      "1\n",
      "eta=0.69\n",
      "1\n",
      "eta=0.695\n",
      "1\n",
      "eta=0.7\n",
      "1\n",
      "eta=0.705\n",
      "1\n",
      "eta=0.71\n",
      "1\n",
      "eta=0.715\n",
      "1\n",
      "eta=0.72\n",
      "1\n",
      "eta=0.725\n",
      "1\n",
      "eta=0.73\n",
      "1\n",
      "eta=0.735\n",
      "1\n",
      "eta=0.74\n",
      "1\n",
      "eta=0.745\n",
      "1\n",
      "eta=0.75\n",
      "1\n",
      "eta=0.755\n",
      "1\n",
      "eta=0.76\n",
      "1\n",
      "eta=0.765\n",
      "1\n",
      "eta=0.77\n",
      "1\n",
      "eta=0.775\n",
      "1\n",
      "eta=0.78\n",
      "1\n",
      "eta=0.785\n",
      "1\n",
      "eta=0.79\n",
      "1\n",
      "eta=0.795\n",
      "1\n",
      "eta=0.8\n",
      "1\n"
     ]
    }
   ],
   "source": [
    "#wp: Once *all* the data has been fit with pca, then start getting the transformed components using *all* the fitted data\n",
    "randomize=False\n",
    "frames = None\n",
    "features = None\n",
    "raw_OPs = []\n",
    "\n",
    "split=1;\n",
    "for etas_, file_base in file_data:\n",
    "    for eta in etas_:\n",
    "        print 'eta={}'.format(eta)\n",
    "        filename = \"{}/trajectory_{:.4f}.gsd\".format(file_base, eta)\n",
    "        frames = ReadGSD(filename, shuffle_data=True, randomize=randomize)\n",
    "        \n",
    "        frames_split = array_split(frames, split)\n",
    "        frames = None\n",
    "        count = 0\n",
    "        means = []\n",
    "        #wp: Array of transformed/projected pca features. OP=order parameter\n",
    "        raw_OPs_single = []\n",
    "        for frames_partial in frames_split:\n",
    "            #wp: returns nearest sorted neighbor distances for traj_frames*Npart/particle_inc examples\n",
    "            features_partial = TrajectoryToFeatures(frames_partial, \n",
    "                                                    N_nn=N_nn, \n",
    "                                                    method=method, \n",
    "                                                    particle_inc=particle_inc, \n",
    "                                                    nn_inc=nn_inc)\n",
    "            #wp:regenerate the *hard disk* projection on the ideal gas space to subtract from the hard disk feature vector\n",
    "            features_partial_ig_projection = incpca_ig.inverse_transform(incpca_ig.transform(features_partial))\n",
    "            transformed_features_partial = incpca.transform(features_partial - features_partial_ig_projection)\n",
    "            #transformed_features_partial = incpca.transform(features_partial)\n",
    "            raw_OPs_single.append(transformed_features_partial)\n",
    "            count = count + 1\n",
    "            print count\n",
    "        #wp: Order parameters for each eta: list of arrays of size (# of etas)x(traj_frames*Npart/particle_inc x N_nn/nn_inc (or n_components))\n",
    "        #wp: So for a run of Npart=400,particle_inc=10, with 100 traj frames and choosing N_nn=30, keeping every nn_inc=2 we get (# etas)x(4000x15)\n",
    "        raw_OPs.append(raw_OPs_single)  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Compute order parameter stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 269,
   "metadata": {},
   "outputs": [],
   "source": [
    "from numpy import array, mean, std, abs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 293,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The feature matrix input for each eta is of dimensions 8000 20\n",
      "8000 20\n",
      "Number of rows:  51\n",
      "Number of columns:  1\n",
      "Array size for each eta:  (8000, 20)\n",
      "[ 2.85546684e-17 -4.91631399e-18 -4.77054209e-19  0.00000000e+00\n",
      "  0.00000000e+00  0.00000000e+00  0.00000000e+00  0.00000000e+00\n",
      "  0.00000000e+00  0.00000000e+00  0.00000000e+00  0.00000000e+00\n",
      "  0.00000000e+00  0.00000000e+00  0.00000000e+00  0.00000000e+00\n",
      "  0.00000000e+00  0.00000000e+00  0.00000000e+00  0.00000000e+00]\n",
      "Shape of average order parameter (51, 20)\n"
     ]
    }
   ],
   "source": [
    "OPs = []\n",
    "OPs_std = []\n",
    "OPs_abs = []\n",
    "\n",
    "for single_eta in raw_OPs:\n",
    "    merged_data = []\n",
    "    \n",
    "    for split in single_eta:\n",
    "        merged_data.extend(split)\n",
    "    OPs.append(mean((split), axis=0))\n",
    "    OPs_std.append(std(split, axis=0))\n",
    "    OPs_abs.append(mean(abs(split), axis=0))\n",
    "\n",
    "OPs = array(OPs)\n",
    "OPs_std = array(OPs_std)\n",
    "OPs_abs = array(OPs_abs)\n",
    "\n",
    "\n",
    "\n",
    "#wp:figuring things out...\n",
    "print \"The feature matrix input for each eta is of dimensions\",len(features_partial),len(features_partial[0])\n",
    "print len(raw_OPs_single[0]),len(raw_OPs_single[0][0])\n",
    "\n",
    "\n",
    "eta_index=0\n",
    "print \"Number of rows: \", len(raw_OPs)\n",
    "print \"Number of columns: \", len(raw_OPs[eta_index])\n",
    "print \"Array size for each eta: \", raw_OPs[eta_index][0].shape\n",
    "print raw_OPs[eta_index][0][1]\n",
    "\n",
    "#wp: for the averaged value\n",
    "print \"Shape of average order parameter\", OPs.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Check out the results of PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "###Plots the feature matrix for the last eta\n",
    "fpartial=np.asarray(features_partial);\n",
    "plt.plot(fpartial[1250,:],'r-o')\n",
    "plt.xlim(0,n_components+10);plt.ylim(0,20);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Explained variance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 294,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot the principal components\n",
    "colors = 200*['green', 'purple', 'orange', 'teal', 'black', 'red']\n",
    "\n",
    "plt.plot(incpca.explained_variance_ratio_, marker='o', linewidth=1.0, markersize=12, color = 'red', markerfacecolor='None')    \n",
    "\n",
    "#backround grid details\n",
    "axes = plt.gca()\n",
    "axes.grid(b = True, which = 'both', axis = 'both', color = 'gray', linestyle = '-', alpha = 0.5, linewidth = 0.5) \n",
    "#axes.set_axis_bgcolor('white')  \n",
    "\n",
    "#font scpecifications\n",
    "title_font = {'family' : 'arial', 'color'  : 'black', 'weight' : 'heavy','size': 20}\n",
    "axis_label_font = {'family' : 'arial', 'color'  : 'black', 'weight' : 'normal','size': 20}                                                   \n",
    "\n",
    "#figure size and tick style\n",
    "plt.rcParams[\"figure.figsize\"] = [6,6]\n",
    "plt.rc('axes',edgecolor='black',linewidth=1)\n",
    "plt.tick_params(which='both', axis='both', color='black', length=4, width=0.5)\n",
    "plt.rcParams['xtick.direction'] = 'in'\n",
    "plt.rcParams['ytick.direction'] = 'in'\n",
    "\n",
    "#plt.yscale('log')\n",
    "#!!!wp: Why eta? #Shouldn't be features?\n",
    "plt.xlabel(r'component', y=3, fontsize=20, fontdict = axis_label_font)\n",
    "plt.ylabel(r'Explained variance', fontsize=20, fontdict = axis_label_font)\n",
    "\n",
    "#title and axis labels\n",
    "plt.tick_params(axis='both', labelsize=20)\n",
    "plt.title('Explained variance for pricipal components', y=1.05, fontdict = title_font)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Principal component weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 295,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot the principal components\n",
    "colors = 200*['green', 'purple', 'orange', 'teal', 'black', 'red']\n",
    "for i in range(0,3):\n",
    "    p_data = incpca.components_[i]\n",
    "    plt.plot(range(len(incpca.components_[0])), p_data, marker='o', linewidth = 1.0, markersize=5, color = colors[i])  \n",
    "\n",
    "#backround grid details\n",
    "axes = plt.gca()\n",
    "axes.grid(b = True, which = 'both', axis = 'both', color = 'gray', linestyle = '-', alpha = 0.5, linewidth = 0.5) \n",
    "#axes.set_axis_bgcolor('white')  \n",
    "\n",
    "#font scpecifications\n",
    "title_font = {'family' : 'arial', 'color'  : 'black', 'weight' : 'heavy','size': 20}\n",
    "axis_label_font = {'family' : 'arial', 'color'  : 'black', 'weight' : 'normal','size': 20}                                                   \n",
    "\n",
    "#figure size and tick style\n",
    "plt.rcParams[\"figure.figsize\"] = [6,6]\n",
    "plt.rc('axes',edgecolor='black',linewidth=1)\n",
    "plt.tick_params(which='both', axis='both', color='black', length=4, width=0.5)\n",
    "plt.rcParams['xtick.direction'] = 'in'\n",
    "plt.rcParams['ytick.direction'] = 'in'\n",
    "\n",
    "#axis range and labels (also specify if log or not)\n",
    "#plt.xlim(0.1,80)\n",
    "#plt.ylim(-1,0)\n",
    "#plt.ylim(0,0.5)\n",
    "#plt.ylim(-2,1.5)\n",
    "#plt.ylim(5, 13)\n",
    "#plt.ylim(1, 0.001)\n",
    "#plt.ylim(-2000.0, 2000)\n",
    "\n",
    "#plt.xscale('log')\n",
    "#wp:axes labels\n",
    "plt.xlabel(r'$n$th nerest neighbor', y=3, fontsize=20, fontdict = axis_label_font)\n",
    "plt.ylabel(r'$P_{i}$ weight', fontsize=20, fontdict = axis_label_font)\n",
    "\n",
    "#title and axis labels\n",
    "plt.tick_params(axis='both', labelsize=20)\n",
    "plt.title('PCA feature weights', y=1.05, fontdict = title_font)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Order parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 298,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot the principal components\n",
    "colors = 200*['green', 'purple', 'orange', 'teal', 'black', 'red']\n",
    "for i in range(0,3): #12,13\n",
    "    p_data = OPs[:,i]\n",
    "    plt.plot(etas, (p_data), marker='o', linewidth = 1.0, markersize=12, color = colors[i], markerfacecolor='None')\n",
    "    p_data = OPs_std[:,i]\n",
    "    #plt.plot(etas, (p_data), marker='^', linewidth = 1.0, markersize=12, color = colors[i])\n",
    "\n",
    "#backround grid details\n",
    "axes = plt.gca()\n",
    "axes.grid(b = True, which = 'both', axis = 'both', color = 'gray', linestyle = '-', alpha = 0.5, linewidth = 0.5) \n",
    "\n",
    "\n",
    "#figure size and tick style\n",
    "plt.rcParams[\"figure.figsize\"] = [6,6]\n",
    "plt.rc('axes',edgecolor='black',linewidth=1)\n",
    "plt.tick_params(which='both', axis='both', color='black', length=4, width=0.5)\n",
    "plt.rcParams['xtick.direction'] = 'in'\n",
    "plt.rcParams['ytick.direction'] = 'in'\n",
    "\n",
    "#axis range and labels (also specify if log or not)\n",
    "#plt.xlim(0.60, 0.84)\n",
    "#plt.ylim(-0.1,0.1)\n",
    "\n",
    "#plt.ylim(-0.1,0.1)\n",
    "#plt.yscale('log')\n",
    "\n",
    "#plt.yscale('log')\n",
    "plt.xlabel(r'$\\eta$', y=3, fontsize=20, fontdict = axis_label_font)\n",
    "plt.ylabel(r'$\\langle P_{i} \\rangle$ and $\\langle |P_{i}| \\rangle$', fontsize=20, fontdict = axis_label_font)\n",
    "\n",
    "#title and axis labels\n",
    "plt.tick_params(axis='both', labelsize=20)\n",
    "plt.title('PCA ellipsoids', y=1.05, fontdict = title_font)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [],
   "source": [
    "OPs_400_20=OPs;   #wp: N_nn=400, nn_inc=20, using 10000 samples-"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {},
   "outputs": [],
   "source": [
    "OPs_200_10=OPs;   #wp: N_nn=200, nn_inc=10, using 10000 samples-"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [],
   "source": [
    "OPs_400_10=OPs;   #wp: N_nn=400, nn_inc=10, using 10000 samples-"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "metadata": {},
   "outputs": [],
   "source": [
    "OPs_400_40_10k=OPs; #wp: N_nn=400, nn_inc=40, using 10000 samples-"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [],
   "source": [
    "OPs_400_20_ig=OPs;  #wp: N_nn=400, nn_inc=40, using 10000 samples-"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 233,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAFiCAYAAAAZYsRyAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvFvnyVgAAIABJREFUeJzs3XtclVW++PHPYnMTRJS7bFS8ooIohWbjmGmXycaw0bEsZ6YmKwemM+d3Gm06TeWvpmbKM7/pnKlQa+wyjZWaTZFTp6l0yvIG5Q1FFAGFjYAgclPu6/cH7O2+ARsBuX3frxevYT/7eZ69kOn5stb6ru9SWmuEEEIIALeeboAQQojeQ4KCEEIICwkKQgghLCQoCCGEsJCgIIQQwkKCghBCCAv3nm5AZyilJJ9WCCEug9ZaOTvep4MCgKyzEEKIjlHKaTwAZPhICCGEFQkKQgghLCQoCCGEsJCgIIQQwkKCghBCCAsJCkIIISwkKAghhLCQoCCEEMJCgoIQQgiLPr+iWQgh+pOwP4ZRVF3U6vuhvqEUrizsts+XnoIQQvQibQUEV97vLAkKQgghLCQoCCFEB2it2Zu/l3s+uIeA5wMwPG0g4PkA7v3gXvaZ9vX5Ip0SFIQQwkX1jfUsT1nO0q1LiQmOIT0pndrHa0lPSic6OJo737uT5SnLqW+s7+mmXjbVl6OaUkr35fYLIfoOrTXLU5ZTUFnA1ju24uvp63BOdV01izYvwuhnZEPChjZLVLc3odxmW1Z37rmnlGp1PwXpKQghhAv2mfaxI3dHqwEBwNfTl/fveJ8duTtILUht837dPWF8ubokKCilblFKZSqlspRSjzp530sptanl/b1KqciW44FKqR1KqSql1Et21/yr5Z4HWr5CuqKtQghxOdamrSUpPqnVgGDm6+lLYnwia9PWtnpObx5e6nRQUEoZgJeB+cBk4C6l1GS705YDZVrrccALwPMtx2uAJ4CVrdx+mdZ6WstXcWfbKoQQlyslM4Vlsctsjn1b8K3TieVlU5aRkpkCOE5Muz3lxuA/DL7sdoT6hl72ta7oip7CDCBLa52tta4D3gUW2p2zEHiz5fv3gBtU84RAtdb6a5qDgxBCXFEdySQqry0nxDfE5to5b8zh6NmjDvcN8Q2hvKbc6cT0zvt2uvxgjxwaya3jb+Uvt/0FvVqjV+tuXbgGXRMUjECe1ev8lmNOz9FaNwDlQKAL9369ZejoCdXWjI0QQnRQRzOJ/L38Ka6+NGBRXF1MdX01BwoPONy7uLqYIV5DWLFtBQWVBaQnprNq1irC/cJ59dtX+bcZ/+ZSGxPjEymqKiK9OL1rfmgX9OaJ5mVa6ynA7Javnzo7KT4+3vL1yiuvXNEGCiH6Jq210we2u5s74X7hrJq1ivTEdEyVJlZsW4HWmoSoBDYe2mi5R875HACnQWHj4Y18b8T3nE5MOxuGas2yKcvILM3kyNkjnfyJXdcVtY9MwAir1xEtx5ydk6+Ucgf8gdK2bqq1NrX8b6VS6m2ah6n+an9eWlra5bdcCDEgmTOJ0hPT280kilkbQ2pBKonxiSzdupSk6c2Tzdll2QQMCuBAkW1QqKqrIjk1mZiQGKcT0+W15XgaPF1qZ4hvCBfqLlzRoNAVPYVUYLxSarRSyhNYCqTYnZMC3NPy/Y+B7W0tMFBKuSulglq+9wAWAFeu/ySE6NcuJ5NohnEGcyPnsmjzIqrrqskuy+a2CbdxoPCAZe6hqq6KxZsXM2/0PHbl7XLoEaSaUnF3c2f0/4zGy+DV5meH+oZSXF2Mv7c/FbUVlF0s69wP7aJOB4WWOYKHgE+BDGCz1vqIUupppVRCy2kbgEClVBbwMGBJW1VK5QJ/Au5VSuW3ZC55AZ8qpQ4BB2juabza2bYKIQQ4H8LRWvPdme8czjVnEimlWL9gPUY/IzFrY/jw2IdMDpoMGr498y1rvllDTHIMRj8j6xesd5iYBrgv5T6ig6NZee1Kah6vQa/WrLx2Jc/MfQaA9+9432ZCeePhjSycuJDJwZOvWG+hS+YUtNYfa60naK3Haq2fbTn2pNY6peX7Gq31Eq31OK31DK11ttW1kVrrAK31YK11hNb6aEtW0tVa61itdbTW+t+11o1d0VYhhHD2wM4uy+Z7G75HY5Pto8acSQTgYfBgQ8IGNv14E3kVeTyz8xlKL5Yy7815ZJRksGXJFl5b+BoeBg+HienymnJyynL48/w/88bBN6iuqwYgoySDsMFh+Hr4svHwpTkL8zBUYnwiMcExHCnuQ0FBCCH6EvsHNsCJcyeobazldPlpm+PmIRwzpRQzjDPwNHhyKPEQK7+3kt/M+g2vL3yd6cbplvPsJ6ZTC1KJGx7HrBGzbIahjpUc41jJMRKiEvgs+zPKa8pthqGmh08nOiS6b/UUhBCiN3B13YH9Axsg61wWAJmlmTbHNx7eSEJUgs2xusY6iquLiRgSwbSwaQ6TzdCcTpqclmzpEezO2821EdfaDENNTp7MqfJTvH34bR6a8RAzjTN5IOUBm2EopRTRwdFXLC1VgoIQol/oyLoD+wc2wInSE3i4eZBZcikoWA/hWDt1/hTGIUbc3dybg4KTtFT7iend+c1BAS4NQ62esxqA0oulXPf6dezK38XXeV/bDEMBxITESE9BCCFc1dF1B9PDp9s8sKF5+Oj6yOstPQX7IRxr2WXZjBk2BoAJgRMoqCygsrbS5hz7iel/5f6LUf6jqG+sx1Rh4r92/Re/+ew3hPuFU/mflTQ82UDxymIuNlzEOMR2/W+4Xzi1DbWcrT7bXf+EFhIUhBB9XkcrmKadSbN5YK/5Zg3HSo5xy9hbOFR0yCGTyL6gQnZZNmOGNgcFdzd3ooOjOVR0yOEzzT2CNTeuwU25cdPfbmLQs4OIXRdLRkkGiyYt4idTfmLpEQzyGMTtE29nU/omm/sopa7YvIIEBSFEn3c56w6sM4nSi9PJOZ/DI58/wq68XQ6ZRPasewpAq0NI0PxAr66v5rao2yh9pJSGJxsofaSU1xe+TkVdBRODJtqcv2zKMpssJLPo4OgrkoEkQUEI0ec5W3fQpJscMozAtoKpOZPoyTlPEjk0kron6vB29+bF+S/aZBLZyzmf43JQgEuTzPaOlRxjUvAkm2NzI+diqjTZzG3AlZtXkKAghOjznK07+PjExyS8k+BwrvW6A7Osc1mMDxiPm3JjXMA4TpSeaPPzssuyGT1stOV1axlIZrvzdzMzYqbNsSbdxPHS40QFRtkcd1NuzBk1hzu23GGTQfWP4/9gd/7ubt8DuitqHwkhRLfQWrPPtI/ktGQ+yvyI8tpy/L38SYhKIGl6EtPDp6OUsqw7CPcLt1ybVpDGoaJDNDY1YnAzWI7brzuA5syjcQHjgOaJ48zSTOKGxzltk3kbzemvOvYk1FOKUN9Qm/LWlbWVnCw7ybSwaTbnni4/TcCgAPy8/CzH6hvrWbFtBTtP7aSmsYbDiYcJHdxc7mJd6jqe/fpZ7vvwPl657RWnw1pdQXoKQoheqSMpps7WHewv3M/Fhotkl2XbHHe27uDEuROMDxgPQFRglMPQjbX2ttG0f3+faR9xYXEORfAyzmbYzCdYZ1BlPpRJkE8Q+RX5lgyqp+Y+xVDvoeScz7FUbu0OEhSEEL1OR1NMf3H1LxzWHew/s59JQZM4XHzYcqy1dQdZ57IYH9gSFIKiHBawdYb1+gRrGSUZTAq6NJ9gnUE12Gswd8fcbTPhrJQiJiSGldeudGkP6MslQUEI0et0NMVUKWWz7qD0QinlteXcPvF2S6poW+sOTpy7NHwUFdj1QcF+PgGaJ5mtewr2GVTLYpex6cgmGpoaLOdEB0dzsuxku3tAd4YEBSFEr9PRFNN1366zWXfw6BePMjFwIpODJ7PPtK/NdQf1jfXkledZsomigqI4Xnq8S4ZntNbsyd/DtSPa7ynYZ1CNCxhH5NBI/nnyn5Zj5gwk6wyqriZBQQjR6zhLMd1n2uf0QWh+QFqvO9h/Zj+Hig5x7wf38unJT9tcd3Cq/BThfuGWMf+h3kPx8fChoLKg0z/HiXMnGOw52GYC3My+p+Asg+o/Zv4Hqz5bRU1D8zb25hpIzjKouooEBSFEr+PsAbnlyBb+e89/O5xr/YA0rzuYEDiBtQvWcvG3F/EyePHS/JdaXXdgnXlk1lVDSK2tTyi5UEJ9Yz1hg8Msx5xVbr0z+k4mB0/m8e2PA1hWNRdVFTlkUHUVCQpCiF7H2QMyoySD3fm7qWussznuLMV0f+F+4sLi8DB4MCFwAkfPHm31s6wzj8yiApuHkDqrvfkE62EsZxlUSinW/nAt76S/w5e5XxLkE8Qg90EkpyU7ZFB1FQkKQohex9kDMqMkAx8PH9IKbPdlt08xra6r5tT5U0wOngxAbGis07pEZlnnshx7CkHO01KtJ31bM9hjsOX7VjOPzmY4rGR2VrkVIMgniFcWvMLP/v4ztudsB+C5r5/jrwf/6rQseGdJUBBC9Dr2D8iL9RcpqCzgrpi7+OrUV5bznKWYHio6xKTgSZa5gykhU2zSUu2dOHfCko5q1trwUV55HiOGjECv1iyatIhNP96EXq3J/lU2AYMCWHvrWm4ZfwvQvGgt61yW00Vwx0qOMTHQtuaRfaltazePvRlvD28S3kmgpqGG2NBYan5b43TNRmdJUBBC9Dr2D8jjpccZM2wM80bPY+fpnUDrKabmoSOzKaHtBIVSx+Ej86pme9Y1j6yvGz1sND+N/Sk783ZaaiClFqQyNXSqw6I1aMk8susp2JfaXvPNGkwVJuoa6vjp339KU1MTNfU1hA0OY1rYNDwMHk7XbHS2xyBlLoQQvY75Abli2wpi1sZwbcS1RPpHMtM4k3s/uJfnvn6OdWnrmDd6nkOK6f4z+7lq+FWW1+bhI621Qwns+sZ68irybOoYAYwZNgZThYnahlq83L0sx83VUbXWnCw7aTPs9Pbhtzl7oXm/A/XUpc8xf29d/sI+88jMnEGVWpBKcmoysetiOX/xPCi4K+YuVs1axWNfPOYwHGZesxGzNobUglRmGGe49g/thAQFIUSvZP2AfPCjBzleepyR/z0SjWZX3i62LNniNKNof+F+7ou7z/J6+ODhNOkmiqqLbLJ9AHLP52L0Mzr8Ne9h8GDU0FFkncsiOiTactwcFAoqC/Dz9LOpW2QOCK0xl7+4WH+RM1VnbKqsWjNnUJkf7Pd+cC/RwdH8v93/z7LCufRiqU3ggeag8/C1D7M2bW2ngoIMHwkhei3rFNO/JPyFhicbuD/ufm4cc6PTgFDfWM/Rs0eJDY21uceUkCkcLnIcQnI2n2DmbF7BHBTsJ6ddHcuvb6y3DIW5u7n2N7l5zYYrNZe6YlGbBAUhRK9nvfr3ulHX2Uw22583augoh5XQrWUgZZ3LYtywcQ7HwXlhvOyybEYPHW0TTMx1mlyxYtsKjp49arOSuT3O1my0pisWtUlQEEL0ag1NDWSdyyIqqHnfgdmjZvPVqa+cTqjuP2M7yWzWWgbSidI2egpBURw/Z7tWwdxTsJ5kNtdpcsWO3B1sz9nudD6hNc7WbLTG2ZqNjpKgIITo1XLKcggbHIaPhw8AI/1H4uPh4zQ7yD7zyKy1DCRnC9fM7HsKFbUVXGy4SIhvCFlll4aPzHWaXJEYn8jnOZ93qKfgbM1Ga5yVBe8omWgWQvQ488Y1rbGfCL5u1HXsPLXT4S/u5NRk6pvqeeTzR1r9HOsNcJwtXDOzL6GdU9acjqqUsukppGSm8Psbft/qZ1pbNmUZj33xWId6ConxiSzdutSlc5NTk9m8ZLPL93ZGegpCiB7X3iSqfWmL60Zdx1enbecVmnQT9U1tT/haf059Yz35FfkO6ahmwT7BNOkmSi6UAJeGjpp0k006akfG/AMHBdKoGy1DYa4wr9lwhbOy4B0lPQUhRJ8ze+RsfvfV72yO5ZTldOgeOedzaGhqwOsZr1bP8XDzILMkk6CRQZZJ5jOVZ2zSUc1j/qG+oe0Gt0G/HwSA3x8upbLab99pz7xm4/UDr7f7M9mv2bgcEhSEEH3OhMAJ1DTUcOr8KUYNHQU0zyd0RNa5LBp1Y5vn1DfVk1mayayRs8guyyYqKMohjdU85t/ag91+PYG99gIJNK+baC/ohPqGdsm+zRIUhBB9jlKqeV7h9M5LQeFMx4LCidITLp1nnmzOPp/N/PHzHcpimMf8k6a3vylQZ7TVm+hKMqcghOiTrhtpu16hoz2FE+dcDAotk83miWb7yem2Ctn1RRIUhBB9knm9gpmrQUE9pVBPKV5Ofdml84+XHqdJN5F7PpfIoZEOaaytFbKrb6zHVGHq2A/VC8jwkRCiz7FOYW1vzL6zcs7ncLr8NAGDAvDx8HFaGsNZIbvymvJu2x2tO0lQEEL0OFcmUa25MjnbVQzKQHRyNBfqLzDsuWFU1lVSdrHMoeqqfSE7y/FuDlpdTYaPhBBdTmvN3vy93PPBPQQ8H4DhaUObu4QVrixEr9aMDxjPwV8cRK/WNDzRgI+HD8/MfcZmktXV4nN1j9ehV2v06s7tL2D0MzI3ci53T7mbz372GT4ePtyXcl+XbWrT20hQEEJ0qfrGepanLGfp1qXEBMeQnpRO7eO17e4SdqbyDKUXS4kJiQHA4GbAw82DzUc3W4JIR4vPdXbDmRDfEG4Zdwtfn/6accPGUVlbydSwqR3a1Ma+l9PR9680GT4SQnQZ80O7oLKA9MR0mxRN8y5hSdOTWLR5ESu2rWBDwgbLEMyXp75k9sjZuKnmv1XLLpbR2NS8jiAlM4WFExd2uPhcRzacMfco5v9tPqkFqZz6P6fw9fQlOTWZ8tpyS+bR+IDxHdrU5kqlknYV6SkIIbqM+aG99Y6trebsmx+o5oe22Ze5XzJn1BzL64ySDGoaazhUdIjbN92Oekoxc8NMcs/nutSWxPhE1qat7dBeB9Bc+C5iSISl/VGBzSUpxgwbY5N55Ovpa/mM/kSCghCiy5grhra3iMvZA/XLU18yJ9IqKJzNoKGp4bLbsmzKMj489mGHh5vSi9MpqCywDAtNCJwANO/DfOLcCZs1Cl2xqU1vI0FBCNFlzLuEWTPvdWzP+oFaXF3MmaozTA2dank/oySjU20xbzizI3dHuwXrQnxDLD2XqvoqfDx8OFZyDICIIRH8cvovCfcLbx4+skpH7YpNbXobCQpCiC7jrGLoD9/+Ie+mv+twrvUD9atTXzFrxCwMbgbL+50JCqG+oRRXF+NucCcpPomilUWWTKQNCRswKINNdlLRyiJLz8Xfy59rIq6xzF0opXjp1pcAOHnupE1PoSs2teltZKJZCNFlzBVDw/3CgeYhoINFB0krSOOeaffYnGv9QLWfTzBf64rWUk7XfLMGwKHnctXwqxgfON6heNyyKcuIXRdLQlQCNfU17MjdQdL0S5vnmCpM+Hv7M9hzsOVYV2xq09tIT0EI0WXsdwl7N/1d4sLi+PbMtw7nWj9Q7ecTLtZf5EzVmctuR1VdFcmpyTQ0NTj0XKaFTWPv/XsdrjH3XBLjE/km/xu252ynSTdZ3jdnHtl/RmJ84mW3szeSoCCE6DKJ8YkkpyVTXVeN1pp3j7zL8zc+z6GiQzaTxtYP1JILJeSez+Wq4VdZ3s8szWTssLGX1YaquioWb17MvNHzWt3feIjXEIdj5p7LDOMMbhpzEzUNNaSaLmVHWWceWX9GZze16W0kKAghuox1xdDdebtpaGrgxjE3Eu4Xbpm4tX+g7jy1k1kjZ+Hudmk0O+NsBpOCXdvH2Lr43Jpv1hCTHIPRz8j6Besva39jc4G7iCER3Pr2rZYCd5klmYT4hjh8Rmc3teltuiQoKKVuUUplKqWylFKPOnnfSym1qeX9vUqpyJbjgUqpHUqpKqXUS3bXXK2UOtxyzZ9Vf/uXF6Ifsq4YeuvbtzLKfxQFlQXEDY/j85OfO32gfnnKyXxCSQaTgia1u9rX292b2HWxDHp2ELHrYskoyWDLki28tvA1PAweNj2XttgPBXkYPFg9ZzUxITEcPXuU2HWx/GnPn3gp9SWHz+hvVGeXgSulDMBx4CYgH0gF7tJaH7U6JwmI1Vr/Qim1FPiR1vpOpZQvEAfEADFa64esrtkH/ArYC3wM/Flr/YndZ+vOtl8I0fUamxox/snINcZr+Drva87XnMfDzYO7ptxFUnwS042Xhlzi1sfx8vyXMbgZSE5L5qPMjyirKcPXw5cfT/4xSdOTmB4+/bL+ItdaszxlOaZKE+/f8b7T9RPmnovRz2izwvpM5Rmik6M5u+osBjcDMckxbFy0kalhUx3u0dcopdBaO/0H7YqewgwgS2udrbWuA94FFtqdsxB4s+X794AbVPMTvVpr/TVQY9fg4cAQrfWelqf+X4Hbu6CtQogrYK9pL4E+gXyw9ANKHynli599wVXDr+L1ha/bBISyi2WcPHeSV757xaZW0uSgybx/5/tt1kpyRXt7HbQ1FDTcbzihg0M5WHSQJt3EybKTjA24vHmOvqQrUlKNQJ7V63zgmtbO0Vo3KKXKgUCgpI175tvd09gFbRVCXAHvpr/LXTF3WR6ycWFxlslm67mDnad24ufpR2FVoaVWUkNTA9nns5k9cjY3j7251VpJrmprr4OEqAS2LNliE6iszY2cy46cHQT7BDPMe5hNOmp/1efXKcTHx1u+f/DBB3nwwQd7sDVCiIamBjYf2czOn++0HPP39rdMNpuroAJsOrKJCw0XbGolZZdlM3zwcAZ5DALoUPG51rS210F75kbO5c2DbxI3PM5hY53+qiuCggkYYfU6ouWYs3PylVLugD9Q2s49I9q5JwBpaWkdba8Qoht9mfslEUMiHB6iV4dfzazXZlFRW+FwzeA/NP8FHuobyvoF6x0yj6xrJV1OULhc10dez/0f3c/8cfMZN2xc+xf0A10RFFKB8Uqp0TQ/uJcCd9udkwLcA+wGfgxsb2uGWGt9RilVoZSaSfNE88+AF7ugrUKIbmYeOrIXPzzeabkLa0XVRRw9e5RJQY7pqOYVx1eK9ZafD33SnAPz2oHXLO+H+ob2ubLYruh0UGiZI3gI+BQwAK9prY8opZ4G0rTWKcAG4C2lVBZwjubAAYBSKhcYAngqpW4Hbm7JXEoC3gAGAZ+0fAkhehGtNftM+yxZQ+W15WitWTxpMftM+2yyhq4Ov9qle3568lMevvZhh+NXuvhce1t+XsktQa+kLplT0Fp/THPaqPWxJ62+rwGWtHJtZCvH02hOVRVC9EL1jfWs2LaiuUZQfBJ/uOEPpJpSeXbns8wwzuDO9+5kbuRc1i9Yj4fBg7iwOJfue6DwADePvdnheH8sPtcbyYpmIUSH2e+wtmrWKsL9wtlydAv3TruXVbNWOWxZ6eoD/ZZxt+Dt7u1wvD8Wn+uNJCgIITrM2Q5rF+ovsO34Nn48+cdA6zustWfxpMUOx/pr8bneSIKCEKLDnO2w9o/j/2CGcYZNVdLL2bJy/vj5Nq/7c/G53kiCghCiw5ztsPZO+jtOs446umVleU25SyuORffo84vXhBBXnv0Oa8XVxWzP2c7rC193ONc6ayjEN8RpKWszDzePDq04Fl1PgoIQosPsd1h788Cb/GjSj5xOJltnDRWtLCLqpShqG2rZdvc2y+rmytpKjH8ycvo/TjPUe+iV+0HaEOob2mbaaXsVXPsqCQpCiA4z71OwatYqtNb8Zf9feGPhG07Ptc8aigmO4ZOsT2wWqP3jxD/4/sjv95qAAPTLhWmukDkFIUSHWe9T8NWpr/Bw82BmxEyH85xlDQ0bNIxAn0AMbgbLsa0ZW51mHYkrT3oKQogOs95hbZj3MB646gGHSeDWsobqm+qxrnJzof4C/zz5T9b+0PUMJdF9pKcghOgw8z4FQT5BbDm6hbKaMpf3KcivyOfshbOWPZs/zfqU+PB4gnyCeurHEVakpyCEuCweBg9mhM+g5EIJuedzXcoaatJNfFvwLRFDIsg4m8GU0CkydNTLSFAQQlwW8wTzi/Nf5PrI61265tjZYwz2HEyTbuLaDddyseEiWmvqG+uJD4+/7G03RdeR4SMhxGXZa9pLbUMtc0bNcel8cwG98zXniQ6OZsnkJXx454dcHX418eHxndp2U3Qd1Zc3vm/e5rnvtl+IvsJZiWx3N3eadJNlbsAZ854DWmuWpyzns+zP+Pdr/p3p4dN59ItHiQ6OJjo4mv+49j+orqtm0eZFGP2Ml7XtpnCdUgqttdN/YOkpCCHaVN9Yz/KU5SzdupSY4BjSk9I5u+os3u7ebQYEuLTngLmA3lDvocwZNYe44c17Nn+Y+SGLJi0CLr+AnuhaEhSEEK1qrUT25iObuWnMTS7fZ23aWu6bdh/ZZdlMDZvKEK8hRAyJIHJoJKOGjrKcdzkF9ETXkqAghGiVsxLZAK9+9yoPXPWAy/dJyUxhYtBEpoVNw9PgCcDMiJncMfkOh3M7WkBPdC3JPhJCtMpZiezvznxH6YVSbhrrek+hvLac46XHmWm8tOp57Q/XWgKEtSu97aawJT0FIUSrnJXIfvXbV1ketxw35frjw9/Ln52nd9qUwvDx8MHdzfHvUtl2s2dJUBBCtMq+RLbWmvePvc9Pp/60Q/e5bcJtfJP3jdP6SPZk282eJUFBCNEqc4lss4LKArTWjPIf1cZVjm6feDsX6y8yzHtYm+fJtps9T4KCEKJV5hLZZvsL9xM3PM6yhqC9PQXM79c01BAxJILFWxZTXVft9FzZdrN3kKAghGiVdYlsgP1n9hMXFmd5v3BlIXq15ofjf0iQTxB779+LXq25e8rdrJ6z2rInwV7TXlZcvQKjn5GYtTGs+WaNywX0xJUlQUGIfkprzd78vdzzwT0EPB+A4WkDAc8HcO8H97LPtA9XqgFYl8iurqtu7ilYBQVo/gv/X7n/4vsjvm/5C//Zec/y4r4XOVN5BoA9+XuYNXIWGxI2sOnHmzh69iix62IZ9OwgYtfFklGSwZYlW3ht4Wt4GDy6/h9DuEzKXAjRD5nrDO3I3UFSfBLLYpdZ9kfeeGgjyWnJzI2cy/oF69t9CFvfq7ymnA+XfsjMiJnN9zq8kZf3vYwzHcnZAAAgAElEQVSp0kTZb8rw8/IDIOyPYe1uZTlQdzbrDdoqcyFBQYh+xlxnqKCywGHRmVlH6wxprfki5wvmb5zPEK8hNiWyfzjuhzz6xaNk/SrLcr56qv3hH71a/tvtKW0FBVm8JkQ/Y16FnJ6Y7jQgwKU6QzFrY0gtSGWGcUab91RKYVAGrjFew9f3fW3zXkpmClFBUV3WftGzJCgI0c84W4XsjHWdIXNQaG/YZ5D7IIdjx0qOMTFwYucaLXoNmWgWop9xtgoZcDqxbF9nqK2AAHCx4aLDsWMlx5gYJEGhv5CgIEQ/Y78KGeBE6QkmJ092OLcr6gxllmZKUOhHJCgI0c/Yr0IG+CTrE46VHKP0QqnN8c7WGdJak3E2Q+YU+hEJCkL0M/arkAE+y/4MgzJwqOiQzfHO1hkquVACQLBP8GXfQ/QuEhSE6GfsVyHXN9bz1amvWDx5sU1Q6Io6Q+b5BPuUVlfLX4jeR7KPhOhnrFchv3/H+3x35jvGBYxjXuQ89pr2Al1XZyizNNPp0JEsTOu7pKcgRD+jlGL9gvWWOkNPffkUM40zmRw8mbSCtC6tMyTpqP2P9BSE6Ic8DB5sSNhAakEqP3z7h+wz7WP9t+tp1I0cOXuELUu2MN1o20No0k14uHlQ31Tf6n3th32OlRxj9lWzu+VnED1DgoIQ/ZRSigmBE6hpqOHsqrN4u3sz7s/jeHTWo0wKnuRw/trUtcSHx7Pz5zsxuBkAuO2d25gXOY/ktGRO/NsJh2taGz4SfZcEBSH6sR05O/jeiO/h7e4NQGxoLNduuJby2tbXJhj/ZLTMCcwZNYdPsj5xqIwKUNtQS155HmOHje2exoseIXMKQvRjn2V/xk1jbrK8nho6tc2AALarmq+PvJ7vznznNChkncsicmiklLruZyQoCNGP2QeF2NDYDl0/LWwa5bXljB462uE9GTrqnyQoCNFP5Z7PpaK2gimhUyzHpoZN7dA9DKp5bqGqvsrhPck86p8kKAjRT3128jNuHHMjburSf+aRQyM7dI/T5afxdvd2WAkNUgivv5KgIEQ/ZT90BNgECFfsL9zPlJAp/Cv3Xw7vyfBR/yRBQYh+qLGpkS9yvnAICh21/8x+rht5Hbnnc22K6WmtOVZyjKhACQr9TZcEBaXULUqpTKVUllLqUSfveymlNrW8v1cpFWn13n+2HM9USv3A6niuUuqwUuqAUiqtK9opxECxv3A/ob6hGIcYO32feGM83xvxPXae3mk5XlhViJfBi0CfwM42VfQynQ4KSikD8DIwH5gM3KWUsi/cvhwo01qPA14Anm+5djKwFIgGbgGSW+5nNldrPU1rHd/ZdgoxkHx20nHoyCxgUECb1wYMCuCeD+4h4PkAPjr+Efen3M/Z6rO8m/6uZaMemU/ov7qipzADyNJaZ2ut64B3gYV25ywE3mz5/j3gBtVccGUh8K7WulZrnQNktdxPCOEirTV78/daHuSGpw08+a8nST+bzj7TPocd13L/PRd3N3cevvZh9Gpt+ap7vI6fT/s5Q7yGEBMcw1c//wo/Tz/Sk9K5NuJa/n7s7yxPWU59Y33zfIIMHfVLXREUjECe1ev8lmNOz9FaNwDlQGA712rgn0qpb5VSD3ZBO4Xod+ob61mespylW5cSExxDelI6Zb8pw9PgydzIudz53p2WB7mZu5s7TbqJW8beYjmmtWbFthUUVBaQnpjOqlmrKKwqJG54HBFDInjhlhfwNHiSez6XFdtWkFGSIT2Ffqo3TzR/X2t9Fc3DUr9USl3n7KT4+HjL1yuvvHJlWyhED3L2IA/3C2dX3i6uHn41j1/3OOmJ6ZgqTazYtsLSY3g3/V2CfYIpqymz3GufaR87cnew9Y6t+Hr6As2TzOaVzB4GD2ZGzOQX8b9gR+4O9pn2SVDop7oiKJiAEVavI1qOOT1HKeUO+AOlbV2rtTb/bzHwd1oZVkpLS7N8PfigdCjEwOHsQQ628wm+nr68f8f77MjdQWpBKlprXkp9iXmj53Gw8KDlmrVpa0mKT7K5z/7C/TblLa4fdT178veQGJ9IenG6pKP2U10RFFKB8Uqp0UopT5onjlPszkkB7mn5/sfAdt38Z0sKsLQlO2k0MB7Yp5TyVUr5ASilfIGbgfQuaKsQ/YazBzm0rE8Ye2mS2dfTl8T4RNamrWWvaS/lNeUsmrSIQ8WXFqSlZKawLHaZzX32F+4nbviloDAncg5fnvqSRRMXUVVX1eGFcKJv6HSVVK11g1LqIeBTwAC8prU+opR6GkjTWqcAG4C3lFJZwDmaAwct520GjgINwC+11o1KqVDg7y2bf7gDb2ut/7ezbRWiP0nJTOH3N/ze5lhhVSF5FXnEh8cT9scwm+J2AG8ceAOAJVuW2CxkK68tJ8Q3xPL624JvqaitYFLQpRLb08Onc7z0OKfLTwPNcxOi/+mS36rW+mPgY7tjT1p9XwMsaeXaZ4Fn7Y5lAx0r0iLEAGP/IAc4XHSYuLA43N3cHQKCvSbdRNnFMoYNGoa/lz/F1cWE+4UD8OS/nuSx7z9mqYBqHWBueOsGANRTl3ZsC/UNlS04+4nePNEshGiD+UFuzVRp6tCCtcPFhwFIiEpg46GNAOzO2016cTr3X3W/5bz2Akx774u+Q4KCEH2U9YPczFRhwujnelAwTzYnxieSnJZMdV01T+x4gieuewIvd68uba/oGyQoCNFHWT/IzQoqCzoUFMzVT2cYZzA3ci5z35xLdlk290y9p50rRX8lQUGIPsr8IF+0eZElMHR0+OhgUXNPQSnFuh+u43T5aSrrKnlhzwuYKkzUN9ZjqrDPMBf9maQPCNFHKaVYv2A9K7atIGZtDInxieSczyHEN8TlB3lqQSqGpw34e/kTHx6Pl8GLdxa/wyvfvULsuljKa8rx9/bv5p9E9CbKvi5KX6KU0n25/UJ0Ba01qQWpJKcm89eDf8VNueHv7c+5i+favTZiSAT/u+x/Geo9lGs3XEtNQw0LJixg/YL1NnsvW2catdqO1fLfYl+hlEJr7fSXKsNHQvRxSilmGGfwl4S/YHAzcPG3Fyl9pJRQ39A2rwv1DeXq4Vdz9OxR9hfuZ6j3UE7+6qRDWQwxsEhQEKKfKKoqIsgnyPIXfuHKQksF1MnBk/nbj/5G5NBIqv6zCr1aU7iykNjQWA4UHuCJHU/w9Nyn8fPysymLYeZKgBH9g8wpCNFPmCqdp6Nqrck9n8vHJz52KIsxNXQqK7atIHJoJAujmiveW5fFmGFsLjkmC9MGDukpCNFPmCqcZx4VVRfh6+HLJ1mfONQ3ig2NpfRiKU/PfZqWsjIALJuyjJRM+xJmYiCQoCBEP2GqNBE+ONzheE5ZDqOHjXZaFmNswFjeXfwu88fNtzke4htCeU15t7ZX9E4SFIToJ1rrKWSXZTNm2BinZTHclBt3xtxp00sAKK4ullTUAUqCghD9REGV89XMOedzGD10tNOyGK3ZeHgjCVEJXd1E0QdIUBCin2ivp+CsLIYzVXVVJKcmkxif2F1NFb2YBAUh+onWso/MPQVnZTHsVdVVsXjzYuaNnsf08Ond3WTRC0lQEKKfMFWYLPshWDP3FMxlMYx+RmLWxrDmmzU29Y3WfLOGmOQYjH5G1i9Y7zDPIAYGWacgRD9QWVtJQ1MDQ72H2hyva6yjsKqQEf7NW6F7GDzYkLDBUhbDur5RQlQCW5ZsYbpReggDmQQFIfqBgsoCjEOMDn/dny4/jdHPaLN1prkshnlhmhDWZPhIiH6g1fmEljUKQrhKgoIQ/UCbmUdDx/RAi0RfJUFBiH6gzcwj6SmIDpCgIEQ/0F7mkRCukqAgRD/Q3mpmIVwlQUGIfqC91cxCuEqCghD9gLM5hfKacmobagnyCeqhVom+SIKCEH1cY1MjRVVFDPcbbnM853yOZSWzEK6SoCBEH1dcXcywQcPwNHjaHJc1CuJySFAQoo8zVbaeeSSTzKKjJCgI0ccVVLaeeSSTzKKjJCgI0ceZKpwvXJOegrgcUhBPiD4i7I9hFFUXtfr+h5kfUriy0PJaegrickhPQYg+oq2AYP9+k24i93wukUMju7lVor+RnoIQvUx7PQJXFFYV4u/lj6+nbxe1SgwU0lMQopfpbECAlvkESUcVl0GCghD9UE6ZzCeIyyNBQYh+SDKPxOWSoCBEPySZR+JySVAQoh+SnoK4XBIUhOgnQn1DLd9LT0FcLgkKQvRBerVGr9b8dvZv8TJ4Uf5ouWXhWm1DLcXVxUQMiejhVoq+SIKCEL2M9V/87b1fVVfFxKCJvHXwLcuxU+WnGDFkBAY3Q7e1UfRfsnhNiF7G/Bf/nvw9PPTxQ6Q9mNbquaZKEwlRCbyc+jJJ05NQSskaBdEpEhSE6GbtrVAO9Q21qVlklnUui/GB49u8t6nCxK9m/IqtGVv58tSXXB95ffMahaEynyAujwwfCdGN6hvrXapZVN9Y73A861wW44aNa/PagsoCjEOMJMUn8XLqy4CsZhadI0FBiG6itWbFthUunbti2wq01jbHTpw70WZPoUk3cabqDOF+4fx06k/5IvsLCioLJPNIdEqXBAWl1C1KqUylVJZS6lEn73sppTa1vL9XKRVp9d5/thzPVEr9wNV7CtHb7TPtY0fuDpfO3ZG7g9SCVJtjWeeyGBfQek+h5EIJfp5+RP53JP7P+VNWU4bxT0a2ZmzlzvfuRD2lUE8pwv4Y1qmfQwwsnQ4KSikD8DIwH5gM3KWUmmx32nKgTGs9DngBeL7l2snAUiAauAVIVkoZXLynEL3a2rS1JMUnuXRuYnwia9PW2hw7UXqC8QGt9xRMFSaMQ4wdKqktRHu6oqcwA8jSWmdrreuAd4GFducsBN5s+f494AallGo5/q7WulZrnQNktdzPlXsK0aulZKawLHaZS+cum7KMlMwUy+tzF8/RqBsJ8glq9RpTpfMd14TojK4ICkYgz+p1fssxp+dorRuAciCwjWtduacQvVp5bTkhviEunRviG0J5TbnltbmX0Py3k3Ot7c0sRGf0+ZTU+Ph4y/cPPvggDz74YA+2RohL/L38Ka4udunc4upi/L39La/bm0+A5uGjcL/wTrVRCHtdERRMwAir1xEtx5ydk6+Ucgf8gdJ2rm3vngCkpbW+sEeInpQQlcDGQxsJ9Q1td53CxsMbSYhKsBw7ce5E+0Gh0sQM44wua68QAMo+Da7DN2h+yB8HbqD5wZ0K3K21PmJ1zi+BKVrrXyillgKLtNZ3KKWigbdpnkMIB74AxgOqvXu23Fd3tv1CdJe9+XtZunUp6Ynp+Hr68uxXz/LcN8/xs9if4Wnw5J30d3jrR2/x+oHX2XRkE9Dcu0iISqCoqog7Y+7k3mn3Wu6ntWafaR/Jacl8lPkRZTVlDPYcTFVdVbtt0avlvxNxiVIKrbXTsclOzym0zBE8BHwKZACbtdZHlFJPK6XMf/psAAKVUlnAw8CjLdceATYDR4H/BX6ptW5s7Z6dbasQV9IM4wzmRs5l0eZFVNdVk3YmjeGDhzMtbBolF0oY6j2URZsXkVaQxpLJS6j5bQ3pSelMDprMFzlf8NvtvyXg+QAMTxsY9twwJrw4gR9t+hHRwdGkJ6UTExzD1ju29vSPKfqZLlmnoLX+WGs9QWs9Vmv9bMuxJ7XWKS3f12itl2itx2mtZ2its62ufbbluiit9Sdt3VOIvkQpxfoF6zH6GYlZG8Ou07uIGBJBTHAMHx3/iIamBqrrqokKjOKtH72Fh8GDYJ9gjpUeo1E34mXw4uaxN3PxsYv8YNwPcHNzw8PgwbGSYwT7BFNQVUBcWFyHCugJ0Z5ODx/1JBk+En2B1prUglS+/9r3MbgZqG2oRaNZGr2UScGT+Puxv5P6QCoGZWB5ynJyz+eSVpBGwcMFTFk3hSeue4LfffU70hPTAVi0eRFhvmG8k/4ONY/X4Kaa/7arrqsmZm0Mm368SeYaRJu6dfhICNE2pRRTQ6fSqBvZsmQLP5v6M0J9Q1k1axVPXPcEAYMCeGnfS5YV0KvnrGZ84HgGew0mMT6R3+/8PUnxSfh6+uLr6cv7d7zP9tztBA4KtAQEAF9PX6eL4IToCAkKQlwBWeeycFNuTA6eTEpmCrNGzuJA4QGUUiTfmswzXz3Df+36L5LikzBVmiyZR8umLCO7LNtmEdyF+gvMCJ9Bo250+Bz7RXBCdFSfX6cgRF9wqOgQTU1NjPQfSXltOTONMzlQeACAqKAofjPrNzzy+SN8nv05gzwG4efpx68++RVGPyMazRPbnyCzNJNjJcdo0k2MGTaGC/UXHD7HfhGcEB0lPQUhroDd+bsZ6j0Udzd3/L38GTV0FPsL91veXzVrFW7KjWMPHSMuLI4fjP0BY4eN5WTZSQDGBYzjmXnPcCTpCKWPlLJ7+W5qGmocPsd+EZwQHSVBQYgr4FDRIUb4N6/HTIhK4GjxUQ4WHqRJN1nO8ffyp0k3ce7iOZbGLOXfZ/474wLGMXbYWNzd3Lk+8npCB4eilGr14W+/CE6IjpKgIMQVcPLcSSYFTQKaK6K+eehN/L39ySnLsZxjXgFtLnFRVVdFcmoyj81+jOS0ZKrrqi3nOnv4m89PjE+8Mj+U6JckKAjRzbTWFFUXER/eXKfLvKittqGW3fm7Leclxify4r4XqWusw8fDh8WbFzNv9DzunXqvzSI4Zw//qroqy/nTw6df8Z9R9B8SFIToZmeqzgAQNzwOsF3U9tDHD7HmmzWYKkzEhcURFRRFXWMd0cnRGP2MrF+wHjc3N8v50cnRXLX+Kq4xXsO00GmYKkys+WYNMckxlvPbqqwqRHsk+0iIbnas5BiAzYY5HgYPnpjzBH/a/SeOnj1K7LpYymvK8Xb3xsPgQUNTAxODJlJcXUyIbwjF1cVEBUbxyYlP8HDz4POcz/H5vQ/+3s21krYs2cJ0o/QQROdJUBCimx0oPECTbsI4xHbvg6uGX0XO+Ry+vu9ry7Gnv3yamoYabp94O8mpyZZgYX74p9yVIg9/0a0kKAjRzVJNqYT4htisPgYY5T+K6rpqS28Amhe5zY2cywzjDClVIXqEzCkI0c2OnD3C6GGjHY4rpZgWNo2DhQctx06cO8H4wNb3ZRaiu0lQEKKbnS4/TWxIrNP34sLibBaxubLjmhDdSYKCEN3InEJ6dfjVTt+fFjbNUu7ifM15ahpqpNS16FESFIToRsdLj+Pl7kVUYJTT9+OGX+opmHsJklIqepIEBSG60bGSYzQ0NTAhcILT9ycGTeTU+VNU11WTdS7LJm1ViJ4gQUGIbrS/cD9oLNlF9jwNnkwMmkh6cTonSk/IfILocRIUhOhG3535DuMQY5tDQubJ5qwy6SmInidBQYhudLz0eLsppubJZukpiN5AFq8J0U2adBOFVYXcPeVup++H/TGMouoim2PXvXGd5ftQ31AKVxZ2axuFsCc9BSG6yeny07i7uRMTEuP0ffuA0NH3hegOEhSE6CbHSo7h7uYu8wSiT5HhIyE6yNmwjzXzsE/G2QxqGmqkbIXoU6SnIEQHuTrsc6DwAJ4GTwIGBVyJZgnRJSQoCNFNDhUdYqT/yJ5uhhAdIkFBiG6SfT6b6JDonm6GEB2itNY93YbLppTSfbn9om9ST11ebSL7FFNX7qNXy/+/RddTSqG1dvp/QOkpCHGF2M9FtFcNVaqlip4g2UdC9BBZmCZ6I+kpCHEF1TfW93QThGiTBAUhOqgzwzortq1A5sFEbyZBQYgOMj1sImJIBAd/cRC9WrNn+R4ih0a6dO2O3B2kFqR2bwOF6AQJCkJ00Pac7YT4hhAb2rzv8tq0tSTFJ7l0bWJ8ImvT1nZn84ToFAkKQnTQGwff4N6p91pep2SmsCx2mUvXLpuyjJTMlG5qmRCdJ0FBiA4orynnH8f/wV1T7rp0rLacEN8Ql1JMQ3xDKK8p7+5mCnHZJCVViA7YfGQzN4y5gSCfIMsxfy9/iquLLSmmm9I38d97/5td9+1y2HHNVGHC39v/irZZiI6QnoIQHWA/dASQEJXAxkMbgeaNdZ7+6mlWz1ntdAvOjYc3khCVcCWaKsRlkaAghIsySzI5ee4kt4y7xeZ4YnwiyWnJVNdV897R9/Dz9OMHY3/gcH1VXRXJqckkxideqSYL0WFS+0iIVri6b4LWmuUpy8mvyCevIo8XfvCCQ+Coqqti8ebFGP2MbEjY4LQXIcSVIrWPhLgMru6boJRi/YL11DfVk12WzcHCg5gqTNQ31mOqMLHmmzXEJMdg9DOyfsF6CQiiV5OeghCtcKWK6TDvYZTXluPv5U9TUxM/j/s5ZTVlfHT8I8pryvH39ichKoGk+CSmG6dfgVYL0b62egqSfSREJ6QnpRPiG8KG7zbw7M5n+SDzA+ZGzqXw14V4GDx6unlCdJgMHwnRCeF+4bgpN17c9yLrF6wnPTEdU6VJahyJPqtTQUEpFaCU+kwpdaLlf4e1ct49LeecUErdY3X8aqXUYaVUllLqz6plsFUp9X+VUial1IGWr1s7004hutPWo1vx8fDhlnG34Ovpy/t3vC81jkSf1dnho0eBL7TWzymlHm15/RvrE5RSAcBqIB7QwLdKqRStdRmwFngA2At8DNwCfNJy6Qta6z92sn0ucTXLRAh7TbqJp758ijU3rbFMIPt6+lpqHM0wzujhFgrRMZ0dPloIvNny/ZvA7U7O+QHwmdb6XEsg+Ay4RSk1HBiitd7TMlv811au73auZpmIgcPVfQ82H9mMj4cP88fNtzkuNY5EX9XZoBCqtT7T8n0h4Kz4ixHIs3qd33LM2PK9/XGzh5RSh5RSr7U2LCVEd9Bas2LbCjwNnm2e5+7mzs8//Dm/m/s7hzRTqXEk+qp2g4JS6nOlVLqTr4XW57X8td9VM2trgbHANOAM8P9aOzE+Pt7y9corr3TRx4uBbJ9pHztyd3DukXNUPFqBj4cPtY/XMsx7GKaHTRT+upDFkxYzZtgYhnkPY9ggx79ZiquLpcaR6JPanVPQWt/Y2ntKqSKl1HCt9ZmW4aBiJ6eZgOutXkcA/2o5HmF33NTymZbxGqXUq8C21tqQlpbW3o8gRIeY90fw9fRlR84OpoVNw9PgSUJUAr/+9Ndsz93O8rjl/G3R3/jz3j87nTuQGkeir+rs8FEKYM4mugf40Mk5nwI3K6WGtQwD3Qx82jLsVKGUmtmSdfQz8/UtAcbsR0B6J9sphMus90fYnb+bmcaZmCpMZJ3LYmvGVrYu2crvb/g93u7eTucOpMaR6Ms6GxSeA25SSp0Abmx5jVIqXin1FwCt9Tngd0Bqy9fTLccAkoC/AFnASS5lHq1pSVU9BMwF/qOT7RTCZeb9EQD25O9hhnEGs1+fzbzIedw95W5+t/N3VNdVA45zB+YaR/NGz2N6uKxgFn1Pp1JStdalwA1OjqcB91u9fg14rZXzYpwc/2ln2iVEZ5j3Rxg+eDi783dz67hbiRgSwdPznqa+sZ4V21YQszaGxPhE5o2exxCvIZgqTGw8vJHk1GTmjZ4nNY5EnyUrmsGlHbPEwGHeHyG7LBsvgxdvHX6L/zPz/wDgYfBgQ8IGNv14E0fPHmXOG3M4X3Oe2HWxZJRksGXJFl5b+JqUuBB9lhTEs/KX7/7CrrxdfJL1CakPpBIxJKL9i0Sf4eoixb35e1m6dSlPzH6CjenNwSHr37IwuBlszq+qqyImOYbNSzbLIjXRp0jpbBfllecxYsgIJgROILMks6ebI7qYq4sUZxhnMDdyLv/3y/9LyYUS/m3GvzkNCDJ3IPojCQpW8iryGOE/ggkBEzheerynmyN6iHl/hKq6Kg4XHaaqrkr2RxADhpTOtpJXkUfEkAjOB50ns1R6CgNZfVM9VXVV/Gjij8guyyZ2XazN/ghblmyR/RFEvyRBwYp5+KiusY7tOdt7ujmiB31z+hs0mudufI7xgeN7ujlCXDESFGiudbM3fy8ny07y/de+T3ltOUop7v3gXpKmJzE9fLoMEQwQAc8HUF5bjkEZ8Hb3pqymDK21/P7FgDHg5xTqG+tZnrKcO967A3flzpFfHqHqsSo83DyICozizvfuZHnKcperZoq+LT0pnZrf1uBp8GRG+Az5/YsBZ0AHBXM1zILKArYs2cLYgLGE+4Xj4+HDSP+R3D7xdtlJa4AJ9wtne852ahtrefW2V+X3LwacAR0UzNUwt96xlZILJYzwH2F5b0LgBDJLM2UnrX7E1UWKz339HD4ePoweNlp+/2LAGdBzCtbVMPMqmieZzaICoyxpqfY7aclObX3TugXr+OOuP/L1fV8DcO8H91LTUENxdTHb72lOLDhWcozvCr/julHXyU5qYkAa0D0F62qY5syjXXm7OHfxnMMCNutqmLJTW9/06nev8uDVD1pep2Sm8PyNz5Ndls3e/L0A/Hnvn5kUNIlZI2bZXCs7qYmBYkD3FKyrYeZV5DE3ci4/ef8nhPuF8+ScJ/nb4b9Zzu0NO2ldTg9Fa80+0z6S05L5KPMjymvL8fdqzrXvy5lVHf23OF1+mj35e3hvyXuWY+W15RiHGFn5vZU8981zvLv4Xd4+/DZjh43l2ohrbe7XG37/QlwJA7qnYK6GCc1Bwdvdm8q6Sob7DWdt2loySzLZm7+Xez64h+D/CqZRNxLwfECPtbejPRRzZtXSrUuJCY4hPSmd2sdrSU9KJzo4uk9n1nT032LDdxtYNmUZgzwGWY6Zf//3xd3H7rzd/PXgX5kUPImMkgziw+Ntrped1MRAMaCDgrkaJkB+RT4FlQXMHjmbv97+V/LK8yi9WMqSLUuICY7hoRkP8bOpPyM9qW/s92OdWZWemM6qWasI9wvH3c2dcL9wVs1aNWAyaxqaGtiwfwMPXPWAzXHz73/M/4yhqLqIB7c9yJ78PVxsuMjgPwxGPaVQTynC/hgmO6mJAWNAB4XE+ESS01cnhiUAACAASURBVJKpqq0ivyKfjJIMZo+cjbe7N+MDxqO15iexPyFxeiJ/O/Q3fjn9l4T7hfd0s11inVnl6+nr9JyBklnzyYlPGOE/gimhU2yOm3//rvQ6ZCc1MVAM6DkFczXM2969DR93H/bk7yExPpF9pn3sMe3hpjE38eLeF/k8+/M+Vw3TOrOqLf05s0Y9pRxeW881mH//rx94vd179bXfvxCXa0AHBXM1zMWbF1NZV0n1uWqCfIJ4YscTLJuyjL35e9FoDhcf5o2Fb/SpCdmUzBR+f8PvbY61Vq5h2ZRlxK6LvVJNc+pKpflaf4b59+9KUJBqqGKgGNDDR9C8k9Z9cfcxLmAcgT6BXPXKVbx16C1eTn2ZuqY6ZkbM5M+3/JnbN93OuYvNW0v3hZ3arDOrAGobajH+yUjphVKHc3tDZk1Ppfm6ukOa7KQmBooBHxSgeZLZ0+BJUnwSpY+U4qbcOLvqLH+6+U+UXCjhgasfYLpxOm8ceAOAwpWFFDxcwDDvYVx47AIGZaDmtzUY/YykPpDaKxauWWdWAaQVpHGm6gy78nY5nHslM2vC/hhmmcC1/hJC9A4SFGheuFZyoYTrRl0HXHqgjg8cz4lzJ2jSTSyMWsiO3B2Wa8zzDOcunsPf2x8vdy9+M+s3PPPVM93Wzo70UKwzqwC+OvUV3u7efH36a4frrmRmTVf9xd8bemNC9EcSFICc8zmcvXDWsmmK+YE6xGsI/l7+mCpMXB95PTtP7aShqQGAf2b/k5vG3GTzQL3/qvvZZ9rHwcKD3dLOwpWF6NWaopVFDPUeStOTTejVmgUTFvDyrS/b9FDMmTXVddUAfHX6K+6Pu5+v82yDQlVdVZ/LrKlvrOfW8bcyyn8UAYMC+Pynn1P/RD2mh0093TQh+jwJCkBGSQZRgVF4u3sDtg9Uc2G8EN8QIoZEsP/MfrTWfHbyM2aNnGXzQB39P6M5U3WGaeunOR0iCftjWJe091DRIaaGTrVMfD4++3Ge/+Z56hrrLOeYM2sWbV5ERU0Fu/J28evv/ZoDhQe4WH8R6Jv7DFuvv9h6x1Z8PXyZN3qeZf1FRzU2NaJoe/hKeiViIBnQ2UdmeeV53B1zt+W19QN1zLAxHC89zo1jbmTe6Hlsz9mOl7sXvh6+/Pqfv7Z5oF6pydKDhQeZGjrV8vqaiGuICozirYNvsfyq5cClzJoV21YwKXkS3gZvPNw8mBQ0iU9OfEJWWRbJqcnMGz2vT2XWmNdfpCemk5yazIIJCzrV9n2mfUSHRHM48XAXtlKIvmvA9xSadBMVtRXMHz/fcsz8QDX6Gfng2Ae8ffhtTBUmZo+czSdZn/DYF49RfKG4xzZuP1h0kKlhU22OPX7d4/zh6z9YhregOWNmQ8IG7ph8B0MHDSV2XSzfnfmOn/z9J2SUZLBlyRZeW/han8qssV5/se3ENhZMWGDzfkczw/5x4h8sGL+glbOFGHgGfE/h9PnTaDRzIufYHDc/UKODo3lhzwuWjdsbdSPDBw/nyeueZNWsVT3S5oNFB/nVNb+yvLbO8ff4neMD3svgxWsLX+PuKXfzfsb7vLb/NV5f2H5ufm9ifpib11+UXSxj/5n9zI2cy5nKMzz39XP8z/z/sZlXMVWYiF0XS+kjjmm4ZtuOb+OlW1/q9vYL0VcM+J7CZ9mf4ePhwxCvIQ7vKaW4Leo2vNy9KH2klIYnG5gaOpXzNedtSjBfSXWNdRwvPU50cLTlWHvDUrWNtcweORuAWSNmsStvF026qVvb2Rl6tUav1vzj7n8wPXw6erW2POzN6y8+PfkpcyLnMMhjEDtyd/DKd6/Y9JKg/fUX+RX55FXkMTNiZrf+PEL0JQM+KPwr918MHzy81fdHDx2NqcJEbUMtAOMDxhPkE9RjFTMzzmYweuhom2qfrjDvKhc6OJQgnyCOnj3aHc1rV0eGd34w9gcUVRex/8x+yzFzuvC249sswz77TPuoaajhROkJm3vZr7/QWluq3gY8H8DIF0ZSVVfF/SnNWWP9uSigEK4a8EHhu8LviAqMavV9D4MHo4aO4mTZSaB5DsKgDFeqeQ6czSd01PdHft/peoUrwZxW++lPPmVu5FxLr8D8ZT38Y3AzcH/c/az/dr3lWEJUAm8dfIv/zfpfbh1/KwCpBakE+wRzqOiQzWdZpws7KyN+6/hbeeHm/9/emUdHVaT9/1PprBAI6UACgRB2IYQAQ4BBFHAEhFEBxQXUY0DHkxHF+ekM+irquA6CjvsLyry4vqCgYgZQnFc0CEHFBEJCEnZISJosGLIQIKTTqd8f6dve3rKQDoFOfc7pk+7quvdW9c2p59bz1POt1y57GXGFwpO0a6NgqbNwtOwooyJHNVhPvwvboVOHKKoqorq2+mI00QltOWpLaEujoJFzMoeYbjGN1rtn5D2szV5LVU0VUL9c+PWdrxPZKZKokChq62rJKMrgrri77IyCPv/ClYx4aGAo2/K2MWfYnHYlI65QNEa7NgqZxZkEGAIY0nVIg/W0/ZqLq4rJr8wnLiKOn/J/cqp3MTSRMoozvMIo7Du5r9HfHaBn555MjJ7IJ3s/AeqXC3fr0I2qmirO1JwhuySbqJAoru59NRnF9UmDjvkXrmTEt+ZuZUT3ERiD6jdNai8y4gpFY7Sr1UeOW1OWV5cD8FHGR/Q39ne7NeWgsEH8XPAzW45uYVKfSQwOG0xybjLX9L3Grp6j5lF5dTlRr0VR8reSZscA3LU/o6jl7qOBxoEcrzjeoOaQp1RJ3ZHzaw5zYuc0qW7iqESeSn6K+0bdhxCCOllHTLcYYlfEMqrHKIZ2G8qQrkPYXbibZTuWOeVfuJIR33RwE9cPvN7uOt4sI65QNJV2YxTMFjOJmxJJzk1mQfwClly7hAe/fpDk3GRGdh/J7Z/fzjV9ruHdG961rdt3lHN2lFheuWtlgwNnl8AujOg+gq25W+3yIC6Uoqoi6mRdg4HxpiCEQNKwi6S1VEmh3rhll2Q3yX2kvwd6I5Z9MhuoTzwM8A0gaX8SFmlhT9EePrv1M5tkCTjLiEsp+erQV2y6Y5PT9S4FGXGFoi1pF+4jVz7lHsE92J63nTM1Z3j2mmdd+pQ9kaE8fcB0Nh/e7JF+aEFmx9nM5SDlrUdTb9VLe7ujsd/YIi18d/d31D5dy7he40gclWhnEMBZRvxY+TFqLDV2y3o1LgUZcYWiLWkXMwW9NILmQjhYehB/X3+6+XbDz+CHn8GP9betJ3ZFLKknUj3mPpg+YDq3fX6bR87lKG+h4ThbkVLS/Z/d6ejXkS9v/7LF7iZPIqXki31f4OvjS9iyMCrOVxASEMKMK2awYPQCty68hhjRfQQAcRFxZBZnOiUiastYNW2kHcd3ML73eJfXuZgy4grFpUi7mCm48ilvy9vGsPBhRHWOspXpfcqeYkT3EVTVVHH41OEWnyuzpGkrj4QQDAsfRvGZYmLDY1t8XU+hLQt9Kvkp+ob2JWtBFuefPE/WgixiusYw69NZDHxrIMalRgzPGTAuNTbpvJqQ4fCI4U7LUsFZRnxH/g7GR413ea6LKSOuUFyKtAujsOHABu6Mu9OubPvx7USFRNmSujTuHHYnGw5s8Ni1hRBMGzCNbw5/0+JzNSfIHBoYSmSnSAw+F55T4UmlV70L75YhtzA3di6RnSLx9fGlW4du7C/dj7/BHx/hw9T+U6leXE3WgqxmXSMuIs62AkmPo4y4O6NwOcqIKxSepl0YBUefMkB+ZT6hgaH06tTLrry5PuXQpaG2p9p5SfNcZsZ6Iq5QXVvN0bKjTVrGqdVvTBL6QrjQALR+WejhssO2fuiNRfaCbNIT09lp2kl6UXqzpbCHRQwj+2Q2ljqLXble9dZUaSK3PNfmctK4HGXEFYrWoF0YBcetKQGSE5KpsdQ4zRSa61N+/brXbS4Qd5mxU/pNYXve9hYlvOWczGGAcQABvgFO3znKNxieM/D1oa/JLc9lZ8HOSyIZS+/C0yeuOeYQtMSF1zmgMxEdI2zZ5xp61dtRK0cR3jGckjMlmC1mTJUmlu1YRuzy2DZTvVUoLiXahVFw9Clr5Ffm28UUoPk+5YziDNsGL+4yY0ODQomLiOOH3B8uuA8ZRRnERTgvlXQl31D8t2IC/QLp4NeB2etmOxkps8Vs88NfLDQX3qlzpzhTc4ZenXtReraUp7c+zW1Db8PX57c1Dy1x4WnBZkc01dvpA6YT7BdM3DtxBL0YRNw7cZetjLhC0Rq0i9VH98ffz5wv5rBgtH2wOb8i326moPmU1926DqhfytmQu8QYZLTbtxl+y4x1XMWkuZCuG3DdBfXBVSaz41JbrW9fHfyKcb3G0TmgMzOumMEnWZ+QuCmRVTNWAZC4KZGJ0ROZccUMFn+/mM13brYphTaU0Kbhqk5jyW6aC29nwU6GdBuCEIIXtr3AlqNbyDmZw2s/vUZYhzCiQ6KJComyJRY2dg8cl9vGRcSRUZTBLTG3OLdbCI6VH2PplKVMGzCt0X4qFO2RdmEU9D7l9bettw2e+pmCK5+y4yA3L2keQ7sNte2jYLaYCVsWRunZUsI6hNnqucqMnT5wOnO/mMvrvH5BfcgozrAJwGm4WmoL9SurJkRPwNfHl6ySLDsjJaW0HZO4KZGR3UeScjylxfLRjcUaNBfevl/r5S2klCQdSALg2F+OIRAUVhWSV57Hj/k/8kXOF1TXVtvdA7PFTNCLQdQ+XcuS7Us4efYkr173qt11hkcM56PMj1y2wWwxk3YijXG9xrWorwqFN9Mu3Ed6n3LsiliW7VjGsbJjnDp3Ckudpck+ZcdVTH4GP8b3Hs/W3K1OdR1dICO6j6CiuoKjZUeb3X4ppUshPFdLbQG2Hd/G1b2vZnTkaFJPpNoZKe2YDn4dSM5NZuYVMy+KDpLmwtPiCZnFmQiEzVgYfAz06tyL8b3HI5FEdoq06R1p6OM9qSdSXQaE3bmPANKL0ukX2k/lISgUDdAujAL85lNee8tack7mMPLdkdRYahi5cmSTfcquVjHdNPgmPs782Kmu4yomH+HDtAHT2Hyo+auQTKdN+Pn4ERFs7ypxtdS27FwZe4v3MrbXWEZFjiK9MB1LncVmpLRjDp06hI/w4eYhN5NyPKXVg9HastC9xXuJ6RZD0v4kZg2e5RTv0Vx4i65cxJu/vGnXLn28x12CYb/Qfpw8c9LlCrIdx93nJygUinpaZBSEEEYhxLdCiEPWv6Fu6iVY6xwSQiToyl8UQuQLIaoc6gcIIdYKIQ4LIXYKIfq0pJ268zKm5xg+mPUBG+ZuYHzUeEofLeX9me87SSO4wtUqpjuH3cn249vJLc+1K9c/1XZ/pTviWcGHGR/y4OYHm732311+gisj9U7aO8yOmU0Hvw4Yg4yEdwznQOkBm5HSjkk+lsw1fa4hKiSKTgGd2P/rfqD1JDE0F15KfgrRIdEkHag3CvocAr0L78ExD3LOfI7tx7cD9jkEhacLOWs+S7/Qfk7XMfgYiA2PZW/JXqfvduTXZzIrFAr3tHSm8F/Ad1LKgcB31s92CCGMwN+BscAY4O8647HRWubIvUCZlHIA8BqwtIXtdKKgsoBenXs1XlGHq1VMHf07kjA8geWpy+3K9U+1LdVQcieX7Wikztee561f3uKv4/5qK4uPjCftRJrNSGnHJOfWGwWACdET2Ja3DfhtExz5d0nBwwUYg4zIv7d8FiGE4JUpr2C2mJm+ejqHSg/RJ6QPI7uPZGzPsYx8dyQx/x1jc+EZfAw8NPYh3tj5hlO8J/VEKvGR8W7dfK5cSFJKUo6ncFXvq1rcF4XCm2lpoHkmMMn6/kNgK/CYQ53rgG+llKcAhBDfAtOAT6SUP1vLXJ33Gev7z4G3hRBCttDH4ah6CrA2e63tfWMraNytYnpg9AP8ftXveWbSM3Tw6+C0iskT7QR4+ceX7dqpGSkt8L1672riIuLslq6OjhxNqimVoqr6+lJK/jfjf9mau5WXJr8EwITeE/g+93sS4xPtrudpyYcjZUcYFjGMa6KvYePBjYxcOZKK6go6B3QmrEMYtXW1DO46mJIzJYR3DGdyv8ks+r9FDH57MFP7T7XFe1JNruMJGtoKJD1Hy45i8DEQHRLtsf4oFN5IS2cKEVLKQuv7IsCV76EnkK/7XGAtawjbMVLKWqACCGvwiCbQ0id2/SomTTIBoL+xP2N7juWTvZ94JDO2qe3Uu17qZB2v/PgKi65cZFc3PjKenaadNtfL/fH388YvbxDoG0ifLn0AmNhnIj/k/mDnv28NyQctyJxenM6r171K6aOl1D5dy6nHTnHwwYMkzUki52SOLYdg3Kpx9A3ty6Q+k+ziPY0JFsZFxJFZYj9T0KQtVGKaQtEwjRoFIcQWIUSWi9dMfT3rU/xFT52Nj4+3vVauXNmq13K1islUacJsMTM3di5PJT91UTNj9Ubqy31fEuAbwB/6/sGuzqCwQewq3MWkPpMYHTmaMT3H0Ltzb8x1Zpth6x/anzpZx7HyY4Dr5bmekOfe9+s+okOi2V24m8n9Jtt9p4/3aMai9NFSNt2xiW8Of2Nrq5TS7cojjbiIOLJKsqiTdbYyFWRWKJpGo+4jKeVkd98JIYqFED2klIVCiB5AiYtqJn5zMQH0ot7N1BAmIAooEEL4AiFAqauKaWlpjZzKs2irmFJPpLI8dTlx78RRUV1BSGAIZouZf079J/eNuu+itEUzUombErlr/V3MGjKLE6dP2GQcVu9dzfLU5XTy78QDox+wGakenXrgZ/AjdkUs98ffz53D7uSq3lfx7/3/xlxndtq5DJxzNgpPFxKzPIbivxXjb/BvUntzTuYQHRLN5H6Tm7wTXXRINEPDh3LtR9dysPQgFecrkFLy+HePu5Xa7hLYBWOQkaNlRxlgHADUzxQu1n1RKC5nWhpT2AAkAC9Z//7bRZ3/AP/QBZenAo838bw/AbcA37c0nuBJtKdaRxfGmzvfZMuxLXaDj15eoiHMFvMFSSz4Gfz486g/8/Whr/Hz8bMzUjOumMFnt37GW7+8xZ6iPYzuOZo6Wce2vG2kJ6ZTWFVoM2zl1eVsOLCBucPmOu1cBs5bmVacr0AguH719bx47YtN2gch52QOZ81nSRie0GA9/W+SuCmRg6UHsdRZyLw/kx9yf+DjzI9tOlOOu+VpaMHmAcYBlJ0rI68ir8V7WysU7YGWxhReAqYIIQ4Bk62fEULECyH+B8AaYH4eSLW+ntMFnZcJIQqADkKIAiHEM9bzrgLChBCHgUdwsarpUiRheAKbD23m1nW3YlxqxOdZH4KXBDfpWL1WUlPQ7zlw+xe3Myd2Dh/O+tDO9aIttR0dOZq0E/UzqqySLEKDQokKibJz12T+OZOenXu6XJ7rSl/p/JPneeLqJzhdc9qlCKAj58znMJ028YvpF64fdL3behp6CY9DDx6iR6ce5JzMYXfhbiZET3CrM6Wh31vhx/wfGR05WukaKRRNoEVGQUpZKqW8Vko5UEo5WRvspZRpUso/6eq9J6UcYH29ryt/VErZS0rpY/37jLW8Wkp5q7X+GCll89OALzJmi5mH//MwAGXVZWQtyGL7PduJ6BhBsF/DhiG8YzjJucmknkht8vW0gfmrO76ipKqEpP1Jbgfm+Mh427m1/ARHYrrFUHm+koLKArtyV1uZavsgJAxPILc8l4zEDLeDs8bB0oN069CN0T1HYwxqfPMcvXpqcEAwfxn7F97Y+YZdkFnTmXL12+n3VtiRv0MtRVUomki7yWhuTfQDZ3JCMlklWYQFhfGvXf9i4ZiFnH7iNFWPVzG1/1Tmj5iP5SkL0SHRZC/IRv5dUvy34mbLRWsD85q9a3ho7ENkL8h2OzCP6D6C/b/up7q22i4/QY8Qgqt7X23LV9BwlLbW09/Yn4jgCDJLMt0Ozho5J3PwET7MumJWk/rnKOFxx7A7SDuRxi+mX4iPjLfVc5Ta1mTE1+xdw4YDGzA8Z+DlH1/mp4KfXO51oVAo7GlXRqG1NrjXD5yjIkcxLGIYn+d8bidD4fhUO6XfFL498q3tHHqtpKa2s/RsKav3rmbh2IUNPjUH+QUxKGwQ6YXpbMvbxqQ+k1yed2L0RCej4E5fSePmwfUrn9wNztoeD3esv4O8ijy25W1r0uDsKOER6BvIfb+7j94hvekS2MWurvbb6d1c43qNw9/gT+FfC/Hz8WNC7wlNcnMpFO0eKeVl+6pvftuT8GWCXJayzPY5aV+SHPuvsdLnWR9ptphltblafn3wa3nfhvtk8IvBsuuyrvKD9A/kDWtusB1TU1sjDc8a7D53eamLLDpd5HR+jed/eF7OT5pvV7Y0ZamclzTP9jni5QjJM7h9RbwcYau7+8RuOfjtwXbnC30pVJoqTXZl2SXZsrK6UkopZUZRhuzzeh9ZV1cnCyoKpHGpUdbU1sj5SfNln9f7yGUpy6Sp0iSveu8qGfVqlFyWskz2eb2PnJ80X9bU1rj9TbXfTk/ZuTL5Q+4PTnVramukzzM+cn7SfHndx9fJqvNVUkop41fGy9d+ek3GrYiTUkpZdb5KTv14qpyfNF/W1dW5vbZC4e1Yx06X42q7mim0Fo5PtTcMuoHiM8UEGAKYvXY23f/ZnX+k/IPBXQfzzV3fUHm+knXZ69iWt8321Oq441vK8RQGGAcQERzhUvjurPksb//ytp2kBTirszYnYS8uIo7C04V20hmO+ko1lhqmfDyF+zbWr7AaFj4MgzCQUZxBeMdwys+Vu4xB5JTkcOOgGxsNEGu40pnqEtiFCdETnOqWnCkhOCDYyc0VFx7HirQVtvyEhmZTCoWiHmUUPIDjwGnwMfD0hKcJCwrD3+DP/gf2s33+dh4Z9whjeo7BUmehylxFoCGQ99LfIyEpgYFvDaTsXJltr+dV6av444A/ujw/wEObH2JK/ykMDR9qV97cPab1GHwMXNX7KrbnbbeVOQ7Oa/auYaBxIJnFmXya9SlCCG4afBPr9623G5zTi9IJXhJsE/07VX2K5WnLEc8KgpcEs6doT4ODs7vd8lyxeu9qunXo5uTmGt59OAdLD9olrbVku0+Foj2gjIIHcPVUO3/kfD6/7XPSCtMI9v9t9ZE2I1g7ey3l1eU8/J+HGWgciDHISMo9Kba9ntdlr2N30W7MFrPT+T/O+JiU4yks/6O9CJ/+/BfKhOgJ/JD327ah+sFZk9JYfPViPrrpIx7a/BCmShM3D7mZL/d/aTc4O/4ertrZ0OCsl/BoCE2O4+TZk06zKU0DylEZtSXbfSoU3o4yCh7A3VOtK62k1XtXc+OgG3ni+ycYFDYIc52ZpP1JTO0/lXG9xhHZKZLZQ2YT7B9MmimN4CXBlFWXMeDNAcxLmsenWZ/y8H8eZt2t6+gU0Mnpmi0VsXMMNusH582HNuNn8GNyv8nER8bzwOgH6PdmP65870qySrJ4bMtjHCk7wqNbHm3StRoanN3pTOnRy3FU1VQ5zaZGdh/JLTG3OIngtWQ2pVB4O8ooeAB3T7WOWknPb3ueN3e+yfio8Ww5uoWbh9yMpc5CVnEWC8csRAiB2WLmjvV3UGOp4cExDxLeMZyVN6yka4euRIdEc/eXdzO462CGdB3i1A5PiNj9rsfvOFJ2hLJzZYD94PxSykssunKRLXN50ZWLqLHUXPC1GhqcG9KZMlWanHbLczVbCwkM4bNbP3PKtG7pbEqh8GaUUfAADT3ValpJ7898n3fS3uHXs7+SuCmRX8/+yvHK48RHxrNgzAJuXnczRaeLSNyUyJFTR1hx/QoWT1jMwjEL2ZG/g2v7Xcu/dv+LmYNn0tGvo1OQ1hPqrFp7f9/r97YtOrXB2d/Hn58KfiKvPA9TpYma2hru2XDPBV8HGh+cHXfL09RT496Jc9otr7kxCE9KgisU3oRwt/rjcsADWyx4DE2nJzk32SYy5yhMp4nMRbwSQdaCLCI7RbI0ZSkFlQV07dCV5anLqbHUYK4zY3rEREhgCKZKE3HvxLHk2iU8+u2jdAnswp9+9yfe3fUua29ZS3RItNP59XIO4tnGlVodN9F5YdsLlFeX88rUV2xls9fOJjokmlPVp9h4cCPl58pBYKdE2lyWTl7Kvl/38f7M9xuv3Ag7C3Yy54s5ZN2f5TanAuqNZ+zyWNbduq5B+W2FwpsRQiCldDk4KKPgQaRV1nl56nI2HtxoJ0y3IH6BTVPI8JyB80+ex9fHl10ndnHXl3ex74F93LDmBvaf3E9eZR4JwxNYOGYhMd1iCHoxiNCgULYmbOWM+QzLU5fzWc5nnDOfIzQo1On8etxt2KPhuLGQlJJ3d73Lk98/SZ2so+J8BZ38O1FdW803d33DxOiJCCGYlzSPod2GNjl+4IrokGiPDc5SSu7dcC+m0ybW37bepWHQZlM9O/Vk1YxVam8FRbtFGYVLDONSo22mUCfrCH85nD1/3kPcijhuGHQDfbv0xd/gz4q0FUR2imRX4S7en/k+dw+/23YObQZR+qhLRfELQpvtfH/se1ueQT9jP+5JugdTlYmjZUdtqqTabKfnq43tl+Se+SPme3Rwbs5sTYnjKdozDRkFFVNoA/T+bx/hwx/6/oEtR7dQXl3O1tyt3B57O4snLCb3/+VyRdgVxPeItzMI4PkVNFKn35S9IJtxUeM4Vn6MU+dOsfHQRtbcvMYu6cxV7kRzCPQN9PhGRM2JQSgUCteomUIboPm/z5rPNrieP7xjOEG+QS5dLJ6eKTj65J9OfhpLnQWDj4HiqmLevfFdAM7UnCF2RSynzp1i3wP7mjRT6BLYhfLqcoxBxgZdXQqF4uLQ0EyhpZvsKC4AbbXS+3saDrCWnClh/oj5LlcTeXoFjaPw3YToCTy25THyK/JJuSfFVk/LCF65ayWrM1cT0TGi0ZjFI+Me8VhAWaFQtC5qptBGmC1m/F9ofBvLmidrnNwdrbGCRh/ngPoZC/uTfwAABNpJREFUQZelXbhx0I2sv329XV1TpYmY5TEYg4xqtY9CcRmiYgqXIE31a7syCJ7IR3DEMUbQ0b8jd8fdzeKrFzvVDe8YTtX5qmZlHHuyrQqFovVQRuESp6EsXk8GaV1lBK+auYpRkaOc6pacKaFLUJdmZRyr5Z8KxeWBch+1IU1JLDMGGd3mO3gSLe9g0fhFjdZdtmOZLUbQ1NwMhUJx6aDyFC5RLiTbuLVQGcEKRftBxRQUjdJcVVIVI1AovBNlFNqQ1toz+kJoriqpihEoFN6Jch8p7FAxAoXC+1ExBYVCoVDYUDEFhUKhUDQJZRQUCoVCYUMZBYVCoVDYUEZBoVAoFDaUUVAoFAqFjXZtFFauXNnWTbjotLc+t7f+gupze6G1+qyMQjujvfW5vfUXVJ/bC8ooKBQKhaLVueyT19q6DQqFQnE54pUZzQqFQqHwLMp9pFAoFAobyigoFAqFwoZXGgUhxDQhxAEhxGEhxH+5+H6eEOKkEGKP9fUn3XcWXfmGi9vyC6exPlvr3CaEyBFCZAsh1ujKE4QQh6yvhIvX6pbRwj575X0WQrym69dBIUS57juvvM+N9Nlb73NvIUSyECJdCJEphPij7rvHrccdEEJc1+yLSym96gUYgCNAP8AfyABiHOrMA952c3xVW/ehlfo8EEgHQq2fw61/jcBR699Q6/vQtu5Ta/bZm++zQ/2FwHvefp/d9dmb7zOwErjf+j4GyNW9zwACgL7W8xiac31vnCmMAQ5LKY9KKWuAT4GZbdym1qYpfb4P+G8pZRmAlLLEWn4d8K2U8pT1u2+BaRep3S2hJX2+XGnu//Zc4BPre2++z3r0fb5caUqfJdDZ+j4EOGF9PxP4VEp5Xkp5DDhsPV+T8Uaj0BPI130usJY5Mts67fpcCBGlKw8UQqQJIX4WQsxq1ZZ6jqb0eRAwSAixw9q3ac049lKkJX0G773PAAghoql/Uvy+ucdeYrSkz+C99/kZ4C4hRAHwNfUzpKYe2yC+zansRWwEPpFSnhdCJAIfAn+wfhctpTQJIfoB3wsh9kopj7RZSz2HL/XulElAL2CbEGJYm7ao9XHZZyllOd57nzXmAJ9LKS1t3ZCLiKs+e+t9ngt8IKX8pxBiHPCxECLWEyf2xpmCCdA/+feyltmQUpZKKc9bP/4PMEr3ncn69yiwFRjZmo31EI32mfonhg1SSrN1WnmQ+gGzKcdeirSkz958nzXmYO9G8eb7rOHYZ2++z/cC6wCklD8BgUDXJh7bMG0dVGmFII0v9UG0vvwWpBnqUKeH7v1NwM/W96FAgPV9V+AQDQS1LpVXE/s8DfhQ17d8IIz6wOMxa99Dre+Nbd2nVu6z195na73BQC7W5FRrmdfe5wb67LX3GdgMzLO+H0J9TEEAQ7EPNB+lmYHmNv8BWulH/SP1T4VHgMXWsueAGdb3S4Bs64+XDAy2ll8J7LWW7wXubeu+eLDPAngVyLH2bY7u2HuoD0gdBua3dV9au8/efJ+tn58BXnJxrFfeZ3d99ub7TP0qox3Wvu0BpuqOXWw97gAwvbnXVjIXCoVCobDhjTEFhUKhUFwgyigoFAqFwoYyCgqFQqGwoYyCQqFQKGwoo6BQKBQKG8ooKBQKhcKGMgoKhUKhsKGMgkKhUChs/H+ajX3ML+is+QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot the principal components\n",
    "colors = 200*['green', 'purple', 'orange', 'teal', 'black', 'red']\n",
    "for i in range(0,1): #12,13\n",
    "    p_data = OPs_400_40_10k[:,i]\n",
    "    plt.plot(etas, (p_data), marker='o', linewidth = 1.0, markersize=12, color = colors[i], markerfacecolor='None')\n",
    "    #p_data = OPs_200_10[:,i]\n",
    "    #plt.plot(etas, (p_data), marker='^', linewidth = 1.0, markersize=8, color = colors[i])\n",
    "    p_data = OPs_400_10[:,i]\n",
    "    plt.plot(etas, (p_data), marker='s', linewidth = 1.0, markersize=8, color = colors[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "2.7.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}