Untitled1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import networkx as nx\n",
    "import os\n",
    "from PIL import Image\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "https://scikit-learn.org/stable/auto_examples/cluster/plot_cluster_comparison.html#sphx-glr-auto-examples-cluster-plot-cluster-comparison-py\n",
    "\n",
    "https://people.maths.ox.ac.uk/porterm/papers/conglong.pdf\n",
    "\n",
    "https://www.pnas.org/content/102/20/7057\n",
    "\n",
    "https://www.sciencedirect.com/science/article/pii/S0378437107007844?via%3Dihub"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/.pyenv/versions/3.5.6/lib/python3.5/site-packages/IPython/core/interactiveshell.py:2785: DtypeWarning: Columns (0) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  interactivity=interactivity, compiler=compiler, result=result)\n"
     ]
    }
   ],
   "source": [
    "edges = pd.read_table(\"./all_withoutnull.csv\",index_col=0, encoding=\"utf-8\",sep=',')\n",
    "#edges"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Партії :\n",
    "<img src=\"2.png\" align=\"center\"  alt=\"Drawing\" />"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[2, 2, 2, 2, 2, 2, 2, 2, 1, 8]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#list(edges.loc['abscd', : ])\n",
    "list(edges.loc['abscd', : ])[0:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "deputies=edges.columns.values\n",
    "deputies_fractions=list(edges.loc['abscd', : ])\n",
    "len(deputies)==len(deputies_fractions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "name    Порошенко Олексій Петрович\n",
       "Name: 3, dtype: object"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "deputies_names = pd.read_table(\"./plenary_deputies-skl8.csv\",index_col=0, encoding=\"utf-8\",sep=',')\n",
    "deputies_names=deputies_names.drop(['sex'], axis=1)\n",
    "deputies_names.loc[3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "#abscd is the column which has information about fraction of each particular deputy\n",
    "edges=edges.drop(['abscd'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>11</th>\n",
       "      <th>...</th>\n",
       "      <th>478</th>\n",
       "      <th>479</th>\n",
       "      <th>480</th>\n",
       "      <th>481</th>\n",
       "      <th>482</th>\n",
       "      <th>483</th>\n",
       "      <th>484</th>\n",
       "      <th>485</th>\n",
       "      <th>486</th>\n",
       "      <th>487</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id_event</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>21429</th>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21430</th>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21431</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21432</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21433</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21434</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21435</th>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21436</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21437</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21438</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 467 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          2  3  4  5  6  7  8  9  10  11 ...   478  479  480  481  482  483  \\\n",
       "id_event                                 ...                                  \n",
       "21429     3  4  4  4  1  4  4  4   4   4 ...     4    4    3    4    4    4   \n",
       "21430     4  4  4  4  1  4  4  4   4   4 ...     4    4    3    4    4    4   \n",
       "21431     1  4  1  3  3  4  1  4   4   0 ...     4    1    1    1    1    1   \n",
       "21432     1  4  1  1  3  4  1  4   1   0 ...     4    1    1    1    1    1   \n",
       "21433     1  4  1  1  3  1  1  4   1   0 ...     4    1    1    1    1    1   \n",
       "21434     1  4  1  1  3  1  4  4   1   0 ...     1    1    1    1    1    1   \n",
       "21435     4  4  1  1  3  1  4  4   1   0 ...     1    1    1    1    4    1   \n",
       "21436     1  4  1  1  1  1  1  4   1   0 ...     1    1    1    1    1    1   \n",
       "21437     1  4  1  4  3  4  4  4   1   0 ...     1    1    1    1    1    1   \n",
       "21438     1  1  4  1  1  1  1  4   1   0 ...     1    1    1    1    4    1   \n",
       "\n",
       "          484  485  486  487  \n",
       "id_event                      \n",
       "21429       0    4    4    3  \n",
       "21430       0    4    3    3  \n",
       "21431       0    1    1    1  \n",
       "21432       0    1    1    3  \n",
       "21433       0    1    1    1  \n",
       "21434       0    1    1    3  \n",
       "21435       0    4    1    1  \n",
       "21436       0    1    1    1  \n",
       "21437       0    1    1    4  \n",
       "21438       0    4    1    1  \n",
       "\n",
       "[10 rows x 467 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "edges[0:10]\n",
    "#edges"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(15556, 467)\n",
      "[[0 0 0 ... 0 0 0]\n",
      " [0 0 0 ... 0 0 0]\n",
      " [1 0 1 ... 1 1 1]\n",
      " ...\n",
      " [1 1 1 ... 0 0 0]\n",
      " [1 1 1 ... 0 0 0]\n",
      " [1 1 1 ... 0 0 0]]\n"
     ]
    }
   ],
   "source": [
    "df=edges.values\n",
    "print(df.shape)\n",
    "for i in range(15556):\n",
    "    for j in range(467):\n",
    "        if (df[i][j]==2):\n",
    "           df[i][j]=-1\n",
    "        elif df[i][j]==1:\n",
    "           df[i][j]=1\n",
    "        else: df[i][j]=0\n",
    "print(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(467, 15556)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2=np.transpose(df)\n",
    "df2.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SVD\n",
    "\n",
    "https://nbviewer.jupyter.org/github/fastai/numerical-linear-algebra/blob/master/nbs/2.%20Topic%20Modeling%20with%20NMF%20and%20SVD.ipynb\n",
    "\n",
    "https://docs.scipy.org/doc/numpy-1.15.1/reference/generated/numpy.linalg.svd.html\n",
    "\n",
    "https://www.hindawi.com/journals/cin/2016/1096271/\n",
    "\n",
    "just svd and Data Compression :http://andrew.gibiansky.com/blog/mathematics/cool-linear-algebra-singular-value-decomposition/\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "After compression: \n"
     ]
    }
   ],
   "source": [
    "print(\"After compression: \")\n",
    "U, S, Vt = np.linalg.svd(df2, full_matrices=False) #single value decomposition\n",
    "sizes = []\n",
    "k = (1, 2, 5, 10, 20, 50, 60, 80, 500, 960)\n",
    "frobenius_dist=[]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(467, 15556) (467, 467) (467,) (467, 15556)\n"
     ]
    }
   ],
   "source": [
    "print(df2.shape, U.shape, S.shape, Vt.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fe0ac59f828>]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(S)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fe0ac507b70>]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(S/sum(S))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fe0ac4c6940>]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(S[:10])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.05669192 0.01391838 0.01125741 0.00976212 0.00827893 0.00701024\n",
      " 0.00626542 0.00570239 0.0055968  0.00505656 0.00493456 0.00483198\n",
      " 0.00467234 0.00438458 0.00427035 0.00409018 0.00395431 0.00393598\n",
      " 0.00386141 0.00383565 0.00369257 0.00363863 0.00362645 0.00353669\n",
      " 0.00352469 0.00348749 0.00344692 0.00343865 0.00337904 0.00335065]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'.........'"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#print(S/sum(S))\n",
    "print((S/sum(S))[0:30])\n",
    "\".........\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "15556"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(list(edges.index))\n",
    "#list(edges.columns.values)\n",
    "#list(edges.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_top_bills=8\n",
    "\n",
    "def show_clusters(a):\n",
    "    top_bills = lambda t: [list(edges.index)[i] for i in np.argsort(t)[:-num_top_bills-1:-1]]\n",
    "    cluster_bills = ([top_bills(t) for t in a])\n",
    "    return [t for t in cluster_bills]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True\n"
     ]
    }
   ],
   "source": [
    "print(np.allclose(np.identity(U.shape[0]), U.T@U))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[1155, 16779, 12701, 12702, 587, 12699, 7864, 3837],\n",
       " [7260, 7259, 16004, 10115, 7258, 10111, 10113, 3462],\n",
       " [221, 84, 85, 925, 117, 296, 1385, 927],\n",
       " [696, 3277, 2933, 3462, 5678, 16012, 16096, 16097],\n",
       " ['21470', '19441', '21467', '21471', '21466', '21463', 994, 231],\n",
       " [10966, '21467', 696, 6027, '18894', 13067, 10709, 6397],\n",
       " [6025, 1450, 2071, 588, 5950, 3191, 2856, 1568],\n",
       " [6943, 6947, 6962, 6942, 6944, 6946, 6945, 6967]]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "show_clusters(Vt[:8])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "After compression: \n"
     ]
    }
   ],
   "source": [
    "print(\"After compression: \")\n",
    "U2, S2, Vt2 = np.linalg.svd(df, full_matrices=False) #single value decomposition\n",
    "sizes = []\n",
    "k = (1, 2, 5, 10, 20, 50, 60, 80, 500, 960)\n",
    "frobenius_dist=[]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(15556, 467) (15556, 467) (467,) (467, 467)\n"
     ]
    }
   ],
   "source": [
    "print(df.shape, U2.shape, S2.shape, Vt2.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fe0ac4859b0>]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(S2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fe0ac4be780>]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(S2[:10])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_top_deps=5\n",
    "\n",
    "def show_clusters_2(a):\n",
    "    top_deps = lambda t: [list(edges.columns.values)[i] for i in np.argsort(t)[:-num_top_deps-1:-1]]\n",
    "    cluster_deps = ([top_deps(t) for t in a])\n",
    "    return [t for t in cluster_deps]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[['398', '182', '191', '187', '249'],\n",
       " ['162', '219', '159', '236', '235'],\n",
       " ['234', '292', '189', '244', '390'],\n",
       " ['349', '36', '392', '337', '410'],\n",
       " ['360', '365', '350', '364', '362'],\n",
       " ['173', '181', '114', '375', '383'],\n",
       " ['64', '271', '63', '118', '71'],\n",
       " ['273', '181', '114', '454', '173']]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = show_clusters_2(Vt2[:8])\n",
    "a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Німченко Василь Іванович'"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "deputies_names.loc[int(a[0][0])][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Німченко Василь Іванович', 'Яценюк Арсеній Петрович', 'Петренко Павло Дмитрович', 'Аваков Арсен Борисович', 'Квіт Сергій Миронович']\n",
      "[4, 1, 1, 1, 2]\n",
      "['Бурбак Максим Юрійович', 'Горбунов Олександр Володимирович', 'Федорук Микола Трохимович', 'Кривенко Вадим Валерійович', 'Кремінь Тарас Дмитрович']\n",
      "[1, 1, 1, 1, 1]\n",
      "['Сторожук Дмитро Анатолійович', 'Кубів Степан Іванович', 'Гриневич Лілія Михайлівна', 'Гройсман Володимир Борисович', 'Кутовий Тарас Вікторович']\n",
      "[1, 2, 1, 0, 2]\n",
      "['Шурма Ігор Михайлович', 'Матвієнков Сергій Анатолійович', 'Мартовицький Артур Володимирович', 'Воропаєв Юрій Миколайович', 'Сажко Сергій Миколайович']\n",
      "[4, 4, 4, 4, 4]\n",
      "['Ленський Олексій Олексійович', 'Кириченко Олексій Миколайович', 'Ляшко Олег Валерійович', 'Купрієнко Олег Васильович', 'Кошелєва Альона Володимирівна']\n",
      "[5, 5, 5, 5, 5]\n",
      "['Іллєнко Андрій Юрійович', 'Левченко Юрій Володимирович', 'Осуховський Олег Іванович', 'Івченко Вадим Євгенович', 'Кириленко Іван Григорович']\n",
      "[0, 0, 0, 6, 6]\n",
      "['Соловей Юрій Ігорович', 'Паламарчук Микола Петрович', 'Тимошенко Юрій Володимирович', 'Євтушок Сергій Миколайович', 'Різаненко Павло Олександрович']\n",
      "[2, 2, 1, 6, 2]\n",
      "['Пацкан Валерій Васильович', 'Левченко Юрій Володимирович', 'Осуховський Олег Іванович', 'Юзькова Тетяна Леонідівна', 'Іллєнко Андрій Юрійович']\n",
      "[2, 0, 0, 5, 0]\n"
     ]
    }
   ],
   "source": [
    "for i in range(8):\n",
    "    print([deputies_names.loc[int(a[i][j])][0] for j in range(num_top_deps)])\n",
    "    print([deputies_fractions[int(np.where(deputies == a[i][j])[0][0])] for j in range(num_top_deps)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "якщо кластеризувати закони відносно депутатів, то оцей список депутатів для кожного закону показує найрішаючих депутатів. в даному випадку перші 5 депутатів належать до різних партій"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import decomposition\n",
    "m,n=edges.shape\n",
    "d=11  # num topics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(467, 15556)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "for i in range(15556):\n",
    "    for j in range(467):\n",
    "        if (df[i][j]==-1):\n",
    "           df[i][j]=2\n",
    "        #elif df2[i][j]==1:\n",
    "        #   df2[i][j]=1\n",
    "        #else: df2[i][j]=0\n",
    "        \n",
    "df2=np.transpose(df)\n",
    "df2.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-6.34452038e-02,  2.48615290e-02, -1.96113599e-02, ...,\n",
       "        -3.54738573e-05,  2.37313647e-05, -2.39827227e-20],\n",
       "       [-6.97330915e-02,  4.12590926e-02,  7.92514601e-03, ...,\n",
       "         1.33318967e-05,  1.77688676e-05,  1.37449537e-16],\n",
       "       [-6.32582629e-02,  2.62020770e-02, -1.49258121e-02, ...,\n",
       "         1.70090006e-05,  2.26602517e-05,  1.00120030e-16],\n",
       "       ...,\n",
       "       [-1.98113887e-03,  2.46122739e-03, -7.67406618e-03, ...,\n",
       "         1.74246601e-04, -1.19207148e-04,  4.68051962e-17],\n",
       "       [-3.93101324e-03,  8.25667595e-03, -1.98185879e-02, ...,\n",
       "         1.47930612e-04,  2.36654998e-04,  6.63796419e-17],\n",
       "       [-1.20385022e-03,  1.83111947e-03, -4.95992188e-03, ...,\n",
       "         6.13264573e-04, -9.19319527e-05,  5.22225663e-17]])"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "U"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#import seaborn as sns\n",
    "#https://people.maths.ox.ac.uk/porterm/papers/conglong.pdf\n",
    "colors = ['#630C3A', '#39C8C6', '#D3500C', '#FFB139', 'red', 'green', 'blue', 'black', 'gray', 'orange', 'yellow', 'rose']\n",
    "for i in range(len(U)):\n",
    "    x = U[i][0]\n",
    "    y = U[i][1]\n",
    "    plt.scatter(x, y, c=colors[int(deputies_fractions[i])])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"3.png\" align=\"center\"  alt=\"Drawing\"/>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "467"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(U[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.manifold import TSNE\n",
    "tsne = TSNE(n_components=2, perplexity=20)\n",
    "points = tsne.fit_transform(df2)\n",
    "fig = plt.figure(figsize=(10,10))\n",
    "for i in range(len(U)):\n",
    "    plt.scatter(y = points[i, 0], x = points[i, 1], c = colors[int(deputies_fractions[i])])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## NMF\n",
    "https://arxiv.org/pdf/1401.5226.pdf\n",
    "\n",
    "https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1000029"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[15342, 2406, 1155, 2407, 16779, 4542, 4282, 282],\n",
       " [1155, 58, 1135, 60, 63, 539, 85, 84],\n",
       " [3462, 6048, 14647, 7284, 7973, 3277, 16004, 7260],\n",
       " ['20250', 17398, 17397, '20247', '20323', 17396, 9855, 16610],\n",
       " [1134, 3126, 3462, 3124, 3277, 587, 696, 2831],\n",
       " [15660, 15661, 16098, 16099, '20996', 16095, 16097, '20886'],\n",
       " [3108, 2855, 3122, 2632, 2928, 12173, 2626, 3112],\n",
       " [6909, 6936, 6981, 6917, 6943, 6986, 6916, 6974],\n",
       " [3462, 6048, 4408, 10968, 10967, 226, 3929, 4414],\n",
       " [7864, 13337, 18413, 10628, 13333, 13338, '19942', 18219],\n",
       " ['20542', '19243', '19244', '19241', '19242', '19925', '19245', 14650]]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d=11\n",
    "clf = decomposition.NMF(n_components=d, random_state=1)\n",
    "\n",
    "W1 = clf.fit_transform(df2)\n",
    "H1 = clf.components_\n",
    "\n",
    "show_clusters(H1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(11, 467)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf = decomposition.NMF(n_components=d, random_state=1)\n",
    "\n",
    "W1 = clf.fit_transform(df)\n",
    "H1 = clf.components_\n",
    "\n",
    "H1.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "467"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "H1.shape[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(H1[:,1])\n",
    "list(H1[:,1]).index(max(H1[:,1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.34881869, 2.83419442, 0.        , 0.39102076, 0.        ,\n",
       "       0.        , 0.15406573, 0.11188409, 0.        , 1.07744737,\n",
       "       3.83485932])"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "H1[:,1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "clusters=[list(H1[:,i]).index(max(H1[:,i])) for i in range(H1.shape[1])]\n",
    "\n",
    "for i in range(len(deputies)):\n",
    "    if clusters[i]==10:\n",
    "        clusters[i]=2\n",
    "    elif clusters[i]==2:\n",
    "        clusters[i]=3\n",
    "    elif clusters[i]==5:\n",
    "        clusters[i]=10\n",
    "    elif clusters[i]==4:\n",
    "        clusters[i]=5\n",
    "    elif clusters[i]==3:\n",
    "        clusters[i]=4\n",
    "    elif clusters[i]==8:\n",
    "        clusters[i]=6\n",
    "    elif clusters[i]==6:\n",
    "        clusters[i]=8\n",
    "    elif clusters[i]==0:\n",
    "        clusters[i]=1\n",
    "    elif clusters[i]==1:\n",
    "        clusters[i]=0\n",
    "#for i in range(len(deputies)):\n",
    "    #print(deputies[i],clusters[i], deputies_fractions[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s = (11,11)\n",
    "choises=np.zeros(s)\n",
    "choises[0,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Недава Олег Анатолійович 35\n",
      "Сабашук Петро Павлович 50\n",
      "Кривохатько Вадим Вікторович 58\n",
      "Гарбуз Юрій Григорійович 81\n",
      "Жолобецький Олександр Олександрович 94\n",
      "Корнацький Аркадій Олексійович 97\n",
      "Голубов Дмитро Іванович 100\n",
      "Сугоняко Олександр Леонідович 120\n",
      "Гордєєв Андрій Анатолійович 143\n",
      "Хлань Сергій Володимирович 145\n",
      "Куліч Валерій Петрович 163\n",
      "Третьяков Олександр Юрійович 177\n",
      "Луценко Юрій Віталійович 242\n",
      "Джемілєв Мустафа 245\n",
      "Ковальчук Віталій Анатолійович 248\n",
      "Павленко Ростислав Миколайович 260\n",
      "Матвієнко Анатолій Сергійович 268\n",
      "Пацкан Валерій Васильович 273\n",
      "Тригубенко Сергій Миколайович 281\n",
      "Насіров Роман Михайлович 285\n",
      "Загорій Гліб Володимирович 286\n",
      "Гвоздьов Михайло Олексійович 287\n",
      "Чепинога Віталій Михайлович 289\n",
      "Грановський Олександр Михайлович 291\n",
      "Кубів Степан Іванович 292\n",
      "Климпуш-Цинцадзе Іванна Орестівна 293\n",
      "Нестеренко Вадим Григорович 387\n",
      "Кутовий Тарас Вікторович 390\n",
      "Насалик Ігор Степанович 396\n",
      "Савченко Олексій Юрійович 408\n",
      "Сольвар Руслан Миколайович 412\n",
      "Матіос Марія Василівна 424\n",
      "Барна Степан Степанович 426\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD8CAYAAABn919SAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvqOYd8AAAEzBJREFUeJzt3V9sXOWZx/HfYyfQBLIhUUxUkhDnAqUEEKUZUf6sEFtTYDeFsFygVqFhEcjlP42oqnR7AVyw4oKWlpZ/ViglwgIhmi2kWZFULgh126aMod0kdiMQgcThT4wCbgoRCfazFzMJyYwdZ2bOOa/PO9+PZI3Pk7Hf58yZ8+Nwzus55u4CAORfS+gGAADJINABIBIEOgBEgkAHgEgQ6AAQCQIdACJBoANAJAh0AIgEgQ4AkZiU5WCzZs3y9vb2LIcEgNzr7e39wN3bxntepoHe3t6uYrGY5ZAAkHtm9vbRPI9TLgAQCQIdACJBoANAJAh0AIgEgQ4AkSDQASASBDoARIJAB4BIjBvoZvYLM9tlZpsPqc00s9+a2evlxxnptgkAeWGjfGXjaI7Qfynp0oraSkk97n6KpJ7yMgA0ubHCO5tQHzfQ3f1lSbsrykslPVH+/glJVyTcFwCgRvWeQ5/t7u+Wv39P0uyxnmhmnWZWNLPi4OBgncMBAMbT8EVRd3dJfoR/73L3grsX2trG/bAwAECd6g30983si5JUftyVXEsAgHrUG+jPS7qm/P01kp5Lph0AyLOxTlaMeRIjUUczbfEpSX+UtNDMBszsOkn3Svq6mb0u6aLyMgBAPspXNsa9wYW7f2uMf+pIuBcAQAP4S1EAiASBDgCRINABIBIEOgBEgkAHgEgQ6AAQCQIdACJBoANAJAh0AIgEgQ4AkSDQASASBDoARIJAB4BIEOgAEAkCHQAiQaADQCQIdACIBIEOAJEg0AEgEgQ6AESCQAeASBDoABAJAh0AIkGgA0AkCHQAiASBDgCRINABIBIEOgBEgkAHgEg0FOhmtsLMtpjZZjN7ysy+kFRjAIDaTKr3B81sjqTbJC1y971m9oykb0r6ZUK9VVm3bp16e3vl7jIzLV68WEuWLElruOBWr16tbdu2HVxesGCBli9fHrAj5N2mTZvU09OjoaEhTZ8+XR0dHTrjjDOiHTeEu+++u6p25513ZjJ2o6dcJkmaYmaTJE2V9E7jLY1u3bp1KhaLcndJkrurWCxq3bp1aQ0ZVGWYS9K2bdu0evXqQB0h7zZt2qS1a9dqaGhIkjQ0NKS1a9dq06ZNUY4bwmhhfqR60uoOdHffKek+SdslvStpyN03JNVYpd7e3prqeVcZ5uPVgfH09PRo//79h9X279+vnp6eKMdtRnUHupnNkLRU0gJJJ0k6zsyuHuV5nWZWNLPi4OBg3Y0eODI/2jqAwx04Qj7aet7HbUaNnHK5SNI2dx909/2S1kg6r/JJ7t7l7gV3L7S1tdU9mJnVVAdwuOnTp9dUz/u4zaiRQN8u6Rwzm2qlVO2Q1J9MW9UWL15cUz3vFixYUFMdGE9HR4cmT558WG3y5Mnq6OiIctxm1Mg59I2SnpX0qqRN5d/VlVBfVZYsWaJCoXDwiNzMVCgUop3lsnz58qrwZpYLGnHGGWfosssuO3hkPH36dF122WWpzzYJNW4IY81myWqWi2V5DrpQKHixWMxsPACIgZn1unthvOfxl6IAEAkCHQAiQaADQCQIdACIBIEOAJEg0AEgEgQ6AESCQAeASBDoABAJAh0AIkGgA0AkCHQAiASBDgCRINABIBIEOgBEgkAHgEgQ6AAQCQIdACJBoANAJAh0AIgEgQ4AkSDQASASBDoARIJAB4BIEOgAEAkCHQAiQaADQCQIdACIBIEOAJFoKNDN7AQze9bM/mZm/WZ2blKNAQBq0+gR+k8lveDuX5J0pqT+xlsaW3e31N4utbSUHru70xwNSI9Z9RfiEDKnJtX7g2Y2XdIFkv5Dktx9n6R9ybRVrbtb6uyUPvmktPz226VlSVq2LK1RgeSNFd5mknu2vSBZoXPKvM53kJl9WVKXpD6Vjs57Jd3u7h+P9TOFQsGLxWJd47W3l16cSvPnS2+9VdevBII40tE4gZ5vaeWUmfW6e2G85zVyymWSpK9Ietjdz5L0saSVozTSaWZFMysODg7WPdj27bXVASBroXOqkUAfkDTg7hvLy8+qFPCHcfcudy+4e6Gtra3uwU4+ubY6AGQtdE7VHeju/p6kHWa2sFzqUOn0SyruuUeaOvXw2tSppToATAShc6rRWS63Suo2s/+T9GVJ/9V4S6Nbtkzq6iqdizIrPXZ1cUEU+TPWeXLOn+df6Jyq+6JoPRq5KAoAzSqLi6IAgAmEQAeASBDoABAJAh0AIkGgA0AkCHQAiASBDgCRINABIBIEOgBEgkAHgEgQ6AAQibrvWBTCaQ+epr4PPv9Ax0WzFmnLzVsCdoS8s7ur7zbhd/IpWajfTetuUldvl4Z9WK3Wqs7FnXpoyUOZjJ2bI/TKMJekvg/6dNqDpwXqCHk3WpgfqQ6M56Z1N+nh4sMa9mFJ0rAP6+Hiw7pp3U2ZjJ+bQK8M8/HqAJC1rt6umupJy02gA8BEd+DI/GjrSSPQASAhrdZaUz1puQn0RbMW1VQHgKx1Lu6sqZ603AT6lpu3VIU3s1zQiLFmszDLBfV6aMlDurFw48Ej8lZr1Y2FGzOb5cIt6ABgguMWdADQZAh0AIgEgQ4AkSDQASASBDoARIJAB4BIEOgAEAkCHQAiQaADQCQIdACIBIEOAJFoONDNrNXMXjOz3yTREJqUWfUXkEfd3VJ7u9TSUnrs7s5s6CTuKXq7pH5J/5TA70IzGiu8zaQMPzwOaFh3t9TZKX3ySWn57bdLy5K0bFnqwzd0hG5mcyUtkbQqmXYAIMd++MPPw/yATz4p1TPQ6CmXn0j6vqSRsZ5gZp1mVjSz4uDgYIPDAcAEtn17bfWE1R3oZvYNSbvcvfdIz3P3LncvuHuhra2t3uEAYOI7+eTa6glr5Aj9fEmXm9lbkp6W9DUzezKRrgAgj+65R5o69fDa1KmlegbqDnR3/4G7z3X3dknflPQ7d786sc7QPMa68MkFUeTNsmVSV5c0f37pov78+aXlDC6ISsnMcgEaR3gjFsuWZRbglRIJdHd/SdJLSfwuAEB9+EtRAIgEgQ4AkSDQASASBDoARIJAB4BIEOgAEAkCHQAiQaADQCQIdACIBIEOAJHI12e5DKyX+h+R9u6SppwonXqDNPeS0F3F5/nzqmuX/yH7PhAP9t1M5OcIfWC99Nd7pb3vS/LS41/vLdWRnNHC/Eh1YDzsu5nJT6D3PyINf3p4bfjTUh3AxMW+m5n8BPreXbXVAUwM7LuZyU+gTzmxtjqAiYF9NzP5CfRTb5Bajz281npsqQ5g4mLfzUx+An3uJdKZK6UpsyVZ6fHMlVwpT9pYs1mY5YJ6se9mxjzDW38VCgUvFouZjQcAMTCzXncvjPe8/ByhAwCOiEAHgEgQ6AAQCQIdACJBoANAJAh0AIgEgQ4AkSDQASASBDoARIJAB4BIEOgAEIm6A93M5pnZi2bWZ2ZbzOz2JBsDANSmkXuKfibpDnd/1cymSeo1s9+6e19CvVXZ/O/Tqmqn//eetIYLLtT67nx0hT7c8Lg0Miy1tGrGxddqznfuT33cZrL1trO1f0f/weXJ807Vwgf+nPq4obZtM+27fdct1Mjudw4ut8w8SYse25rJ2HUfobv7u+7+avn7PZL6Jc1JqrFKo70hjlTPu1Dru/PRFfrwhVWlHV6SRob14QurtPPRFamO20wqw1yS9u/o19bbzk513FDbtpn23cowl6SR3e+o77qFmYyfyDl0M2uXdJakjUn8PoTz4YbHa6qjdpVhPl49KWzb9FWG+Xj1pDUc6GZ2vKRfSfquu/99lH/vNLOimRUHBwcbHQ5pO3D0drR15AfbNnoNBbqZTVYpzLvdfc1oz3H3LncvuHuhra2tkeGQhZbW2urID7Zt9BqZ5WKSHpPU7+4/Tq4lhDTj4mtrqqN2k+edWlM9KWzb9LXMPKmmeuLjN/Cz50v6tqSvmdlfyl//llBfVca6Ih7rlfJQ6zvnO/drxqXXf37U1tKqGZdezyyXBC184M9V4Z3FLJdQ27aZ9t1Fj22tCu8sZ7lwT1EAmOC4pygANBkCHQAiQaADQCQIdACIBIEOAJEg0AEgEgQ6AESCQAeASBDoABAJAh0AIkGgA0AkGrkFXeb+ZeMf9ekhy8dKevGr56Y+7nkb/1hV+0MG44ZaXyBpt/VtVnHP5x/GVZg2TQ8sOj1gR+kJud/m5gi98kWSpE/L9TSNFuZHqicl1PoCSasMc0kq7tmj2/o2B+ooPaH329wEeuWLNF4975ptfRGvyjAfr55noffb3AQ6AODICHQAiERuAv3YGut512zri3gVpk2rqZ5noffb3AT6i189t+pFyeLq8VizWdKe5RJqfYGkPbDo9KrwjnWWS+j9llvQAcAExy3oAKDJEOgAEAkCHQAiQaADQCQIdACIBIEOAJEg0AEgEgQ6AESCQAeASBDoABCJhgLdzC41s61m9oaZrUyqKQBA7eq+BZ2ZtUp6UNLXJQ1IesXMnnf3vqSaq3T99H+WRg757JkW06qh36c13EF/emaD1tz1iHYP7NLMuSfqyrtu0DlXXZz6uLfOu0R7P/rHweUpJxyvn+1Yn/q431u4VB+988HB5RNOmqX7tj6X6phPrviRXn781xoZHlFLa4suuPYKXX3/HamOKYXbttdPO7+qtmrP/6Y+bohtK4V7nUOMG2q/lRo7Qj9b0hvu/qa775P0tKSlybRVrSrMJWnES/UU/emZDVp9y73aveN9yV27d7yv1bfcqz89syHVcSvfFJK096N/6NZ5l6Q6buUOL0kfvfOBvrcwtU2rJ1f8SC+tWqOR4RFJ0sjwiF5atUZPrvhRamNK4bbtaGF+pHpSQmxbKdzrHGLcUPvtAY0E+hxJOw5ZHijX0lEZ5uPVE7Lmrke0b+/hN5Dat/dTrbnrkVTHrXxTjFdPSuUOP149CS8//uua6kkJtW1DCbFtpXCvc4hxQ+23B6R+UdTMOs2saGbFwcHBtIdL3O6BXTXVUbsDR+ZHW08K2zYboV7nZty+jQT6TknzDlmeW64dxt273L3g7oW2trYGhgtj5twTa6qjdi2to78Nx6onhW2bjVCvczNu30b2mFcknWJmC8zsGEnflPR8Mm2NosVqqyfkyrtu0DFTDr8HyTFTjtWVd92Q6rhTTji+pnpSTjhpVk31JFxw7RU11ZMSatuGEmLbSuFe5xDjhtpvD6g70N39M0m3SFovqV/SM+6+JanGKq0a+n11eGcwy+Wcqy7W8p+v1Mx5syUzzZw3W8t/vjL1K+U/27G+6k2QxdXy+7Y+V7WDpz0T4ur779CF11958Ii8pbVFF15/ZeqzXEJt27Fms6Q9yyXEtpXCvc4hxg213x7ALegAYILjFnQA0GQIdACIBIEOAJEg0AEgEgQ6AESCQAeASBDoABAJAh0AIpHpHxaZ2aCktxP4VbMkpfsRcRNLM61vM62rxPrGLMl1ne/u434YVqaBnhQzKx7NX03FopnWt5nWVWJ9YxZiXTnlAgCRINABIBJ5DfSu0A1krJnWt5nWVWJ9Y5b5uubyHDoAoFpej9ABABVyF+hmdqmZbTWzN8xsZeh+0mJm88zsRTPrM7MtZnZ76J6yYGatZvaamf0mdC9pM7MTzOxZM/ubmfWb2bmhe0qLma0ov483m9lTZvaF0D0lycx+YWa7zGzzIbWZZvZbM3u9/Dgj7T5yFehm1irpQUn/KmmRpG+Z2aKwXaXmM0l3uPsiSedIujnidT3U7SrdAasZ/FTSC+7+JUlnKtL1NrM5km6TVHD30yW1qnTLypj8UtKlFbWVknrc/RRJPeXlVOUq0CWdLekNd3/T3fdJelrS0sA9pcLd33X3V8vf71FpZ58Ttqt0mdlcSUskrQrdS9rMbLqkCyQ9Jknuvs/dPwrbVaomSZpiZpMkTZX0TuB+EuXuL0vaXVFeKumJ8vdPSEr3JrnKX6DPkbTjkOUBRR5ykmRm7ZLOkrQxbCep+4mk70saCd1IBhZIGpT0ePkU0yozOy50U2lw952S7pO0XdK7kobcfUPYrjIx293fLX//nqTZaQ+Yt0BvOmZ2vKRfSfquu/89dD9pMbNvSNrl7r2he8nIJElfkfSwu58l6WNl8L/kIZTPHS9V6T9iJ0k6zsyuDttVtrw0nTD1KYV5C/SdkuYdsjy3XIuSmU1WKcy73X1N6H5Sdr6ky83sLZVOpX3NzJ4M21KqBiQNuPuB/+t6VqWAj9FFkra5+6C775e0RtJ5gXvKwvtm9kVJKj/uSnvAvAX6K5JOMbMFZnaMShdWng/cUyrMzFQ6v9rv7j8O3U/a3P0H7j7X3dtV2q6/c/doj+Lc/T1JO8xsYbnUIakvYEtp2i7pHDObWn5fdyjSC8AVnpd0Tfn7ayQ9l/aAk9IeIEnu/pmZ3SJpvUpXyn/h7lsCt5WW8yV9W9ImM/tLufaf7v4/AXtCsm6V1F0+OHlT0rWB+0mFu280s2clvarS7K3XFNlfjJrZU5IulDTLzAYk3SnpXknPmNl1Kn3K7FWp98FfigJAHPJ2ygUAMAYCHQAiQaADQCQIdACIBIEOAJEg0AEgEgQ6AESCQAeASPw/hSuJVQIMH8IAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(len(U)):\n",
    "    x = clusters[i]\n",
    "    y = deputies_fractions[i]\n",
    "    choises[x,y]+=1\n",
    "    plt.scatter(x, y, c=colors[int(deputies_fractions[i])])\n",
    "    if (x==0 and y==2):\n",
    "        print(deputies_names.loc[int(edges.columns.values[i])][0],edges.columns.values[i] )\n",
    "    #if int(edges.columns.values[i]) in deputies_udar['number'].values:\n",
    "        #print(\"удар\",edges.columns.values[i],clusters[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[  9.   7.  33.   1.   0.   4.   1.   2.   0.]\n",
      " [  1.  69.   0.   0.   0.   0.   0.   0.   0.]\n",
      " [  6.   7. 103.   0.   0.   0.   0.   3.   0.]\n",
      " [  5.   0.   2.  25.   0.   0.   0.   0.   0.]\n",
      " [  2.   0.   0.   0.  42.   0.   0.   0.   0.]\n",
      " [  1.   0.   0.   0.   0.  20.   0.   0.   0.]\n",
      " [ 11.   0.   4.   0.   0.   0.  20.   0.   0.]\n",
      " [  1.   3.  12.   1.   0.   0.   0.   1.   0.]\n",
      " [  6.   7.   5.   0.   1.   2.   1.  15.  25.]]\n",
      "<class 'numpy.ndarray'>\n"
     ]
    }
   ],
   "source": [
    "choises_=np.delete(choises, 9, 0)\n",
    "choises_=np.delete(choises_, 7, 0)\n",
    "choises_=np.delete(choises_, 9, 1)\n",
    "choises_=np.delete(choises_, 7, 1)\n",
    "print(choises_)\n",
    "print(type(choises_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe098e8c668>"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sn\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "df_cm = pd.DataFrame(choises_, index = ['cluster '+str(i) for i in (range(9))],\n",
    "                  columns = ['fraction '+str(i)+\" \" for i in range(9)])\n",
    "plt.figure(figsize = (10,7))\n",
    "sn.heatmap(df_cm, annot=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"2.png\" align=\"center\"  alt=\"Drawing\" />"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "33 відділені від Порошенка\n",
      "Недава Олег Анатолійович 35\n",
      "Сабашук Петро Павлович 50\n",
      "Кривохатько Вадим Вікторович 58\n",
      "Гарбуз Юрій Григорійович 81\n",
      "Жолобецький Олександр Олександрович 94\n",
      "Корнацький Аркадій Олексійович 97\n",
      "Голубов Дмитро Іванович 100\n",
      "Сугоняко Олександр Леонідович 120\n",
      "Гордєєв Андрій Анатолійович 143\n",
      "Хлань Сергій Володимирович 145\n",
      "Куліч Валерій Петрович 163\n",
      "Третьяков Олександр Юрійович 177\n",
      "Луценко Юрій Віталійович 242\n",
      "Джемілєв Мустафа 245\n",
      "Ковальчук Віталій Анатолійович 248\n",
      "Павленко Ростислав Миколайович 260\n",
      "Матвієнко Анатолій Сергійович 268\n",
      "Пацкан Валерій Васильович 273\n",
      "Тригубенко Сергій Миколайович 281\n",
      "Насіров Роман Михайлович 285\n",
      "Загорій Гліб Володимирович 286\n",
      "Гвоздьов Михайло Олексійович 287\n",
      "Чепинога Віталій Михайлович 289\n",
      "Грановський Олександр Михайлович 291\n",
      "Кубів Степан Іванович 292\n",
      "Климпуш-Цинцадзе Іванна Орестівна 293\n",
      "Нестеренко Вадим Григорович 387\n",
      "Кутовий Тарас Вікторович 390\n",
      "Насалик Ігор Степанович 396\n",
      "Савченко Олексій Юрійович 408\n",
      "Сольвар Руслан Миколайович 412\n",
      "Матіос Марія Василівна 424\n",
      "Барна Степан Степанович 426\n"
     ]
    }
   ],
   "source": [
    "deputies_udar = pd.read_table(\"./udar2.csv\",index_col=0, encoding=\"utf-8\",sep=',')\n",
    "#print(deputies_udar)\n",
    "\n",
    "print(\"33 відділені від Порошенка\")\n",
    "for i in range(len(U)):\n",
    "    x = clusters[i]\n",
    "    y = deputies_fractions[i]\n",
    "    if (x==0 and y==2):\n",
    "        print(deputies_names.loc[int(edges.columns.values[i])][0],edges.columns.values[i] )\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "позафракційні як Порошенко\n",
      "Герега Олександр Володимирович 151\n",
      "Геращенко Ірина Володимирівна 247\n",
      "Гопко Ганна Миколаївна 296\n",
      "Савченко Надія Вікторівна 371\n",
      "Констанкевич Ірина Мирославівна 471\n",
      "Шевченко Віктор Леонідович 472\n"
     ]
    }
   ],
   "source": [
    "print(\"позафракційні як Порошенко\")\n",
    "for i in range(len(U)):\n",
    "    x = clusters[i]\n",
    "    y = deputies_fractions[i]\n",
    "    if (x==2 and y==0):\n",
    "        print(deputies_names.loc[int(edges.columns.values[i])][0],edges.columns.values[i] )\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "позафракційні як Тимошенко\n",
      "Ярош Дмитро Анатолійович 26\n",
      "Дідич Валентин Володимирович 27\n",
      "Тарута Сергій Олексійович 37\n",
      "Балога Віктор Іванович 45\n",
      "Дерев'янко Юрій Богданович 62\n",
      "Міщенко Сергій Григорович 72\n",
      "Добродомов Дмитро Євгенович 82\n",
      "Матківський Богдан Миронович 86\n",
      "Мусій Олег Степанович 89\n",
      "Береза Борислав Юхимович 171\n",
      "Чумак Віктор Васильович 172\n"
     ]
    }
   ],
   "source": [
    "print(\"позафракційні як Тимошенко\")\n",
    "for i in range(len(U)):\n",
    "    x = clusters[i]\n",
    "    y = deputies_fractions[i]\n",
    "    if (x==6 and y==0):\n",
    "        print(deputies_names.loc[int(edges.columns.values[i])][0],edges.columns.values[i] )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.8622152172114952\n"
     ]
    }
   ],
   "source": [
    "correct=0\n",
    "for i in range(len(U)):\n",
    "    for j in range(len(U)):\n",
    "        if (clusters[i]==clusters[j] and deputies_fractions[i]==deputies_fractions[j]):\n",
    "            correct+=1\n",
    "        if (clusters[i]!=clusters[j] and deputies_fractions[i]!=deputies_fractions[j]):\n",
    "            correct+=1\n",
    "print(correct/(len(U)*(len(U)-1)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## NMF for 9 fractions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(9, 467)"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf = decomposition.NMF(n_components=9, random_state=1)\n",
    "\n",
    "W1 = clf.fit_transform(df)\n",
    "H1 = clf.components_\n",
    "\n",
    "H1.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "clusters=[list(H1[:,i]).index(max(H1[:,i])) for i in range(H1.shape[1])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "clusters=[list(H1[:,i]).index(max(H1[:,i])) for i in range(H1.shape[1])]\n",
    "for i in range(len(deputies)):\n",
    "    if clusters[i]==0:\n",
    "        clusters[i]=2\n",
    "    elif clusters[i]==2:\n",
    "        clusters[i]=3\n",
    "    elif clusters[i]==3:\n",
    "        clusters[i]=4\n",
    "    elif clusters[i]==4:\n",
    "        clusters[i]=5\n",
    "    elif clusters[i]==5:\n",
    "        clusters[i]=10\n",
    "    elif clusters[i]==10:\n",
    "        clusters[i]=0\n",
    "    elif clusters[i]==8:\n",
    "        clusters[i]=6\n",
    "    elif clusters[i]==6:\n",
    "        clusters[i]=8\n",
    "#for i in range(len(deputies)):\n",
    "#    print(deputies[i],clusters[i], deputies_fractions[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s = (11,11)\n",
    "choises=np.zeros(s)\n",
    "choises[0,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(len(U)):\n",
    "    x = clusters[i]\n",
    "    y = deputies_fractions[i]\n",
    "    choises[x,y]+=1\n",
    "    plt.scatter(x, y, c=colors[int(deputies_fractions[i])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.  0.  0.  0.  0.  0.  0.  0.  0.]\n",
      " [ 6.  8. 20.  1.  0.  2.  1.  0.  0.]\n",
      " [ 7. 72. 73.  0.  0.  1.  0.  2.  0.]\n",
      " [ 4.  0.  2. 25.  0.  0.  0.  0.  0.]\n",
      " [ 2.  0.  0.  0. 43.  0.  0.  0.  0.]\n",
      " [ 1.  0.  2.  0.  0. 21.  0.  0.  0.]\n",
      " [11.  0. 10.  0.  0.  0. 20.  0.  0.]\n",
      " [ 1.  4. 26.  1.  0.  0.  0.  0.  0.]\n",
      " [11.  9. 25.  0.  0.  2.  1. 19. 25.]]\n",
      "<class 'numpy.ndarray'>\n"
     ]
    }
   ],
   "source": [
    "choises_=np.delete(choises, 9, 0)\n",
    "choises_=np.delete(choises_, 7, 0)\n",
    "choises_=np.delete(choises_, 9, 1)\n",
    "choises_=np.delete(choises_, 7, 1)\n",
    "print(choises_)\n",
    "print(type(choises_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe096c2f128>"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_cm = pd.DataFrame(choises_, index = ['cluster '+str(i) for i in range(9)],\n",
    "                  columns = ['fraction '+str(i)+\" \" for i in range(9)])\n",
    "plt.figure(figsize = (10,7))\n",
    "sn.heatmap(df_cm, annot=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7832709928224169\n"
     ]
    }
   ],
   "source": [
    "correct=0\n",
    "for i in range(len(U)):\n",
    "    for j in range(len(U)):\n",
    "        if (clusters[i]==clusters[j] and deputies_fractions[i]==deputies_fractions[j]):\n",
    "            correct+=1\n",
    "        if (clusters[i]!=clusters[j] and deputies_fractions[i]!=deputies_fractions[j]):\n",
    "            correct+=1\n",
    "print(correct/(len(U)*(len(U)-1)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## NMF for 3 fractions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3, 467)"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf = decomposition.NMF(n_components=3, random_state=1)\n",
    "\n",
    "W1 = clf.fit_transform(df)\n",
    "H1 = clf.components_\n",
    "\n",
    "H1.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "clusters=[list(H1[:,i]).index(max(H1[:,i])) for i in range(H1.shape[1])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s = (3,11)\n",
    "choises=np.zeros(s)\n",
    "choises[0,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(len(U)):\n",
    "    x = clusters[i]\n",
    "    y = deputies_fractions[i]\n",
    "    choises[x,y]+=1\n",
    "    plt.scatter(x, y, c=colors[int(deputies_fractions[i])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 14.  69. 107.   0.  38.  13.   0.  13.  20.]\n",
      " [ 16.  22.  35.   1.   4.  12.   3.   7.   5.]\n",
      " [ 18.   2.  19.  26.   1.   1.  19.   2.   0.]]\n",
      "<class 'numpy.ndarray'>\n"
     ]
    }
   ],
   "source": [
    "choises_=np.delete(choises, 9, 1)\n",
    "choises_=np.delete(choises_, 7, 1)\n",
    "print(choises_)\n",
    "print(type(choises_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe098ff24e0>"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x216 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_cm = pd.DataFrame(choises_, index = ['cluster '+str(i) for i in range(3)],\n",
    "                  columns = ['fraction '+str(i)+\" \" for i in range(9)])\n",
    "plt.figure(figsize = (10,3))\n",
    "sn.heatmap(df_cm, annot=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5862504710001746\n"
     ]
    }
   ],
   "source": [
    "correct=0\n",
    "for i in range(len(U)):\n",
    "    for j in range(len(U)):\n",
    "        if (clusters[i]==clusters[j] and deputies_fractions[i]==deputies_fractions[j]):\n",
    "            correct+=1\n",
    "        if (clusters[i]!=clusters[j] and deputies_fractions[i]!=deputies_fractions[j]):\n",
    "            correct+=1\n",
    "print(correct/(len(U)*(len(U)-1)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## PCA\n",
    "http://cs229.stanford.edu/notes/cs229-notes10.pdf\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/.pyenv/versions/3.5.6/lib/python3.5/site-packages/IPython/core/interactiveshell.py:2785: DtypeWarning: Columns (0) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  interactivity=interactivity, compiler=compiler, result=result)\n"
     ]
    }
   ],
   "source": [
    "edges = pd.read_table(\"./all_withoutnull.csv\",index_col=0, encoding=\"utf-8\",sep=',')\n",
    "edges=edges.drop(['abscd'])\n",
    "#edges\n",
    "for i in range(len(deputies_fractions)):\n",
    "    k=sum(df[:,i])\n",
    "    if k==0:\n",
    "        print(edges.column.vales[i])\n",
    "for j in range(len(edges)):\n",
    "    k=sum(df[j,:])\n",
    "    if k==0:\n",
    "        print(j)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(15556, 467)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[0, 0, 0, ..., 0, 0, 0],\n",
       "       [0, 0, 0, ..., 0, 0, 0],\n",
       "       [1, 0, 1, ..., 1, 1, 1],\n",
       "       ...,\n",
       "       [1, 1, 1, ..., 0, 0, 0],\n",
       "       [1, 1, 1, ..., 0, 0, 0],\n",
       "       [1, 1, 1, ..., 0, 0, 0]])"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_pca=edges.values\n",
    "for i in range(15556):\n",
    "    for j in range(467):\n",
    "        if (df_pca[i][j]==2):\n",
    "            df_pca[i][j]=-1\n",
    "        elif df_pca[i][j]==1:\n",
    "            df_pca[i][j]=1\n",
    "        else: df_pca[i][j]=0\n",
    "print(df_pca.shape)\n",
    "df_pca"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Zero-mean matrix:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[-0.54769864, -0.59346876, -0.57321934, ..., -0.01799949,\n",
       "        -0.04204166, -0.01054256],\n",
       "       [-0.54769864, -0.59346876, -0.57321934, ..., -0.01799949,\n",
       "        -0.04204166, -0.01054256],\n",
       "       [ 0.45230136, -0.59346876,  0.42678066, ...,  0.98200051,\n",
       "         0.95795834,  0.98945744],\n",
       "       ...,\n",
       "       [ 0.45230136,  0.40653124,  0.42678066, ..., -0.01799949,\n",
       "        -0.04204166, -0.01054256],\n",
       "       [ 0.45230136,  0.40653124,  0.42678066, ..., -0.01799949,\n",
       "        -0.04204166, -0.01054256],\n",
       "       [ 0.45230136,  0.40653124,  0.42678066, ..., -0.01799949,\n",
       "        -0.04204166, -0.01054256]])"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "row_means = np.mean(df_pca, axis=0)\n",
    "#print(row_means)\n",
    "print(\"\\nZero-mean matrix:\")\n",
    "df_pca_normalized = df_pca - row_means\n",
    "df_pca_normalized"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(467, 467)\n",
      "Empirical covariance matrix :\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "matrix([[3875.60761121, 1752.64618154, 1073.17125225, ...,   59.64438159,\n",
       "           61.80509128,   44.1774235 ],\n",
       "        [1752.64618154, 3951.09642582, 1156.0390846 , ...,  -19.17125225,\n",
       "          -27.12856776,  -17.32887632],\n",
       "        [1073.17125225, 1156.0390846 , 3877.60317562, ...,   83.49858575,\n",
       "           70.11455387,   38.9920288 ],\n",
       "        ...,\n",
       "        [  59.64438159,  -19.17125225,   83.49858575, ...,  300.960144  ,\n",
       "          249.22833633,  114.04808434],\n",
       "        [  61.80509128,  -27.12856776,   70.11455387, ...,  249.22833633,\n",
       "          656.50475701,  157.10516842],\n",
       "        [  44.1774235 ,  -17.32887632,   38.9920288 , ...,  114.04808434,\n",
       "          157.10516842,  188.27102083]])"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "C=np.matrix(df_pca_normalized).transpose()*np.matrix(df_pca_normalized)\n",
    "print(C.shape)\n",
    "print(\"Empirical covariance matrix :\")\n",
    "C"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Eigenvalues:\n",
      "[275933.72872228  80519.05527416  54856.3805157   43478.52442308\n",
      "  29715.64552926  22498.57114215  15102.66764085  14545.02952342\n",
      "  13076.54526706  11540.15543275  10879.82545817  10306.5667732\n",
      "   8877.08385538   8465.70053447   7686.89994644   7385.73276224\n",
      "   7146.63083609   6870.32599848   6820.99473125   6468.9306356\n",
      "   6218.26516142   6073.04218613   6005.64981988   5729.77441812\n",
      "   5620.22069648   5498.28899894   5445.53613052   5351.48680301\n",
      "   5231.71965025   5149.13112313   5110.8939735    5022.05731887\n",
      "   4918.9395288    4885.02648794   4826.12138512   4708.55489545\n",
      "   4659.23711613   4618.84080639   4552.57508256   4498.41960535]\n",
      "Eigenvectors:\n",
      "[[ 0.06864691  0.01396696 -0.02517853 ...  0.00842258  0.00640981\n",
      "   0.00817381]\n",
      " [ 0.08090768  0.02340621  0.00450904 ...  0.01286883 -0.09324044\n",
      "   0.03033991]\n",
      " [ 0.05500868  0.02608918 -0.0064855  ...  0.00460768  0.00129636\n",
      "   0.04888784]\n",
      " ...\n",
      " [ 0.0017828   0.00320455 -0.00795783 ...  0.01239602  0.00248064\n",
      "  -0.01245441]\n",
      " [ 0.00032249  0.01386895 -0.01495404 ...  0.00187553  0.0068585\n",
      "  -0.01208879]\n",
      " [ 0.00123915  0.00239192 -0.00473774 ... -0.00175429  0.00169753\n",
      "   0.00184787]]\n"
     ]
    }
   ],
   "source": [
    "from numpy import linalg as LA\n",
    "\n",
    "w, v = LA.eig(C)\n",
    "print(\"Eigenvalues:\")\n",
    "print(w[0:40])\n",
    "print(\"Eigenvectors:\")\n",
    "print(v)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(467, 467)"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "v.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(15556,)\n"
     ]
    }
   ],
   "source": [
    "s = (15556,)\n",
    "Y1=np.zeros(s)\n",
    "print(Y1.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'numpy.float64'>\n",
      "<class 'numpy.ndarray'>\n"
     ]
    }
   ],
   "source": [
    "print(type(v[0,1]))\n",
    "print(type(df_pca_normalized[:,1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The variance along the first component\n",
      "0.6131148329331308\n",
      "The variance along the second component\n",
      "0.38882807295698424\n"
     ]
    }
   ],
   "source": [
    "for i in range(len(deputies)):\n",
    "    Y1+=v[0,i]*df_pca_normalized[:,i]\n",
    "#print(Y1)\n",
    "Y2=np.zeros(s)\n",
    "for i in range(len(deputies)):\n",
    "    Y2+=v[1,i]*df_pca_normalized[:,i]\n",
    "#print(Y2)\n",
    "print(\"The variance along the first component\")\n",
    "print(np.std(Y1))\n",
    "print(\"The variance along the second component\")\n",
    "print(np.std(Y2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[-0.84056034 -0.9297729  -0.01448281 ...  1.34913216  1.27240651\n",
      "  1.00328583]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "15556"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(Y1)\n",
    "len(Y1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "colors = ['#630C3A', '#39C8C6', '#D3500C', '#FFB139', 'red', 'green', 'blue', 'black', 'gray', 'orange', 'yellow', 'rose']\n",
    "for i in range(len(deputies)):\n",
    "    x = v[i,0]\n",
    "    y = v[i,1]\n",
    "    plt.scatter(x, y, c=colors[int(deputies_fractions[i])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pandas import DataFrame\n",
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<iframe id=\"igraph\" scrolling=\"no\" style=\"border:none;\" seamless=\"seamless\" src=\"https://plot.ly/~anastasykhab/2.embed\" height=\"525px\" width=\"100%\"></iframe>"
      ],
      "text/plain": [
       "<plotly.tools.PlotlyDisplay object>"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import plotly \n",
    "plotly.tools.set_credentials_file(username='anastasykhab', api_key='1A5tXzke659VuNNyl0YF')\n",
    "import plotly.plotly as py\n",
    "import plotly.graph_objs as go\n",
    "\n",
    "x, y, z = (v[:,0], v[:,1], v[:,2])\n",
    "s=(len(deputies), 1 )\n",
    "c=np.zeros(s)\n",
    "for i in range(len(deputies)):\n",
    "    c[i,0]=int(deputies_fractions[i])\n",
    "trace1 = go.Scatter3d(\n",
    "    x=x,\n",
    "    y=y,\n",
    "    z=z,\n",
    "    mode='markers',\n",
    "    marker=dict(\n",
    "        size=5,\n",
    "        color=c, \n",
    "        colorbar=dict(\n",
    "                title='Colorbar'\n",
    "            ),# set color to an array/list of desired values\n",
    "        colorscale=[[0,'#e6194b'], [0.1,'#3cb44b'], [0.2, '#ffe119'], [0.3,'#4363d8'], [0.4,'#f58231'], \n",
    "                    [0.5,'#911eb4'], [0.6, '#46f0f0'], [0.7, '#f032e6'], [0.8,'#bcf60c'], \n",
    "                    [0.9, '#fabebe'],[1,'#008080']],  # choose a colorscale\n",
    "        opacity=0.8\n",
    "    )\n",
    ")\n",
    "\n",
    "data = [trace1]\n",
    "layout = go.Layout(\n",
    "    margin=dict(\n",
    "        l=0,\n",
    "        r=0,\n",
    "        b=0,\n",
    "        t=0\n",
    "    )\n",
    ")\n",
    "fig = go.Figure(data=data, layout=layout)\n",
    "py.iplot(fig, filename='3d-scatter-colorscale')\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PCA(copy=True, iterated_power='auto', n_components=None, random_state=None,\n",
       "  svd_solver='auto', tol=0.0, whiten=False)"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import decomposition\n",
    "pca = decomposition.PCA()\n",
    "pca.fit(df_pca_normalized)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "#print(pca.explained_variance_)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(15556, 3)"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca.n_components = 3\n",
    "df_pca_reduced = pca.fit_transform(df_pca_normalized)\n",
    "df_pca_reduced.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 7.69877053,  1.24879452, -0.51430799],\n",
       "       [ 7.4718291 ,  2.95074828,  1.1384467 ],\n",
       "       [ 0.79576579, -0.50485595, -1.3607198 ],\n",
       "       ...,\n",
       "       [-6.92106776, -2.92898272,  2.64179696],\n",
       "       [-6.69503129, -2.75546864,  3.11244029],\n",
       "       [-6.81120402, -4.16747651,  1.94616682]])"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_pca_reduced"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "clusters_pca=np.zeros((len(deputies),1))\n",
    "for dep in range(len(deputies)):\n",
    "    #print(\"_________________________________________________________________\")\n",
    "    i=0\n",
    "    error=np.sqrt(sum((df_pca_reduced[:,0]-df_pca_normalized[:,dep])**2)/(467*w[0]))\n",
    "    for j in range(3):\n",
    "        error_j=np.sqrt(sum((df_pca_reduced[:,j]-df_pca_normalized[:,dep])**2)/(467*w[j]))\n",
    "        if (error_j<error):\n",
    "            i=j\n",
    "            error=error_j\n",
    "            #print(error)\n",
    "            #print(i)\n",
    "    clusters_pca[dep,0]=i      "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.]])"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clusters_pca"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "s = (11,11)\n",
    "choises_pca=np.zeros(s)\n",
    "choises_pca[0,0]\n",
    "for i in range(len(U)):\n",
    "    x = int(clusters_pca[i])\n",
    "    y = deputies_fractions[i]\n",
    "    choises_pca[x,y]+=1\n",
    "    plt.scatter(x, y, c=colors[int(deputies_fractions[i])])  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe08de652b0>"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_pca_cm = pd.DataFrame(choises_pca, index = ['cluster '+str(i) for i in range(11)],\n",
    "                  columns = ['fraction '+str(i) for i in range(11)])\n",
    "plt.figure(figsize = (10,7))\n",
    "sn.heatmap(df_pca_cm, annot=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"2.png\" align=\"center\"  alt=\"Drawing\" />"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(15556, 11)"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca.n_components = 11\n",
    "df_pca_reduced = pca.fit_transform(df_pca_normalized)\n",
    "df_pca_reduced.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "clusters_pca=np.zeros((len(deputies),1))\n",
    "for dep in range(len(deputies)):\n",
    "    #print(\"_________________________________________________________________\")\n",
    "    i=0\n",
    "    error=np.sqrt(sum((df_pca_reduced[:,0]-df_pca_normalized[:,dep])**2)/(467*w[0]))\n",
    "    for j in range(11):\n",
    "        error_j=np.sqrt(sum((df_pca_reduced[:,j]-df_pca_normalized[:,dep])**2)/(467*w[j]))\n",
    "        if (error_j<error):\n",
    "            i=j\n",
    "            error=error_j\n",
    "            #print(error)\n",
    "            #print(i)\n",
    "    clusters_pca[dep,0]=i   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "s = (11,11)\n",
    "choises_pca=np.zeros(s)\n",
    "choises_pca[0,0]\n",
    "for i in range(len(U)):\n",
    "    x = int(clusters_pca[i])\n",
    "    y = deputies_fractions[i]\n",
    "    choises_pca[x,y]+=1\n",
    "    plt.scatter(x, y, c=colors[int(deputies_fractions[i])])  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe08d7c9710>"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_pca_cm = pd.DataFrame(choises_pca, index = ['cluster '+str(i) for i in range(11)],\n",
    "                  columns = ['fraction '+str(i) for i in range(11)])\n",
    "plt.figure(figsize = (10,7))\n",
    "sn.heatmap(df_pca_cm, annot=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SPECTRAL CLUSTERING\n",
    "http://papers.nips.cc/paper/2092-on-spectral-clustering-analysis-and-an-algorithm.pdf\n",
    "https://scikit-learn.org/stable/modules/generated/sklearn.cluster.SpectralClustering.html\n",
    "\n",
    "https://arxiv.org/pdf/0711.0189.pdf\n",
    "\n",
    "One  standard  approach  is  based  on  generative  models,  in  which  algorithms  such  as  EM  are  used  to  learn  a  mixture  density.  These  approaches  suffer  from  several  drawbacks.  First,  to  use  parametric  density  est imators,  harsh  simplifying  assumptions  usually  need  to  be  made  (e.g.,  that  the  density  of  each  cluster  is  Gaussian) .  Second ,  the  log  likelihood  can  have  many  local  minima  and  therefore  multiple  restarts  are  required  to  find  a  go\n",
    "Many  empirical  successes  of  spectral  clustering  methods cluster  points  using  eigenvectors  of  matrices  derived  from  the  data. But, many  of  these  algorithms  have  no  proof  that  they  will  actually  compute  a  reasonable  clustering.  \n",
    "This  spectral  clustering  algorithm  works using  tools  from  matrix  perturbation  theory.\n",
    "\n",
    "<img src=\"1.png\" align=\"center\"  alt=\"Drawing\" />"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/.pyenv/versions/3.5.6/lib/python3.5/site-packages/IPython/core/interactiveshell.py:2785: DtypeWarning:\n",
      "\n",
      "Columns (0) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "edges_spec = pd.read_table(\"./all_withoutnull.csv\",index_col=0, encoding=\"utf-8\",sep=',')\n",
    "edges_spec =edges_spec .drop(['abscd'])\n",
    "#edges_spec "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(467, 15556)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[ 0,  0,  0,  0,  0,  0,  0,  0,  1,  1,  1,  1,  0,  1],\n",
       "       [ 0,  1,  1,  1,  1,  1,  1,  1,  0,  1,  1,  1,  0,  0],\n",
       "       [ 0,  0,  1,  1,  1,  1,  1,  0,  1,  1,  0,  1,  0,  1],\n",
       "       [ 1,  0,  0,  0,  0,  0,  1,  0,  1,  1,  1,  1,  0,  1],\n",
       "       [ 0,  0,  0,  1,  1,  1,  1,  0,  1,  0,  0,  0,  0,  0],\n",
       "       [ 0,  1,  1,  1,  0,  0,  1,  0,  1,  1,  1,  1,  1,  1],\n",
       "       [ 0,  0,  0,  0,  0,  0,  0,  0,  0,  1,  1,  0,  0,  0],\n",
       "       [ 0,  0,  1,  1,  1,  1,  1,  1,  1,  0,  0,  1,  0,  0],\n",
       "       [ 0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0],\n",
       "       [ 0,  0,  1,  0,  0,  0,  0,  1,  0,  1,  1,  1,  0,  0],\n",
       "       [ 0,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1],\n",
       "       [ 0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0],\n",
       "       [ 0,  0,  0,  0,  0,  0,  0,  0,  0,  1,  0,  0, -1,  0],\n",
       "       [ 0,  0,  0,  1,  1,  1,  0,  0,  0,  0,  0,  0,  0,  0]])"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_spec=edges_spec.values.transpose()\n",
    "for i in range(467):\n",
    "    for j in range(15556):\n",
    "        if (df_spec[i][j]==2):\n",
    "            df_spec[i][j]=-1\n",
    "        elif df_spec[i][j]==1:\n",
    "            df_spec[i][j]=1\n",
    "        else: df_spec[i][j]=0\n",
    "print(df_spec.shape)\n",
    "df_spec[1:15,1:15]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "#np.exp(-d(df_spec,df_spec) ** 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/.pyenv/versions/3.5.6/lib/python3.5/site-packages/sklearn/manifold/spectral_embedding_.py:234: UserWarning:\n",
      "\n",
      "Graph is not fully connected, spectral embedding may not work as expected.\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([ 5,  7,  5,  5,  5,  5,  5,  7,  1,  5,  7,  5,  5,  8, 10,  6, 10,\n",
       "        1,  1,  1,  1,  1,  5,  7,  5,  5,  5,  5,  5,  5,  5,  7,  5,  5,\n",
       "        0,  5,  5,  5,  8,  7,  0,  5,  5, 10,  5,  5,  0,  6,  0,  5,  5,\n",
       "        5,  1,  6,  0,  6,  5,  0,  5,  0,  0,  0,  7,  6,  6,  5,  7,  0,\n",
       "        1,  6,  1,  5,  1,  5,  7,  5,  0,  0,  6,  0,  0,  5,  6,  5,  5,\n",
       "        0,  5,  6,  5,  1,  0,  5,  0,  7,  6,  0,  5,  6,  1,  0,  7,  6,\n",
       "        7,  8,  0,  5,  5,  5,  0,  8,  5,  1,  1,  0,  1,  5,  1,  5,  5,\n",
       "        5,  1,  5,  1,  5,  5,  5,  0,  0,  5,  0,  5,  5,  6,  0,  5,  5,\n",
       "        5,  7,  5,  5,  8,  5,  5,  1,  1,  7,  0,  8,  7,  1,  0,  1,  6,\n",
       "        7,  6,  7,  7,  5,  5, 10,  5,  5,  7,  1,  1,  7,  6,  5,  7,  0,\n",
       "        5,  6,  5,  0,  7,  6,  5,  8,  7,  5,  0,  5,  5,  0,  3,  4,  5,\n",
       "        5,  9,  0,  0,  6,  0,  1,  7,  5,  0,  6,  6,  5,  6,  1,  5,  0,\n",
       "        0, 10,  5,  5,  5,  1,  1,  6,  5,  6,  6,  7,  0,  5,  5,  7,  5,\n",
       "        8,  0,  7,  6,  1,  7,  6,  1,  7,  0,  0,  5,  6,  0,  5,  5,  5,\n",
       "        6,  5,  7,  6,  5,  0,  5,  3,  5,  5,  0,  5,  0,  7,  1,  7,  8,\n",
       "        0,  5,  5,  1,  0,  4,  0,  0,  1,  5,  1,  6,  5,  6,  5,  1,  7,\n",
       "        5,  1,  6,  8,  1,  7,  5,  7,  0,  5,  5,  5,  0,  5,  8,  5,  5,\n",
       "        5,  0,  6,  7,  0,  5,  6,  6,  1,  7,  7,  1,  7,  0,  6,  5,  6,\n",
       "        0, 10,  7,  6,  7,  6,  0, 10,  5,  0,  5,  5,  7,  5,  0,  5,  6,\n",
       "        5,  5,  5,  5,  7,  6,  0,  0,  5,  6,  7,  7,  5,  0,  5,  5,  0,\n",
       "        5,  0,  5,  6,  7,  6,  5,  0,  5,  7,  5,  6,  2,  5,  5,  5,  0,\n",
       "        0,  0,  1,  0,  8,  5,  5,  1,  5,  7,  5,  8,  7,  7,  0,  7,  5,\n",
       "        2,  5,  7,  1,  5,  5,  7,  5,  1,  5,  7,  1,  7,  0,  5,  0,  6,\n",
       "        5,  6,  7, 10,  1,  0,  5, 10,  0,  5,  1, 10,  7,  7,  5,  0,  0,\n",
       "        5,  0,  6,  5,  0,  6,  5,  0, 10, 10,  6,  5,  8,  7,  5,  6,  7,\n",
       "        5,  1, 10,  7,  0,  0,  5,  0,  7,  5,  5,  5,  0,  7,  5,  0,  7,\n",
       "        7,  6,  0,  5,  1,  7,  8,  1,  6,  1,  0,  6,  7,  0,  7, 10,  1,\n",
       "        7,  0,  5,  5,  5,  5,  5,  5], dtype=int32)"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.cluster import SpectralClustering\n",
    "clustering_1 = SpectralClustering(n_clusters=11,\n",
    "        assign_labels=\"kmeans\"\n",
    "                                ,random_state=0\n",
    "                               ).fit(df_spec)\n",
    "clustering_1.labels_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SpectralClustering(affinity='rbf', assign_labels='kmeans', coef0=1, degree=3,\n",
       "          eigen_solver=None, eigen_tol=0.0, gamma=1.0, kernel_params=None,\n",
       "          n_clusters=11, n_init=10, n_jobs=1, n_neighbors=10,\n",
       "          random_state=0)"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clustering_1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "467"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(clustering_1.labels_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "clusters_spec=clustering_1.labels_\n",
    "#for i in range(len(deputies)):\n",
    "#    hghg=1\n",
    "#for i in range(len(deputies)):\n",
    "#    print(deputies[i],clusters_spec[i], deputies_fractions[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "s = (11,11)\n",
    "choises_spec=np.zeros(s)\n",
    "choises_spec[0,0]\n",
    "for i in range(len(U)):\n",
    "    x = int(clusters_spec[i])\n",
    "    y = deputies_fractions[i]\n",
    "    choises_spec[x,y]+=1\n",
    "    plt.scatter(x, y, c=colors[int(deputies_fractions[i])])  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe0979930f0>"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_spec_cm = pd.DataFrame(choises_spec, index = ['cluster '+str(i) for i in range(11)],\n",
    "                  columns = ['fraction '+str(i) for i in range(11)])\n",
    "plt.figure(figsize = (10,7))\n",
    "sn.heatmap(df_spec_cm, annot=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/.pyenv/versions/3.5.6/lib/python3.5/site-packages/IPython/core/interactiveshell.py:2785: DtypeWarning:\n",
      "\n",
      "Columns (0) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "edges_spec = pd.read_table(\"./all_withoutnull.csv\",index_col=0, encoding=\"utf-8\",sep=',')\n",
    "edges_spec =edges_spec .drop(['abscd'])\n",
    "#edges_spec "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(467, 15556)\n"
     ]
    }
   ],
   "source": [
    "df_spec=edges_spec.values.transpose()\n",
    "for i in range(467):\n",
    "    for j in range(15556):\n",
    "        if (df_spec[i][j]==2):\n",
    "            df_spec[i][j]=0\n",
    "        elif df_spec[i][j]==1:\n",
    "            df_spec[i][j]=1\n",
    "        else: df_spec[i][j]=0\n",
    "print(df_spec.shape)\n",
    "#df_spec[1:15,1:15]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/user/.pyenv/versions/3.5.6/lib/python3.5/site-packages/sklearn/manifold/spectral_embedding_.py:234: UserWarning:\n",
      "\n",
      "Graph is not fully connected, spectral embedding may not work as expected.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "clustering_3 = SpectralClustering(n_clusters=11,\n",
    "        assign_labels=\"kmeans\"\n",
    "                                #,random_state=0\n",
    "                               ).fit(df_spec)\n",
    "#clustering_3.labels_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "clusters_spec=clustering_3.labels_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "s = (11,11)\n",
    "choises_spec=np.zeros(s)\n",
    "choises_spec[0,0]\n",
    "for i in range(len(U)):\n",
    "    x = int(clusters_spec[i])\n",
    "    y = deputies_fractions[i]\n",
    "    choises_spec[x,y]+=1\n",
    "    plt.scatter(x, y, c=colors[int(deputies_fractions[i])])  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe08da92b00>"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkgAAAHMCAYAAAAu11f8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvqOYd8AAAIABJREFUeJzs3Xl4VPXZ//H3nQXCvggIYREFXIoLCuJWlKWitQrWh0erD4rt41a1Ravgxq8u1Ipi8QEVrXtUsICiIqKCIoooAoogouyLbAJGkBCELPfvjxliEkLWmZwz+Hld17mumTNn+cw5k+TO93u+Z8zdEREREZGfJQUdQERERCRsVCCJiIiIFKMCSURERKQYFUgiIiIixahAEhERESlGBZKIiIhIMSqQRERE5IBiZslmNt/MJkef9zKzz83sCzP7yMzal7UNFUgiIiJyoBkIfF3o+WPA/7h7J2AsMKSsDahAEhERkQOGmbUCfgc8VWi2A/WjjxsAG8raTkrso4mIiIgE5v+AwUC9QvOuAKaY2S7gR+Dksjbyiy+Qph18Uai/a+XYX20KOkKZGk94NugIpepz/HVBRyjT7Tl1go5QqhE1dgcdoUyr92QGHaFMX2auDjpCqc48+NigI5Rp2ncLg45Qqi5NOgQdoUyzN8yw6txfztaVMfs7W6Npu6uBqwrNesLdn9j7xMzOBTa7+2dm1r3QcjcC57j7p2Y2CBhBpGjar198gSQiIiKJIVoMPVHKIqcBfczsHCANqG9mbwJHuvun0WXGAW+XtS9dgyQiIiLxk58Xu6kM7n6bu7dy97bAH4DpQF+ggZkdHl3sTIpewF0itSCJiIjIAcvdc83sSuAVM8sHfgD+VNZ6KpBEREQkfjw/mN26zwBmRB+/CrxakfVVIImIiEj85AdTIFWVCiQRERGJGw+oBamqdJG2iIiISDFqQRIREZH4URebiIiISDHqYhMRERE5MKgFSUREROKnHDd4DCMVSFVQM/0gjn7kOmo0aQDurHvxPb598i06/P1/aNq7M/k5uexa/R1fDXyM3B+zqz1fUtOmNLjtDpIaNQKc7MlvsOuVV0hp1456N96E1apF/qZNbL93KJ5d/fkKy8vL46L//SvNmjZh9PC7mT1vPv969Gny853atdO4946baNMqPdCMe/X9U1/OvuRsDOPtl97mtadfCzRPjfSDOOLhv1CjaQPcYdML09jw1BQOGfwHDjr7RDw/n5ytP7J04CPs+e6HQDJeP/yvdOl1Itu/387AM68HoG6Dutw0ejDNWh3M5nXf8eC197Nz+85A8hVXt35d7hxxK+2OOAx35+4b/8nCz74KOlYRZ/XuzogR95CclMQzz77EA8MfDTpSEWH7OSlJ2I9hm3at+cfjdxY8b9mmBU8Mf5ZxT70cYKpKUBdbhJndZWY3V2K9hmZ2bQz2f6iZfWpmy81snJnVqOo298dz81h65wt8cvpNzDlnCK3/2Js6h7fk+w++5JMzbmZ2j8HsXLGRtn89P14RSpeXx47HHuX7Pw4g89o/U7vv70k+5BDq3zyYrCf/Teb//pGfPppJ7Yv+EEy+Ql6c8DqHtW1T8Hzog48y7M7BvJLxKL87swf/fu6lANP97JAjDuHsS87mhnNv4NqzrqVrr660aNsi0Eyem8fKuzL47PQbWXDObbT449nUPrwV60a/zuc9b2L+bwaROe0z2vztvwPLOH3Ce9xz2V1F5l1wXT++nLWQ6864mi9nLeSCa/sFE64Eg/9xAx9P/5QLul3CRb0GsHLZmqAjFZGUlMSokfdy7nn9Oea4Hlx00fkcdVR4viQ1jD8nxYX9GAKsXfEtl515BZedeQWXn3UVP+3azQdvzQw61i9GmK5BaghUqECyiOLv4X7gIXdvT+R24v8bo3z72LN5Gzu+XAVA3s6f2LlsPTWbNybzg4V4XqRi3v7ZMtLSD4pXhFLlZ2aSu2wZAL5rF7lr15DcpCnJrVqRs2BB5D3Mm0va6WcEkm+vTZu38OHHc/iv884qmGfAzp2RVq0dWTtp2iSYY1hc6/atWTJ/Cbt/2k1+Xj5ffvolp519WqCZcjZvY2ehz+GuZeup0bwxeVm7CpZJql0TiNkXalfY4jlfsWPbjiLzup55Eu+//B4A77/8Hif1PjmIaPuoW68OJ5x8HK+OfQOA3Jxcsn7MCjhVUV1PPJ4VK1azatVacnJyGD/+dfoU+vkJWhh/TooL+zEsrku3E1i/Zj2b1n8XdJSKy8+P3VSNqlQgmdllZrbQzBaY2QslvD7DzLpEHzcxs9XRxx3NbI6ZfRFdvwMwDGgXnTc8utwgM5sbXebu6Ly2ZrbEzJ4HFgGtC+3PgJ7A3vbHDKBamm/SWjel3tGHsv3z5UXmt7ykB1vfm18dEUqVdHBzUtt3IOfrxeSuXk3N034NQFr3HiQ1axZotvtH/pu/Xfu/FK517771Bv5889/pdX5/3njnPa64NLjWj8LWLFlDx64dqdewHjXTanJijxNpmt406FgFarZuSp2j27Lj80hhfMitF9P1s8dp9l/dWPPAuIDTFdWwSUN+2Bzp8vth8w80bNIw4EQR6W3S+eH7bdw98g5emvYsf//XraTVTgs6VhHpLZvz7boNBc/Xrd9IenrzABMVFfafEwj/MSzuzL49mfra9KBjVIp7fsym6lTpAsnMOgJDgJ7ufhwwsAKrXwOMdPdOQBdgHXArsMLdO7n7IDPrDXQAugKdgM5mdnp0/Q7AaHfv6O6F274PAra5e270+TqgZSXfYrkl167JcU//jaX/L6PIf+2H3vB7PDePTa98FO8IpbK0WjS85x52PPownp3Njw/cT62+59P4309gtWpBTk5g2WbM+pTGjRrS8ciiTdvPj3uVxx68h/dee5Hzz+nNA6OeDChhUd8u/5YJoydw75h7GfriUFYuXkl+Xjj615Nqp3HUUzez8u/PFXwO1wx7iTmdr2HzKzNp8aezA05YuuDat4pKSUnmyGMOZ8Jzr3LxmX9kV/Yu/nT9pUHHSihh/jlJRCmpKXTrfRrT35gRdJRflKpcpN0TmODuWwHcPbMC634C3GFmrYCJ7r4s0vhTRO/otLf5pS6RwmgtsMbdZ1c2uJldBVwFMLBeZ35Xq11lN4WlJHPsMzex8ZWP2DxlTsH8FhedQZMzT+CzfkMrve2YSE6mwT338NO777J7ZqTvOu/btWwbHLlMLLlVK2qefEpg8eYvXMyMj2Yz85O57N6Tw86d2fz55r+zas23HNvxSAB+2+t0rr5pSGAZi5s6bipTx00FYMAtA9i6cWvAiSKfw189fTNbJs7k+ymf7vP6lokz6TjmdtYOHx9AupJt27qNRs0a8cPmH2jUrBHbt24LOhIA323YzOaNW1g0fzEA706ewR//0j/gVEVtWL+J1oUGLbRq2YINGzYFmGhfYfw5KSwRjuFep/Q8iSVfLiVzazCDLKosQW8UGe9rkHIL7aOgjdrdxwJ9gF3AFDPrWcK6BtwXbVHq5O7t3f3p6Gv7G+ryPdDQzPYWfq2A9cUXcvcn3L2Lu3epSnEE8KuHrmHnsvWs/febBfMO6nEcba/rwxeXPUD+rj1V2n5V1R98C7lr1pA94ec/jNYw2pVhRp1LLyP7jUkBpYMb//xH3nvtRaa+ksHwu2+la+fjeHjYnWTtzGb12nUAfDx3Pocd0qaMLVWfBgc1AKBpelNOO/s0Zrw2I9hAQIeHriV72TrW/3tywby0Q3/uLjjo7BPZtXyfH4VAzZ02hx79egHQo18v5kzbt7ALwvdbMtm0fjOHtIt85rp268zKpauDDVXM3Hlf0L79obRt25rU1FQuvLAvb0yeGnSsIsL4c1JYIhzDvXqf34upr70XdIzK8/zYTdWoKi1I04FXzWyEu39vZo1LaEVaDXQG5gAFQ1TM7DBgpbuPMrM2wLHAAqBeoXXfAYaa2Rh3zzKzlkCpfUHu7mb2fnRf/wEGAK9X4T2WqmHXI0i/8HR2LF7Dye/dD8Dyf77EEff+kaQaKXQeH2n12P7ZMr4e/FS8YuxX6tHHUKv3WeSsWEHjJyP7z3rqSZJbtaJ2398DsHvmh/z01pRqz1aalJRk7rrlr9x4x71YklG/Xl2G3nZj0LEKDHliCPUb1ic3N5fRQ0az88dgh6bX73okB//3GexcvIbj3x0OwOr7xtL84l7Uap8O+c5P67awfPATgWX828M30/GUY6jfqD5Pfvos/xkxlomjX+bmx26h10VnsmX9Zh788/2B5Svu/jse4p+j7yQlNYX1azZw5w3/DDpSEXl5eQy8YQhT3hxLclISz2WMY/HipUHHKiJsPyfFJcIxBEirlUbXbp0ZNvhfQUf5xTH3yvf8m9kAYBCQB8x398vN7C4gy90fNLMjgfHR198E+rt7WzO7FbiUSMGzCbjE3TPNbCyRYumt6HVIA4ErorvLAvpHtzXZ3Y/eT6bDiBRHjYl0z/V39937ew/TDr4oLJc+lOjYX4WzybewxhOeDTpCqfocf13QEcp0e06doCOUakSN/f4IhcbqPRXp5Q/Gl5mrg45QqjMPPjboCGWa9t3CoCOUqkuTcN0qoCSzN8zY55qWeNr9zQcx+ztb88gzqi17lW4U6e4ZREaKFZ53V6HH3xApePYaEp0/jMioteLbu6TY85HAyBJ2XWJxFF1nJZELu0VERCRoulGkiIiIyIFBXzUiIiIi8ZOgo9hUIImIiEj8qItNRERE5MCgFiQRERGJH3WxiYiIiBTlnhd0hEpRF5uIiIhIMWpBEhERkfhJ0Iu0VSCJiIhI/CToNUjqYhMREREpRi1IIiIiEj/qYhMREREpJj8xR7GZe6i/zD7uUmq0/GUfAJFySE3W/1KxkJOXG3QEEXL3rLfq3N9Pc1+J2d/ZtBP/q9qy67eeiIiIxI+62ERERESK0Sg2ERERkQODWpBEREQkftTFJiIiIlKMuthEREREDgxqQRIREZH4SdAWJBVIIiIiEjfuiXmjSHWxiYiIiBSjFiQRERGJH3WxiYiIiBSToMP81cUmIiIiUowKpBg6q3d3vlr0Id8s/ojBg64LOs4+wp4PlDEWwp7v8ceHs2bNZ8ybNzXoKPuVCBnDfp4h/BnDng8SI2OZ8vNjN1WjmBZIZnaXmd1cifUamtm1Mdj/9Wa23MzczJpUdXsVkZSUxKiR93Luef055rgeXHTR+Rx1VIfqjFCqsOcDZYyFsOcDeOGFCfTtOyDoGKUKe8ZEOM9hzxj2fJAYGcvF82M3VaOwtCA1BCpUIFlE8fyzgN8Aa2IVrLy6nng8K1asZtWqteTk5DB+/Ov0Oe+s6o6xX2HPB8oYC2HPBzBr1hwyM7cFHaNUYc+YCOc57BnDng8SI+OBrNIFkpldZmYLzWyBmb1QwuszzKxL9HETM1sdfdzRzOaY2RfR9TsAw4B20XnDo8sNMrO50WXujs5ra2ZLzOx5YBHQuvA+3X2+u6+u7HuqivSWzfl23YaC5+vWbyQ9vXkQUUoU9nygjLEQ9nwSG4lwnsOeMez5IDEylkuCdrFVahSbmXUEhgCnuvtWM2tcgdWvAUa6+xgzqwEkA7cCR7t7p+j2ewMdgK6AAZPM7HRgbXT+AHefXZnsIiIiUo0SdBRbZYf59wQmuPtWAHfPrMC6nwB3mFkrYKK7LzOz4sv0jk7zo8/rEimM1gJrqlocmdlVwFUAltyApKQ6VdkcABvWb6J1q/SC561atmDDhk1V3m6shD0fKGMshD2fxEYinOewZwx7PkiMjAeyeF6DlFto+2l7Z7r7WKAPsAuYYmY9S1jXgPvcvVN0au/uT0df21nVYO7+hLt3cfcusSiOAObO+4L27Q+lbdvWpKamcuGFfXljcnhGwIQ9HyhjLIQ9n8RGIpznsGcMez5IjIzl8kvqYgOmA6+a2Qh3/97MGpfQirQa6AzMAfrtnWlmhwEr3X2UmbUBjgUWAPUKrfsOMNTMxrh7lpm1BHIqmbVa5OXlMfCGIUx5cyzJSUk8lzGOxYuXBh2rQNjzgTLGQtjzAWRkjKJbt1No0qQRy5fPZujQh8jIGBd0rCLCnjERznPYM4Y9HyRGxnJJ0Dtpm7tXbkWzAcAgIA+Y7+6Xm9ldQJa7P2hmRwLjo6+/CfR397ZmditwKZGCZxNwibtnmtlYIsXSW+4+yMwGAldEd5cF9I9ua7K7H72fTH8FBgPNgc3AFHe/oqRl90qp0bJyB0DkFyQ1WTfdj4WcvNygI4iQu2f9Pte1xNOuN/8vZn9na/3uhmrLXukC6UChAkmkbCqQYkMFkoRBtRdIk0fErkA692/Vll2/9URERCR+ErSLTQWSiIiIxE+CDvMPy520RUREREJDBZKIiIjETwDD/M0s2czmm9nk6PNDzezT6Pe1joveqLpUKpBEREQkfoL5stqBwNeFnt8PPOTu7YEfgP8tawMqkEREROSAEf2mjt8BT0WfG5FvAHk5ukgGcH5Z29FF2iIiIhI/MRzFVvirwqKecPcnii32f0Tuibj3BtQHAdvcfe99NtYBLcvalwokERERiZ8YFkjRYqh4QVTAzM4FNrv7Z2bWvSr7UoEkIiIiB4rTgD5mdg6R74GtD4wEGppZSrQVqRWwvqwN6RokERERiR/32E1l7spvc/dW7t4W+AMw3d3/B3ifn78XdgDwelnbUoEkIiIi8RPAMP8S3AL8zcyWE7km6emyVlAXm4iIiBxw3H0GMCP6eCXQtSLrq0AKOX1JaNXpC0KrTsdQRCpN38UmIiIiUoy+i01ERETkwKAWJBEREYkfdbGJiIiIFFOO4flhpC42ERERkWLUgiQiIiLxoy42ERERkWIStEBSF5uIiIhIMWpBEhERkfhJ0PsgqUASERGRuPF8jWITEREROSCoBUlERETiRxdpy1m9u/PVog/5ZvFHDB50XdBx9vH448NZs+Yz5s2bGnSUEoU9315hP89hzwfhzxj2fKCMsRD2fJAYGcvk+bGbqlFMCyQzu8vMbq7Eeg3N7NoY7H+MmS0xs0Vm9oyZpVZ1m+WVlJTEqJH3cu55/TnmuB5cdNH5HHVUh+rafbm88MIE+vYdEHSM/Qp7Pgj/eQ57Pgh/xrDnA2WMhbDng8TIeCALSwtSQ6BCBZJFFM8/BjgSOAaoBVwRm3hl63ri8axYsZpVq9aSk5PD+PGv0+e8s6pr9+Uya9YcMjO3BR1jv8KeD8J/nsOeD8KfMez5QBljIez5IDEylku+x26qRpUukMzsMjNbaGYLzOyFEl6fYWZdoo+bmNnq6OOOZjbHzL6Irt8BGAa0i84bHl1ukJnNjS5zd3Re22gL0fPAIqB14X26+xSPAuYArSr7/ioqvWVzvl23oeD5uvUbSU9vXl27l2oS9vMc9nwQ/oxhzwfKGAthzweJkbFc8vNjN1WjSl2kbWYdgSHAqe6+1cwaV2D1a4CR7j7GzGoAycCtwNHu3im6/d5AB6ArYMAkMzsdWBudP8DdZ5eSLxW4FBhY8XcnIiIiv3SVHcXWE5jg7lsB3D2zAut+AtxhZq2Aie6+zMyKL9M7Os2PPq9LpDBaC6wprTiKGg186O4zS3rRzK4CrgKw5AYkJdWpQPySbVi/idat0guet2rZgg0bNlV5uxIuYT/PYc8H4c8Y9nygjLEQ9nyQGBnLRaPY9pFbaPtpe2e6+1igD7ALmGJmPUtY14D73L1TdGrv7k9HX9tZ2k7N7E6gKfC3/S3j7k+4exd37xKL4ghg7rwvaN/+UNq2bU1qaioXXtiXNyaHezSWVFzYz3PY80H4M4Y9HyhjLIQ9HyRGxnJxj91UjSrbgjQdeNXMRrj792bWuIRWpNVAZyLXAvXbO9PMDgNWuvsoM2sDHAssAOoVWvcdYKiZjXH3LDNrCeSUFcrMrgDOAnq5V+94wLy8PAbeMIQpb44lOSmJ5zLGsXjx0uqMUKaMjFF063YKTZo0Yvny2Qwd+hAZGeOCjlUg7Pkg/Oc57Pkg/BnDng+UMRbCng8SI+OBzLySFZmZDQAGAXnAfHe/3MzuArLc/UEzOxIYH339TaC/u7c1s1uJXB+UA2wCLnH3TDMbS6RYesvdB5nZQH4ehZYF9I9ua7K7H72fTLnAGmBHdNZEd7+ntPeRUqNlqO+Bnpqse3lWVU5ebtARRERCI3fP+n2ua4mn7BFXxuzvbO2/PVlt2StdIB0oVCAd+FQgiYj8rNoLpAeviF2BdPNT1ZY9LPdBEhEREQkNNU+IiIhI/FTzV4TEigokERERiZ9qvgN2rKiLTURERKQYtSCJiIhI3HiC3ihSBZKIiIjEj7rYRERERA4MakESERGR+NEoNhEREZFi1MUmIiIicmBQC5KIiIjEj0axJab2DdODjlCq42u1DDpCmbp43aAjlGrIlplBRyhTWnJq0BHKlJZSI+gIpUqEY/jtjq1BR5A4a1q7QdARwkddbCJyoAp7cSQiEmu/+BYkERERiSONYhMREREpRl1sIiIiIgcGtSCJiIhI3Oi72ERERESKUxebiIiIyIFBLUgiIiISPwnagqQCSUREROJHw/xFREREiknQFiRdgyQiIiJSjFqQREREJG48QVuQVCDF0HvzXmdnVjZ5+fnk5ebSr/eAoCNx9fDrOb5nF378fjuDew8E4KRzTqXfjX8gvX0r/l+fQaz8ckVg+c4afiWH9epE9vc/knHmbQCc++j1NDqsBQA169dm94/ZvPDbOwLLWNjjjw/nt7/tyZYt39OlS++g4+yjZcsWPPbkcJo2a4K7k/Hsf/j36IygYxVRs2YNXp3yPDVq1iAlOYXJk6by4H2PBB1rH0lJSUx67yU2bdzMFZf8Jeg4+zird3dGjLiH5KQknnn2JR4Y/mjQkfYR9oxhz5coPytlUoEUYWZ3AVnu/mAF12sIXOLuo6u4/6eBLoABS4HL3T2rKtusiMsuuIZtmdura3dl+mDCdN7JmMK1IwYWzPt26VpGXD2MK/55bYDJIhZN+JD5GdP47UNXF8ybfN3PvwDOGHIJu3dkBxGtRC+8MIHHH8/gqadGBB2lRLm5uQy57T4WLviKunXr8P7M15gxfRZLvlkedLQCu3fvoV+fP5G9M5uUlBRef/tFpk/7kM/nLQw6WhF/vPp/WL50JXXr1Q06yj6SkpIYNfJezj7nYtat28jsT6bwxuSpfP31sqCjFQh7xrDng8T5WTlQhekapIZAhf5iW0Tx93Cjux/n7scCa4HrYxUwEX0zZzFZ24rWhxuWr2Pjyg0BJSpq/Zwl/LRt//XrEeeexDevf1KNiUo3a9YcMjO3BR1jv777bgsLF3wFQFbWTpYuWUGLFgcHnGpf2TsjRW9qagqpqSl4yP7BbJ7ejB69uzHuxVeDjlKiricez4oVq1m1ai05OTmMH/86fc47K+hYRYQ9Y9jz7RX2n5Vyyc+P3VSNqlQgmdllZrbQzBaY2QslvD7DzLpEHzcxs9XRxx3NbI6ZfRFdvwMwDGgXnTc8utwgM5sbXebu6Ly2ZrbEzJ4HFgGtC+/T3X+MLmdALaDaPk7uztPjH+GVac9z4aW/r67dHrBadj2CnVu3s231d0FHSUit27Tk2ON+xWfzFgQdZR9JSUlMmzmRL5d9xAfvf8z8z8L1H/Hf7x3MsLseIj+kX5GQ3rI53677+Z+cdes3kp7ePMBE+wp7xrDn2yvsPyvlku+xm6pRpbvYzKwjMAQ41d23mlnjCqx+DTDS3ceYWQ0gGbgVONrdO0W33xvoAHQl0l02ycxOJ9Iq1AEY4O6z95PtWeAcYDFwU6XeYCVcct6VbN60hcZNGvHMhEdYuWw182bPr67dH3CO7HtKqFqPEkmdOrV5fsyj3HbLP9ixo9p6mMstPz+fM7tdQP0G9XjmxVEccVR7lnwdjm7Anr1PZ+vWTBYt+JqTTusSdBz5hQvzz8qBriotSD2BCe6+FcDdMyuw7ifA7WZ2C3CIu+8qYZne0Wk+8DlwJJHCCGDN/oqjaJY/AunA18BFxV83s6vMbJ6Zzdu2a0sFYpdu86bItjK3/sC7U2Zw7AkdY7btXxpLTqLD2Sey5I1Pg46ScFJSUsgY8ygTxk1i8qSpQccp1Y/bdzBr5hx69OoWdJQCnU/qxG/O7s7M+VN4+Mn7ObXbiTz0+D+DjlXEhvWbaN0qveB5q5Yt2LBhU4CJ9hX2jGHPV1wYf1bKLUFbkOJ9DVJuoX2k7Z3p7mOBPsAuYIqZ9SxhXQPuc/dO0am9uz8dfW1nWTt29zzgP8B/lfDaE+7exd27NKzVtGLvaD9q1U6jTp3aBY9P634yS78ObnRYojvk10eTuWIDWZsqUncLwMOj72PpkuWMfuSZoKOU6KCDGlG/QT0A0tJqckb3U1m+bGXAqX42fOgoTj2mN92OP4e/XHkLH8+cy43X3B50rCLmzvuC9u0PpW3b1qSmpnLhhX15Y3K4iuGwZwx7Pgj/z0p5uXvMpupUlVFs04FXzWyEu39vZo1LaEVaDXQG5gD99s40s8OAle4+yszaAMcCC4B6hdZ9BxhqZmPcPcvMWgI5pQWKXnfUzt2XRx/3Ab6pwnsst4OaHsQjzz0AQHJyCpMnvs1H7wffPfSXUX/jqFOOpl6j+jwy+ylefug/ZG3bweV3X0n9xg0Y/Oz/Y/XiVQy77O5A8v3u4etodcpR1GpUl6s+HcXHI15h0bgPOKLPyXwzKfjjV1xGxii6dTuFJk0asXz5bIYOfYiMjHFBxypw8imd+cMlv+erRd/w4ceTABh617+YNvWDgJP9rFnzpox87D6Sk5NIsiQmvfY2774TnnyJIC8vj4E3DGHKm2NJTkriuYxxLF68NOhYRYQ9Y9jzgX5WKsPM0oAPgZpEapyX3f1OMxtDZIR7DpGa5Gp3L72mqEpFZmYDgEFAHjDf3S8vPMzfzI4ExkdffxPo7+5tzexW4NJo0E1EhvdnmtlYIsXSW+4+yMwGAldEd5cF9I9ua7K7H11CniRgJlCfSAvUAuDPey/cLsmRzU4M9ZiA42u1DDpCmbp4+IZBFzZky8ygI5QpLTk16AilSkupEXSEMoX9GAJ8u2Nr0BEkzprWbhB0hDJt3LbYqnN/P17ZO2Z/Z+s/ObXU7NHGkTrRhpVU4CNgINCgmKPQAAAgAElEQVQYeCu62FjgQ3d/rLRtVek+SO6eAWQUm3dXocffECl49hoSnT+MyKi14tu7pNjzkcDIEna9T3EUXT4fOK186UVERCTuqvHaIY+0+uwdmZIandzdp+xdxszmAK3K2laY7oMkIiIiUiVmlmxmXwCbgWnu/mmh11KJ9GC9XdZ2VCCJiIhI3Hi+x2wqPAo9Ol21z/7c86K3DGoFdDWzwr1Oo4l0r5V57YW+i01ERETiJ4ZdbO7+BPBEOZfdZmbvA2cDi8zsTqApcHXpa0aoBUlEREQOCGbWNPrdrphZLeBM4BszuwI4C7g4er1ymdSCJCIiIvFTvd/Y0wLIMLNkIo1A4919spnlAmuATyID3Zjo7veUtiEVSCIiIhI3Xr2j2BYCx5cwv8L1jrrYRERERIpRC5KIiIjETzV/h1qsqEASERGR+Knea5BiRl1sIiIiIsWoBUlERETipjov0o4lFUgiIiISPwnaxfaLL5CWb9sQdIRSrdmxOegIZXot6ABlyMnLDTpCmcKecceeXUFHEEkIW7K3Bx1BYuQXXyCJiIhI/KiLTURERKS4BO1i0yg2ERERkWLUgiQiIiJxU76vhg0fFUgiIiISPwlaIKmLTURERKQYtSCJiIhI3KiLTURERKS4BC2Q1MUmIiIiUoxakERERCRu1MUmIiIiUkyiFkjqYhMREREpRgVSDJ3VuztfLfqQbxZ/xOBB1wUdZx+PPz6cNWs+Y968qUFHKVHY8+0V9vMc9nwQ/oxhzwfKGAthzweJkbEsnh+7qTrFvEAys7vM7OZKrNfQzK6NYY5RZpYVq+2VJSkpiVEj7+Xc8/pzzHE9uOii8znqqA7VtftyeeGFCfTtOyDoGPsV9nwQ/vMc9nwQ/oxhzwfKGAthzweJkbFc3GI3VaMwtSA1BCpUIFnEPu/BzLoAjWIVrDy6nng8K1asZtWqteTk5DB+/Ov0Oe+s6oxQplmz5pCZuS3oGPsV9nwQ/vMc9nwQ/oxhzwfKGAthzweJkfFAVqUCycwuM7OFZrbAzF4o4fUZ0WIFM2tiZqujjzua2Rwz+yK6fgdgGNAuOm94dLlBZjY3uszd0XltzWyJmT0PLAJaF9tnMjAcGFyV91ZR6S2b8+26DQXP163fSHp68+qMINUg7Oc57Pkg/BnDng+UMRbCng8SI2N5JGoXW6VHsZlZR2AIcKq7bzWzxhVY/RpgpLuPMbMaQDJwK3C0u3eKbr830AHoChgwycxOB9ZG5w9w99klbPt6YJK7bzSr3uY4ERERKcrzE/NvcVWG+fcEJrj7VgB3z6zAup8Ad5hZK2Ciuy8roZjpHZ3mR5/XJVIYrQXWlFQcmVk68N9A99J2bmZXAVcBWHIDkpLqVCB6yTas30TrVukFz1u1bMGGDZuqvF0Jl7Cf57Dng/BnDHs+UMZYCHs+SIyMB7J4X4OUW2gfaXtnuvtYoA+wC5hiZj1LWNeA+9y9U3Rq7+5PR1/buZ/9HQ+0B5ZHu/Nqm9ny4gu5+xPu3sXdu8SiOAKYO+8L2rc/lLZtW5OamsqFF/bljcnhHo0lFRf28xz2fBD+jGHPB8oYC2HPB4mRsTx+cV1swHTgVTMb4e7fm1njElqRVgOdgTlAv70zzewwYKW7jzKzNsCxwAKgXqF13wGGmtkYd88ys5ZATmmB3P1NoKCD1syy3L195d9i+eXl5THwhiFMeXMsyUlJPJcxjsWLl1bHrsstI2MU3bqdQpMmjVi+fDZDhz5ERsa4oGMVCHs+CP95Dns+CH/GsOcDZYyFsOeDxMhYHl7No89ixdy98iubDQAGAXnAfHe/3MzuArLc/UEzOxIYH339TaC/u7c1s1uBS4kUPJuAS9w908zGEimW3nL3QWY2ELgiurssoH90W5Pd/ehy5Mty97qlLZNSo2XlD0A1SE3Wzc6rKicvN+gIIiKhkbtnfbVWLOtP6Rmzv7MtP5lebdmrVCAdCFQgHfhUIImI/Ky6C6R1J8WuQGr1afUVSPrrKyIiInGTqKPYwnSjSBEREZFQUAuSiIiIxE2iXsmjAklERETiRl1sIiIiIgcItSCJiIhI3CRqC5IKJBEREYkbXYMkIiIiUkyitiDpGiQRERGRYtSCJCIiInGTqN/FpgJJRERE4sbzg05QOb/4Aqlp7QZBRyhT9waHBx2hVI2sRtARyvTqtkVBRyjVT7l7go5QprSU8J/nLdnbg46Q8OrVqBV0hFLt2LMr6Ahl0ndoHhh0FkMu7MVRIgh7cZQIVBz9MoS9OEoEKo72la8uNhEREZGiEvUaJI1iExERESlGLUgiIiISN4l6HyQVSCIiIhI3iXonbXWxiYiIiBSjFiQRERGJG3WxiYiIiBSTqMP81cUmIiIiUoxakERERCRuEvU+SCqQREREJG40ik1ERETkAKECKUZq1qzBlPf+w7sfTWTGJ5O4+bbrg44EwNXDr+fxz57jgakjC+addM6pDJ82ijGrJnLYMe0CTAf9H/gz9897kiHvPFgw79y/XcQdbw3ntikP8Jfn76BBs0YBJiwqrOd5r5YtWzBpyot8Mu9tPp77FldfOyDoSPsI+zEEOKt3d75a9CHfLP6IwYOuCzpOicKeMRE+i2E/ho8/Ppw1az5j3rypQUepkny3mE3VKaYFkpndZWY3V2K9hmZ2bQz2/5yZrTKzL6JTp6pus7x2795Dvz5/4je/voDfdLuAHr1+zQldjq2u3e/XBxOmM2zAPUXmfbt0LSOuHsY3ny4OKNXPZr88g0cG/LPIvHefmMS9vx3EfecM5svpn3POwH4BpdtXWM/zXrm5uQy57T5O6XI2vXv044or+3PEke2DjlVE2I9hUlISo0bey7nn9eeY43pw0UXnc9RRHYKOVUQiZAz7ZzERjuELL0ygb9/wFZYV5W4xm8piZq3N7H0zW2xmX5nZwGKv32RmbmZNytpWWFqQGgIVKpAsoqT8g9y9U3T6Ijbxyid7ZzYAqakppKamhKLf9Zs5i8nallVk3obl69i4ckNAiYpaPudrdm4vmu+nrF0Fj2vWromH4UAWEsbzvNd3321h4YKvAMjK2snSJSto0eLggFPtK8zHsOuJx7NixWpWrVpLTk4O48e/Tp/zzgo6VhGJkDHsn8VEOIazZs0hM3Nb0DESTS5wk7v/CjgZuM7MfgWR4gnoDawtz4YqXSCZ2WVmttDMFpjZCyW8PsPMukQfNzGz1dHHHc1sTrSFZ6GZdQCGAe2i84ZHlxtkZnOjy9wdndfWzJaY2fPAIqB1ZfPHQ1JSEtNmTuTLZR/xwfsfM/+zhUFHSlh9bv4D9348mhP7/prJI8YFHaeIRDnPrdu05NjjfsVn8xYEHWUfYT6G6S2b8+26n/+BWLd+I+npzQNMtK9EyFhYGD+LiXYME5l77Kay9+Ub3f3z6OMdwNdAy+jLDwGDgXL9S1apAsnMOgJDgJ7ufhwwsIxVCrsGGOnunYAuwDrgVmBFtNVnkJn1BjoAXYFOQGczOz26fgdgtLt3dPc1JWz/3mhR9ZCZ1azM+6us/Px8zux2ASd07MHxnY/hiKPC05ycaCY9+B/uOPVa5r7+EWcMODvoOEUkwnmuU6c2z495lNtu+Qc7dmSVvUI1S4RjKLER9s+ixF9Q1yCZWVvgeOBTM+sLrHf3clfplW1B6glMcPetAO6eWYF1PwFuN7NbgEPcfVcJy/SOTvOBz4EjiRRGAGvcffZ+tn1bdNkTgcbALSUtZGZXmdk8M5uXveeHCkQvnx+372DWzDn06NUt5tv+pZnz2kyOP/ukoGOUKKznOSUlhYwxjzJh3CQmTwr3xZ1hPIYb1m+idav0guetWrZgw4ZNASbaVyJkhHB/FhPlGEpRhf9+R6er9rNcXeAV4AYi3W63A3+vyL7ieQ1SbqHtp+2d6e5jgT7ALmCKmfUsYV0D7it0LVF7d386+trO/e0w2rTm7r4beJZIC1RJyz3h7l3cvUvtGrEZIXXQQY2o36AeAGlpNTmj+6ksX7YyJtv+pWna9udm7uPOPJFNK8JxvRQkxnl+ePR9LF2ynNGPPBN0lBKF/RjOnfcF7dsfStu2rUlNTeXCC/vyxuRw/XFPhIwQ7s9iohzDA0EsL9Iu/Pc7Oj1RfH9mlkqkOBrj7hOBdsChwILo5T6tgM/NrNQ+1creKHI68KqZjXD3782scQmtSKuBzsAcoGAYkpkdBqx091Fm1gY4FlgA1Cu07jvAUDMb4+5ZZtYSyCkrlJm1cPeNZmbA+USuU6oWzZo3ZeRj95GcnESSJTHptbd5950Pqmv3+/WXUX/jqFOOpl6j+jwy+ylefug/ZG3bweV3X0n9xg0Y/Oz/Y/XiVQy77O5A8v1x1EAOP/lX1G1Uj3s/eYw3HxpPxx4ncPBhLfB8J3P9Vsbesc/nPzBhPc97nXxKZ/5wye/5atE3fPjxJACG3vUvpk0NT8awH8O8vDwG3jCEKW+OJTkpiecyxrF48dKgYxWRCBnD/llMhGOYkTGKbt1OoUmTRixfPpuhQx8iIyNc12SWR3UOz4/+/X8a+NrdRwC4+5dAs0LLrAa67O0F2++2KjtCyMwGAIOAPGC+u19uZncBWe7+oJkdCYyPvv4m0N/d25rZrcClRAqeTcAl7p5pZmOJFEtvRa9DGghcEd1dFtA/uq3J7n70fjJNB5oSaYH6ArjG3Uvt9G7R8FchGj+zr+4NDg86QpkaWY2gI5Tq1W3VVidX2k+5e4KOUKq0lHCfY4At2duDjpDw6tWoFXSEMu3YU9JVGeGRmhz+L6jYtWtNtd5Q6NP0C2L2d/akDRNLzW5mvwZmAl8C+dHZt7v7lELLrCaeBdKBQgVS1alAqjoVSFWnAqnqVCBVnQqkfc2OYYF0chkFUiyF/0yKiIhIwqruO2DHSlhuFCkiIiISGmpBEhERkbgpz1eEhJEKJBEREYmb/LIXCSV1sYmIiIgUoxYkERERiRtHXWwiIiIiReSH+mY6+6cuNhEREZFi1IIkIiIicZOvLjYRERGRohL1GiR1sYmIiIgUoxYkERERiZtEvQ/SL75AOqR2s6AjlOonzws6QpnuPOy7oCOU6q2FqUFHKFPYP4drsjcHHaFM7RumBx2hTMu3bQg6QqkS4UuJw/5ltc1rNww6Quioi01ERETkAPGLb0ESERGR+FEXm4iIiEgxiVogqYtNREREpBi1IImIiEjcJOpF2iqQREREJG7yE7M+UhebiIiISHFqQRIREZG40XexiYiIiBTjQQeoJHWxiYiIiBSjFiQRERGJm0S9D5IKJBEREYmbfNM1SL9obdq15h+P31nwvGWbFjwx/FnGPfVygKng+uF/pUuvE9n+/XYGnnk9AHUb1OWm0YNp1upgNq/7jgevvZ+d23cGki+paVMa3HYHSY0aAU725DfY9corpLRrR70bb8Jq1SJ/0ya23zsUz84OJGNxSUlJTHrvJTZt3MwVl/wl6DhFhPVzWFjNmjV4dcrz1KhZg5TkFCZPmsqD9z0SdKwi3pv3OjuzssnLzycvN5d+vQcEHWkfZ/XuzogR95CclMQzz77EA8MfDTpSEYlwnsN+DCHcv28OdDEtkMzsLiDL3R+s4HoNgUvcfXQV92/AP4D/BvKAx9x9VFW2WV5rV3zLZWdeAUQ+0G98/jIfvDWzOnZdqukT3mNKxpsMfOjGgnkXXNePL2ctZOLol7ng2n5ccG0/XrgvI5iAeXnseOxRcpctw2rVovG/n2TPvHnUv3kwOx4fTc6CBaT99hxqX/QHdj77TDAZi/nj1f/D8qUrqVuvbtBR9hHWz2Fhu3fvoV+fP5G9M5uUlBRef/tFpk/7kM/nLQw6WhGXXXAN2zK3Bx2jRElJSYwaeS9nn3Mx69ZtZPYnU3hj8lS+/npZ0NEKhP08J8IxhHD/vikvXaRdNQ2BayuygkUUz3850Bo40t2PAv4Tm3gV06XbCaxfs55N678LYvdFLJ7zFTu27Sgyr+uZJ/H+y+8B8P7L73FS75ODiAZAfmYmucsiv5B81y5y164huUlTklu1ImfBAgD2zJtL2ulnBJaxsObpzejRuxvjXnw16ChlCtPnsLjsnZHWwNTUFFJTU/BE/Q0akK4nHs+KFatZtWotOTk5jB//On3OOyvoWPsI83lOhGOYSL9vSpMfw6k6VbpAMrPLzGyhmS0wsxdKeH2GmXWJPm5iZqujjzua2Rwz+yK6fgdgGNAuOm94dLlBZjY3uszd0XltzWyJmT0PLCJSDBX2Z+Aed88HcPfNlX1/VXFm355MfW16ELsul4ZNGvLD5h8A+GHzDzRs0jDgRBFJBzcntX0Hcr5eTO7q1dQ87dcApHXvQVKzZgGni/j7vYMZdtdD5OeH/7LDMH8Ok5KSmDZzIl8u+4gP3v+Y+Z+Fo1VhL3fn6fGP8Mq057nw0t8HHWcf6S2b8+26DQXP163fSHp68wATlSzM5zkRjmEi/b4pTb7FbqpOlSqQzKwjMATo6e7HAQMrsPo1wEh37wR0AdYBtwIr3L2Tuw8ys95AB6Ar0AnobGanR9fvAIx2947uvqbYttsBF5nZPDN7K1p8VauU1BS69T6N6W/MqO5dV1oY/qmztFo0vOcedjz6MJ6dzY8P3E+tvufT+N9PYLVqQU5O0BHp2ft0tm7NZNGCr4OOUqawfw7z8/M5s9sFnNCxB8d3PoYjjmofdKQiLjnvSv7rN5dy5cUDueRP/ehy8vFBR0pIYT/PYZZIv28OVJVtQeoJTHD3rQDunlmBdT8BbjezW4BD3H1XCcv0jk7zgc+BI4kURgBr3H32frZdE/jJ3bsATwIlXrRiZldFi6h5m7M3lLRIpZ3S8ySWfLmUzK0/xHS7sbRt6zYaNWsEQKNmjdi+dVuwgZKTaXDPPfz07rvsnhm5Xibv27VsG3wzmVdfxU/T3yNvQ2zPU2V0PqkTvzm7OzPnT+HhJ+/n1G4n8tDj/ww6VokS4XMI8OP2HcyaOYcevboFHaWIzZu2AJC59QfenTKDY0/oGHCiojas30TrVukFz1u1bMGGDZsCTFS6MJ7nsB/DRPp9U5Z8LGZTdYrnNUi5hbaftnemu48F+gC7gClm1rOEdQ24L9qi1Mnd27v709HXShtutQ6YGH38KnBsSQu5+xPu3sXduzSrnV7SIpXW+/xeTH3tvZhuM9bmTptDj369AOjRrxdzpn0aaJ76g28hd80asieML5hnDaPdfmbUufQyst+YFFC6nw0fOopTj+lNt+PP4S9X3sLHM+dy4zW3Bx2rRGH+HB50UCPqN6gHQFpaTc7ofirLl60MONXPatVOo06d2gWPT+t+Mku/XhFwqqLmzvuC9u0PpW3b1qSmpnLhhX15Y/LUoGMVEfbzHPZjmEi/b8riMZyqU2VHsU0HXjWzEe7+vZk1LqEVaTXQGZgD9Ns708wOA1a6+ygza0OkiFkA1Cu07jvAUDMb4+5ZZtYSKE8fy2tAD2AVcAawtHJvr3LSaqXRtVtnhg3+V3XutlR/e/hmOp5yDPUb1efJT5/lPyPGMnH0y9z82C30uuhMtqzfzIN/vj+wfKlHH0Ot3meRs2IFjZ98CoCsp54kuVUraveNXPuxe+aH/PTWlMAyJpowfg4La9a8KSMfu4/k5CSSLIlJr73Nu+98EHSsAgc1PYhHnnsAgOTkFCZPfJuP3v8k4FRF5eXlMfCGIUx5cyzJSUk8lzGOxYur9dddmcJ+nhPhGEqwzCs5rMDMBgCDiAynn+/ulxce5m9mRwLjo6+/CfR397ZmditwKZGCZxOR4f2ZZjaWSLH0VvQ6pIHAFdHdZQH9o9ua7O5H7ydTQ2AM0Ca6zjXuvqC093FyevcwXIKzXy1S6pW9UMAeb/dj0BFKdfLCgLsQy+Hgmo2CjlCqNdmBjHeokAY16gQdoUzLtwXfVVyaprUbBB2hTFuyw3nrhb1a12sSdIQyrfp+QbX2VT3fsn/M/s5etv7Faste6fsguXsGkFFs3l2FHn9D0S6uIdH5w4iMWiu+vUuKPR8JjCxh1yUWR9F1tgG/Kzu9iIiIVIdEHYMXlvsgiYiIiISGvmpERERE4ibU17GUQgWSiIiIxE113+AxVtTFJiIiIlKMWpBEREQkbhL1Im0VSCIiIhI3iVogqYtNREREpBi1IImIiEjceIJepK0CSUREROJGXWwiIiIiBwi1IImIiEjcJGoLkgokERERiRvdSTtBbcvdGXSEUl2e3CroCOXwY9ABSvVTXk7QEcoU9s/hT7l7go5QpgY16gQdIeFt2x3uz2EiaJhaN+gIEiO/+AJJRERE4idRv2pEBZKIiIjETaJeg6RRbCIiIiLFqAVJRERE4iZRW5BUIImIiEjcJOooNnWxiYiIiBSjFiQRERGJm0QdxaYWJBEREYmb/BhO5WFmz5jZZjNbVGz+X8zsGzP7ysweKGs7KpBERETkQPIccHbhGWbWA+gLHOfuHYEHy9qIuthEREQkbqr7Im13/9DM2hab/WdgmLvvji6zuaztqAVJRERE4iYfj9lUBYcD3czsUzP7wMxOLGsFFUgiIiKSEMzsKjObV2i6qpyrpgCNgZOBQcB4Myv18nF1scXQe/NeZ2dWNnn5+eTl5tKv94CgI9H9wSs5pFcndn3/I+N/cxsAB/2qDaff9yeSa6aSn5fHR3c8x+YvVgaSL6lpUxrcdgdJjRoBTvbkN9j1yiuktGtHvRtvwmrVIn/TJrbfOxTPzg4kY2E1a9bg1SnPU6NmDVKSU5g8aSoP3vdI0LGKCOPnsLCWLVvw2JPDadqsCe5OxrP/4d+jM4KOVUTYjyHAWb27M2LEPSQnJfHMsy/xwPBHg45UxOOPD+e3v+3Jli3f06VL76DjlCjsxxCgbv263DniVtodcRjuzt03/pOFn30VdKwKieWNIt39CeCJSqy6Dpjo7g7MMbN8oAmwZX8rxLRAMrO7gCx3L/Pip2LrNQQucffRVdz/TKBe9GkzYI67n1+VbVbUZRdcw7bM7dW5y1ItmfAhi56bRs//u7pg3sl3XMy8hyby7YyFtOlxHCfffjGTLrw3mIB5eex47FFyly3DatWi8b+fZM+8edS/eTA7Hh9NzoIFpP32HGpf9Ad2PvtMMBkL2b17D/36/InsndmkpKTw+tsvMn3ah3w+b2HQ0YoI2+ewsNzcXIbcdh8LF3xF3bp1eH/ma8yYPosl3ywPOloRYT6GSUlJjBp5L2efczHr1m1k9idTeGPyVL7+elnQ0Qq88MIEHn88g6eeGhF0lBIlwjEEGPyPG/h4+qcMumIIKakppNVKCzpShYXkRpGvAT2A983scKAGsLW0FcLSxdYQuLYiK1hEkfzu3s3dO7l7J+ATYGIMMyakjZ8uYfe2rKIz3alRrxYANerXZud3PwSQLCI/M5PcZZFfSL5rF7lr15DcpCnJrVqRs2ABAHvmzSXt9DMCy1hc9s5IS1ZqagqpqSl4SH76E8V3321h4YLIf8BZWTtZumQFLVocHHCqxNL1xONZsWI1q1atJScnh/HjX6fPeWcFHauIWbPmkJm5LegY+5UIx7BuvTqccPJxvDr2DQByc3LJ+jGrjLXEzF4iUgMcYWbrzOx/gWeAw6JD//8DDIi2Ju1XpQskM7vMzBaa2QIze6GE12eYWZfo4yZmtjr6uKOZzTGzL6LrdwCGAe2i84ZHlxtkZnOjy9wdndfWzJaY2fPAIqD1frLVB3oSqRirjbvz9PhHeGXa81x46e+rc9cVMuuuFzn5jovp/+lIThlyMZ8OGxd0JACSDm5OavsO5Hy9mNzVq6l52q8BSOveg6RmzQJO97OkpCSmzZzIl8s+4oP3P2b+Z+FqPUqUzyFA6zYtOfa4X/HZvAVBRyki7McwvWVzvl23oeD5uvUbSU9vHmCixJMIxzC9TTo/fL+Nu0fewUvTnuXv/7qVtNqJ14JU3fdBcveL3b2Fu6e6eyt3f9rd97h7f3c/2t1PcPfpZW2nUl1sZtYRGAKc6u5bzaxxBVa/Bhjp7mPMrAaQDNwKHB1t+cHMegMdgK6AAZPM7HRgbXT+AHefXco+zgfec/cf95P/KuAqgIPrHkLDWk0rEH//LjnvSjZv2kLjJo14ZsIjrFy2mnmz58dk27HU8dJefHz3GFa9NZd2555E9+FXMvmSYYFmsrRaNLznHnY8+jCenc2PD9xPvb/8lTqXXcbuWbMgJyfQfIXl5+dzZrcLqN+gHs+8OIojjmrPkq/D0z2UKJ/DOnVq8/yYR7ntln+wY0e4/itOlGMoB7aUlGSOPOZw7r/9IRbNX8ygoQP50/WXMvqBJ4OOViG/tDtp9wQmuPtWAHfPrMC6nwC3m9ktwCHuvquEZXpHp/nA58CRRAojgDVlFEcAFwMv7e9Fd3/C3bu4e5dYFUcAmzdFrvXK3PoD706ZwbEndIzZtmPp8H7dWPXWXABWTP6UZp3aBRsoOZkG99zDT+++y+6ZMwHI+3Yt2wbfTObVV/HT9PfI27ChjI1Uvx+372DWzDn06NUt6ChFJMLnMCUlhYwxjzJh3CQmT5oadJx9hP0Ybli/idat0guet2rZgg0bNgWYKPEkwjH8bsNmNm/cwqL5iwF4d/IMjjz28IBT/XLE8xqk3ELbL2gTdPexQB9gFzDFzHqWsK4B9+29nsjd27v709HXdpa2UzNrQqTl6c2qvoGKqFU7jTp1ahc8Pq37ySz9ekV1Rii37O9+IP3kowBoeVpHtq8K9pdC/cG3kLtmDdkTxhfMs4YNow+MOpdeRvYbkwJKV9RBBzWifoPIOIC0tJqc0f1Uli8LZgRgSRLlc/jw6PtYumQ5ox8J/sL74hLhGM6d9wXt2x9K27atSU1N5cIL+/LG5PAVmmGWCMfw+y2ZbFq/mUPatQGga7fOrFy6OthQlRCS+yBVWGVHsU0HXjWzEe7+vZk1LqEVaTXQGZgD9Ns708wOA1a6+ygzawMcC1IEGPMAACAASURBVCzg59FnAO8AQ81sjLtnmVlLoLx9LP2Aye7+U6XeWSUd1PQgHnku8tUuyckpTJ74Nh+9///bu/t4q+a8/+OvzzmnpKIidEuIybiLmogrQ4iZ34wwXW6aiBmaGczFXL9pMFw/ZhjMxLiEuGIuYmpGjJsklJvclHQjpaIURSUkmZLhdPr8/tirOmd37jpr7b2+++z302M/7L322Wu/zt7ntL9nrfXd+7V8JlTr+DsupsORB9Bsl5YMmj6cmbf8g5cu/wtHX3sOVlZCxdflvHTFX+peUY40Oehgdux3EuVLlrDLPfcCsP7eeyjt1Inm/TPHfnz9ysv86+kJqTVWtnu73bjtrhspLS2hxEoY9/gzPPfsS2lnbRHqz2FlR/buwVkDT2P+vHd4eWpm4HvdtbcwaWIYj2MhPIYVFRVcetnVTHhqDKUlJdw/6iEWLFiUdlYVo0YNp0+f3rRt24bFi6dx3XW3MmpUGMc7QmE8hgB/vOpWbhhxDWVNylixbCXXXHZD2knbrVDnsVgdB3HXfEOzwWTebKkCmO3u51We5m9m3YCx0fVPAYPcvYuZXQGcQ2bAs4rM9P41ZjaGzGDpaXcfamaXAhdEd7ceGBSta7y7H1RL12Qybyf+TH2+j267fyfo5+6yHb6VdkKdTuv6YdoJteo+u8a3uQhGq6Yt0k6o1ccb0pvpWF97NG+TdkKdFq8Nb1dxZU1Kw39rvPKKjWkn1OrgXbqknVCn2aum5PWooKu6DEzsdfYPS8fkrb3Bvw3uPgoYlbXs2krn3yEz4Nns6mj5TWRmrWWvb2DW5duA26q56xoHR9Htjq29XERERPIlyTeKzKfw/1wQERGRgpXvY4eSEsobRYqIiIgEQ1uQREREJGcKc/uRBkgiIiKSQ4V6DJJ2sYmIiIhk0RYkERERyZlCPUhbAyQRERHJmcIcHmkXm4iIiMg2tAVJREREcqZQD9LWAElERERyxgt0J5t2sYmIiIhkKfotSF9XfJN2Qq2etS/STqjTZ+91SjuhVq2abkg7oU6F8GGwzcqapp1Qq2XrPkk7oeB1axX27zLAW2uWpp1Qq7Xl69NOCI52sYlIoxX64EhEwqVp/iIiIiJZCnN4pGOQRERERLahLUgiIiKSM9rFJiIiIpKlUA/S1i42ERERkSzagiQiIiI5U6hvFKkBkoiIiOSMdrGJiIiINBLagiQiIiI5o11sIiIiIlm0i01ERESkkdAAKWElJSWMf/Eh7h1ze9opAFwy7D+4/40HuW3SHVuWtWzVkmtG/547X/ofrhn9e1q0apFa3w+GXchls0Zw4cSbqizveV4/fvb8MIZM+iN9rzw7pbrqPT/zCcZN/huPvTCaRyaOSjunio4d2zNuwl95beYzTJ3xND+7aHDaSdvYYYemTHj+7zz36qNMfm0cv77ykrSTtnH33cNYtmwWM2dOTDulRif1O5b5817mnQWv8puhF6edU62WO7dk2L3X8+grY/jHy6M5pMeBaSdVUQiPYWivKQ2xyT2xUz4lPkAys2vN7NcNuF1rM7sogfs/3szeMLM3zexVM+sad53b4/yf/ZjFi97L513W6oWHn+f3515bZdnpFw/grSlzufi7P+OtKXM5/aIB6cQBcx5+hb8P/lOVZXv1/jb7n9iDe793JSNPvJxpI59Kqa5m557+c07r+2MG9AtrALJx40auvvJGevc8mX7HDeCCCwfxrW55/RWo09dff8OAU37CCf92Oif0OZ3jjv83Du95SNpZVTz44MP07x/Wc1tZSUkJw2/7Az/44SAOPvQ4zjzzVA44YL+0s7bxm+svY+oLr3N6n4Gcefxg3nt3WdpJWxTKYxjaa0pDeIKnfAppC1JrYLsGSJaR/T3cBfzY3bsDY4CrE+qrU7sOu3Ncvz489NfH8nWXdVowfT7r1q6rsqzXiUfw4iPPA/DiI89zRL8j00gD4MPp7/DV2vVVlh0+6HimjhhHxTcbAdjw2T/TSCtIH3/8KXPnzAdg/fovWbRwCe3b75Fy1bY2fLkBgCZNymjSpIw8/2FYpylTprNmzdq0M2rU6zuHsWTJUt5//wPKy8sZO/YJTvnhSWlnVdFypxYcfuShPDbmSQA2lm9k/T/X13Gr/CmExzDE15RiEmuAZGbnmtlcM5tjZg9Wc/1kM+sZnW9rZkuj8wea2fRoK89cM9sPuAnYN1o2LPq6oWY2I/qa30XLupjZQjN7AJgHdM66Wwd2js63AlbG+R63x//7w2+46dpb2bQp7EPSWrdtzeeffA7A5598Tuu2rVMuqmrXvduzZ69unPf47xj00NW0P2SftJOqcHf+MvYO/jHpAc4457S0c2rUec+OHHLot5k1c07aKdsoKSlh0iuP8ta7r/LSi1OZPWtu2kkFpUPHdny4fOs/bctXfESHDu1SLNpWhz078Plna/ndbVfxt0n38f9uuYJmzZulnbVFITyGhfKaUpdNeGKnfGrwAMnMDiSzdaavux8KXLodN/85cFu0lacnsBy4Alji7t3dfaiZ9QP2A3oB3YEeZnZMdPv9gBHufqC7Z2+zvQCYYGbLgXPIDLxyrm+/Y1i9eg3z5rydj7tLVGB/vGNlJTRr3YL7T72GF24Yw+kjfpl2UhUDf3ghPzrhHC48+1IG/mQAPY88LO2kbbRo0ZwHRt/JlZdfz7p14fzVvtmmTZs4sc/pHH7gcRzW42C+dUBYuwElvrKyUrodvD8P3/8YZ594Pl9t+IqfXHJO2lkFo5BfU7J5gv/lU5wtSH2Bh919NYC7r9mO274G/NbMLgf2cvevqvmaftFpNvAG0I3MwAhgmbtPq2HdvwK+7+6dgPuAP2d/gZkNMbOZZjZz3b8+247smvU4ojsnnHwsr8yewO33/JGj+nyHW+++IZF1J23t6rW02b0NAG12b8MXq8PalbDuozUsfGYmACvnvIdvcprvslPKVVt9supTANas/pznJkzmkMPDOvC0rKyMUaPv5OGHxjF+XLgHGQP884t1THllOscd3yftlIKycsUqOnfqsOVyp47tWblyVYpF2/p45Sd88tGnzJu9AIDnxk+m2yH7p1y1VeiPYSG9pjRWuT4GaWOl+9iybdXdxwCnAF+R2drTt5rbGnBjtEWpu7t3dfe/RNd9Wd2dmdluwKHu/nq06CHgqOyvc/eR7t7T3Xvu1GzXBn1j2YZdN5yjDu5Hn8O+zy8vvJypr8zgVz//bSLrTtqMSdM5bsDxABw34HimT3q9jlvk16KJs9ir9wEA7LJ3O0qblLFhzbo6bpUfOzZvRosWzbecP/rYI1n09pKUq6q6fcSNLFq4mBF3/G/aKdXaddc27NwqM+Bt1mwHvnvsUSx+t7APQs23GTPfpGvXvenSpTNNmjThjDP68+T4sAbDn326hlUrPmGvffcEoFefHry3aGm6UZWE/hgW0mtKXTYleMqnOG8U+QLwmJn92d0/M7NdqtmKtBToAUwHtkyVMrN9gPfcfbiZ7QkcAswBKm8meBa4zsxGu/t6M+sIlNfR9DnQysz2d/dFwIlA4W+fjOE/b/81B/Y+mJ3b7Mw9r9/H3/88hkdHPMKv77qc4888kU9XfMLNv/hjan2nDr+YvXofwI5tduKX027n5Vsf4c2xk/nBsCFcOPEmNpVvZNz/vTu1vmy77rYrd9yfmXVXWlrG+Eef4dUXX0u5aqsje/fgrIGnMX/eO7w8dRwA1117C5MmvpRy2Va7t9uN2+66kdLSEkqshHGPP8Nzz4bTBzBq1HD69OlN27ZtWLx4GtdddyujRj2UdtYWFRUVXHrZ1Ux4agylJSXcP+ohFixYlHbWNv541a3cMOIaypqUsWLZSq65LJwtIIXyGDYG+T52KCnmMaaPmNlgYChQAcx29/PM7FpgvbvfbGbdgLHR9U8Bg9y9i5ldQeb4oHJgFTDQ3deY2Rgyg6Wno+OQLiVzTBHAemBQtK7x7n5QDU2nAb8nM9j8HPiJu9f45+neux4a9DPXvcWeaSfUqaftXPcXpejBr8L/R+/jDZ+nnVCrZmVN006o09qvq92wHJTyio1pJ9Tq4F26pJ1Qp7fWLE07oVadd2qbdkKd3v9sjuXz/v59r/6Jvc4+vOyJvLXHGiA1BhogxacBUnwaIMWnAVJ8GiDFpwHStgbsdUpir7OPLBuXt3Z9FpuIiIjkTKG+SUFIbxQpIiIiEgRtQRIREZGcKdRDeTRAEhERkZwp1Fls2sUmIiIikkVbkERERCRnCvUgbQ2QREREJGfy/RlqSdEuNhEREZEs2oIkIiIiOVOoB2lrgCQiIiI5U6jT/LWLTURERCSLtiCJiIhIzmgWW4EK/QMun9kwN+2EOn3UZu+0E2oV+gfBFoLQf08g/A+CLQTvfLE87YSC9+G61WknBEez2EREREQaCQ2QREREJGc24Ymd6sPMfmVm881snpn9zcyaNaRbAyQRERHJGXdP7FQXM+sI/AfQ090PAkqBsxrSrQGSiIiINCZlwI5mVgY0B1Y2dCUiIiIiOZHPN4p09xVmdjPwAfAVMNHdJzZkXdqCJCIiIjnjCf5nZkPMbGal05DK92VmbYD+wN5AB6CFmQ1qSLe2IImIiEhBcPeRwMhavuQE4H13/xTAzB4FjgL+ur33pQGSiIiI5Mym/H7UyAfAkWbWnMwutuOBmQ1ZkQZIIiIikjP5HB65++tm9gjwBrARmE3tW5xqpAGSiIiINBrufg1wTdz1aIAkIiIiOZPPWWxJ0gBJREREcqZQB0ia5p+Qjh3bM27CX3lt5jNMnfE0P7tocNpJ27j77mEsWzaLmTMb9JYQObfnvp15YNK9W07PL3yKMy8YkHZWFaE/z6H3Qfg/hwAn9TuW+fNe5p0Fr/KboRennVOt0Bv1PCejEBobK6vPW3fXe2Vm1wLr3f3m7bxda2Cgu4+Ief99gZuBpsAs4KfuXutHfLdp2TWRB2CPPXZjj3a7M3fOfFq2bMGLrzzOoLN/wcJ3Fsda778qypPIA+Doo3vx5ZcbuPfeP9OzZ7/E1ntom70TW9dmJSUlPPnGI/z0//yCVSs+jrWuhf9M7hPKc/U8J6WYfw7LK2r9Va+3kpIS3p7/Cid//2yWL/+Iaa9NYNA5F/H22+8msv4k5KqxSWlyOxX0PMeXq8aN36ywhBLr5cgOxyY20Ji2cnLe2kPZgtQauGh7bmAZJZUulwCjgLOiz19ZBuTtz+ePP/6UuXPmA7B+/ZcsWriE9u33yNfd18uUKdNZs2Zt2hn10rPP4axYtiL24ChpoT/PofdB+D+Hvb5zGEuWLOX99z+gvLycsWOf4JQfnpR2VhWF0KjnOb5CaKyPfH9YbVIaPEAys3PNbK6ZzTGzB6u5frKZ9YzOtzWzpdH5A81supm9Gd1+P+AmYN9o2bDo64aa2Yzoa34XLetiZgvN7AFgHtC50l3uCnzj7ouiy5OAHzX0+4uj854dOeTQbzNr5pw07r5ROLF/XyY+/kLaGbUK/XkOvS9UHTq248PlWz+6afmKj+jQoV2KRdsqhMbQFcJjWAiNjVmDtqea2YHA1cBR7r7azHbZjpv/HLjN3UebWVMyn7R7BXCQu3eP1t8P2A/oBRgwzsyOIfMGUPsBg919WtZ6VwNlZtbT3WcCA6g6gMqLFi2a88DoO7ny8utZt259vu++UShrUkaffkdz1w33pJ1So9Cf59D7RKR4eIEepN3QHc59gYfdfTWAu6/Zjtu+BlxlZp2AR939XbNtdin2i06zo8styQyMPgCWVTM4wt3dzM4CbjWzHYCJQEV1AdFntwwB2LHpbuzQZOftyK9ZWVkZo0bfycMPjWP8uHAPTAxd775HsPCtRaxZ/XnaKdUK/XkOvS90K1esonOnDlsud+rYnpUrV6VYtK1CaAxdITyGhdBYH0ke65xPuTwGaWOl9TfbvNDdxwCnkHkL8AnRgdXZDLjR3btHp67u/pfoui9rukN3f83d+7h7L+BlYFENXzfS3Xu6e8+kBkcAt4+4kUULFzPijv9NbJ3FqN+pxzPx8efTzqhR6M9z6H2hmzHzTbp23ZsuXTrTpEkTzjijP0+OD2ugWQiNoSuEx7AQGhuzhg6QXgD+3cx2BahhF9tSoEd0fstcbTPbB3jP3YcDTwCHAOuAnSrd9lngJ2bWMrpNRzPbva6ozV8TbUG6HLh7+76thjuydw/OGngax3y3Ny9PHcfLU8dxYr/v5uvu62XUqOFMnvwY+++/D4sXT2Pw4DPTTtpGsx2b0atPDyZPeCXtlGqF/jyH3gfh/xxWVFRw6WVXM+GpMcybO5lHHnmSBQuq/VsrNYXQqOc5vkJorI9CPUi7wdP8zWwwMJTMbqzZ7n5e5Wn+ZtYNGBtd/xQwyN27mNkVwDlAObCKzPT+NWY2hsxg6Wl3H2pmlwIXRHe3HhgUrWt8NEutuqZhwA/IDPzucvf/ruv7SGqaf64kOb06V3IxzT9JSU7zL1aF8HOY1PTvYpbkNP9c0fMcX76n+R/W7ujEXmdnr5qSt/ZE3wepEGmAFJ8GSI1fIfwc6oUzPg2QikO+B0iHtjsqsdfZOaumFt37IImIiIgEI/w/F0RERKRgFds0fxEREZE6bSrQQ3m0i01EREQki7YgiYiISM5oF5uIiIhIFu1iExEREWkktAVJREREcka72ERERESyaBebiIiISCOhLUgiIiKSM9rFVqDWffNV2gm1KoTPRpq5+t20E0QkAfqcM8kF7WITERERaSTC3zwhIiIiBUu72ERERESyuG9KO6FBtItNREREJIu2IImIiEjObNIuNhEREZGqXLPYRERERBoHbUESERGRnNEuNhEREZEs2sUmIiIi0khoC5KIiIjkTKF+1IgGSCIiIpIzhfpO2trFlqCT+h3L/Hkv886CV/nN0IvTztnG3XcPY9myWcycOTHtlBqF/hhC+I2h90H4jaH3gRqTEHofFEZjY9WgAZKZXWtmv27A7Vqb2UUNuc+s9VxiZovNzM2sbaXlZmbDo+vmmtnhce+rvkpKShh+2x/4wQ8HcfChx3HmmadywAH75evu6+XBBx+mf//BaWfUqBAew9AbQ++D8BtD7wM1JiH0PiiMxvpw98RO+ZTvLUitge0aIEWDnuzOKcAJwLKs5d8D9otOQ4C7Gti53Xp95zCWLFnK++9/QHl5OWPHPsEpPzwpX3dfL1OmTGfNmrVpZ9SoEB7D0BtD74PwG0PvAzUmIfQ+KIzG+tiEJ3bKpzoHSGZ2brQ1Zo6ZPVjN9ZPNrGd0vq2ZLY3OH2hm083szej2+wE3AftGy4ZFXzfUzGZEX/O7aFkXM1toZg8A84DOle/T3We7+9JqcvsDD3jGNKC1mbXfjsejwTp0bMeHy1duubx8xUd06NAuH3fdaBTCYxh6Y+h9EH5j6H2gxiSE3geF0diY1XqQtpkdCFwNHOXuq81sl+1Y98+B29x9tJk1BUqBK4CD3L17tP5+ZLb29AIMGGdmxwAfRMsHRwOd+uoIfFjp8vJo2UfbsQ4RERFJSKG+D1Jds9j6Ag+7+2oAd1+zHet+DbjKzDoBj7r7u2aW/TX9otPs6HJLMgOjD4Bl2zk4qjczG0JmFxxW2oqSkhax17lyxSo6d+qw5XKnju1ZuXJV7PUWk0J4DENvDL0Pwm8MvQ/UmITQ+6AwGuujUKf5J3EM0sZK62m2eaG7jwFOAb4CJphZ32pua8CN7t49OnV1979E133ZgJYVVN0d1ylaVoW7j3T3nu7eM4nBEcCMmW/StevedOnSmSZNmnDGGf15cny4s8VCVAiPYeiNofdB+I2h94EakxB6HxRGY2NW1xakF4DHzOzP7v6Zme1SzVakpUAPYDowYPNCM9sHeM/dh5vZnsAhwBxgp0q3fRa4zsxGu/t6M+sIlMf4fsYBl5jZ34EjgC/cPS+71yoqKrj0squZ8NQYSktKuH/UQyxYsCgfd11vo0YNp0+f3rRt24bFi6dx3XW3MmrUQ2lnbVEIj2HojaH3QfiNofeBGpMQeh8URmN9FOouNqsr3MwGA0OBCmC2u59nZtcC6939ZjPrBoyNrn8KGOTuXczsCuAcMgOeVcBAd19jZmPIDJaedvehZnYpcEF0d+uBQdG6xrv7QTU0/QfwG6Ad8Akwwd0vsMw+vDuAk4ENwPnuPrO276+sacegn7kmpeG/l2d5xca0E0REpJ42frNim+NdcqlVy30Te539Yv2SvLXXOUBq7DRAik8DJBGRwqEBUv2E/+orIiIiBatQN8RogCQiIiI5U8yz2EREREQaFW1BEhERkZzxPH9ESFI0QBIREZGc0S42ERERkUZCW5BEREQkZzSLTURERCRLoR6DpF1sIiIiIlk0QBIREZGccffETvVhZieb2UIzWxx97FmDaBebiIiI5Ew+j0Eys1LgTuBEYDkww8zGufuC7V2XtiCJiIhIY9ELWOzu77n7N8Dfgf4NWZEGSCIiIpIznuCpHjoCH1a6vDxatt2Kfhdb0p9qbGZD3H1kkutMmhrjC70Pwm8MvQ/Cbwy9D9SYhND76pLk66yZDQGGVFo0MlePjbYgJW9I3V+SOjXGF3ofhN8Yeh+E3xh6H6gxCaH35Y27j3T3npVO2YOjFUDnSpc7Rcu2mwZIIiIi0ljMAPYzs73NrClwFjCuISsq+l1sIiIi0ji4+0YzuwR4FigF/tfd5zdkXRogJa8Q9hOrMb7Q+yD8xtD7IPzG0PtAjUkIvS8o7j4BmBB3PVaon5EiIiIikis6BklEREQkiwZIIiIiIll0DFJMZrYHW9+EaoW7f5xmT32ZWUt3X592RyEys13cfU3aHbUxs1PcvUEzN/Ih9MfQzLoChwJvN+QjCnLQ09rd16bdURczK3P3jdH5lkA34L3Qnmsz243M9O8KMn3B/FtoZt3IvPPzltcVYJy7v51eVXHSFqQGMrPuZjYNmAz8KTq9ZGbTzOzwVOPqJ/V/9AHM7ODoMfvQzEaaWZtK101Psy1qONrM3jaz+WZ2hJlNIvPZPh+aWe+0+wDM7PSs04+AkZsvB9B3daXz3zazRcAsM1tqZkekmLaFmb1oZm2j8+eQOcDze8BDZvbLVOMyVpvZc2b2UzNrnXZMdczsPOBjM1tkZt8D5gJ/BOaY2dmpxkWin7/ngNeA14F7gLfM7H4za5VuHZjZ5WQ+GsOA6dHJgL/F+dBVaRgdpN1AZvYm8DN3fz1r+ZHA/7j7oemUVWn5z5quAq5y913y2VNtiNmrwPXANOAC4HzgFHdfYmaz3f2wlPumAz8FWgJPAqe6+6vRIPh2dz86zT4AMysnM6X1EzLPLcAA4BHA3f0nabUBmNkb7n54dP4p4A53f9rMegH/7e5HpdkXdc1z94Oi8zOAk939MzNrDkxz90NS7nsLuBI4GzgZeBX4G/CEu3+VZttmUeNxwE7AHOCw6Pd4D2BS2o8hQPRH7WB3Xxj9/F3s7oPN7ELgJHcfkHLfIuBAdy/PWt4UmO/u+6VTVpy0BanhWmQPjgDcfRrQIoWe6twAtCHzD1blU0vCee53cvdn3H2tu98MXAI8Ew00Qxi9N3H3t9z9NeBTd38VwN3fAHZMN22Lo8i0zHD38939fGB1dD7VwVE1Orj70wDuPp1wHsNyM9u8S2M98GV0/msy76WStnJ3H+/uPyaza2g0cAaw3MzGpJu2RYW7r3b394H17r4EILDDDnZ094Ww5efv4Oj8PcCBaYZFNgEdqlnePrpO8kjHIDXc09Ffww+w9YPxOgPnAs+kVlXVG8Dj7j4r+wozuyCFnmqZWSt3/wLA3V+MdhH9A0h9CxdVB5JXZl3XNJ8hNXH3GWZ2IvBLM3sRuJwwBpeb7WNm48hs3epkZs3dfUN0XZMUuyr7FTDRzP4BzAdeMLNngX8D7ku1LGPLZ1lFW4zGAmOj3UKnplZV1QdmdiOZP8LeMbNbgEeBE4CPUi3baomZ/RfwAnA68CaAmTUhjD8aLwOeN7N32fq6sifQlcwfj5JH2sUWQ7SfvbqD6WK/QVUSzOxbwGfuvrqa6/YI4S87MxtI5iDJaVnL9wT+y90vTKdsS8cpwHOVXtA3L98X+JG7/ymdsuqZWQfgv4Ge7r5P2j0AZvbdrEWz3H19tOtlgLvfmUZXtmiwMRDYn8wfj8vJ7MJ6J9UwwMx+HW1hDZaZ7QxcTGZwfgdwEpld5suA69099UFSdPzWb4Fvk9kNeJO7r4ue+wOy/x1Kg5mVAL2o+royw90r0qsqThogiYiIBE4zj/MvhE2KIiIiUrsgZh4XEx2DJCIiEoA6Zh63zGeLaAuSiIhIKAph5nHR0BakmKJ3ZL0Q6EKlxzOk6dVqjC/0Pgi/MfQ+CL8x9D5QY0wFMfO4WGiAFN8TwCvAc2Tetj5Eaowv9D4IvzH0Pgi/MfQ+UGMc5wOf1XBdz3yGiGaxxWZmb7p797Q7aqPG+ELvg/AbQ++D8BtD7wM1SuOhfZrxjTez76cdUQc1xhd6H4TfGHofhN8Yeh+oURoJbUGKyczWkflokW+AzZ+f4+6+c3pVVakxvtD7IPzG0Psg/MbQ+0CN0nhogCQiIiKSRQdpJyD6OIpjoouT3X18mj3VUWN8ofdB+I2h90H4jaH3gRrjCniWXVHRFqSYzOwm4DtkPl0b4Gxgprtnf7BpatQYX+h9EH5j6H0QfmPofaDGJJjZVDKz7GZRaZadu/8jtagipAFSTGY2F+ju7puiy6XAbHc/JN2yrdQYX+h9EH5j6H0QfmPofaDGJGiWXRg0iy0ZrSudb5VaRe3UGF/ofRB+Y+h9EH5j6H2gxrg0yy4AOgYpvhuB2Wb2IpnPyzkGuCLdpG2oMb7Q+yD8xtD7IPzG0PtAjUm4FPitmWmWXYq0iy0BZtaezP5sgOnuvirNnuqoMb7Q+yD8xtD7IPzG0PtAjdI4aIDUQGbWzd3fMbPDq7ve3d/Id1M2NcYXeh+E3xh6CN3JGgAABLtJREFUH4TfGHofqDFpIc+yKxYaIDWQmY109yHRJtps7u598x6VRY3xhd4H4TeG3gfhN4beB2pMUuiz7IqFBkgxmVkzd/9XXcvSpMb4Qu+D8BtD74PwG0PvAzUmIfRZdsVCs9jim1rPZWlSY3yh90H4jaH3QfiNofeBGpMS8iy7oqBZbA1kZu2AjsCOZnYYmZkQADsDzVMLq0SN8YXeB+E3ht4H4TeG3gdqTFjos+yKggZIDXcScB7QCbiFrb9o/wR+m1JTNjXGF3ofhN8Yeh+E3xh6H6gxMe7+NzObzNZZdpdrll3+6RikmMzsR6G//bsa4wu9D8JvDL0Pwm8MvQ/UGEchzbIrBjoGKb4eZrZlX7GZtTGz69MMqoYa4wu9D8JvDL0Pwm8MvQ/UGMd/Rv+/pZrTzWlFFSttQYrJzGa7+2FZy95w92r/AkiDGuMLvQ/Cbwy9D8JvDL0P1JiE0GfZFQttQYqv1Mx22HzBzHYEdqjl69OgxvhC74PwG0Pvg/AbQ+8DNSahEGbZNXo6SDu+0cDzZnZfdPl8YFSKPdVRY3yh90H4jaH3QfiNofeBGhusgGbZFQXtYkuAmX0POD66OMndn02zpzpqjC/0Pgi/MfQ+CL8x9D5QY0OZ2WAys+x6AjOoOstulLs/mlJaUdIASUREJCChzrIrNjoGKSYzO9LMZpjZejP7xswqzOyfaXdVpsb4Qu+D8BtD74PwG0PvAzUmJNRZdkVFA6T47iDzQYLvAjsCFwB3plq0LTXGF3ofhN8Yeh+E3xh6H6gxCd9z97WbL7j758D3U+wpShogJcDdFwOl7l7h7vcBJ6fdlE2N8YXeB+E3ht4H4TeG3gdqTEDos+yKgmaxxbfBzJoCb5rZn4CPCG/gqcb4Qu+D8BtD74PwG0PvAzUmIchZdsVGB2nHZGZ7AR8DTYFfkfnU5RHRXydBUGN8ofdB+I2h90H4jaH3gRqTEuIsu2KjAVIMZlYKPODuP067pSZqjC/0Pgi/MfQ+CL8x9D5QozQuIW1SLDjuXgHsFW2qDZIa4wu9D8JvDL0Pwm8MvQ/UmJQCmGVXFHQMUnzvAVPMbBzw5eaF7v7n9JK2ocb4Qu+D8BtD74PwG0PvAzUm4Q7gLOBhMm8aeS6wf6pFRUgDpPiWRKcSYKeUW2qixvhC74PwG0Pvg/AbQ+8DNSbC3RebWWm0xes+M5sNXJl2VzHRAKmBzOxBdz8HWOvut6XdUx01xhd6H4TfGHofhN8Yeh+oMWGhz7IrCjpIu4HMbAFwAvA0cCxbPzMHAHdfk0JWFWqML/Q+CL8x9D4IvzH0PlBjkgphll0x0BakhrsbeB7YB5hF1V80j5anTY3xhd4H4TeG3gfhN4beB2pMRDTL7oZolt2/gN+lnFS0tAUpJjO7y91/kXZHbdQYX+h9EH5j6H0QfmPofaDGJJjZq0Bfd/8m7ZZipgGSiIhIQMzsAeAAINRZdkVBu9hERETCEvwsu2KgAZKIiEgACmiWXVHQLjYREZEAFMosu2KhLUgiIiJhCH6WXTHRFiQREZGAhD7LrlhogCQiIiKSRW9dLiIiIpJFAyQRERGRLBogiYiIiGTRAElEREQkiwZIIiIiIln+P3m7pu/yQcfrAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_spec_cm = pd.DataFrame(choises_spec, index = ['cluster '+str(i) for i in range(11)],\n",
    "                  columns = ['fraction '+str(i) for i in range(11)])\n",
    "plt.figure(figsize = (10,7))\n",
    "sn.heatmap(df_spec_cm, annot=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}