bond_gan_mk1-Copy1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook addresses the question, \"Can we represent a molecule as a graph via a 1D column vector or a 2D matrix of fixed length, with maximum number of atoms n_rows?\" Then, can we use this representation to learn neural fingerprints? E.g., can we make an aromatic ring detector? \n",
    "\n",
    "Scheme:\n",
    "feature_matrix = X\n",
    "for each ligand:\n",
    "    choose a central atom. this can be the atom (node) that minimizes distance to furthest heavy atom in graph.\n",
    "    set first row of X to be this central atom\n",
    "    set next four rows to be the atoms bonded to that centrl atom\n",
    "        set zeros for rows where row ind > n_bonds of atom\n",
    "    for each of those atoms:\n",
    "        repeat. find their neighbors. add to matrix.\n",
    "\n",
    "algorithm: breadth-first search:\n",
    "1. create networkx graph based on molecule\n",
    "2. find \"central\" atom (different strategies)\n",
    "3. define atom matrix of size (1+4+4*3^(L-1)) x (n_features_per_atom)\n",
    "4. start atom queue q\n",
    "5. central_atom.layer = 0; central_atom.row_idx = 0;\n",
    "6. q.enqueue(central_atom)\n",
    "7. define adjacency matrix of size (1+4+4*3^(L-1)) x 4\n",
    "\n",
    "def get_row_idx(curr_layer, prev_row_idx, curr_neighbor_idx):\n",
    "    if curr_layer == 0:\n",
    "        return(0)\n",
    "    if curr_layer == 1:\n",
    "        row_idx = 1 + curr_neighbor_idx\n",
    "    if layer == 2:\n",
    "        last_max = 5\n",
    "        row_idx = last_max + (3*(prev_row_idx-last_max)) + curr_neighbor_idx\n",
    "    if layer > 2:\n",
    "        last_max = 5 + 4*3^(curr_layer-2) \n",
    "        row_idx = last_max + 3*(prev_row_idx-last_max) + curr_neighbor_idx\n",
    "    return(row_idx)\n",
    "    \n",
    "\n",
    "while q.is_not_empty():\n",
    "    a = q.dequeue()\n",
    "    a.visited = True\n",
    "    for n_idx, n in enumerate(a.neighbors()):\n",
    "        if not n.visited:\n",
    "            row_idx = c\n",
    "            n.layer = a.layer + 1\n",
    "            row_idx = get_row_idx(n.layer, a.row_idx, n_idx)\n",
    "            n.row_idx = row_idx\n",
    "            adj_matrix[a.row_idx][n_idx] = n.row_idx\n",
    "            atom_matrix[row_idx][elem_to_idx[n.elem]] = 1\n",
    "\n",
    "input_matrix = tf.concat([atom_matrix, atom_matrix[adj_matrix[:,0]], atom_matrix[adj_matrix[:,1]], atom_matrix[adj_matrix[:,2]], atom_matrix[adj_matrix[:,3]]\n",
    "\n",
    "neural net:\n",
    "h1 = relu([tf.zeros([n_features_per_atom, 4]) * input_matrix + bias))\n",
    "h1_conc = tf.concat([h1, h1[adj_matrix[:,0], ..., h1[adj_matrix[:,3])\n",
    "\n",
    "repeat h1 to get h2\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "dihedral predictor pseudocode:\n",
    "\n",
    "get bonds for molecule\n",
    "create networkx graph out of molecule (use atom indices)\n",
    "\n",
    "for each edge:\n",
    "   for neighbor_i in atom_i.neighbors():\n",
    "       if neighbor_i == atom_j: continue\n",
    "       for neighbor_j in atom_j.neighbors():\n",
    "           if neighbor_j == atom_i: continue\n",
    "           dihedrals.append((neighbor_i, atom_i, neighbor_j, atom_j))\n",
    "           check to make sure (atom_j, neighbor_j, atom_i, neighbor_i)) not already in list\n",
    "\n",
    "for dihedral in dihedrals:\n",
    "    angle =  rdMolTransforms.GetDihedralDeg(c, 0,1,2,3)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import matplotlib\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from rdkit import Chem\n",
    "from rdkit.Chem import AllChem, rdMolTransforms\n",
    "import os\n",
    "import fnmatch\n",
    "import numpy as np\n",
    "import deepchem as dc\n",
    "from scipy.sparse import csr_matrix\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "from sklearn.mixture import GMM\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_torsions_angles(mol):\n",
    "    c = mol.GetConformer(0)\n",
    "\n",
    "    torsions = []\n",
    "    torsion_tuples = []\n",
    "    \n",
    "    torsion_matrix = np.zeros((100,3))\n",
    "    torsion_indices = np.zeros((100,200,12)).astype(np.uint8)\n",
    "    bond_idx = 0\n",
    "    for idx, bond in enumerate(mol.GetBonds()):\n",
    "        atom_i = bond.GetBeginAtom()\n",
    "        atom_j = bond.GetEndAtom()\n",
    "        if bond.IsInRing(): \n",
    "            continue\n",
    "        \n",
    "        exist_dihed = False\n",
    "        for neighbor_j in atom_j.GetNeighbors():\n",
    "            if neighbor_j.GetIdx() == atom_i.GetIdx():\n",
    "                continue\n",
    "            \n",
    "            dihed_idx = 0\n",
    "            for neighbor_i in atom_i.GetNeighbors():\n",
    "                if neighbor_i.GetIdx() == atom_j.GetIdx():\n",
    "                    continue\n",
    "                \n",
    "                exist_dihed=True\n",
    "                torsion_tuple = (neighbor_i.GetIdx(), atom_i.GetIdx(), atom_j.GetIdx(), neighbor_j.GetIdx())\n",
    "\n",
    "                torsion_matrix[bond_idx][dihed_idx] = np.cos(rdMolTransforms.GetDihedralRad(c, *torsion_tuple))\n",
    "                torsion_indices[bond_idx][torsion_tuple[0]][dihed_idx*4] = 1\n",
    "                torsion_indices[bond_idx][torsion_tuple[1]][dihed_idx*4+1] = 1\n",
    "                torsion_indices[bond_idx][torsion_tuple[2]][dihed_idx*4+2] = 1\n",
    "                torsion_indices[bond_idx][torsion_tuple[3]][dihed_idx*4+3] = 1\n",
    "                dihed_idx += 1\n",
    "            #torsion_matrix[bond_idx][3] = 3.14159 - np.abs(np.abs(torsion_matrix[bond_idx][2]-torsion_matrix[bond_idx][1])-3.14159)\n",
    "            #torsion_matrix[bond_idx][4] = 3.14159 - np.abs(np.abs(torsion_matrix[bond_idx][2]-torsion_matrix[bond_idx][0])-3.14159)\n",
    "            #torsion_matrix[bond_idx][5] = 3.14159 - np.abs(np.abs(torsion_matrix[bond_idx][1]-torsion_matrix[bond_idx][0])-3.14159)\n",
    "            break\n",
    "        if exist_dihed:\n",
    "            bond_idx += 1\n",
    "            break\n",
    "    return((torsion_indices, csr_matrix(torsion_matrix)))\n",
    "                "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def featurize_mols(mol_files):\n",
    "    featurizer = AdjacencyFingerprint(max_n_atoms=200)\n",
    "    features = []\n",
    "    for idx, mol_file in enumerate(mol_files):\n",
    "        if idx % 100 == 0:\n",
    "            print(idx)\n",
    "        if 1==1:\n",
    "            mol = Chem.MolFromMol2File(mol_file)\n",
    "            if mol is None:\n",
    "                features.append(None)\n",
    "                continue\n",
    "            torsions = get_torsions_angles(mol)\n",
    "            graph_feat = featurizer.featurize([mol])[0]\n",
    "            features.append((mol_file, torsions, graph_feat))\n",
    "        else:\n",
    "            features.append(None)\n",
    "    return(features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from deepchem.feat.graph_features import ConvMolFeaturizer\n",
    "from deepchem.feat.adjacency_fingerprints import AdjacencyFingerprint"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "feature_file = \"./dihedral_features_pdbbind.pkl\"\n",
    "if not os.path.exists(feature_file):\n",
    "#if 1== 1:\n",
    "    pdbbind_dir = \"/home/evan/Documents/deep_docking/datasets/v2015/\"\n",
    "    def find_files(directory, pattern):\n",
    "        for root, dirs, files in os.walk(directory):\n",
    "            for basename in files:\n",
    "                if fnmatch.fnmatch(basename, pattern):\n",
    "                    filename = os.path.join(root, basename)\n",
    "                    yield filename\n",
    "    ligand_files = [f for f in find_files(pdbbind_dir, \"*ligand.mol2\")][:1]\n",
    "    features = featurize_mols(ligand_files)\n",
    "    with open(feature_file, \"wb\") as f:\n",
    "        pickle.dump(features, f, protocol=2)\n",
    "else:\n",
    "    with open(feature_file, \"rb\") as f:\n",
    "        features = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "features = [f for f in features if f is not None]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.61143659, -0.99900387,  0.48252666],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ]])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features[0][1][1].toarray()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "alpha = 0.01\n",
    "n_layers = 1\n",
    "S = 32\n",
    "B = 200\n",
    "L_list = [50, 50, 50, 50]\n",
    "p = 75\n",
    "n_bonds=100\n",
    "\n",
    "x = tf.placeholder(tf.float32, shape=[S, B, p])\n",
    "\n",
    "non_zero_inds = tf.placeholder(tf.int32, shape=[None, S*n_bonds])\n",
    "\n",
    "adj_matrix = tf.placeholder(tf.float32, shape=[S, B, B])\n",
    "dihed_indices = tf.placeholder(tf.float32, shape=[S, n_bonds, B, 12])\n",
    "\n",
    "label_placeholder = tf.placeholder(\n",
    "    dtype='float32', shape=[None,3], name=\"label_placeholder\")\n",
    "\n",
    "phase = tf.placeholder(dtype='bool', name='phase')\n",
    "\n",
    "z = tf.placeholder(tf.float32,\n",
    "                          [None, 200, 75], name='z')\n",
    "\n",
    "z_ini = tf.placeholder(tf.float32,\n",
    "                          [None, 200, 75], name='z_ini')\n",
    "\n",
    "ones_like = tf.placeholder(tf.float32, shape=[None,1])\n",
    "zeros_like = tf.placeholder(tf.float32, shape=[None,1])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "#DISCRIMINATOR\n",
    "\n",
    "\"\"\"\n",
    "D_W_list = [None for i in range(n_layers)]\n",
    "D_b_list = [None for i in range(n_layers)]\n",
    "D_h_list = [None for i in range(n_layers)]\n",
    "\n",
    "\n",
    "\n",
    "for layer_idx in range(n_layers):\n",
    "    if layer_idx == 0:\n",
    "      L_in = p\n",
    "      L_out = L_list[0]\n",
    "      atom_matrix = x\n",
    "    else:\n",
    "      L_in = L_list[layer_idx-1]\n",
    "      L_out = L_list[layer_idx]\n",
    "      atom_matrix = D_h_list[layer_idx-1]\n",
    "\n",
    "    D_W_list[layer_idx] = tf.Variable(tf.truncated_normal([L_in, L_out], seed=2017), name=\"D_W_list%d\" %layer_idx)\n",
    "    D_b_list[layer_idx] = tf.Variable(tf.zeros([1, L_out]))\n",
    "\n",
    "    \n",
    "D_W2 = tf.Variable(tf.truncated_normal([L_list[-1]*12+3, 100]))\n",
    "D_b2 = tf.Variable(tf.zeros([1, 100]))\n",
    "\n",
    "D_W3 = tf.Variable(tf.truncated_normal([100, 3]))\n",
    "D_b3 = tf.Variable(tf.zeros([1, 3]))\n",
    "\n",
    "D_W4 = tf.Variable(tf.truncated_normal([3, 10]))\n",
    "D_b4 = tf.Variable(tf.zeros([1, 10]))\n",
    "\n",
    "D_W5 = tf.Variable(tf.truncated_normal([10, 1]))\n",
    "D_b5 = tf.Variable(tf.zeros([1, 1]))\n",
    "\n",
    "\"\"\"\n",
    "\n",
    "D_W0 = tf.Variable(tf.truncated_normal([3, 100]))\n",
    "D_b0 = tf.Variable(tf.zeros([1,100]))\n",
    "\n",
    "D_W1 = tf.Variable(tf.truncated_normal([100, 100]))\n",
    "D_b1 = tf.Variable(tf.zeros([1,100]))\n",
    "\n",
    "D_W2 = tf.Variable(tf.truncated_normal([100, 1]))\n",
    "D_b2 = tf.Variable(tf.zeros([1,1]))\n",
    "\n",
    "def adjacency_conv_layer(atom_matrix, W, b, L_in, L_out, layer_idx, is_training=True):\n",
    "    print(\"layer_idx: %d\" %(layer_idx))\n",
    "    h = tf.matmul(adj_matrix, atom_matrix, name=\"adj_conv1\")\n",
    "    h = tf.reshape(h, shape=(S*B, L_in))\n",
    "\n",
    "    h = tf.matmul(h, W) + b\n",
    "    h = tf.maximum(alpha*h, h)\n",
    "    \n",
    "    h = tf.reshape(h, (S, B, L_out))\n",
    "    #h = tf.contrib.layers.batch_norm(h, \n",
    "    #                                    center=True, scale=False, \n",
    "    #                                    is_training=is_training)\n",
    "\n",
    "    return(h)\n",
    "\n",
    "def discriminator(angles, is_training=True):\n",
    "    D_h0 = tf.matmul(angles, D_W0) + D_b0\n",
    "    D_h0 = tf.maximum(alpha*D_h0, D_h0)\n",
    "\n",
    "    D_h1 = tf.matmul(D_h0, D_W1) + D_b1\n",
    "    D_h1 = tf.maximum(alpha*D_h1, D_h1)\n",
    "    \n",
    "    D_h2 = tf.matmul(D_h1, D_W2) + D_b2\n",
    "    #D_h2 = tf.maximum(alpha*D_h2, D_h2)\n",
    "    \n",
    "    return(D_h2)\n",
    "    \n",
    "    \"\"\"\n",
    "    \n",
    "    for layer_idx in range(n_layers):\n",
    "        if layer_idx == 0:\n",
    "          L_in = p\n",
    "          L_out = L_list[0]\n",
    "          atom_matrix = x\n",
    "        else:\n",
    "          L_in = L_list[layer_idx-1]\n",
    "          L_out = L_list[layer_idx]\n",
    "          atom_matrix = D_h_list[layer_idx-1]\n",
    "        \n",
    "        D_h_list[layer_idx] = adjacency_conv_layer(atom_matrix, D_W_list[layer_idx], D_b_list[layer_idx], L_in, L_out, layer_idx)\n",
    "\n",
    "    L_final = L_list[n_layers-1]\n",
    "    h_final = tf.reshape(D_h_list[layer_idx], (S, B, L_final))\n",
    "\n",
    "    #add dihedral regressor layers\n",
    "\n",
    "    d0 = []\n",
    "    for i in range(0, S):\n",
    "        mol_tuple = []\n",
    "        for j in range(0, 12):\n",
    "          entry = h_final[i]\n",
    "          indices = dihed_indices[i][:,:,j]\n",
    "          atom_list = tf.matmul(indices, entry, name=\"disc1\")\n",
    "          atom_list = tf.reshape(atom_list, (n_bonds, L_final))\n",
    "          mol_tuple.append(atom_list)\n",
    "        mol_tuple = tf.reshape(tf.stack(mol_tuple, axis=1), (n_bonds, L_final*12))\n",
    "        d0.append(mol_tuple)\n",
    "\n",
    "    d0 = tf.concat(d0, axis=0)\n",
    "    d0 = tf.concat([d0, angles], axis=1)\n",
    "    d0 = tf.matmul(tf.cast(non_zero_inds, tf.float32), d0)\n",
    "\n",
    "\n",
    "    d2 = tf.matmul(d0, D_W2, name=\"disc2\") + D_b2\n",
    "    d2 = tf.maximum(d2, alpha*d2)\n",
    "\n",
    "    #d2 = tf.contrib.layers.batch_norm(d2, \n",
    "    #                                    center=True, scale=False, \n",
    "    #                                    is_training=is_training)\n",
    "    \n",
    "    d3 = tf.matmul(d2, D_W3, name=\"disc3\") + D_b3\n",
    "    d3 = tf.maximum(d3, alpha*d3)\n",
    "    \n",
    "    #d3 = tf.nn.tanh(d3)\n",
    "\n",
    "    #d3 = tf.contrib.layers.batch_norm(d3, \n",
    "    #                                    center=True, scale=False, \n",
    "    #                                    is_training=is_training)\n",
    "    \n",
    "                                                                                                                                                               \n",
    "    d4 = tf.matmul(d3, D_W4, name=\"disc3\") + D_b4\n",
    "    \n",
    "    #\n",
    "    output = tf.matmul(d4, D_W5, name=\"disc4\") + D_b5\n",
    "    \n",
    "    #D_prob = tf.nn.sigmoid(D_logit)\n",
    "    return(output)\n",
    "    \"\"\"\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "def atan2(y, x, epsilon=1.0e-12):\n",
    "  # Add a small number to all zeros, to avoid division by zero:\n",
    "  x = tf.where(tf.equal(x, 0.0), x+epsilon, x)\n",
    "  y = tf.where(tf.equal(y, 0.0), y+epsilon, y)\n",
    "\n",
    "  angle = tf.where(tf.greater(x,0.0), tf.atan(y/x), tf.zeros_like(x))\n",
    "  angle = tf.where(tf.logical_and(tf.less(x,0.0),  tf.greater_equal(y,0.0)), tf.atan(y/x) + np.pi, angle)\n",
    "  angle = tf.where(tf.logical_and(tf.less(x,0.0),  tf.less(y,0.0)), tf.atan(y/x) - np.pi, angle)\n",
    "  angle = tf.where(tf.logical_and(tf.equal(x,0.0), tf.greater(y,0.0)), 0.5*np.pi * tf.ones_like(x), angle)\n",
    "  angle = tf.where(tf.logical_and(tf.equal(x,0.0), tf.less(y,0.0)), -0.5*np.pi * tf.ones_like(x), angle)\n",
    "  angle = tf.where(tf.logical_and(tf.equal(x,0.0), tf.equal(y,0.0)), tf.zeros_like(x), angle)\n",
    "  return angle\n",
    "\n",
    "#GENERATOR\n",
    "\n",
    "G_W_list = [None for i in range(n_layers)]\n",
    "G_b_list = [None for i in range(n_layers)]\n",
    "G_h_list = [None for i in range(n_layers)]\n",
    "\n",
    "\n",
    "for layer_idx in range(n_layers):\n",
    "    if layer_idx == 0:\n",
    "      L_in = p\n",
    "      L_out = L_list[0]\n",
    "      atom_matrix = x\n",
    "    else:\n",
    "      L_in = L_list[layer_idx-1]\n",
    "      L_out = L_list[layer_idx]\n",
    "      atom_matrix = G_h_list[layer_idx-1]\n",
    "\n",
    "    G_W_list[layer_idx] = tf.Variable(tf.truncated_normal([L_in, L_out], seed=2017), name=\"G_W_list%d\" %layer_idx)\n",
    "    G_b_list[layer_idx] = tf.Variable(tf.zeros([1, L_out]))\n",
    "\n",
    "    \n",
    "G_W2 = tf.Variable(tf.truncated_normal([L_list[-1]*12, 100]))\n",
    "G_b2 = tf.Variable(tf.zeros([1, 100]))\n",
    "\n",
    "G_W3 = tf.Variable(tf.truncated_normal([100, 3]))\n",
    "G_b3 = tf.Variable(tf.zeros([1, 3]))\n",
    "\n",
    "#G_W4 = tf.Variable(tf.truncated_normal([100, 3]))\n",
    "#G_b4 = tf.Variable(tf.zeros([1, 3]))\n",
    "\n",
    "\n",
    "def gen_adjacency_conv_layer(atom_matrix, W, b, L_in, L_out, layer_idx, z, is_training):\n",
    "    print(\"layer_idx: %d\" %(layer_idx))\n",
    "    h = tf.matmul(adj_matrix, atom_matrix, name=\"gen0_%d\" %layer_idx)\n",
    "    h = tf.reshape(h, shape=(S*B, L_in))\n",
    "\n",
    "    h = tf.matmul(h, W, name=\"gen1\") + b\n",
    "    h = tf.maximum(alpha*h, h)\n",
    "\n",
    "    \n",
    "    h = tf.reshape(h, (S, B, L_out))\n",
    "    #h = tf.contrib.layers.batch_norm(h, \n",
    "    #                                    center=True, scale=False, \n",
    "    #                                    is_training=is_training)\n",
    "\n",
    "    return(h)\n",
    "\n",
    "def generator(z, is_training=True):\n",
    "    for layer_idx in range(n_layers):\n",
    "        if layer_idx == 0:\n",
    "          L_in = p\n",
    "          L_out = L_list[0]\n",
    "          atom_matrix = tf.add(x, z)\n",
    "        else:\n",
    "          L_in = L_list[layer_idx-1]\n",
    "          L_out = L_list[layer_idx]\n",
    "          atom_matrix = G_h_list[layer_idx-1]\n",
    "        \n",
    "        G_h_list[layer_idx] = gen_adjacency_conv_layer(atom_matrix, G_W_list[layer_idx], G_b_list[layer_idx], L_in, L_out, layer_idx, z, is_training)\n",
    "\n",
    "    L_final = L_list[n_layers-1]\n",
    "    g_h_final = tf.reshape(G_h_list[layer_idx], (S, B, L_final))\n",
    "\n",
    "    #add dihedral regressor layers\n",
    "\n",
    "    g_d0 = []\n",
    "    for l in range(0,12):\n",
    "        indices = dihed_indices[:,:,:,l]\n",
    "        g_d0.append(tf.matmul(indices, g_h_final))\n",
    "    g_d0 = tf.reshape(tf.concat(g_d0, axis=2), (S*n_bonds, L_final*12))\n",
    "    \n",
    "    #for i in range(0, S):\n",
    "    #    mol_tuple = []\n",
    "    #    for j in range(0, 12):\n",
    "    #      entry = g_h_final[i]\n",
    "    #      indices = dihed_indices[i][:,:,j]\n",
    "    #      atom_list = tf.matmul(indices, entry, name='gen2')\n",
    "    #      atom_list = tf.reshape(atom_list, (n_bonds, L_final))\n",
    "    #      mol_tuple.append(atom_list)\n",
    "    #    mol_tuple = tf.reshape(tf.stack(mol_tuple, axis=1), (n_bonds, L_final*12))\n",
    "    #    g_d0.append(mol_tuple)\n",
    "\n",
    "    #g_d0 = tf.concat(g_d0, axis=0)\n",
    "    \n",
    "    g_d2 = tf.matmul(g_d0, G_W2, name='gen3') + G_b2\n",
    "    g_d2 = tf.tanh(g_d2)\n",
    "    #g_d2 = tf.maximum(alpha*g_d2, g_d2)\n",
    "\n",
    "    \n",
    "    #g_d2 = tf.contrib.layers.batch_norm(g_d2, \n",
    "    #                                    center=True, scale=False, \n",
    "    #                                    is_training=is_training)\n",
    "    \n",
    "    g_d3 = tf.matmul(g_d2, G_W3) + G_b3\n",
    "    #g_d3 = tf.nn.tanh(g_d3)\n",
    "\n",
    "    #g_d3 = tf.contrib.layers.batch_norm(g_d3, \n",
    "    #                                    center=True, scale=False, \n",
    "    #                                    is_training=is_training)\n",
    "                                                                                                                                                               \n",
    "    #g_d4 = tf.matmul(g_d3, G_W4) + G_b4\n",
    "    \n",
    "    output = tf.cos(g_d3)\n",
    "    output = tf.matmul(tf.cast(non_zero_inds, tf.float32), output)\n",
    "\n",
    "    #d3_cos = tf.cos(g_d4)\n",
    "    #d3_sin = tf.sin(g_d4)\n",
    "    #output = atan2(d3_sin, d3_cos)\n",
    "    \n",
    "    #output = tf.concat([output,\n",
    "    #                    tf.reshape(3.14159 - tf.abs(tf.abs(output[:,2]-output[:,1])-3.14159), (-1,1)),\n",
    "    #                    tf.reshape(3.14159 - tf.abs(tf.abs(output[:,2]-output[:,0])-3.14159), (-1,1)),\n",
    "    #                    tf.reshape(3.14159 - tf.abs(tf.abs(output[:,1]-output[:,0])-3.14159), (-1,1))],\n",
    "    #                   axis=1)\n",
    "    print(\"output\")\n",
    "    print(output)\n",
    "    \n",
    "    G_logit = tf.nn.sigmoid(output)\n",
    "\n",
    "    return(output, G_logit)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "theta_D = [D_W0, D_W1, D_W2, D_b0, D_b1, D_b2]\n",
    "#theta_D = [D_W_list[0], D_W2, D_W3, D_W4, D_W5, D_b_list[0], D_b2, D_b3, D_b4, D_b5]\n",
    "\n",
    "theta_G = [G_W_list[0], G_W2, G_W3, G_b_list[0], G_b2, G_b3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "gmm = GMM(2, n_iter=1., covariance_type=\"spherical\")\n",
    "\n",
    "gmm.means_ = np.array([[-10], [10]])\n",
    "gmm.covars_ = np.array([[0.0001], [0.0001]])\n",
    "\n",
    "def construct_feed_dict(X, start=None,\n",
    "                      stop=None, y=None,\n",
    "                      keep_prob=1.0, train=False):\n",
    "    if start is None:\n",
    "      start = 0\n",
    "      stop = len(X)\n",
    "\n",
    "    adj = [X[idx][2][0].toarray().astype(np.float32) for idx in range(start, stop)]\n",
    "    A_batch = [X[idx][2][1].toarray() for idx in range(start, stop)]\n",
    "    D_batch = [X[idx][1][0] for idx in range(start, stop)]\n",
    "    y_batch = [X[idx][1][1].toarray() for idx in range(start, stop)]\n",
    "\n",
    "    y_batch = np.concatenate(y_batch, axis=0)\n",
    "    non_zero_batch = np.where(y_batch[:,0] != 0.)[0]\n",
    "    y_batch = y_batch[non_zero_batch]\n",
    "    \n",
    "    ones_batch = np.random.uniform(0.7, 1.2, size=(len(non_zero_batch),1))\n",
    "    zeros_batch = np.random.uniform(0., 0.3, size=(len(non_zero_batch),1))\n",
    "\n",
    "    #y_batch = gmm.sample(len(y_batch))\n",
    "    \n",
    "    y_batch = y_batch + np.random.normal(0, 0.03, size=y_batch.shape)\n",
    "    \n",
    "    #y_batch = np.squeeze(gmm.sample(len(y_batch)))\n",
    "    \n",
    "    #adj = [np.random.uniform(-1., 1., size=a.shape) for a in adj]\n",
    "    #A_batch = [np.random.uniform(-1., 1., size=a.shape) for a in A_batch]\n",
    "    #D_batch = [np.random.uniform(-1., 1., size=a.shape) for a in D_batch]\n",
    "\n",
    "\n",
    "    onehotter = OneHotEncoder(n_values = S*X[0][1][1].shape[0])\n",
    "    non_zero_onehot = onehotter.fit_transform(non_zero_batch).toarray().reshape((len(non_zero_batch),S*X[0][1][1].shape[0]))\n",
    "    \n",
    "    #z_batch = np.zeros((S, adj[0].shape[0], 75))\n",
    "    z_batch = np.random.uniform(-0.1, 0.1, size=(S, adj[0].shape[0],75))\n",
    "    #y_batch = np.random.random(size=(S*250))\n",
    "    \n",
    "    feed_dict = {x: A_batch,\n",
    "                 adj_matrix: adj,\n",
    "                 phase: train,\n",
    "                 label_placeholder: y_batch,\n",
    "                 non_zero_inds: non_zero_onehot,\n",
    "                 dihed_indices: D_batch,\n",
    "                 z:z_batch,\n",
    "                 ones_like: ones_batch,\n",
    "                 zeros_like: zeros_batch\n",
    "                }\n",
    "    return(feed_dict)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "layer_idx: 0\n",
      "output\n",
      "Tensor(\"MatMul_49:0\", shape=(?, 3), dtype=float32)\n",
      "layer_idx: 0\n",
      "output\n",
      "Tensor(\"MatMul_63:0\", shape=(?, 3), dtype=float32)\n",
      "Training epoch 0\n",
      "Iter: 0\n",
      "D loss: -8.379e-05\n",
      "G_loss: 6.028e-05\n",
      "()\n",
      "[-0.00389653  0.05083536 -0.02650795]\n",
      "Training epoch 100\n",
      "Iter: 100\n",
      "D loss: -0.0008525\n",
      "G_loss: -0.0001191\n",
      "()\n",
      "[ 0.02856161 -0.02196538 -0.07060545]\n",
      "Training epoch 200\n",
      "Iter: 200\n",
      "D loss: -0.001678\n",
      "G_loss: -0.0005358\n",
      "()\n",
      "[ 0.01692576  0.04972874  0.15451401]\n",
      "Training epoch 300\n",
      "Iter: 300\n",
      "D loss: -0.003159\n",
      "G_loss: -0.0009463\n",
      "()\n",
      "[-0.02929551  0.03558866 -0.18427068]\n",
      "Training epoch 400\n",
      "Iter: 400\n",
      "D loss: -0.004271\n",
      "G_loss: -0.0008361\n",
      "()\n",
      "[-0.1436801  -0.05769541 -0.08762261]\n",
      "Training epoch 500\n",
      "Iter: 500\n",
      "D loss: -0.006829\n",
      "G_loss: -0.002026\n",
      "()\n",
      "[-0.15769298 -0.05677586  0.14737701]\n",
      "Training epoch 600\n",
      "Iter: 600\n",
      "D loss: -0.007988\n",
      "G_loss: -0.002042\n",
      "()\n",
      "[-0.18912064  0.10309114 -0.04047532]\n",
      "Training epoch 700\n",
      "Iter: 700\n",
      "D loss: -0.009675\n",
      "G_loss: -0.003133\n",
      "()\n",
      "[-0.09364519  0.20757742  0.07753406]\n",
      "Training epoch 800\n",
      "Iter: 800\n",
      "D loss: -0.01175\n",
      "G_loss: -0.004659\n",
      "()\n",
      "[-0.0625753   0.21866946  0.02456829]\n",
      "Training epoch 900\n",
      "Iter: 900\n",
      "D loss: -0.01271\n",
      "G_loss: -0.004524\n",
      "()\n",
      "[-0.41501123  0.54549098 -0.28174976]\n",
      "Training epoch 1000\n",
      "Iter: 1000\n",
      "D loss: -0.02061\n",
      "G_loss: -0.01096\n",
      "()\n",
      "[-0.61834377  0.67376649 -0.64141583]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAD8CAYAAABkbJM/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2c2dd33e50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAD8CAYAAABkbJM/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3XeYVOXZ+PHvPVtZeu/dRQGRtiBYkSZgwZIoGiMxyc+U\n16gxbdXEbkRjTDXFGhOjxDfqiwoqgg0rRVC6VOm9w/Z5fn/MmdkzM2fmnCnbZu/Pde21c848z5nn\nLHrueboYY1BKKaXc+Oq6AEoppRoGDRhKKaU80YChlFLKEw0YSimlPNGAoZRSyhMNGEoppTzRgKGU\nUsoTDRhKKaU80YChlFLKk+y6LkA6tWvXzvTq1auui6GUUg3KkiVL9hlj2ruly6iA0atXLxYvXlzX\nxVBKqQZFRL7ykk6bpJRSSnmiAUMppZQnGjCUUkp5ogFDKaWUJxowlFJKeaIBQymllCcaMJRSSnmS\nloAhIpNEZK2IrBeRYof3RUT+aL3/hYgMc8srIm1E5C0RWWf9bp2OssZTWlHFvFW7qfIbSsqr8Ptj\nb1978Hg5s5Ztr+kiKaVUvZHyxD0RyQIeBSYA24BFIvKKMWaVLdlkoND6OR34K3C6S95iYL4xZoYV\nSIqBX6RaXidVfsOsZdu55YXPo977w7QhbNx7nE837eeqkT2YMKAj2w6WMPF37wPQuiAXgMHdWyEC\n2T4hN8vHkdJK1u0+yul92tZEkZVSqtalY6b3SGC9MWYjgIjMBKYC9oAxFfinMcYAn4hIKxHpDPSK\nk3cqMMbK/wzwLjUUMD7ZuN8xWADcNHOZLd2BqPevfWph3GvfedEATpRXUVpRxQ/G9KUgN/An/8O8\ndQzp0Ypz+7nOxldKqXohHQGjK7DVdryNQC3CLU1Xl7wdjTE7rde7gI5pKKujM09qx3PfPZ3Rfdsi\nIryweCsl5VV8tuUgH67fz3knt+doaSVvrNyV8LXvfrU6bm7ce5zZy3fSvnkee4+WkZvt4+lvjaC0\nooq87CzeXLmLhZsO8JdrhvH0h5u4YFAXRvfVGopSqn5oEGtJGWOMiDh2KIjI9cD1AD169Ej6M844\nqV3o9RVF3QG4+vQeZPsEEQm9t2L7YW6auZTLhnXj2tE92XLgBBf88QNPnzF7eSD+7T1aBkB5pZ9v\nPPFpVLoZr6/hrVW7efaTLXx+x0Te/XIPN81cxie3jqNTy/y4n7HjUAlPLNjE7Rf0J8sncdMqpVQi\n0tHpvR3objvuZp3zkiZe3t1WsxXW7z1OH26MecwYU2SMKWrfPr3NOzlZvrBgAXBq15bM/8kY/ue8\nk2ien8PALi1Zcff5/PnqofztmuFcOrQrQ3u0AuC+S06lTdNcLhvaNaHPfWvV7tDrwffMDTWLjXpg\nPv+7eCu3v7yca2yBZufhEv723gaMMfz8v1/w1IebWPLVwbBr7jlayg+eXcKR0oqEylLb7nl1Fa99\nsYPH3t9Ar+LZlFf667pISilLOmoYi4BCEelN4GE/Dbg6Is0rwA1WH8XpwGFjzE4R2Rsn7yvAdGCG\n9XtWGspaI5rlZXPhaV0AmHRqJ7YfKuH/lm7nG6f34JpRPQFo0SSHf3y0mV9e0J/vnNWbvUfLmPyH\nBUwc2InnF24BoHPLfA6XVHCivCrmZ/3sv1+EXr++fCdFvdow+oG3ARjQuUXoAVtlG+G1/VAJF//p\nA/YfL2d4z9Z89+w+Kd1vRZUfYyA3O/2jsp/6cBNPfQjN8wL/aZZUVNXI5ySitCLw75Gfk1Wn5VCq\nrkmgHzrFi4hMAX4PZAFPGWPuF5HvAxhj/iaBr+l/BiYBJ4DrjDGLY+W1zrcFXgB6AF8BVxhjonud\nbYqKikx9Xd7cGIMx4HNoJtp+qIRmudm0LMhhwiPvsW7PMbq1boII7D5SFvdb9oQBHcNqJCN7t2Hh\npuo/09s/OZc7X1nJgnX7APjpxH5879y+/Ovjr1i18wjXjOrJkO6tErqX0Q/M53BJBavumcT6PUdZ\nvv0wd85ayb+/O4pB3VomdK2gu15ZydHSSl78bBsATXKyKKmoYumvJtC6aW5S10yXPrfOJj8ni1X3\nTKrTcihVU0RkiTGmyC1dWvowjDFzgDkR5/5me22A//Ga1zq/HxiXjvLVByKCxOhS6NqqSej1U98a\nwdKth7h4cJewNL2KZzvmtQcLICxYAIz97Xthx6UVfp5fuIV7Xgt0xs9ZvjPmg/B4WSXFLy3nVxf2\np0PzfLYeOMHlf/2IPVYfDMD4R94PvV6wfq9rwPD7Df9ZvJXLhnUlLzvwjf3jDfv5x0ebw9JV+gNB\ncufhUg6cKKdv+2Zxr1uT/Ia4tT6lGgud6V3PdG9TEBUs0mnjvmPcMWtl6PhEeRXllX7eWbOHTzbu\n56E31mCMocpveOmzbbz6+Q4emfslAM8v3BIWLD7asC/s2s3zcxw/c/XOI6FJkC9+to1bX1rOEws2\nhd6/6vFPovJUWumn/HEB4yKCnlKqbjSIUVIq0KZ/tKySz++YyOB75gKBmkmbprks337Y83XmLI8e\nGtzvl6+HHbdrlsc/PtrMlgMngOr+kMimsasfDx/hVVYR/i38gddX07lFPne9uoqfTzqZK4q68/Dc\ntYG0Lp3ZkS2lD7y+muJJp0QNQjh8ooLB98zlyelFjOsfPvJ6x6ESXl66nR+O6RuWzxjDwRMVtKnB\npq4dh0roYqs5KpUJtIbRQHxy2zhW3H0+LQtyWPzL8Xx532Q+LB7Ls98JTFvp0DyP1350FoMdmoRu\nnXxKQp91z2urQsECAgHjeFklT3ywKU4uWLPrKPfPXkV5pZ8F6/by9/c2cpc1D2X5tsPc8sLn7D4S\nqKH8cf46Fqzb67lMf39vI5v2HefBN9Zwy3+qJ1Ou3BkIlo+9vzEqzw+eXcJv3lzLpn3Hw87/Z9FW\nht37Fmt2HQGgpLyKuUnMsYnl5aXbOGPG2yzaHLfLzVFJeRW9imfzktWXo1R9ogGjgWial00za+RQ\nu2Z5oZFDLZpkc+O4Qp797umc2rUls244i5+dfzIA3z6zN4t/OZ5vn9U7dJ1rR/fksqFdSWSKxt5j\nZQy8803XdP9dso3HF2zi1peW880nw2fA+0Q4fKI87Nw3n1zI51sPeS7HzEVb+eu7G3hpafWo7Yqq\nQFXEaSTVkdJKx+u892UgUG3YEwgkd76yguv/tYQV2w/zf0u3M+P1Na5lmbVsOyti1OyCAwwiA5UX\nu46UAoGAqlR9owGjgRMRbpnQj34dm4fOnXdyBwCG9mhFu2Z55GRV/zPfM/VUHrlyCG/dcm7YdS48\nrXPMzwg+AL160eHbsc8njsNSP1jv/dqRtYg3Vuzky11HAcLuMSjYlGYIDEGOHBEYbKUK1qaOlFZw\n83+W8bf3NriW5aaZy7jwT84TNo+XBQJVMMAnIlhmX6wREkrVIQ0YGWhAlxZ8cdfEuEGgb/tmbJ5x\nQVieoB+M6eupfb9D8zzPZXr18x1s3h/9jXvj3sS/hQfyHeP7z37G/XNWA5DrEDD8VoB4/P2N/ODf\nn/HGikCzU2T/SPDh7DbCPN7qxXbHywJ9OU2TCBjBoKbxQtVH2umdoVo4jFg6y7b8SaRvjOxJyyY5\nTD61M22a5vLe2r0cOF7umPZPVw3lR88vpWWTHI6WVlJS4W3IabD/ws6pNuJF5HBhpyapYADYaDUN\nHS4JzHI/aDWNBZ/JwYDhd4kYZZV+muS6T947ZtUwcpJYmiUYk7J8whMLNjKiVxsGJzhPRqmaojWM\nRmLNvZP4x3UjYr7fsiCHb5zeM1SzOC2i8/ybo3ry5PQiFv9yPN3bFACQl+Nj9b2TmPU/Z3LZMG/L\nn9TUyKS31+zh7IfeZtfhQB+A32/YfqgEgBPl1gM8y8eK7Yf51Jqr8pw1wz74bT6yArF82+GwZqwK\nv/PIrj1HS8PSBZuk7n51FRv2HotKv/dodODceuAEvYpns/irQNk27z/BfbNX83PbzH6l6poGjEYi\nPyeLbIdmmzsvGsDvrhwcdf6uiweGXn9532Tuvngg4/p3pF2zPNpYe4CcXRhYu2tw91ahiXVdXYaS\nxqq1pOpYWSVbD5SEmr2etI3oWrE9MBqqypiwfodg34w41DAWrNvLRX/+gN+8uTZ07qbnl4ZqKUFr\ndx1l5P3zefaTr0LngpP81u4+ynf+sSgs/Wtf7GDE/fOiJlgGO+L//UkgiHlZQ+tYWSVF982Lmg+j\nVE3RgNHIXXdmby4d2i3qvL2DOjfbF7akSY+2Bcy75Rx+OvHk0LngN+wCW5PNFUXR13UzfXTPhPPY\nTXssMAlwjdUZbhfr23rw1q57uvrhvtbK/5d3qzvA31m7NywQAazbE0j3q1krQ8Ek2CQFgZn1dsFA\nsWpH+AirYC2nosr7Yourdhxh37EyHrYFNaVqkvZhqJh+87XTaFXg3IR0UofmYcfBL+cFto7eaSN7\n8MJi9z6Kswvbcevk/gzo0oIT5ZU88/FXrnncNM3zvlCg04ik+2avdkybHdEvYV/k8cq/f8y2gyVh\nASMYAKr8hnteXclWa0RW5ATEyPRB8Tq/r/j7x1aZ9Hufqh36X5qK6etF3ZkwwNu+VcHnZoGtZtI0\ntzp4zP3xOTHz5mb5QqO0CnLDv8O0yM9mQOcWTtlieuz9DaH5GV4k0jednRWeuNL2OWt2HQ0LFlAd\nAFbtOMIzH3/FO2vjT1ZMZjn3dMWLyio/vYpn848P40/QVI2X1jBUWhgCD87gpk1DurcKa57q17E5\nb/34HCr9hnV7jnHj80tD7108JPbaWd8+qzcTB3Riyh8XeC7Lr+e4T7yzi/Vt30lkDWPO8p0xUgYE\nA5eX0VUAux06xN2ka6OsE9Zot4fnfsm3zuztklo1RlrDUGlx5YjunNShGVeNDOx62Dw/OyxgABR2\nbE7/zi1CneZBU4c4j7D6w7Qh3Di2sEZ3Dty87zgfJDAxcc+RMn47t7rPYP4ax329QoKr7rrFpFeW\n7QDCm7gC+aozllZU8c7a6M/zMsnv4PFy3nEpq/EHP9P1cqqR0hqGSovOLZsw75Zz8fsNN5x3Et8c\n3TNm/0fX1vFHUp15Uls+XL+fId1b4fMJZZU1t7T4mIffTSi923pakSqqDGWVVVz66Idh5yMfyp9u\ncl936u5XV/L8wq3889sjw/poFqzbhzEmbk3p+n8tZtHmg3x+x0RaFjivKhysJeoscxWLBgyVVj6f\n8NPzT46bpkur+PuS//2bRby1ahc92zYFoHWMwNNQ3D97dcx1rQAOnYg91Nj+6A6O3Lr2qYVR6coq\n/XF3BAzOqC+PMworWLnRreBVLNokpWpUm6a59G3fNOxcXnYW//rOSMB5vaVmedlhQ327tylgzo1n\nh6W5/pw+PHj5IM/lqMuH4KodR+K+fzROMAl+2V+29RCfbYm9UKPbBk/B2seI++exdMtBxzRu61hN\n+v37jLh/XtzPUZlNaxiqRi28bZxjU8nZhe1Zc6/3LU9P7lQ9jPfRq4cxZVAnVu8Mn2tx6dCuvGxb\nydbuzJPahSbqndKpueM8jZpypLQi6pz9LxJvSZKVO47E3G3R7nhZZdxZ9PaA+ernOxnao3VUGr/L\nOla1+TdT9ZPWMFSNys7yxey0zs/JituMYme/xgWndUZEotaPmn5Gr7Djx6+t3qJ4vG1zpZrcOMlJ\n5OxwCEz022MtZR7Z0Z2M4+WxaykQHgSOlzmnDZYjkVFjqnHRGoZqsPIiAkb/zs05p197bhx7EkW9\n2rBud/U3YvuSJbnZPs4ubJfwsu3Jclp0EeDL3cd4d+1ez0Nu43EKAu+s2UPz/GyKerVBbHWayLki\nQdVNUikXR2UorWGoBiuyhpGXncU/vz2Sol5tgPBaSb+OzUO1jK6tmqTlW32qKv1+fv7iF/zINicl\nWcfKovswrvvHIr72t8BscHulYfuhEnoVz46aQxKqYaARQznTgKEaLKc9MOzsS2ZkZQmPXzuch78+\nmF9eMIDKiIDRtpabqSA9TVFBpS5LzNs7soN7wEduSBX8m+w6UsqnG/enrWwqc2jAUA1G8/xsvnF6\nj9Cxz6XtxP6tOtsniAhfG96NJrlZUQ/r1lbAqM3m+8igldK1ElgKJVagsne+X2kt4hj0wuKtyRVM\nZRTtw1ANxvK7zg87bp6Xzeg+bZkwoCPnn9opKn1w3w6IHioa+dBskR/4X6FNQS77a2gJ9khutYJE\nVPr9fLpxP1k+CTXJ2XkJhPFqPPNW7U6leCpDaMBQDZbPJzx//ShPaSPXgIocyhocGdS1dRPXgDGw\nSwtWusyt8MJt7kQiyiv9oVrB5hkXRD38vczejhcwSpNYFFFlHm2SUo1CVlb8GkanloHZ55fEWNfq\noctPC7122ogqGRv2RO/Gl6zbX14RdnwiYphtqjWMsjTWhlTDpQFDNQpZLk1SAzq3YOFt47juzF6h\nc9NGdAfg15cO4grrNUTXVtzESh65LtW9Uwc6J/QgcskPe+3lq/3H+Wr/ibj5H5izmpmLYvdTlGkN\nQ5FiwBCRNiLyloiss35HTx8NpJskImtFZL2IFLvlF5EJIrJERJZbv8emUk6lIicPRgaMvGwfHVrk\nh01aC74MLsoXlGjA8Do5sW+HZgldNx77vIyfuewLvn7PUf7+/kaet/Y4d6IBQ0HqNYxiYL4xphCY\nbx2HEZEs4FFgMjAAuEpEBrjk3wdcZIwZBEwH/pViOVUjF/mQv9oabfXsd04HYFx/p42iAnkiV+7I\nSbBJymvA8InwYfFYbptySkLXj/TV/uMsta075XcZjXXtk9GLGUZKdMXgjzfsjzmjXDVcqXZ6TwXG\nWK+fAd4FfhGRZiSw3hizEUBEZlr5VsXKb4yxz2RaCTQRkTxjTOK7yyhFdA3jujN7c521SdDmGRc4\n5qmuYcS/lhuvy4Vn+YSurZpQ2LG5e+I4zv3Nu2HHVXHWqoLqVWrjKavwXsPYc6SUqx7/hIkDOvKY\nbXmWROw9WsbR0gr6tE9frUulLtUaRkdjTHC66C7A6WtaV8DeOLrNOuc1/+XAZ7GChYhcLyKLRWTx\n3r3xt79Ujc9z/+90rizqntT6SKEcUSOqErtOfo63/82CgSWyvyVVS+OscgvR2846SWSSYXDpkS93\nJ79Y4dkPvc3Y376XdH5VM1xrGCIyD4ge5A632w+MMUZEkp6J5JRfRAYCDwIT4+R7DHgMoKioqO7X\ne1D1yhl923FG33ZJ5Y1Vw4hXY7hlQj86tcznsqFd2bD3OCd1aMaQe+Z6+rxgzaW2NjAK3peXJrbI\nfpx4gkOW3SZWxlOaQI1G1R7XgGGMGR/rPRHZLSKdjTE7RaQz4LQH5Hagu+24m3UOIGZ+EekGvAxc\na4zZ4OFelEoridGHEesxePmwblx/Tp9Qn0VwSfZ4+13YBWsWvloeuxjrfg4cLw+t7BurVetoaQX7\njpXTu131nifByojXmlJJeRU+X2AtMFW/pfqf5isEOqWxfs9ySLMIKBSR3iKSC0yz8sXMLyKtgNlA\nsTHmQ5SqA8GmmsjnXqzn4G+vGBy3g/uyYc5zPIKCgSLdTVJuYu3H8chbax3P2139+KecF7HNrdtG\nTJH63/EGYyL6XVT9lGrAmAFMEJF1wHjrGBHpIiJzAIwxlcANwJvAauAFY8zKePmt9CcBd4jIMuun\nQ4plVSohN4/vxzWjenBFUfeId5J7oN8yoV/c90NNUrW8vnis7gn7eXuSu19dyaA73wSqFzIMz1fd\nJPXnt9dx3dMLee7TLVGr41bY5o7sPFyaXOFVrUopYBhj9htjxhljCo0x440xB6zzO4wxU2zp5hhj\n+hlj+hpj7veQ/z5jTFNjzBDbj1Nzl1I1pmWTHO67ZFBUrcH+PB/VJ3rdpljyc7KY/5Nzo843t9ax\nCjVJ2b6Zv/zDM+Jes0+7pnHf9yJWh7a9pmOvhDz94WaORgyZtV/Db8UBn8DDc7/knbV7ue3l5fzw\n35+F0iz56gCFt7/OR+vj70liXEZ4qdqlM72VSlDwOdqtdROe+tYIz/nysn30dRgmGpwjEhzJlciw\n3SpjeO1HZ3lOH+lwSQXbD5U4vmcvhtuD215bCNUwYjRJlVf6+e+SbQAscAkY6VzRV6VOA4ZSCWqW\nlwPAjWMLKcj1PpUpVv9GltV5EQwUYd/sXa7pN4aOLfI9lyHSbS8tj/meuJTjVlte+9IkVaGA4Xzd\nX7z4Bc8vDIy0t/efHDxezl/f3RAWnCqqdLRUfaIBQymPgs1PrQsCAePgicSWQY+1pEh2RKBIpM/b\n73ffSCqevUdjz4X1hTVJRYcM+1Ii5balQ4zLsNq31+yxpa0+f/N/lvHgG2vCVgKuqNQaRn2iAUMp\nj5777ig2/npKaLOlgycqEsofa/Jgli88UCTSJGWMidqqNhGxRkhBYnt7hzdJBfNHX6Cyyh+2zIi9\n72PD3sDqvfaaWOSiion4aMM+/vLu+qTzq2gaMJTyyOcTfD7h3H7tATjdqnHcedEAnkhgCYwlvxzP\nR8XV62kGA0Tw4Z1IwPAbyPEwU9uJMSZuwAhL6/K+vYYRXLvKaXjwhX/6IGxSnj3QBGs78ZqkVmw/\nzNYD8VfeDbr68U956I212nGeRrqBklIRurVuwuE4tYdTu7bky/smh77ZB9ek8qpts7yw42CACH7b\ntjcx2Z91Z57Ulg/Xh++17Tcmof05OrfMDw1h/WJb9JBYuyc+2ESl33DXxQNdFzCsqPJTWlHFHbNW\ncI4VUJ0qVGt2hS8XUmJbhj24Iq597avyiFVyL/zTB0Ds9b+cHC6poFVB7e/Znom0hqFUhPd/dh7L\n7oy5Gg1AQs1AD14+iG+O6hnz/SHdWwHQJDfQFBPczAnglE7VCxFmOUwBT3QQ0c3jCxNK/4+PNgPu\nNYyySj8vL93OC4u3MeP1NYC3mlKJw8ZMv3mjesKgvUmqMk7zVHmln1tfWs4u23yOZnmB78O7jugc\nj3TRgKFUBJ9PEl6RNp4rR/Tg3ktOjTrfskkO5w/syAOXDeKlH55B55ZNgPA2/KZ52fz60kGAc6e5\n1yaloESXZg9x+Zjl2w7zxbbAIoeJzPR2WjNqvq1T3N7HEe/B/96Xe3l+4RZ+Nat658ECKwAfOFY7\ne7Q3BtokpVQd+dxWixnWI3zvsd9+fTBdWgUCSIsmgf9NOzQPb8qCxAJGbpYv6Q5yt08ptg2xDc6d\n8DJj3a1TOxgwyiv9nP+7912v5/cbjpRWcNpd1Qs+Xv3Epwk1YanYNGAoVQ9dPrxb6PWUUzvz4OWV\nnNuvQ9Q2qm59C3Z52b6UhuB65Q/VMNzTxmtmguo+nPV7jnG8PPYmTsHP8hvDut3p2ytdhdMmKaXq\nOZ9PuHJEDwryoif+JdIilZfjIyfZGkYCH7T/eKAJyMsiipVV8a8b7AB3m/MSbP4yJLecyM7DJaze\necQ9YSOnAUOpBqJFfg5/+cawsHNj+3tfk7MgN5s8jzWMyJpIMit0eNm0qsylhhFscjtw3KUfIrh3\niUmurKMfeJvJf1iQeMZGRpuklGpApgzqTLZPOOOkdsy4bBBtmwWGi147uif/u3ib46ijoKevG8FB\ntwevpUWTbPZZncVVfuO6zasTL7HJrUkq2Lx1pLR6mLM9Dn20YR9dWjYJrR9sSGx3QJUYDRhKNTDr\nfz0l6tw9U0/lrosG0ue2OY55zi5sR9/2zfiiLP52rUH2GkbfGNd0s+Srg65p3Jqkgs9++2gqe1PX\n1Y9/CsDj1sRJY4xO1KtB2iSlVIaINyopuJtdvGG1VxRVd7TvSMP+FPs8DGd1W1wwWFsoKa9eTl0E\nth44waZ9x0PnZlrrWiXbJKW80YChVIb6xaRTuHfqQACG9QxMDow3rPahrw2Oe70/XTWUq0b2SF8B\ncR9WG+zDsI+QqqgynP3QO2E7/QVniRuclzvRZqr00CYppTLILRP68fj7G1l+9/kAlFZUUV5luHZ0\nYKZ5rGG1vSM2YnrkisHc8sLnYedG923rqZkpEW41DL8xvL58p+vnVlq7Nn24fj9j+kUPBCiv9Idm\n0tvZF0JU7jRgKJVBbhxXyI3jqpf/yM/J4jtnVa915VTD+MO0IZxT2D7s3GBruRK7ds3yyMtJb6NE\nhUsfxuGSCm54bqnrdew1iPvnrI56P1bAGHTXXHJqeUvchkwDhlKNiNPyIlOHdI06165p9KxyqO4L\nSRe34bJeggW478wXq+mrvNKPLhzinfZhKNWIZFsLGAYX5oulWb7z+14m49UFt9FWwYBhjOHJDzax\n/1jsjaNUbBowlGpEmlqzxW8aF3/V2iyfML5/x6jz9bX1xq0vJDjfY+WOI9z72ipu/s8y12tuO3gi\n5n7njZUGDKUakewsH5tnXMD/O6ePa1qn0UZeFhSsC5H7bEQK3kqw6epwiftuiWc9+A5nzng75bJl\nEg0YSilHTv0CXpYsd/OTCf1SvkaiDHDd0wu55NEPgcDmUb2KZ9d6ORo6DRhKKUdV/uhmnnQsdntK\n5xapXyRBfmN4Z+3eWv/cTKMBQynlaHjEHh3gXMMY3L0VP53ovdaQ7B7kqdDVQtJDA4ZSytFN4/vx\nwvdGh51zWoF27MkduGGs961fa2NPjki6vlR6pPQvJyJtROQtEVln/Y7+ShJIN0lE1orIehEp9ppf\nRHqIyDER+Wkq5VRKJS7LJ5zcsXn4OYfKQaVD01U8ye76lwoNF+mR6r9cMTDfGFMIzLeOw4hIFvAo\nMBkYAFwlIgM85n8EeD3FMiqlYji1a3h/wrQR3Zk0sFPo2BfxhHAaJeW2HlSk7DqoYSS697mdfWn1\nxi7Vmd5TgTHW62eAd4FfRKQZCaw3xmwEEJGZVr5V8fKLyCXAJuA4Sqm0+/yOiVFLfcy4/LSw46yI\nAOHUh1FRmdjD2Kkz3U1eti+0wGAyUmmROu2uubonuCXVUN/RGLPTer0LiJ7pA10B+0bE26xzMfOL\nSDMCgePuFMunlIqhZUEO+Tnxl/qIDBBOAaO8KrEF/MoqEn/wr71vMm/efE7C+YJSqWGoaq4BQ0Tm\nicgKh5+p9nQm0KuU9L9KRP67gN8ZY1x3cxeR60VksYgs3rtXh80plU6RAcKpNSnRGkbk9I7IfpJY\nTu7U3HGTJX0CAAAX/ElEQVT2uRepxgtdHj3ANWAYY8YbY051+JkF7BaRzgDW7z0Ol9gOdLcdd7PO\nESf/6cBDIrIZuBm4TURuiFG+x4wxRcaYovbt2zslUUolKbJJymmUVLxlOb5rWyk36Iy+bcOOE+kE\n/+s1w0IbPRU4rD4bS6oBozyF5rBMkmqT1CvAdOv1dGCWQ5pFQKGI9BaRXGCalS9mfmPM2caYXsaY\nXsDvgV8bY/6cYlmVUgmK7ON2bpIKf5g2ty1s+P0xfaOvGXHRvAQCRk6Wj9ZNA/uYJ7IQotcmqVjD\nbzVgBKQaMGYAE0RkHTDeOkZEuojIHABjTCVwA/AmsBp4wRizMl5+pVT9EFmjcFpKKnKl2Hd+Nib0\nOlbtwT7RLzvBiXw51tCtrATyea1gxGp5KkuwnyZTpTRKyhizHxjncH4HMMV2PAeI2kk+Vv6INHel\nUkalVPrY48cz3x7JzTOX8r1zAwsZvvPTMeRl+2jXrHovjViT9G4YW8j6Pcf4v2U7EOI/+CNX1g02\nkznt7RFLYjUMh1pUjBqGMYZfzVrB5cO6MdRhZnym0ZneSinPgs/dS4Z04dx+7Vl6x8TQg7J3u6Z0\nadUkLL19FNb3zunDN06v3hM81iP8smHhGzr9OGKxwmCgSKRfwmvaWDWMWAGjtMLPs59s4arHP/Fe\nmAZMA4ZSKmFOnd+xdLWCyEWDu3D/pYNC52M9xIf39PZNPZF+bK9Lg8SqicSanBj8MzSWQVQaMJRS\nnqUy2ihyxFWsS7k9fCusBIUdmoWdn3n9qJh5Un2eT/r9AvbZdunbc7SUvUfLqv8eHj9gza4jfLk7\n/t4d9ZkGDKVUwhLppg5+C4+1l4aJeNr6XSJGcBjvOf3Ch9EP7dEqZh63awYdLa1k/R7nB/qH6/eF\nXo+8fz4j7p8XKrvfGMoqq3jmo81x52xM+v0CJv7ufU9lqY80YCilPEvmm3p1wAg/P8Z64J/SKXw9\nK7cO6gqrPyFymfS87Czm/+RcxzxfbDvsqazf/edixj/yfswmLL/fhM07mblwa6jMf313A3e+spIX\nl2zz9FkNUaprSSmlGpHQgzSJLS0i519cPrwb4/t3ZM6KnWHnXZukqoIBI/r7bt/2zaLOAdw/Z7Wn\nMn6+9RAAC9btc3z/+n8tZt7q6vnJTyzYCATKHNz2NZMXK9QahlIqYW5DYe2CTVFOOVoW5ESdc+ug\nLrfmfTgFjHT53bwvHc/bgwXAjsOlodfBv8m63cdiNms1dBowlFKeJdUk5fL+pUO7MmFA9RpRbus2\nBWsYseZ4fFQ8lvd/dh5/u2Z4QuW0KylPfKJesAL1n8VbGf9Iw+2niEcDhlIqYQmMqg2JFQbyc7J4\n9OphoeMqlxpG8KMjl2YP6tKqCT3aFoQFoUSt2ZV4DSGZv0lDo30YSinvkqhiBJuk4sUBe/eG29Dd\n4smn0DQvm8mnduYmlnm6Zjos3nww7vuJzE1pqLSGoZRKmNuj0Se21WRDiWNHAvuQ2wtP6xz3Yd+2\nWR53XTzQdZXbdD/Adx8pjft+5ocLrWEopTxokR94VETOmYhl1T2TQq+DD9J4NQf7s71n26ZsfOAC\nehXPTrSYNWruqt1x328MNQwNGEqpuBbeNo48a02o0Khal2ejfQ2p4IM0XqjJhIdtBtyCK22SUkrF\n1aFFPi2bhA9/TWRYbfGkU2iel0331gXpLpqr8wcm3/GdqEYQLzRgKKW8S2ZY7fgBHVl+9/k08bBD\n3oheyS0RPrpPW8fzPxpb6Hi+JjSGGoY2SSmlPOvXMTCTeniSD/Z43v7JuXRskZ9QnmkjujNz0Vae\nvm5E2suTqERqXQ2VBgyllGfDe7bhg1+cF1qyPJ36xFjWI54Zl5/GjMtPi/l+rAUPa0K6h/HWRxow\nlFIJ6VYHfRHJqs1mosawJ4b2YSilMlZt1jD+/M76WvusuqIBQymVsSLjRWNoNqpJGjCUUhnLHiAm\nDuhIYYfmdVeYDKABQylVL2WnpTpQfY3bL+jP76cN4ZROzSmefEoart34aKe3UqpeWnbnRNfd99z4\nIpYcAXjj5nMAmHJqZ46VVTLljwtS+oxYRj8wnytHdOfm8f0Awnbqa6i0hqGUqpea5WXTIj96g6VE\nxFtypEfbAgZ0aRHz/VTtPFzK7+etCx0/9MaaGvus2qIBQymVsepDJ/fhE4EtW1dsP1LHJUmdBgyl\nVMaqydnXWR6j0btfBrZ19brSb32mAUMplbFqchpGrC1io8sQKETkxL5KW5/GjkMlrN9zLG1lqykp\nBQwRaSMib4nIOuu34wIzIjJJRNaKyHoRKfaSX0ROE5GPRWSliCwXkcQWmVFKNXo1GjBcNnAKevTt\n9ew5Uhq2cuOc5Ts56fbXQ0HijBlvM/6R92qimGmVag2jGJhvjCkE5lvHYUQkC3gUmAwMAK4SkQHx\n8otINvAs8H1jzEBgDFCRYlmVUo1MTc70zvMYMNbuPspP/vfzsCap11fsAmDljsM1UraakmrAmAo8\nY71+BrjEIc1IYL0xZqMxphyYaeWLl38i8IUx5nMAY8x+Y0xVimVVSjUyqcQLtz4KrzUMgJLyzHh8\npRowOhpjdlqvdwFOu5V0BbbajrdZ5+Ll7wcYEXlTRD4TkZ+nWE6lVCOUTA2jX8dm3HnRAPq2bxo3\nXSIBI8sncbeobShcJ+6JyDygk8Nbt9sPjDFGRJL+k0TkzwbOAkYAJ4D5IrLEGDPfoXzXA9cD9OjR\nI9mPV0ploGQqGG/cdA4+n/Dcp1vipvPa6Q2QnSVU2nq968Fo36S4BgxjzPhY74nIbhHpbIzZKSKd\ngT0OybYD3W3H3axzALHybwPeN8bssz5nDjCMQD9HZPkeAx4DKCoqyoAYrpRKl2T2Cvf53PcgB+99\nGABZPh+Z0KqeapPUK8B06/V0YJZDmkVAoYj0FpFcYJqVL17+N4FBIlJgdYCfC6xKsaxKqUamPoyS\ngsC6WJnwbTbVgDEDmCAi64Dx1jEi0sWqFWCMqQRuIBAEVgMvGGNWxstvjDkIPEIg2CwDPjPGzE6x\nrEqpRiaVUVLGpdOhtML72lBZPmHplkNJl6W+SGnxQWPMfmCcw/kdwBTb8Rxgjtf81nvPEhhaq5RS\nKWmWF/tR98erhjJr6XbmrwlvUQ+Gi19e0J/7Zq+Oyrd6p/elPiIXHly46YDnvPWJzvRWSmW8ol6O\nc4oBuHhwF5781ojoN6yIMapPW8d8vgQWqtp3rCzseNeRUgCOl1VRWtFw+jZ0eXOlVMZq0zSX5/7f\n6Qzu1irpa8Rq1UpkYcM9R8ocz9/28nIenrs2iVLVDQ0YSqmMdkbfdp7SfVQ8lsqq6n6L4KtY/SCJ\n9I9E1jDsDhwv93yduqYBQymlgC6tmoQdBzu90xEwIhcebKi0D0MppRxU1zCc368Pe23UNg0YSinl\nIDiqNtbkv2QmBTZ0GjCUUspBcHVZrWFU04ChlFIOgjWMWH0VZ57krTM9k2jAUEopB24BY8blp9Vi\naeoHDRhKKRVHrK6Kgpys2i1IPaABQymlkpDITO9MoQFDKaUcBOdhJDsYKhPjiQYMpZRyEJyHkezw\n2Wxf5j1eM++OlFIqDULzMDyk/cO0IVHnMjBeaMBQSiknwXkY8SoYD399MAD9OjZPy2cu+eoAf5i3\nLi3Xqgm6lpRSSjmormHEjhhfG96NCf070rIgJ+q9prnZlFYktrDg5X/9GICbxhcmlK+2aA1DKaUc\nVPdhxE/nFCwAWjfNTf6zXXb7qysaMJRSKg4BFt0+PuF8rWMEEi8q6+nythowlFLKgf1LfvvmeWHv\n9W3fNCp9i/zwFv7pZ/RK+rP/d/E2Zi3bHjo+UV7JifLKpK+XLhowlFLKQddW+QDkZEU/Jq8a2SPq\n3PyfjOHGcYG+h/H9O3LhaV2S/uzbXl7OTTOXsfXACQAG3PEmg+6am/T10kU7vZVSysGT3xrBJxv3\nO/ZFOM3NaN88j6E9AlvBllf501IG+37fVfWgmUprGEop5aBds7yEawl52YFHanlllUtKb+rbnhta\nw1BKKQ+uLOpOWWUVX+4+xtQhzoEkGDDKKtNTw1iz6wgndWiWlmulgwYMpZTy4MGvuS9nnpsVWMG2\nrCI8YHxUPJb1e45x7VMLE/rMG55bypRTOyeUpyZpwFBKqTTJy7GapCL6MLq0akJ5krWOuu+5qKZ9\nGEoplSYFuYEaRuQQW0h+1Vt/PZrEpwFDKaXSpFvrAn596SD+ds3wtF3THjD8fsOgO9/kP4u2pO36\niUgpYIhIGxF5S0TWWb9bx0g3SUTWish6ESl2yy8iOSLyjIgsF5HVInJrKuVUSqnacvXpPejQIj/q\nfLw1qeKxVzC+/+wSjpZV8qtZK5MtXkpSrWEUA/ONMYXAfOs4jIhkAY8Ck4EBwFUiMsAl/9eBPGPM\nIGA48D0R6ZViWZVSqs50aJHnnsiBvYYxd9XudBUnKakGjKnAM9brZ4BLHNKMBNYbYzYaY8qBmVa+\nePkN0FREsoEmQDlwJMWyKqVUncnPyeKlH56RcL56MF8vJNWA0dEYs9N6vQvo6JCmK7DVdrzNOhcv\n/3+B48BOYAvwsDHmgFMBROR6EVksIov37t2b/J0opVQNy0liV6X61OntOqxWROYBnRzeut1+YIwx\nIpL0nUXkHwlUAV2A1sACEZlnjNnokO8x4DGAoqKi+vOXVUqpCMnswuevR1UM14BhjIm5rq+I7BaR\nzsaYnSLSGdjjkGw70N123M06BxAr/9XAG8aYCmCPiHwIFAFRAUMppRqK4NpQp3VrSW6Wj8VfHXTN\nU4/iRcpNUq8A063X04FZDmkWAYUi0ltEcoFpVr54+bcAYwFEpCkwCliTYlmVUqpOtWsW6Pi+fFg3\nmlhzNtzUpyapVAPGDGCCiKwDxlvHiEgXEZkDYIypBG4A3gRWAy8YY1bGy09gVFUzEVlJIOA8bYz5\nIsWyKqVUnerZtilLfzWBa0f3xGscWLXDYbxPHcWQlJYGMcbsB8Y5nN8BTLEdzwHmJJD/GIGhtUop\n1WD96zsjKSkPX7k2uFy68fjUT3T9qZqkM72VUqqGnF3YnokDncYM4bmGEcu2gye4+M8fsP9YWWoX\nSoAGDKWUqgOp9k08sWATX2w7zPD75vHJxv1pKlV8GjCUUqoOOMWLF38wmvwcb4/lCtuKuD954fN0\nFSsuDRhKKVUHIuNFXraP4T3beF5zyr5la22NpNKAoZRSdcBEPOSDR16WQTcYKm0Bo7b2+9aAoZRS\n9YH1zPe6pm14DSP9xXGiAUMppepAZCtScJiteNxpyd6HoU1SSimVwSIf8Yk+87UPQymlGonIh3zw\n2GuTlPZhKKVUIxHdJGXx0ultoNLWJFVby01pwFBKqToQq0lKaxhKKaXCxagWtLHWmnKjfRhKKdVI\nxKoUnFXYzlP+korqRQ01YCilVAaLtVrtHRcO5Idj+rrmX7rlUOi1zsNQSqlGKDfbx0WDu8RNUxkR\nIbQPQymlMli8ViSfx8l7tU0DhlJK1YF4lQJf/YwXGjCUUqouRC4+aFdPKxgaMJRSqr6prYl4idKA\noZRSdWBUn7Yx36utUU+J0oChlFJ14PYL+nPDeSc5vldbo54SpQFDKaXqQE6Wj17tmjq+V1sT8RKl\nAUMppepIrI7vehovNGAopVRdsceFM/pW92loDUMppZSjM/q25cnpI0LHPdoUOKY77+T2rL5nUm0V\nK4oGDKWUqmNdWjWhSW5W6Lh101w2z7iAF39wRli6H0/oF5bO7mhpRY2WEVIMGCLSRkTeEpF11u/W\nMdJNEpG1IrJeRIpt578uIitFxC8iRRF5brXSrxWR81Mpp1JK1UsuLU/De1Y/Uvu0b8pp3VrFTHvv\na6vSVaqYUq1hFAPzjTGFwHzrOIyIZAGPApOBAcBVIjLAensFcBnwfkSeAcA0YCAwCfiLdR2llMoY\nwRVrPU3sdgkux8ur4idIg1QDxlTgGev1M8AlDmlGAuuNMRuNMeXATCsfxpjVxpi1Ma470xhTZozZ\nBKy3rqOUUhnHy1Igbt3gedk138OQ6id0NMbstF7vAjo6pOkKbLUdb7POxZNMHqWUalDO7deBLJ/w\nzVG9XNPGW3sKaidgZLslEJF5QCeHt263HxhjjIjU+lgwEbkeuB6gR48etf3xSimVtE4t89nw6yme\n0ro9XHOz6kHAMMaMj/WeiOwWkc7GmJ0i0hnY45BsO9DddtzNOheP5zzGmMeAxwCKiorq5+BlpZRK\nkdvUjCxf/W+SegWYbr2eDsxySLMIKBSR3iKSS6Az+xUP150mInki0hsoBBamWFallGqwYm3pGlQL\nFYyUA8YMYIKIrAPGW8eISBcRmQNgjKkEbgDeBFYDLxhjVlrpLhWRbcBoYLaIvGnlWQm8AKwC3gD+\nxxhT80MAlFKqnmqelxP3/bbN8mq8DK5NUvEYY/YD4xzO7wCm2I7nAHMc0r0MvBzj2vcD96dSPqWU\nyhRPfqso6tyCn59Hq4IcBt01t1ZWuE0pYCillKodnVs2iTrXoUUeWdaYXA0YSimlYsr2+UL7f1fW\nQsDQtaSUUqqBKZ58CgA+AREhyyf4tYahlFIq0vfP7cv3z+0bOs7yidYwlFJKucv2CVV+f41/jgYM\npZRq4LJEaxhKKaU8yMqqnT4MDRhKKdXAZWsfhlJKKS+yfFIr8zA0YCilVAOnfRhKKaU8qa0+DJ2H\noZRS9dhPJ/Zj074TcdNk+3y1UsPQgKGUUvXYDWMLXdOMObk93VoX1HhZNGAopVQDd+dFA2vlc7QP\nQymllCcaMJRSSnmiAUMppZQnGjCUUkp5ogFDKaWUJxowlFJKeaIBQymllCcaMJRSSnkixtT8dPLa\nIiJ7ga9SuEQ7YF+aitMQNLb7Bb3nxkLvOTE9jTHt3RJlVMBIlYgsNsYU1XU5aktju1/Qe24s9J5r\nhjZJKaWU8kQDhlJKKU80YIR7rK4LUMsa2/2C3nNjofdcA7QPQymllCdaw1BKKeWJBgxARCaJyFoR\nWS8ixXVdnnQRke4i8o6IrBKRlSJyk3W+jYi8JSLrrN+tbXlutf4Oa0Xk/LorffJEJEtElorIa9Zx\npt9vKxH5r4isEZHVIjK6Edzzj63/pleIyPMikp9p9ywiT4nIHhFZYTuX8D2KyHARWW6990cRkaQL\nZYxp1D9AFrAB6APkAp8DA+q6XGm6t87AMOt1c+BLYADwEFBsnS8GHrReD7DuPw/obf1dsur6PpK4\n71uA54DXrONMv99ngO9ar3OBVpl8z0BXYBPQxDp+AfhWpt0zcA4wDFhhO5fwPQILgVGAAK8Dk5Mt\nk9YwYCSw3hiz0RhTDswEptZxmdLCGLPTGPOZ9foosJrA/2xTCTxksH5fYr2eCsw0xpQZYzYB6wn8\nfRoMEekGXAA8YTudyffbksCD5UkAY0y5MeYQGXzPlmygiYhkAwXADjLsno0x7wMHIk4ndI8i0hlo\nYYz5xASixz9teRKmASPwAN1qO95mncsoItILGAp8CnQ0xuy03toFdLReZ8Lf4vfAzwG/7Vwm329v\nYC/wtNUM94SINCWD79kYsx14GNgC7AQOG2PmksH3bJPoPXa1XkeeT4oGjEZARJoBLwI3G2OO2N+z\nvnVkxFA5EbkQ2GOMWRIrTSbdryWbQLPFX40xQ4HjBJoqQjLtnq12+6kEgmUXoKmIXGNPk2n37KQu\n7lEDBmwHutuOu1nnMoKI5BAIFv82xrxknd5tVVWxfu+xzjf0v8WZwMUisplA0+JYEXmWzL1fCHxj\n3GaM+dQ6/i+BAJLJ9zwe2GSM2WuMqQBeAs4gs+85KNF73G69jjyfFA0YsAgoFJHeIpILTANeqeMy\npYU1GuJJYLUx5hHbW68A063X04FZtvPTRCRPRHoDhQQ6zBoEY8ytxphuxpheBP4d3zbGXEOG3i+A\nMWYXsFVETrZOjQNWkcH3TKApapSIFFj/jY8j0D+XyfcclNA9Ws1XR0RklPW3utaWJ3F1PRKgPvwA\nUwiMINoA3F7X5UnjfZ1FoMr6BbDM+pkCtAXmA+uAeUAbW57brb/DWlIYTVHXP8AYqkdJZfT9AkOA\nxda/8/8BrRvBPd8NrAFWAP8iMDooo+4ZeJ5AH00FgZrkd5K5R6DI+jttAP6MNWE7mR+d6a2UUsoT\nbZJSSinliQYMpZRSnmjAUEop5YkGDKWUUp5owFBKKeWJBgyllFKeaMBQSinliQYMpZRSnvx/hysj\npLgy5v4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2c28bd7a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training epoch 1100\n",
      "Iter: 1100\n",
      "D loss: -0.0229\n",
      "G_loss: -0.01219\n",
      "()\n",
      "[-0.73315543  0.77764291 -0.81948227]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAD8CAYAAABkbJM/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2c28b20210>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAD8CAYAAABkbJM/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3XeYVOX1wPHv2dnC0tuCCwgLSldAXClREQQUMAnGFjUa\nk5hgoqaqcQ2plgQ1MSbRmJgYgyZKiJWfoIiIokalifQO0nsv28/vj7kzO+VOn+3n8zw8O/fe973z\nXsF79u2iqhhjjDGxZNR2AYwxxtQPFjCMMcbExQKGMcaYuFjAMMYYExcLGMYYY+JiAcMYY0xcLGAY\nY4yJiwUMY4wxcbGAYYwxJi6ZtV2AdGrfvr0WFBTUdjGMMaZeWbx48X5VzYuVrkEFjIKCAhYtWlTb\nxTDGmHpFRD6LJ501SRljjImLBQxjjDFxsYBhjDEmLhYwjDHGxMUChjHGmLhYwDDGGBOXtAQMERkn\nImtFZIOIFLlcFxH5o3N9mYgMjpVXRNqKyBwRWe/8bJOOshpjjElOygFDRDzA48B4oB9wnYj0C0k2\nHujp/JkEPBFH3iJgrqr2BOY6x9Vq79FiZq/cTVlFJTsOn6K8otI13eb9J5i3dm91F8cYY+qUdEzc\nGwJsUNVNACIyDZgIrApIMxF4Rr0biH8kIq1FJB8oiJJ3IjDSyT8VeAe4Ow3lDbP3WDFPvLORpz/Y\nEnS+WbaHKwZ34dmPPqNN0yweumogqsqkZxcDcPuoMzn/zPbsO17CkIK27D5aTF6LHJ77+DO+e3FP\nmmR5wr5LVRGR6ngMY4ypVukIGJ2BbQHH24GhcaTpHCNvR1Xd5XzeDXRMQ1ldHTheyr8/3hp2/kRp\nBc9+5J0AeehkGd96JngW+WPzNvDYvA2u93x83kYAsjMzeOP7F/LJ1sNsO3SS/y7aztw7LvIHk+Ky\nCvYdK6FLm1w27jvOmR1apPPRjDEmberF0iCqqiKibtdEZBLeZi66du2a1P375rfko3tG07ZZNgAf\nbjxAfqsmPPHORr44qBOVqryzdh9Pvb854XuXlldy8e/eDTrX52dvhKX74sBOzPh0J09//TxG9e7A\nU+9vZmj3tpzVuVXU+y/bfphu7ZrRKjcr4bIZY0wi0hEwdgCnBxx3cc7FkyYrSt49IpKvqruc5ivX\nTgNVfRJ4EqCwsNA1qMTDFywAhp/RDoAHrxrgP3dhzzwmT+jLm6t2c263tuS1yEFV+cWMlWzcd5wP\nNhxI9qsBmPHpTgC+/vTCoPP3TexPXosmbN5/gu+MPAOAT7cdxpMh9M1vyRcf+4DCbm144Tuf8+f5\n7MAJ5qzaw80XdK9TzV8zPt1JXvMcth48wYodR7nv8rNqu0jGmASkI2AsBHqKSHe8L/trgetD0swA\nbnf6KIYCR5xAsC9K3hnATcAU5+eraShrSjIyhHFn5fuPRYR7J3pfesVlFZwqreCuF5bxwJfOokmm\nB0UZdO8cAB798iAAnv5gM59uPxL3d/7s1ZX+z706Nicn08MNT30MwMLJYwBY9NkhTpaW0zQ7k9W7\njjL+D+8BMHFQZ/Ja5CT1rMdLyjlWXEZ+q1zA2/dysrSCZjnJ/5P53vOfBB3XVsDYsPcYXds2IzvT\nRpUbkwjx9kOneBORCcCjgAf4h6o+ICLfBlDVv4j319zHgHHASeDrqrooUl7nfDtgOtAV+Ay4RlUP\nRitHYWGh1rXVak+UlJMhQm62t8+irKKSikrleEk5p0oraN88h4H3vklpeSU9OzRn/d7jcd97VO88\n5q3d5z++deQZ/Pmdjf7j1757AWt2H+MXr64gr0UO79w1yvU+FZVKWUWlv19l95Firv/bR2zaf4It\nUy7juY+38tjb69l5pJg1941z7cyP5uHZa/x9OoG2TLksofukw75jJZz3wFtcN+R0fnPFgNgZjGkE\nRGSxqhbGSpeWPgxVnQXMCjn3l4DPCtwWb17n/AFgdDrKV5tCfyPP8mSQ5SHopfvuXSM5UVIe1OE9\n7tH5rNl9LOq9A4MFEBQsAD7adID7Z64G4MSBk/4RWqXllew+UkzXdk0BuP25Jby+YjdbplyGqjLs\nN3P99zhRUs5PXl7uPz5yqixqwHhpyXYembOO+XeNIiNDeHvNHtdgAXDPS8u4tP9pjOzdIepzptOp\n0goA5q/bX2PfaUxDYXXyOiC/VW7Y6KiOLZsAcHGf5F+mvmDh0/2eWRQUzaTXT19nxMPzWLv7GA/M\nXMXrK3YD3manwyfLgvIU3v9W0PFJ54ULsGrnURZtOcige9/kzZW7OXKqjB9N/5Tth05xqsybbvrC\n7RHL9/yCbXzt6YVsO3gy7FpJeQW/mbWa4yXlYdc+2nSAkvKKsPPxyPSIc3/3OTbGmMgsYNRRv7tm\nIPdO7M9TNxXyu6sH8tPL+rJw8hiG9/B2yA/sEn30VDwufXQ+f3uvauTX0eJyHn1rXVAa34vf54XF\n23h//X62HTzJhD++x1V/+ZDDJ8uY8sYarnrif/50X/n7xxw5VRaW382FD81jy/4TfH/aJ/5AMG3B\nNv46fxN/Dhm2vHrXUa598iN+M2tN0PmfvrKcfj+vGn22aMtBDp8sDfuuSqcJtjTOgPOFP73v7xNK\nxPenfcK0BeFDtY2pz+rFsNrGqH3zHL46vACAK8/t4j//yJcHct9rq3j4qoE0y8lk6bbDXP74BwA8\n840hFLRrxoiH5wHQvX0zWuZm8em2w3F9503/WMDSGGkfn7fRtYkpQySo/2XptsPcMX0p767bF5bW\nzfg/vMepsgpuGNaN8wra+msWoT1s+4+XALB+b3Bz3b8+qno5qypX/eVD+ndqyXdGnsHf39vMK7ed\nD0ClU7EIrGHMW7OX7YdPceOwbmHlWr4j/gEKgV5dupNXl+7k2iHJDfU2pi6yGkY9k98qlz9/5Vx/\n38ig01vTuXUu3734TEb0yqNZTlX/wrw7R/Lqbedz16W9/ecGRKmZxAoW0WzefyLs3OLPDsWd31cT\naZadya9nreZfzoTJbE/wP9FKJ4IUl1Xyw/8sZe+x4rB7lVV4E63ceZTbn/uEpdsOU+FkrHBqGIEB\n4+v/XMjPXlkRd1mNaawsYDQAHxRdzB2XeIOCbzRW++ZV80puG3Umn3PmllxwZnv/+Ved37rdZCQ4\nfcP3Qg50KKQ/JB73vLycJ+dvYtcRbyDIyQoNGN7vWfzZIV7+ZAe/n7M+7B5lLmuAlToBwq2c1emz\nAyc44dIPY0x9ZAGjgWmanclPL+vL9FuGB53f7byAR/ftyNr7x7FlymUMPL11WP7cLA9Du7flvbsv\nBiDLI9x/+Vmc261mFgsObT57Z+0+/vJuVRPYkZAglOURFm6pGm2952gx5RVVQcE3b7G0vJKDJ0pj\nBozAYeaVAWnfWx9f09qzH25hzqo9/uOLHn6He/9vVeQMxtQjFjAaoG9e2IMeec2Dzt1xSW86tMhh\nQJdW5GRWNVsVje8DeGskz3xjCKvvG8d/bhlOa2epkWE92nHDsG68GDCTvCYt2HyQKa+v8dcafvCf\npUHXi8squPovH/qPb/rHAsoqq2oYvorSy59sZ/B9c5j64Rb/tf8sDO6U3rL/BN3vmcXMZd4lzI4F\n1AxufGpBUNrpi7ZRUDTTH4h9fvbqyrA1x6I19X372cW88knowgjG1E0WMBqJywbks2DyGLJC+gQu\n7OltojqnaxtG9Mrzn2+Wk8l/vz2cP39lMJEk2mwVeP9E3f3CMtfl5qcvCh62u+tIcVCTlK+S8IRT\nS3kuYJHJu19cHpR3mdPBfdtzS9h+6CRbD4QP9/V5aYn3ezftjz3R0jeUN9Sx4jLeWLk7LAgaU1fZ\nKKlGrl9+S+6b2J8vDuwcdu28grZR8/5kQt+wuR5uPii6mNNaNqGiUun109eTKudLn+zgZ58P3WYl\nnKoGNUn57DlaEjNvRUDN5IIH57mmKauopOfkxJ4h0nJeoTWRVGzad5zjJeUM6BLezGhMulgNo5ET\nEW4cXkCrpomvdnvzBd39nxdOHkPvji343uieYek6t87FkyFhazcFjt6Kxzn3zYmZRoHSCBtfxXK8\nJPrcjMpKTWrCn+AeMRIZRRbLxb97ly8+9kHa7meMGwsYJm5DuwfXOAJXwm3fPJvZPxzBj8b2oml2\nVR/J4K7uv/FednY+37qwR/oLqfDu2vg6qKFqlWAg5tDakvLKqJ3mR07FHhW27eBJ9h719nvEu4yb\nqvK/jftJx7pvxqTCmqRM3P71zaGUVVQyY+lOth86FXQtOHjksNVZ7iO036Jz61x2HD7Fg1cNqJbV\nYo+VlHPva/GPSgpdQTeajzYf4MXFwX0mgbWHbzs7MYYKbJK68CFvU9dDVw2gPCD4HC8pp3mElYBf\nWLydu15YxiPXDOSKwV1c0xhTE6yGYeKW5cmgaXYm1w7pyp1RmpPGn3Wa/7MnpAF/xu3n88frzvG/\nHMf0rbaNFNPu608v5LVluyJe/2iz+54obg1SP35hWdDxsu2RR1L51tr676LtEfeZj2bB5oMcLU58\nTowxoSxgmJQ8eOXZQX0ZAD8e14cnbzyXTq2acMW5wb8Rt2uewxcHdvIf/+WGwTQLaMK6fmjXqCOz\n6rJILUZbDpxk3pq9FBTNjJi3pCxyIPDV3j7cdIApr6+JmM7N8ZJyrvnrhxFrP8YkwpqkTEq+fF74\nWkmeDOGS/qdxSf/TXHIEy/RkkJ2ZwYnSCt74wYX07tgibE8QT4a49h2M6JXHfGetqrbNsjl4Inyx\nwer2yJy1/PcM7xyVSOU8cqqMr/9zYdj5QNEWacwIqKV9kuDyLb4aycqdRxPKZ4wbq2GYWud7xbbO\nzUZEwtaP8o2m+u+3h7PmvnH+831Pq1oS/trzTqc2LNxyiBMl5Tz1/uaUOqWLQwJGcVkFz3y4hcpK\nDZrvcizBpiVfH4t1mJt0sBqGqXW+JTh8neChneG3jOjBzRd0D5t0eHaXVrTIyeRYSTld2jStmcK6\n+PWs1fz749SWMg+tYfx+zjr+On8TbZtlB3War9tznIKimTz/rWH+veejUSccW7gw6WA1DFPrcpwd\n/HwzojNDppCLSFiwAO9Ktp/+4hL+/JXBtVbDAO/6VakKbcryLb5YXqFBI9B83ljh3vl+xZ+D52Ks\n9e3aaBHDpIHVMEyte+6bQ5m5fBctnJFTOQFbwN468oyI+TI9QkaGMOHs/GovYzTp2L2volI5fLKU\nJlkemmR5/DWOJlmeiDPF3SzZGtzH8eUnPwIsXpj0sIBhal3Pji34Qceq/ohWuVm8dOvn6N2xRdie\n6AAje+fxztp9cU98q26h/Q/JKK9QBt07h/6dWjLzexf675mb7Qnq9PZJ9NGtD8OkgzVJmTppcNc2\nrsECqjZVKnNZMypU6whLntw2yltzSXQBRTefbk9uV75AD8zyrsnlG810ytk7PcsTaWGRxFi4MOmQ\nUsAQkbYiMkdE1js/XTdNEJFxIrJWRDaISFGs/CIyVkQWi8hy5+fFqZTTNCxZ/oARvSno3btG8s6d\nI/3HviVLXrr1c/Ro713+3e2390SVBjRJhfa/JMvXJHWypILfRJl78dGmA2HzO6w2YapLqjWMImCu\nqvYE5jrHQUTEAzwOjAf6AdeJSL8Y+fcDX1DVs4GbgGdTLKdpQLKczvFYAaNbu2a0blq186DvVd62\naba/gz0dASPQBT3b872Lz0z5Pr6AEWv59GudPopAbnNBLIaYdEi1D2MiMNL5PBV4B7g7JM0QYIOq\nbgIQkWlOvlWR8qtq4AI/K4FcEclR1dhrVJsGz1fDCF3G/PHrB+PJ8I6qatssOyyfb7RRpkf8gSLN\n8QKPCD8c24sx/TomvXps4f1z2H/cOwmxNIkO9fJKJWCPLKBqeG0k767bx03/WMDCyWPIa5GT0Pep\nKuWV6jqSzTQsqf4Nd1RV3/i+3YDbwkCdgW0Bx9udc/HmvxJYEilYiMgkEVkkIov27Yt/lVJTf00Y\n4B0VFbrF7GUD8hl3Vj6X9j/NdS8PX2zI8mT4A0a6axi7jxZHHAYcL1+wAPjtm+sSzl+eRA3jH+9v\nBmDFjsT7Y556fzM9J79eKzPtTc2KWcMQkbcAtzUeJgceqKqKSNIVX7f8ItIfeBC4JEq+J4EnAQoL\nC63i3QiM6t2Bzb+Z4Do/ISoneWaG+Du7o3U5/GRCH7q2bQYoZRXeJcafX7AtcgaqOq09aerLSIbb\nAoWxAkalkyAjiXK/tMS7xezOw6dca3am4YgZMFR1TKRrIrJHRPJVdZeI5AN7XZLtAAJnVXVxzgFE\nzC8iXYCXga+q6sY4nsU0IgkHC6pqGBki/qaoSDWMH47pxaQRwXNA5qzaE/d3pbvmkojnFmzl1pHB\n/SiVARHjyKkypv5vC7eNOtMf2Hz9HqGrC7t55ZMdjO7bgRZNvCPQfP1BbjUb07Ck2iQ1A2+nNM7P\nV13SLAR6ikh3EckGrnXyRcwvIq2BmUCRqto2YiYtLnD2L8/OzPAHnEjvx++PCd858GRpOQATB3UK\nuxaqumsY0WoMD72xNuxceaUyfeE2Dp8s5f7XVvHInHW8vabq9zt/DUNg9srdHDpRyvo9x8Lus2LH\nEX7wn6VMfrlqs6mqoJP6BEZTt6Xa6T0FmC4iNwOfAdcAiEgn4O+qOkFVy0XkdmA24AH+oaoro+UH\nbgfOBH4uIj93zl2iqm41GGPi8sg1g/jR2FM0y8n01wASebH75n18YUAnDp0s86+UC9ClTW7QplKB\nv6mL1I1RSj9+cRmvr8gjx+kRD2y68r3rl+84EjSMd+3948jJ9LD1wEnueXkZNw0vAGDvsarlUHxD\nieOZF2Pqt5RqGKp6QFVHq2pPVR2jqged8ztVdUJAulmq2ktVz1DVB+LIf7+qNlPVQQF/LFiYlDTJ\n8nBmB+/8i8DmKZ+/f7Uwan7fMN6crAweunJA0LV2IW33GQH/Z13rsgR8oPsuPyvqdTci8GmCS50D\n7D1W4q9NBDbrVTjnjpeUB6Uvr1BKyyu57bklfLDhALNXepvlAjv1M52HjbZ9rWkYbBycaZR8L3Tf\nbPJm2R7G9Iu++59vGG+WJ4PcrOBxqzkh41gTqblcEuN73ajCxMcTb61dufMobzp9MYHNcb6XfWjf\nS3mF8tNXlrPcGT1VUu6dH+LJEB6ft4H56/ZZH0YjYgHDNEqDu7YhJzODn3/eO4f0RGns9aDKnHab\nLI/QMje4NTcnK/h/peDO4+gv0tD9P2pKUAmdGkZooCutqOS99fv9x76FFjMzhIdnr+Wr/1jgb5JK\nZvtY8HbC+wKRqdssYJhGqXXTbNbeP57RfTsAwcNrO7fOdc3jq2FkZng7zcf260gfZxOn0Jd+IqO4\nQoNNulTG+I0/cORURUCnd6BFWw76l1qHqoBx9FRV01WmbyJlyPcVl1XEtUzJwF+9yXUuM9ZN3WMB\nwzRqIsIfrh3E698fAcDHPxnNGz+40DWtrw/D137/t68W8oMxvYDw+Qu+5Usgdod3MjWMZz/6LOr9\nXly8PWYTUUl5JcVlFTz1/mZ/MAwt63f+vSQ4j7NkyeFT3kl62ZkZATWMqswb9h6jz8/e4LVl7vt2\nhApdlt3UTRYwTIP3/t2jePOHIyJenzioM72dmkLHlk388wtCfcfZm6Nru8i7+7Vv7l1Wo2XAPT4/\noGoYbp+AbWV9MuMMGA9dNSB2IrzNSHf899OYndCl5ZX86e313PfaKtY4Gy3Fqg8UOzWMdXuO++/h\nW74kcG2v951mrI82HQjKX1mp/Or/VrJhb/iQXVP32X4YpsFL1/atEwd1ZuKgzkHnzunqXZ7kq8O7\nccuIHnRt6/2uwBrHBT3b+7eSrUxhfG2TkI72WMpjzIuYvXI376wNXk4nVvlKXPb++HjzQSA4YCzf\n4Z3x3imkeW/boZM8/cEW3l6zl3fuHMnGfSf8146XlNM8wpL2pm6wvx1jUtCxZRO2TLnM9dqL3/mc\nfzTV5wd24vkFW/1DUJORaNNV6OKMod5aHT5SPVY8i7a7oC/YvLlyNy8u2Q6EBznfKKzyCmXawm3c\n89Jy/7U7pi/lrzdGH9psapc1SRlTTc7t1oZ+nVoCcN/E/nzys7H+PTmSkWjneDLDXGPlcKthhH7f\n3IBAFG108dKQfgtfM1ckJ0rKkx6JZdLDAoYxNSDTk0GbZtl8d3T4kiPxykmwhpHMRLpYo5pKo7yw\nfaOy9h+vWlg6dF6H7/aqGtb8FWt/k/6/mM2tIZ3wpmZZwDCmBl3UK487xnpHVk2/Zbi/M35M3w5h\nkwEDTZ7Ql+zMxP533bz/ROxEIWKOrCoLf6n7QoIv75FTZf5rvgC0bs8xXlqy3T+XpVKrhvL6xAoY\ngH/Soakd1odhTA27/eIzueWiM4ICwF9vLKS0vJK+P3/DNc+3RvRgeYJ7h1/3t8TnNjzxTvSFoUvc\nXupOxPDVaIoDJuH5ulEu+f18AL5+fgHg3dAptDITq8/F1D6rYRhTw0QkrLbgyRByQ/o37g9ZYypS\nDWNo97ZcMbiz6zWA7u2bJVnScNF2APQFjMBayFur9jDqt+/4j5/+YAvgbZoKbTKLp4ZhapcFDGPq\noPsuP4sbhnWjz2ktuPmC7kDkgPGtC3vQKtc77+Pszq3Crr9y6/nVV1DgWLF31neFKqt3HQ2aGf7h\npgOuTWN7j5X4l4v3sbWo6j5rkjKmDnn/7lFkeTLo2LIJAG/8oGrCYabLkKOrz+3CRb3zmLfWOzLp\nisGd/QsF+rRq6j4RMd2Onipn/B/eizt96LDeSDWM3j99ncucbXlN7bKAYUwdEm2SYaYnOGD0y2/J\nw1cPBKqailIZtpuqv7yb2saYkfbTKCmv9G8Da2qXNUkZU09khS1wWPXZ99t5oiOp6qoJf3iPv7+3\nqbaLYUI0jH9dxjQC7Zplc8fYXvzh2kFAcMAoDVgYcfxZp9VG8dJq1a6j3D9zddQ00xZsZdyj82uo\nRAasScqYekNE+O7onizbHr6yq69JKtuTUa87jzfsPcZ1f/vYf/z9aZ9ETFsUsKyIqRlWwzCmAfCt\n8ZSdmZH2rVK/l8Ls9EQ9Pm8j+45VzRR/denOGvtuE5sFDGPqKQnYMy+wD2NU77y48n/tcwVxpcv2\nxL8ZVKoS2HfK1IKUAoaItBWROSKy3vnZJkK6cSKyVkQ2iEhRvPlFpKuIHBeRO1MppzENidtyT4FN\nUjcM68bz3xoW8z6+7WljqcmOdMEiRl2W6r+EImCuqvYE5jrHQUTEAzwOjAf6AdeJSL848z8CvJ5i\nGY1pUHyjpVoHzK/om+9dFbd98xxEhHbNs2PeJ3SXwFjfVxOshlG3pfovYSIw1fk8FbjcJc0QYIOq\nblLVUmCaky9qfhG5HNgMrEyxjMY0KH3zW/DLL/Tj0S8P8p/7+Rf68cpt51PgLAMSukpsKqLtgZFu\nyZS6oGgmx0vKYyc0KUs1YHRUVd+mvbuBji5pOgPbAo63O+ci5heR5sDdwK9SLJ8xDY6I8LXzu9PO\n2Q4WICfTw6DTW/uPPXHWHuJxMsGX8dKfj+WhK+PbTjZUsnFu+6GTyWU0CYkZMETkLRFZ4fJnYmA6\n9a5jnPTwjJD8vwR+r6rRd1Txlm+SiCwSkUX79u2LldyYRsGTxhpG8yaJjb5v3TSba847nRZJbLea\nbB+GrXRbM2L+jarqmEjXRGSPiOSr6i4RyQfC93yEHcDpAcddnHMAkfIPBa4SkYeA1kCliBSr6mMu\n5XsSeBKgsLDQ/tUYAyS7E6xIeKf6gC6t3RPHsPxXl1JQNDPh709GtI2dTPqk2iQ1A7jJ+XwT8KpL\nmoVATxHpLiLZwLVOvoj5VfVCVS1Q1QLgUeDXbsHCGOMu0Sap+5yl1PMCmrl8VCGvRfj56pBswCir\nwX6WxizVgDEFGCsi64ExzjEi0klEZgGoajlwOzAbWA1MV9WV0fIbY1ITqUmqa1vv4oZj+nq7Gx/9\n8iCe+9ZQRvRsD7h3livKnB+OCDvvxrfMevKSixiRFi58c+Vu/v3xZ6kUyARIaWkQVT0AjHY5vxOY\nEHA8C5gVb/6QNL9MpYzGNEaRahjv3DkSEW/HOcDl53jHn2w76O00FoE/XXcO333euyTH4K6tOef0\nNuRme1ybq574ymC+E7DP9tt3XJRSuZPtq/dt/Rpq0rOLAfjK0G7JFskEsJnexjRAkQJGRob4g4Ub\nAXqf1gKAji1zeOnW88N2AgzUI6958P1T7GxPNvt76/an9L0mPhYwjGmA4p2UF0pE/C/90NqE2wzz\n0D064gkYU78xJPL3J9kk9Y8PNvPeeu8oySMny7jo4Xms3Bm8kdTWAydRt4cApi/cRkHRTE6VVrhe\nN14WMIxpgJIdVitStbNf6KvVbfvXrJDhWJFGZ33j/O7+z93bNeO2UWe4pounv6GgaCa/n7Mu7Pxu\nZ2vYDzbu57MDJ7nsj+/7ry3ZeogRD8/jXx9vdb3nH99eDxC08KEJZwHDmAYo0VFSgb94+/KG/jL+\nzDeG8J9JwWtUJVPD8HiEuy7t43ot3oV2/+S84EP1+unr3BrQp+KzeZ93X/Elnx1yzWdLksTHAoYx\nDVCifQnq1CcyRAKCTfDbu02zbIb2aBd0LnSf8XgCVTomFUYKLKURhtf6vvLlT3bw1qo9KX9/Y2UB\nw5gGKNEahu8FHNQkFeGlPO/Okf7PoR3o8cQCX9mW/nwsq+69lKwaWD49MIB+85lF1f59DZUFDGMa\noET7vH2dwRki/g7zSK1D3ds3o10z72q4oQEitGaTm+Xxp/XxBYzWTbNpmp3JsJBaS7I+2nQw4rV4\nKzWa/OpGjYJt0WpMAxRt6Kyb5s66TwO7tPK/9CsjVTGiCG1u+vQXlwAw5fU1VWnSuDBioBeXbE86\nr290VhKP3KhYDcOYBuz8M+P77b1DyybMuP18plw5wD+wNd6X56p7L/V/Do1T2ZkZYRsw1UQTVKhY\nfTrW6R0fq2EY00D9r+hi2jaLvZGSj2+RQd/+F83jXG22aXZVunhqNrlZwRMBu7Vrynvug57SxgJC\nelgNw5gGqlPrXJpkRZ6lHUmr3Cx+9vl+cW3zmkwTTmhQ+ell/cKCSLrZ1q/pYQHDGBPm5gu607Vd\n04jXp34GqFy5AAAZDUlEQVRjCNcN6Ur7OLaCBeie590JcHSfDmHXmmR5GNvPbe+19Im/09tEY01S\nxjQCb/zgQg4cL03b/c7q3IrfXHF23OlvGNqVnh2aM7R7W9fr1d1kFKsDv6rfxkJGNBYwjGkE+pzW\nsla/X0SiDp+t7vf08u1Hol73NZNZuIjOmqSMMbWuIt41QZL01/mb4kpnFYzoLGAYY2pdYMD49kXu\nCxPWDIsY0VjAMMakLCcztVdJeUDA+MLA/FSLk7RqrujUe9aHYYxJ2bw7R7Lj8Kmk8/s6pe+d2J/+\nnVqxZcplALy/fj8l5RXcPDW96z8VFM3kmW8MYUSvPAA27z8RVA7jzgKGMSZlnVrn0ql1btL5fTWM\n0JrKBc5e49Vhyutr/AHDJ8JOr8ZhTVLGmFpX4bypMyPtwFQNVu06CkBlQDuU1TCiS+lvR0Taisgc\nEVnv/GwTId04EVkrIhtEpCie/CIyQEQ+FJGVIrJcRJqkUlZjTN3l6/QO3ZCpRr47SpA4WlxGSblt\n2+qTajgvAuaqak9grnMcREQ8wOPAeKAfcJ2I9IuWX0QygX8B31bV/sBIoCzFshpj6ihfU1CWp2Yb\nPaYt2Bo0QmvjvuMUFM1kxQ7vvI0Bv3yTL//1oxotU12W6t/ORGCq83kqcLlLmiHABlXdpKqlwDQn\nX7T8lwDLVPVTAFU9oKoW5o1poMr9TVI1W8Moemk5p0qrXi1vOrvxvfLJDv+5pdsO12iZ6rJUA0ZH\nVd3lfN4NuC0I0xnYFnC83TkXLX8vQEVktogsEZEfp1hOY0wd5vstPyuB4bnDergvM5KoU2VVAcO3\nNIj1ZLiLOUpKRN4CTnO5NDnwQFVVRJL+7xySPxO4ADgPOAnMFZHFqjrXpXyTgEkAXbt2TfbrjTG1\nyNePkBVHp3eWR7jr0t5keTKi7rIXr+CAkfLtGrSYfzuqOkZVz3L58yqwR0TyAZyfe11usQM4PeC4\ni3OOKPm3A/NVdb+qngRmAYMjlO9JVS1U1cK8vDy3JMaYOq6s3AkYcXR6n5HXnEkjzkjbyz2wScpG\nSUWXapPUDOAm5/NNwKsuaRYCPUWku4hkA9c6+aLlnw2cLSJNnQ7wi4BVKZbVGFNHHS32jmlp3TT2\ncumeGHuOJ6q4LDBgpOmmDVSqAWMKMFZE1gNjnGNEpJOIzAJQ1XLgdrxBYDUwXVVXRsuvqoeAR/AG\nm6XAElWdmWJZjTF11JFT3oDRpmlWzLS+jvF0LUX+o+mf+j9bBSO6lGZ6q+oBYLTL+Z3AhIDjWXib\nleLK71z7F96htcaYBu6k0ywUqYbx+PWDWb3rKI/N2+CvYfgM7tqaJVuTH8m09eBJ/+e3Vu9J+j6N\ngc30NsbUuvsm9ie/VROyI4ySumxAvn8ZD18aX23gnK6u84VT8tT7m3l+wda037e+s7WkjDG17sbh\nBdw4vCBqmsFdW3PjsG7cclEPANTpxaiumRv3vLS8mu5cf1nAMMbUC5meDO67/Cz/sa+GUd3bu5oq\n1iRljKnXxCJGjbGAYYypl3wDmixc1BwLGMaYesk/yc4iRo2xgGGMqZeq4oVFjJpiAcMYU69ZF0bN\nsYBhjDEmLhYwjDH1km9pkBreQqNRs4BhjKmXKmP0YXzrwu41WJrGwQKGMaZei9SHMfky707Qg7u2\nTvk7HpmzjsWfpb73Rn1nAcMYUy/Fs7Ls/4ou5l/fHBrWbJWb5Unou/44dz1XPvFhQnkaIgsYxph6\nKXAtqQFdWrmm6dQ6l6bZmWGLGnZomVPdxWuQLGAYY+olfw1DhBm3XxB07fJBnYKOczKDaxR9TmuR\n1Heu23OM8orKpPI2BBYwjDH1km8712yXbV1/e/XAoONpk4bRuXWu//ibF/ZI6jsv+f18fjHDu//b\nN/65kAl/eC+p+9RXFjCMMfXSzRf0YNKIHtx8QfjLP9MT/Grrm9+S+y7vD8DI3nlke5J/9b22bBcA\nb6/Zy6pdR5O+T31ky5sbY+ql3GwPP5nQN+70vmapkrLKsF37EtGY531YwDDGNAi/u3ogudkeyiL0\nMeQ4Hd8l5RVBAaNz61x2HD4V9/ccOlnGpn3HUytsPWUBwxjTIFx5bpeo1/01jPJKMp2Akd+qCR8U\nXUxB0cyEvutv721KrpD1nPVhGGMahdxsb8DI9GT4axiV8UzmcJHRSFc8TClgiEhbEZkjIuudn667\nsYvIOBFZKyIbRKQoVn4RyRKRqSKyXERWi8g9qZTTGGPOyGvGnZf04vHrzyEzw/vqSzJeBAWMtbuP\nUVA0k483HUhHMeu0VGsYRcBcVe0JzHWOg4iIB3gcGA/0A64TkX4x8l8N5Kjq2cC5wC0iUpBiWY0x\njZiIcPvFPenSpilOvPCvR5WowI7vSx+dD8DrK3anWMK6L9WAMRGY6nyeClzukmYIsEFVN6lqKTDN\nyRctvwLNRCQTyAVKgcY1fs0YU218NQzfRq/TJg1LKH9GIx0qlWrA6Kiqu5zPu4GOLmk6A9sCjrc7\n56LlfwE4AewCtgK/VVXXlb9EZJKILBKRRfv27Uv+SYwxjYavD8PXJDWsR7uE8jfWPoyYo6RE5C3g\nNJdLkwMPVFVFJMkKXlj+IUAF0AloA7wnIm+patjQBFV9EngSoLCwMOnvN8Y0Hr4KQrKd3qnM46jP\nYgYMVR0T6ZqI7BGRfFXdJSL5wF6XZDuA0wOOuzjnACLlvx54Q1XLgL0i8gFQCDTOsWzGmLQSp4bg\nFi765rdkdYwZ3I20gpFyk9QM4Cbn803Aqy5pFgI9RaS7iGQD1zr5ouXfClwMICLNgGHAmhTLaowx\nAP4tlwIrGGP6duSuS3vTqVWTmPnLK9LTmKGqFBTN5Devr07L/apbqgFjCjBWRNYDY5xjRKSTiMwC\nUNVy4HZgNrAamK6qK6PlxzuqqrmIrMQbcJ5W1WUpltUYYwBomuOdk3HjsG7+c3+/qZDbRp3pWusI\n9dT7m6NeP3KqjFueXcTBE6VR0/kC1l/frR+NJynN9FbVA8Bol/M7gQkBx7OAWQnkP453aK0xxqRd\nTqaHTb+e4Nq0pEn2a6gq//fpTkrKK9l7rJjZK/dQ0H4jReP6AFXNYIEqQr5r77FiXl6yg0kjerim\nr222NIgxplGKNDQ2lcam7z7/SdBxSVkl3e+ZxW2jzuCuS/uEpfd1uvtiw3ef+4SPNx/kot559Dmt\nZQolqR62NIgxxgRwm8z32PXnJHWvEyXlADzzv89cr/sqGL7QdazYmz5dfSTpZgHDGGMChDZJfe1z\nBeS3yo2QOjrfyrkel02eIHxYb90ME1UsYBhjTBSq6l/dNlFlTk3BE6E/IrA2s3Tb4ZjDeWubBQxj\njAkQ2uetJD9Rr9SpYUTqLwmsYXy4sWrxwmQXRaxuFjCMMSaAhjQMqUJmhCal4HzhSsqdJqlINYxK\nX6d38PXQMtQVFjCMMSZAZciGfUryTVIlZRVA5BpKZUCndx0cRRvGAoYxxgQI7Yiu1OQXG/TVMDIi\nvGmTXcuqtljAMMaYAKGvcFVo3iS5KWulTsDYfaSYO//7adh+45ECRl2NIxYwjDEmUNjLWslrnhMz\nW2gwANi477hzTXlh8XY+CtmVLzAw1IMWKQsYxhgTyK3TW0T477eHR833/IJtYed8TVI+GSJs3n/C\n39kdWMOwPgxjjKlnQmd6+17q5xW0Tfnen24/zKjfvsO/P/bO/K7wj5IKTldHW6QsYBhjTKDQmd7p\n7E94f/1+ALYfPhV077IKZc/RkohlqCssYBhjTICwTu803vvIqTIAf59IYJNUrCXT6wILGMYYEyBs\npncaI0bo8Fy3hQ7rMgsYxhgT4JrC04OO09k8FLpESMRhtWn7xvSygGGMMQGuH9qVr59f4D9O58vb\nFy98caKu9lVEYgHDGGOiSGsNI6RJymXqRp1mAcMYY0IExogz8pqn7b6LPzsUdGwzvY0xpoEY1/80\nbh11pv94aPfwuRij+3TgT9cltiOfojwwcxU3/P3jiCnqopQChoi0FZE5IrLe+dkmQrpxIrJWRDaI\nSFHA+atFZKWIVIpIYUiee5z0a0Xk0lTKaYwxyRjSvW3QSrP/uWU4C34yOijNlCsH0LNj4rWQv723\nmQMnSl2v1dXRU6nWMIqAuaraE5jrHAcREQ/wODAe6AdcJyL9nMsrgCuA+SF5+gHXAv2BccCfnfsY\nY0yNcVuuo0PLJkHHeS1yyM1K7PUUq8mpoTZJTQSmOp+nApe7pBkCbFDVTapaCkxz8qGqq1V1bYT7\nTlPVElXdDGxw7mOMMdUu0Y7uJokGjBjXi8sq+OzAiaBzbosb1rRUA0ZHVd3lfN4NdHRJ0xkIXJVr\nu3MummTyGGNMWnx+YCcALjizfVzpm2QmFjDmr9sX9fqPpi/looffodjZgOmNFbvoOfl11u4+ltD3\npFvMRd5F5C3gNJdLkwMPVFVFpMYrUiIyCZgE0LVr15r+emNMA3ReQVu2TLks7vQ5WYn97v2/jQei\nXt9/3Nu3cfRUGTmZGcxZtReAZdsP0/u0Fgl9VzrFDBiqOibSNRHZIyL5qrpLRPKBvS7JdgCBUye7\nOOeiiTuPqj4JPAlQWFhYR1v+jDENWU5m9Qw4HfLrufxwTC9/X0ptv+BSfcoZwE3O55uAV13SLAR6\nikh3EcnG25k9I477XisiOSLSHegJLEixrMYYUy2kGjezeGXpjqrNlWo5YqQaMKYAY0VkPTDGOUZE\nOonILABVLQduB2YDq4HpqrrSSfclEdkODAdmishsJ89KYDqwCngDuE1VK1IsqzHGpNWDV54ddu6h\nKwek9TvKKysDahi1GzGS26jWoaoHgNEu53cCEwKOZwGzXNK9DLwc4d4PAA+kUj5jjKlOEweFj8W5\n5rzT+fGLy9L2HeUVijh1jNoebmszvY0xJkmha0NVh7IK9dcwFmw5yPlT3uZ4SXm1f6+blGoYxhjT\nmAWuVv7INQPp0KJJ5MRJqghoknppiXfsz+pdR9OyZWyiLGAYY0ySAmsYVwzuUi3fUV6hQHBNpvrr\nNe6sScoYY5IUq0Vq3f3jU/6OsoAaRrzfW10sYBhjTJJiDadNx4vd2+kddufUb5wECxjGGFNN0tEp\nXl6pVsMwxpiGLqOaXuzWh2GMMQ1MumaAS2indy1VMSxgGGNMgu4Z34e2zbIjXg/cH+O5bw5N+ft2\nHSkOOvZt7aqqVNbgbks2rNYYYxJ0y0VncMtFZ0S8/tr3LuDjTQcB+FycS6RH89bqPUHHFU6Q+NX/\nreKf/9uS0Mq6qbCAYYwxaXZGXnPOyEt829Z4+QLGP/+3BfDWNGqimcqapIwxpp6prNSgzZRqao0p\nCxjGGFPPbDlwkksfne8/rqyhiGEBwxhj6pmfvLw86LjCAoYxxjRO489y2xU7MmuSMsaYRuqJG85N\nKP2KHUeqqSTBLGAYY0w9d9VfPqyR77GAYYwxNezjn4ymR16z2i5GwixgGGNMNdsy5TKG92jnP+7Y\nsgntm+XUYomSYxP3jDGmBjw/aRiPzFnHoROlACi1vEF3EixgGGNMDfnR2F7+z74loB740llMfnlF\nLZUoMSk1SYlIWxGZIyLrnZ9tIqQbJyJrRWSDiBQFnL9aRFaKSKWIFAacHysii0VkufPz4lTKaYwx\ndc0j1wzky4Wn8+XC07nloh707tiitosUU6p9GEXAXFXtCcx1joOIiAd4HBgP9AOuE5F+zuUVwBXA\n/JBs+4EvqOrZwE3AsymW0xhj6pRu7Zrx4FUDyPRkcM/4vsz+4YiU7ldcVpGmkkWWasCYCEx1Pk8F\nLndJMwTYoKqbVLUUmObkQ1VXq+ra0Ayq+omq7nQOVwK5IlL/eoiMMaaGHD1VVu3fkWrA6Kiqu5zP\nu4GOLmk6A9sCjrc75+J1JbBEVUvcLorIJBFZJCKL9u3bl8BtjTGm/ujWrmnU62U1sC9GzE5vEXkL\ncJunPjnwQFVVRNJaYhHpDzwIXBIpjao+CTwJUFhYWP+GHRhjTBwyY+z3Wl5RWf1liJVAVcdEuiYi\ne0QkX1V3iUg+sNcl2Q7g9IDjLs65qESkC/Ay8FVV3RgrvTHG1HfPfGMIh06Wul47VlweNe8f527g\nd9cMrI5i+aU6rHYG3k7pKc7PV13SLAR6ikh3vIHiWuD6aDcVkdbATKBIVT9IsYzGGFMvjOiVF/Ha\nkO5tUWDmsl2u11furP71pFLtw5gCjBWR9cAY5xgR6SQiswBUtRy4HZgNrAamq+pKJ92XRGQ7MByY\nKSKznfveDpwJ/FxEljp/OqRYVmOMqbfuHteHx68fHPF6Rg3suJdSDUNVDwCjXc7vBCYEHM8CZrmk\nexlvs1Po+fuB+1MpmzHGNCSnt43e6V0D8cLWkjLGmLqoZ4eqPcG/N7pnzPQWMIwxppH65Rf7k5Pp\nfUXHEwtqoknKAoYxxtRB55/ZnlsuOgOIr/ZQAxUMCxjGGFPXxbUFq9UwjDGm8cprng1Ae+dnNDHm\n9aWFLW9ujDF11PVDu9G8SSZfHBh7NaWaaJKygGGMMXWUJ0P40jld4kor1iRljDEm0Izbz3c9b53e\nxhhjgpzVqZX/8+QJff2fbR6GMcaYIL7A0LVt06B9wc8raFvt3219GMYYU4+ICE9/7Tz6d2rJK0u9\nC3+P6p3HHZf0rvbvtoBhjDH1zKg+wWuxntmhOZ4aGFdrTVLGGGPiYgHDGGPqqcwM7ys8y1Mzr3Jr\nkjLGmHrq+qFd2XO0mNtGnVkj32cBwxhj6qkmWR7uCRhaW92sScoYY0xcLGAYY4yJiwUMY4wxcbGA\nYYwxJi4pBQwRaSsic0RkvfOzTYR040RkrYhsEJGigPNXi8hKEakUkUKXfF1F5LiI3JlKOY0xxqQu\n1RpGETBXVXsCc53jICLiAR4HxgP9gOtEpJ9zeQVwBTA/wv0fAV5PsYzGGGPSINWAMRGY6nyeClzu\nkmYIsEFVN6lqKTDNyYeqrlbVtW43FpHLgc3AyhTLaIwxJg1SDRgdVXWX83k30NElTWdgW8Dxdudc\nRCLSHLgb+FWK5TPGGJMmMSfuichbwGkulyYHHqiqikg8W5XH45fA71X1eKxdpERkEjDJOTwuIq41\nlji1B/ankL8us2ervxry89mz1Q3d4kkUM2Co6phI10Rkj4jkq+ouEckH9rok2wGcHnDcxTkXzVDg\nKhF5CGgNVIpIsao+5lK+J4EnYz1HPERkkaqGdb43BPZs9VdDfj57tvol1aVBZgA3AVOcn6+6pFkI\n9BSR7ngDxbXA9dFuqqoX+j6LyC+B427BwhhjTM1JtQ9jCjBWRNYDY5xjRKSTiMwCUNVy4HZgNrAa\nmK6qK510XxKR7cBwYKaIzE6xPMYYY6pJSjUMVT0AjHY5vxOYEHA8C5jlku5l4OUY3/HLVMqYoLQ0\nbdVR9mz1V0N+Pnu2ekRU09VPbYwxpiGzpUGMMcbExQIGkZcuqS9E5HQRmSciq5ylVr7vnI+4dIuI\n3OM871oRubT2Sh8fEfGIyCci8ppz3JCerbWIvCAia0RktYgMbyjPJyI/dP5NrhCR50WkSX1+NhH5\nh4jsFZEVAecSfh4ROVdEljvX/iix5g/UFaraqP8AHmAj0APIBj4F+tV2uRJ8hnxgsPO5BbAO7zIs\nDwFFzvki4EHncz/nOXOA7s7ze2r7OWI844+A54DXnOOG9GxTgW86n7PxDiWv98+Hd4LuZiDXOZ4O\nfK0+PxswAhgMrAg4l/DzAAuAYYDgXf5ofG0/Wzx/rIYRZemS+kJVd6nqEufzMbyj0ToTeemWicA0\nVS1R1c3ABrz/HeokEekCXAb8PeB0Q3m2VnhfQk8BqGqpqh6mgTwf3oE1uSKSCTQFdlKPn01V5wMH\nQ04n9DzOnLWWqvqReqPHM7gvq1TnWMBIYumSukxECoBzgI+JvHRLfXvmR4EfA5UB5xrKs3UH9gFP\nO01ufxeRZjSA51PVHcBvga3ALuCIqr5JA3i2EIk+T2fnc+j5Os8CRgPirMH1IvADVT0aeM35Tabe\nDYkTkc8De1V1caQ09fXZHJl4mzieUNVzgBOErPpcX5/PacufiDcodgKaicgNgWnq67NF0tCeJ5QF\njOSWLqlzRCQLb7D4t6q+5Jze41R/CVm6pT498/nAF0VkC97mwotF5F80jGcD72+X21X1Y+f4BbwB\npCE83xhgs6ruU9Uy4CXgczSMZwuU6PPscD6Hnq/zLGAELF0iItl4ly6ZUctlSogzwuIpYLWqPhJw\nybd0CwQv3TIDuFZEcpwlW3ri7YSrc1T1HlXtoqoFeP9u3lbVG2gAzwagqruBbSLS2zk1GlhFw3i+\nrcAwEWnq/Bsdjbd/rSE8W6CEnsdpvjoqIsOc/y5fxX1Zpbqntnvd68IfvLPS1+EdxTC5tsuTRPkv\nwFsNXgYsdf5MANrh3dhqPfAW0DYgz2TneddST0ZoACOpGiXVYJ4NGAQscv7+XgHaNJTnw7tFwRq8\nm6U9i3fEUL19NuB5vP0xZXhrhzcn8zxAofPfZCPwGM4k6rr+x2Z6G2OMiYs1SRljjImLBQxjjDFx\nsYBhjDEmLhYwjDHGxMUChjHGmLhYwDDGGBMXCxjGGGPiYgHDGGNMXP4f2fpWTvJ0L0kAAAAASUVO\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2c28a4eb90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training epoch 1200\n",
      "Iter: 1200\n",
      "D loss: -0.02672\n",
      "G_loss: -0.01466\n",
      "()\n",
      "[-0.83490461  0.90284789 -0.86862808]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAD8CAYAAABkbJM/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nChgiIhIWBQwREQnL/wcLQc2J9LlejwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2c28afc750>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAD8CAYAAABkbJM/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2c28be2d50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training epoch 1300\n",
      "Iter: 1300\n",
      "D loss: -0.0272\n",
      "G_loss: -0.01501\n",
      "()\n",
      "[-0.91411364  0.91120875 -0.91564703]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAD8CAYAAABkbJM/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niIhISJQwREQkJEoYIiISkv8PSmkwphxZc88AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2c25ad6a10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAD8CAYAAABkbJM/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2c245f5e90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training epoch 1400\n",
      "Iter: 1400\n",
      "D loss: -0.027\n",
      "G_loss: -0.0148\n",
      "()\n",
      "[-0.96182871  0.92571729 -0.93304491]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAD8CAYAAABkbJM/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2c24530250>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAD8CAYAAABkbJM/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2c24594110>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "G_pred, G_pred_logit = generator(z, is_training=False)\n",
    "G_sample, G_sample_logit = generator(z, is_training=True)\n",
    "\n",
    "D_real  = discriminator(label_placeholder)\n",
    "D_fake  = discriminator(G_sample)\n",
    "\n",
    "# D_loss = -tf.reduce_mean(tf.log(D_real) + tf.log(1. - D_fake))\n",
    "# G_loss = -tf.reduce_mean(tf.log(D_fake))\n",
    "\n",
    "# Alternative losses:\n",
    "# -------------------\n",
    "\n",
    "#D_loss_real = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=D_logit_real, labels=ones_like))\n",
    "#D_loss_fake = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=D_logit_fake, labels=zeros_like))\n",
    "#D_loss = D_loss_real + D_loss_fake\n",
    "#G_loss = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=D_logit_fake, labels=ones_like))\n",
    "\n",
    "#D_loss_real = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=D_logit_real, labels=tf.ones_like(D_logit_real)))\n",
    "#D_loss_fake = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=D_logit_fake, labels=tf.zeros_like(D_logit_fake)))\n",
    "#D_loss = D_loss_real + D_loss_fake\n",
    "#G_loss = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=D_logit_fake, labels=tf.ones_like(D_logit_fake)))\n",
    "\n",
    "#WGAN:\n",
    "D_loss = tf.reduce_mean(D_real) - tf.reduce_mean(D_fake)\n",
    "G_loss = -tf.reduce_mean(D_fake)\n",
    "clip_D = [p.assign(tf.clip_by_value(p, -0.01, 0.01)) for p in theta_D]\n",
    "\n",
    "D_solver = tf.train.AdamOptimizer(learning_rate=1e-4).minimize(D_loss, var_list=theta_D)\n",
    "G_solver = tf.train.AdamOptimizer(learning_rate=1e-4).minimize(G_loss, var_list=theta_G)\n",
    "\n",
    "mb_size = S\n",
    "\n",
    "sess = tf.Session()\n",
    "sess.run(tf.global_variables_initializer())\n",
    "\n",
    "train_dataset = [features[0] for i in range(0,32)]\n",
    "n_train = len(train_dataset)\n",
    "preds = []\n",
    "\n",
    "d_losses = []\n",
    "g_losses = []\n",
    "\n",
    "for it in range(100000):\n",
    "    if it % 100 == 0:\n",
    "        print(\"Training epoch %d\" %it)\n",
    "    batch_sched = list(range(0, n_train+1,S))\n",
    "    for j in range(0, len(batch_sched)-1):\n",
    "        start = batch_sched[j]\n",
    "        stop = batch_sched[j+1]\n",
    "        feed_dict = construct_feed_dict(train_dataset, start, stop)\n",
    "\n",
    "        _, D_loss_curr = sess.run([D_solver, D_loss], feed_dict=feed_dict)\n",
    "        \n",
    "        for _ in range(5):\n",
    "            feed_dict = construct_feed_dict(train_dataset, start, stop)\n",
    "\n",
    "            _, D_loss_curr, _ = sess.run(\n",
    "                [D_solver, D_loss, clip_D],\n",
    "                feed_dict=feed_dict)\n",
    "            \n",
    "        \n",
    "        _, G_loss_curr = sess.run([G_solver, G_loss], feed_dict=feed_dict)\n",
    "        \n",
    "        d_losses.append(D_loss_curr)\n",
    "        g_losses.append(G_loss_curr)\n",
    "\n",
    "    if it % 100 == 0:\n",
    "        print('Iter: {}'.format(it))\n",
    "        print('D loss: {:.4}'. format(D_loss_curr))\n",
    "        print('G_loss: {:.4}'.format(G_loss_curr))\n",
    "        print()\n",
    "\n",
    "    if it % 100 == 0:\n",
    "        samples = []\n",
    "        for trial in range(0,100):\n",
    "            feed_dict = construct_feed_dict(train_dataset, start, stop)\n",
    "            samples.append(np.reshape(sess.run(G_sample, feed_dict=feed_dict)[0], (1,3)))\n",
    "        samples = np.concatenate(samples, axis=0)\n",
    "        print(np.mean(samples, axis=0)[:10])\n",
    "        if it > 999:\n",
    "            plt.plot(range(0,it+1,1), d_losses)\n",
    "            plt.show()\n",
    "            plt.plot(range(0, it+1, 1), g_losses)\n",
    "            plt.show()\n",
    "        #print(samples[:10])\n",
    "        #plt.hist(samples, bins=50)\n",
    "        #plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]),\n",
       " array([0, 1, 1, 1, 2, 3, 3, 3, 4, 4, 4, 7]),\n",
       " array([ 0,  1,  5,  9,  4,  2,  6, 10,  3,  7, 11,  8]))"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.where(feed_dict[dihed_indices][0] != 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "array([ 0.61143659, -0.99900387,  0.48252666])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    " [-0.91292151,  3.09695411,  1.06725919,  2.02969492,  1.9801807 , 2.27330438],"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.61143659, -0.99900387,  0.48252666],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ],\n",
       "       [ 0.        ,  0.        ,  0.        ]])"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features[0][1][1].toarray()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "sess.run(G_sample, feed_dict=feed_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "array([[ 0.91292151],\n",
    "       [ 3.09695411],\n",
    "       [ 1.06725919],"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "plt.hist(sess.run(G_sample, feed_dict=feed_dict))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "batch_sched"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "for it in range(20001):\n",
    "    if it % 100 == 0:\n",
    "        print(\"Training epoch %d\" %it)\n",
    "    batch_sched = list(range(0, n_train+1,S))\n",
    "    for j in range(0, len(batch_sched)-1):\n",
    "        start = batch_sched[j]\n",
    "        stop = batch_sched[j+1]\n",
    "        feed_dict = construct_feed_dict(train_dataset, start, stop)\n",
    "\n",
    "        _, D_loss_curr = sess.run([D_solver, D_loss], feed_dict=feed_dict)\n",
    "        _, G_loss_curr = sess.run([G_solver, G_loss], feed_dict=feed_dict)\n",
    "\n",
    "    if it % 1000 == 0:\n",
    "        print('Iter: {}'.format(it))\n",
    "        print('D loss: {:.4}'. format(D_loss_curr))\n",
    "        print('G_loss: {:.4}'.format(G_loss_curr))\n",
    "        print()\n",
    "\n",
    "    if it % 1000 == 0:\n",
    "        samples = sess.run(G_sample, feed_dict=feed_dict)\n",
    "        plt.hist(samples, bins=50)\n",
    "        plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "preds = []\n",
    "for k in range(0,100):\n",
    "    feed_dict = construct_feed_dict(train_dataset, 0, 1)\n",
    "    samples = sess.run(G_sample, feed_dict=feed_dict)[2]\n",
    "    preds.append(samples)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "plt.hist(np.concatenate(preds))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "preds[4]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "train_dataset[0][1][1].toarray()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "res = gan.predict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "plt.hist(res, bins=50)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "features[0][1][1].toarray()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "features[0][1][1].toarray()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "preds = []\n",
    "for i in range(0,100):\n",
    "    preds.append(gan.predict(features)[0][2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "np.mean(preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "features[0][2].atom_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}