Tuto-GUDHI-persistence-diagrams.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# TDA  with Python using the Gudhi Library  \n",
    "\n",
    "# Persistence diagram   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import gudhi as gd  \n",
    "from sklearn import manifold\n",
    "from pylab import *\n",
    "%matplotlib inline\n",
    "# import sklearn_tda"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Persistence homology and persistence diagram"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One strong motivation for defining homology groups is the fact that two shapes can be distinguished by examining their holes.\n",
    "\n",
    "Persistent homology is a powerful tool to compute, study and encode efficiently multiscale topological features  of nested families of simplicial complexes and topological spaces. It encodes the evolution of the homology groups of the nested complexes (holes) across the scales.\n",
    "\n",
    "In the example below we consider the filtration given by a union of growing balls centered on the finite set of points $C$. \n",
    "<img src=\"Images/persistence.png\"  style=\"width: 600px;\"/>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "a) For the radius r = 0, the union of balls is reduced to the initial finite set of point, each of them corresponding to a 0-dimensional feature, i.e. a connected component; an interval is created for the birth for each of these features at r = 0.  \n",
    "\n",
    "b) Some of the balls started to overlap resulting in the death of some connected components that get merged together; the persistence diagram keeps track of these deaths, putting an end point to the corresponding intervals as they disappear.    \n",
    "\n",
    "c) New components have merged giving rise to a single connected component and, so, all the intervals associated to a 0-dimensional feature have been ended, except the one corresponding to the remaining components; two new 1-dimensional features, have appeared resulting in two new intervals (in blue) starting at their birth scale.  \n",
    "\n",
    "d) One of the two 1-dimensional cycles has been filled, resulting in its death in the filtration and the end of the corresponding blue interval.     \n",
    "\n",
    "e) all the 1-dimensional features have died, it only remains the long (and never dying) red interval. \n",
    "\n",
    "The final barcode can also be equivalently represented as **a persistence diagram** where every interval (a,b) is represented by the the point of coordinate (a,b) in $\\mathbb R^2$. \n",
    "\n",
    "Intuitively the longer is an interval in the barcode or, equivalently the farther from the diagonal is the corresponding point in the diagram, the more persistent, and thus relevant, is the corresponding homological feature across the filtration. \n",
    " \n",
    "          "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Protein binding dataset\n",
    "\n",
    "The data we study in this notebook represent configurations of protein binding. This example is borrowed from the paper a paper of  Kovacev-Nikolic et.al [1](https://arxiv.org/pdf/1412.1394.pdf).\n",
    "\n",
    "The paper compares closed and open forms of the maltose-binding protein (MBP), a large biomolecule consisting of 370 amino acid residues. The analysis is not based on geometric distances in $\\mathbb R^3$ but on a metric of *dynamical distances* defined by\n",
    "$$D_{ij} = 1 - |C_{ij}|,$$\n",
    "where $C$ is the correlation matrix between residues.\n",
    "\n",
    "Correlation matrices between residues can be found at this [link](https://www.researchgate.net/publication/301543862_corr), we are greatful to the authors for sharing data !\n",
    "\n",
    "We start from the Vietoris-Rips filtrations of the Protein binding distance matrices. See this [previous tutorial](Tuto-GUDHI-simplicial-complexes-from-distance-matrix.ipynb) for more details on the construction."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "path_file = \"./datasets/Corr_ProteinBinding/\"\n",
    "files_list = [\n",
    "'1anf.corr_1.txt',\n",
    "'1ez9.corr_1.txt',\n",
    "'1fqa.corr_2.txt',\n",
    "'1fqb.corr_3.txt',\n",
    "'1fqc.corr_2.txt',\n",
    "'1fqd.corr_3.txt',\n",
    "'1jw4.corr_4.txt',\n",
    "'1jw5.corr_5.txt',\n",
    "'1lls.corr_6.txt',\n",
    "'1mpd.corr_4.txt',\n",
    "'1omp.corr_7.txt',\n",
    "'3hpi.corr_5.txt',\n",
    "'3mbp.corr_6.txt',\n",
    "'4mbp.corr_7.txt']\n",
    "\n",
    "corr_list = [pd.read_csv(path_file+u,\n",
    "                         header=None,\n",
    "                         delim_whitespace=True) for u in files_list]\n",
    "\n",
    "dist_list = [1- np.abs(c) for c in corr_list]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We first consider the first distance matrix:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "mat_dist0 = dist_list[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Persistence barecodes and persistence diagrams"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We build the Vietoris-Rips complex from the distance matrix `mat_dist0`, see this [notebook](Tuto-GUDHI-simplicial-complexes-from-distance-matrix.ipynb) for more details."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "skeleton_protein0 = gd.RipsComplex(distance_matrix=mat_dist0.values,\n",
    "                                  max_edge_length=0.8) \n",
    "\n",
    "Rips_simplex_tree_protein0 = skeleton_protein0.create_simplex_tree(max_dimension=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we can compute persistence on the simplex tree structure using the <code>persistence()</code> method of the simplex tree class:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "BarCodes_Rips0 = Rips_simplex_tree_protein0.persistence()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The object `BarCodes_Rips0` is the list of barcodes: each element in the list is a tuple (dim,(b,d)) where dim is a dimension, b is birth parameter  and d is death parameter.\n",
    "\n",
    "Let's print the 20 first elements in the list:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, (0.07963602000000003, 0.35798637))\n",
      "(1, (0.12677510000000003, 0.39508646999999997))\n",
      "(1, (0.26003449999999995, 0.5273952))\n",
      "(1, (0.07943339999999999, 0.31429881000000004))\n",
      "(1, (0.08248586999999996, 0.30429980999999995))\n",
      "(1, (0.11378021999999999, 0.31171713999999995))\n",
      "(1, (0.07726765000000002, 0.26078758))\n",
      "(1, (0.09107215000000002, 0.25065161999999996))\n",
      "(1, (0.0709843, 0.22765623000000001))\n",
      "(1, (0.09347402000000005, 0.24999733000000002))\n",
      "(1, (0.07013614000000001, 0.22504734000000004))\n",
      "(1, (0.08752541000000003, 0.20355559))\n",
      "(1, (0.21541215000000002, 0.32814707))\n",
      "(1, (0.06835270000000004, 0.17527247))\n",
      "(1, (0.08857625000000002, 0.19539684000000002))\n",
      "(1, (0.08241111999999995, 0.18353136999999997))\n",
      "(1, (0.10362273, 0.20264340000000003))\n",
      "(1, (0.09289979999999998, 0.19181444000000003))\n",
      "(1, (0.09581541999999998, 0.19310879999999997))\n",
      "(1, (0.09541275000000005, 0.18175165000000004))\n"
     ]
    }
   ],
   "source": [
    "for i in range(20):\n",
    "    print(BarCodes_Rips0[i])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These 20 topolological features have dimension 1, they corresponds to holes of dimension 1."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have access to persistence_intervals per dimension using the `persistence_intervals_in_dimension()` method, for instance for dimension 0:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(0.0, 0.014988169999999967),\n",
       " (0.0, 0.016141790000000045),\n",
       " (0.0, 0.016186290000000048),\n",
       " (0.0, 0.019646410000000003),\n",
       " (0.0, 0.02002598),\n",
       " (0.0, 0.020196210000000048),\n",
       " (0.0, 0.02128167000000003),\n",
       " (0.0, 0.02139837),\n",
       " (0.0, 0.021894140000000006),\n",
       " (0.0, 0.022110460000000054),\n",
       " (0.0, 0.022791349999999988),\n",
       " (0.0, 0.023054660000000005),\n",
       " (0.0, 0.02376383999999998),\n",
       " (0.0, 0.023989579999999955),\n",
       " (0.0, 0.024279770000000034),\n",
       " (0.0, 0.02431161000000004),\n",
       " (0.0, 0.02514183000000003),\n",
       " (0.0, 0.025262099999999954),\n",
       " (0.0, 0.025565999999999978),\n",
       " (0.0, 0.025724140000000006),\n",
       " (0.0, 0.026689060000000042),\n",
       " (0.0, 0.027036240000000045),\n",
       " (0.0, 0.027239420000000014),\n",
       " (0.0, 0.027307149999999947),\n",
       " (0.0, 0.02737215000000004),\n",
       " (0.0, 0.027430060000000034),\n",
       " (0.0, 0.02758413000000004),\n",
       " (0.0, 0.02769299000000003),\n",
       " (0.0, 0.027711200000000047),\n",
       " (0.0, 0.028153890000000015),\n",
       " (0.0, 0.028894520000000035),\n",
       " (0.0, 0.029182599999999947),\n",
       " (0.0, 0.029376019999999947),\n",
       " (0.0, 0.029542529999999956),\n",
       " (0.0, 0.029609910000000017),\n",
       " (0.0, 0.029655129999999974),\n",
       " (0.0, 0.02989476000000002),\n",
       " (0.0, 0.029938699999999985),\n",
       " (0.0, 0.030054809999999987),\n",
       " (0.0, 0.030193550000000013),\n",
       " (0.0, 0.030221199999999948),\n",
       " (0.0, 0.030444119999999963),\n",
       " (0.0, 0.030617319999999948),\n",
       " (0.0, 0.030667760000000044),\n",
       " (0.0, 0.030936469999999994),\n",
       " (0.0, 0.03093904000000003),\n",
       " (0.0, 0.03096135),\n",
       " (0.0, 0.031193849999999967),\n",
       " (0.0, 0.031273030000000035),\n",
       " (0.0, 0.031370710000000024),\n",
       " (0.0, 0.031390559999999956),\n",
       " (0.0, 0.031416629999999945),\n",
       " (0.0, 0.031471049999999945),\n",
       " (0.0, 0.03165437000000004),\n",
       " (0.0, 0.031975549999999964),\n",
       " (0.0, 0.032014100000000045),\n",
       " (0.0, 0.03203754999999997),\n",
       " (0.0, 0.0320416),\n",
       " (0.0, 0.032043829999999995),\n",
       " (0.0, 0.032054090000000035),\n",
       " (0.0, 0.032091230000000026),\n",
       " (0.0, 0.03216505999999997),\n",
       " (0.0, 0.03251925),\n",
       " (0.0, 0.03252582000000004),\n",
       " (0.0, 0.032629699999999984),\n",
       " (0.0, 0.033200180000000024),\n",
       " (0.0, 0.03320555000000003),\n",
       " (0.0, 0.03322223000000002),\n",
       " (0.0, 0.033297669999999946),\n",
       " (0.0, 0.033486980000000055),\n",
       " (0.0, 0.033855549999999957),\n",
       " (0.0, 0.034142549999999994),\n",
       " (0.0, 0.03438534999999998),\n",
       " (0.0, 0.034528450000000044),\n",
       " (0.0, 0.03464100999999997),\n",
       " (0.0, 0.0351205),\n",
       " (0.0, 0.03512744000000001),\n",
       " (0.0, 0.03513681999999996),\n",
       " (0.0, 0.03527347999999997),\n",
       " (0.0, 0.035412629999999945),\n",
       " (0.0, 0.035590760000000055),\n",
       " (0.0, 0.03564263000000001),\n",
       " (0.0, 0.03568636999999997),\n",
       " (0.0, 0.035723210000000005),\n",
       " (0.0, 0.035939420000000055),\n",
       " (0.0, 0.035951949999999955),\n",
       " (0.0, 0.03602799000000001),\n",
       " (0.0, 0.03609680999999998),\n",
       " (0.0, 0.03610732999999999),\n",
       " (0.0, 0.036163060000000025),\n",
       " (0.0, 0.03644183000000001),\n",
       " (0.0, 0.03654824000000001),\n",
       " (0.0, 0.03661895000000004),\n",
       " (0.0, 0.037240709999999955),\n",
       " (0.0, 0.03733498000000002),\n",
       " (0.0, 0.037684680000000026),\n",
       " (0.0, 0.03774493000000001),\n",
       " (0.0, 0.03798142999999998),\n",
       " (0.0, 0.03815908999999995),\n",
       " (0.0, 0.038479739999999985),\n",
       " (0.0, 0.03851494),\n",
       " (0.0, 0.038575020000000015),\n",
       " (0.0, 0.038926959999999955),\n",
       " (0.0, 0.03907594999999997),\n",
       " (0.0, 0.03913962999999998),\n",
       " (0.0, 0.03923370000000004),\n",
       " (0.0, 0.03926191999999995),\n",
       " (0.0, 0.03947111999999997),\n",
       " (0.0, 0.03964137999999995),\n",
       " (0.0, 0.039716399999999985),\n",
       " (0.0, 0.03975846000000005),\n",
       " (0.0, 0.039848520000000054),\n",
       " (0.0, 0.03997198999999996),\n",
       " (0.0, 0.04015835000000001),\n",
       " (0.0, 0.04017048999999995),\n",
       " (0.0, 0.04023692000000001),\n",
       " (0.0, 0.04046704000000001),\n",
       " (0.0, 0.04072275999999997),\n",
       " (0.0, 0.04074774999999997),\n",
       " (0.0, 0.04088822999999997),\n",
       " (0.0, 0.04095618999999995),\n",
       " (0.0, 0.040961950000000025),\n",
       " (0.0, 0.04110269),\n",
       " (0.0, 0.04123368999999999),\n",
       " (0.0, 0.04127687000000002),\n",
       " (0.0, 0.04131072000000002),\n",
       " (0.0, 0.04140789),\n",
       " (0.0, 0.04165061999999997),\n",
       " (0.0, 0.041761539999999986),\n",
       " (0.0, 0.04181858999999999),\n",
       " (0.0, 0.041914959999999946),\n",
       " (0.0, 0.041964760000000045),\n",
       " (0.0, 0.04209390000000002),\n",
       " (0.0, 0.042270290000000044),\n",
       " (0.0, 0.042367350000000026),\n",
       " (0.0, 0.04254986000000005),\n",
       " (0.0, 0.042606880000000014),\n",
       " (0.0, 0.04261733999999995),\n",
       " (0.0, 0.042664409999999986),\n",
       " (0.0, 0.04296580999999999),\n",
       " (0.0, 0.04303701000000004),\n",
       " (0.0, 0.04306739000000004),\n",
       " (0.0, 0.043172690000000014),\n",
       " (0.0, 0.043465089999999984),\n",
       " (0.0, 0.04370370000000001),\n",
       " (0.0, 0.04372465000000003),\n",
       " (0.0, 0.04383413000000003),\n",
       " (0.0, 0.04391403999999999),\n",
       " (0.0, 0.04398658),\n",
       " (0.0, 0.044002929999999996),\n",
       " (0.0, 0.04407296999999999),\n",
       " (0.0, 0.04409098),\n",
       " (0.0, 0.044358110000000006),\n",
       " (0.0, 0.04439727999999998),\n",
       " (0.0, 0.04443403999999995),\n",
       " (0.0, 0.0444774),\n",
       " (0.0, 0.04451368),\n",
       " (0.0, 0.04482898999999996),\n",
       " (0.0, 0.04495304),\n",
       " (0.0, 0.04547506000000001),\n",
       " (0.0, 0.04547827999999998),\n",
       " (0.0, 0.04554780000000003),\n",
       " (0.0, 0.04615431000000003),\n",
       " (0.0, 0.046190469999999983),\n",
       " (0.0, 0.04625533000000004),\n",
       " (0.0, 0.04628038999999995),\n",
       " (0.0, 0.04678061),\n",
       " (0.0, 0.04679401999999999),\n",
       " (0.0, 0.04680114999999996),\n",
       " (0.0, 0.04704525999999998),\n",
       " (0.0, 0.04719843999999995),\n",
       " (0.0, 0.04747816000000005),\n",
       " (0.0, 0.04767326000000005),\n",
       " (0.0, 0.04768366000000002),\n",
       " (0.0, 0.04792481999999998),\n",
       " (0.0, 0.047979419999999995),\n",
       " (0.0, 0.04802757999999996),\n",
       " (0.0, 0.04811273999999999),\n",
       " (0.0, 0.04824103999999996),\n",
       " (0.0, 0.04824236999999998),\n",
       " (0.0, 0.04833752999999996),\n",
       " (0.0, 0.048487550000000046),\n",
       " (0.0, 0.04855529999999997),\n",
       " (0.0, 0.048578189999999966),\n",
       " (0.0, 0.04860536000000004),\n",
       " (0.0, 0.048857710000000054),\n",
       " (0.0, 0.04897733999999998),\n",
       " (0.0, 0.049018510000000015),\n",
       " (0.0, 0.04944188000000005),\n",
       " (0.0, 0.04948697999999996),\n",
       " (0.0, 0.049505270000000046),\n",
       " (0.0, 0.04955799999999999),\n",
       " (0.0, 0.049729880000000004),\n",
       " (0.0, 0.04981983000000001),\n",
       " (0.0, 0.04982386000000005),\n",
       " (0.0, 0.04984922999999997),\n",
       " (0.0, 0.049862760000000006),\n",
       " (0.0, 0.050271049999999984),\n",
       " (0.0, 0.05044280000000001),\n",
       " (0.0, 0.050556950000000045),\n",
       " (0.0, 0.050662969999999974),\n",
       " (0.0, 0.05070344999999998),\n",
       " (0.0, 0.05095380000000005),\n",
       " (0.0, 0.05103036999999999),\n",
       " (0.0, 0.0510853),\n",
       " (0.0, 0.05112470000000002),\n",
       " (0.0, 0.05119141000000005),\n",
       " (0.0, 0.051258079999999984),\n",
       " (0.0, 0.051416099999999965),\n",
       " (0.0, 0.05146225000000004),\n",
       " (0.0, 0.051573149999999957),\n",
       " (0.0, 0.05181603999999995),\n",
       " (0.0, 0.05189840000000001),\n",
       " (0.0, 0.05228542000000003),\n",
       " (0.0, 0.052458019999999994),\n",
       " (0.0, 0.05249015999999995),\n",
       " (0.0, 0.05256426000000003),\n",
       " (0.0, 0.05261344999999995),\n",
       " (0.0, 0.05268700000000004),\n",
       " (0.0, 0.052691350000000026),\n",
       " (0.0, 0.05286959000000002),\n",
       " (0.0, 0.05308782000000001),\n",
       " (0.0, 0.053129810000000055),\n",
       " (0.0, 0.05328801999999999),\n",
       " (0.0, 0.05388744999999995),\n",
       " (0.0, 0.05406628999999996),\n",
       " (0.0, 0.05424359000000001),\n",
       " (0.0, 0.05432115000000004),\n",
       " (0.0, 0.05434061999999995),\n",
       " (0.0, 0.054474369999999994),\n",
       " (0.0, 0.05475445000000001),\n",
       " (0.0, 0.054768700000000003),\n",
       " (0.0, 0.054808140000000005),\n",
       " (0.0, 0.055019629999999986),\n",
       " (0.0, 0.05515901000000001),\n",
       " (0.0, 0.05516030000000005),\n",
       " (0.0, 0.05521653999999998),\n",
       " (0.0, 0.055390720000000004),\n",
       " (0.0, 0.05547035),\n",
       " (0.0, 0.05559307999999996),\n",
       " (0.0, 0.05608519000000001),\n",
       " (0.0, 0.05611107999999998),\n",
       " (0.0, 0.05611354000000002),\n",
       " (0.0, 0.05622495999999999),\n",
       " (0.0, 0.05627249000000001),\n",
       " (0.0, 0.05637188999999998),\n",
       " (0.0, 0.05729700000000004),\n",
       " (0.0, 0.057299710000000004),\n",
       " (0.0, 0.05746996999999998),\n",
       " (0.0, 0.05753706000000003),\n",
       " (0.0, 0.058465679999999964),\n",
       " (0.0, 0.05849420999999999),\n",
       " (0.0, 0.058554910000000016),\n",
       " (0.0, 0.05858432999999996),\n",
       " (0.0, 0.058810150000000005),\n",
       " (0.0, 0.05891398000000003),\n",
       " (0.0, 0.05910733999999995),\n",
       " (0.0, 0.05927985000000002),\n",
       " (0.0, 0.05930276000000001),\n",
       " (0.0, 0.05937791999999997),\n",
       " (0.0, 0.059579099999999996),\n",
       " (0.0, 0.05959093000000004),\n",
       " (0.0, 0.059601050000000044),\n",
       " (0.0, 0.06001639000000003),\n",
       " (0.0, 0.06091707999999996),\n",
       " (0.0, 0.06102110999999999),\n",
       " (0.0, 0.06114719999999996),\n",
       " (0.0, 0.06140900000000005),\n",
       " (0.0, 0.061500650000000046),\n",
       " (0.0, 0.06153713999999999),\n",
       " (0.0, 0.06154331000000002),\n",
       " (0.0, 0.06164122000000005),\n",
       " (0.0, 0.06191270000000004),\n",
       " (0.0, 0.06200894000000001),\n",
       " (0.0, 0.062124570000000046),\n",
       " (0.0, 0.06247820000000004),\n",
       " (0.0, 0.06275173),\n",
       " (0.0, 0.06289644999999999),\n",
       " (0.0, 0.06309262999999998),\n",
       " (0.0, 0.06329755000000004),\n",
       " (0.0, 0.06339085),\n",
       " (0.0, 0.06340604000000005),\n",
       " (0.0, 0.06340869999999998),\n",
       " (0.0, 0.06365606999999995),\n",
       " (0.0, 0.06370852000000005),\n",
       " (0.0, 0.06376314000000005),\n",
       " (0.0, 0.06452473999999997),\n",
       " (0.0, 0.06464462000000004),\n",
       " (0.0, 0.06539583999999998),\n",
       " (0.0, 0.06557880000000005),\n",
       " (0.0, 0.06560913000000002),\n",
       " (0.0, 0.06565494000000005),\n",
       " (0.0, 0.06628835),\n",
       " (0.0, 0.06640049999999997),\n",
       " (0.0, 0.06651978999999997),\n",
       " (0.0, 0.06655549999999999),\n",
       " (0.0, 0.06667502000000003),\n",
       " (0.0, 0.0667084),\n",
       " (0.0, 0.06679813000000001),\n",
       " (0.0, 0.06682169999999998),\n",
       " (0.0, 0.06688167),\n",
       " (0.0, 0.06728802),\n",
       " (0.0, 0.06736142),\n",
       " (0.0, 0.06764263999999998),\n",
       " (0.0, 0.06825241000000004),\n",
       " (0.0, 0.06869391999999996),\n",
       " (0.0, 0.06877336),\n",
       " (0.0, 0.06892966),\n",
       " (0.0, 0.06897429),\n",
       " (0.0, 0.06918575000000005),\n",
       " (0.0, 0.06954788999999995),\n",
       " (0.0, 0.06996146000000003),\n",
       " (0.0, 0.07021496999999999),\n",
       " (0.0, 0.07103846999999996),\n",
       " (0.0, 0.07105813999999999),\n",
       " (0.0, 0.07122810999999996),\n",
       " (0.0, 0.07162656),\n",
       " (0.0, 0.07171408000000001),\n",
       " (0.0, 0.07273788000000003),\n",
       " (0.0, 0.07326412000000004),\n",
       " (0.0, 0.07355539),\n",
       " (0.0, 0.07385251000000004),\n",
       " (0.0, 0.07389020000000002),\n",
       " (0.0, 0.07394714999999996),\n",
       " (0.0, 0.07403272999999999),\n",
       " (0.0, 0.07465659999999996),\n",
       " (0.0, 0.07496113999999998),\n",
       " (0.0, 0.07510989000000001),\n",
       " (0.0, 0.07526754000000002),\n",
       " (0.0, 0.07619469999999995),\n",
       " (0.0, 0.07620041),\n",
       " (0.0, 0.07628743000000004),\n",
       " (0.0, 0.07870332000000002),\n",
       " (0.0, 0.07881972999999998),\n",
       " (0.0, 0.07893212999999999),\n",
       " (0.0, 0.07893759),\n",
       " (0.0, 0.07933292000000003),\n",
       " (0.0, 0.08004957000000001),\n",
       " (0.0, 0.08036911000000002),\n",
       " (0.0, 0.08080198999999999),\n",
       " (0.0, 0.08109694),\n",
       " (0.0, 0.08175500000000002),\n",
       " (0.0, 0.08176781),\n",
       " (0.0, 0.08248694999999995),\n",
       " (0.0, 0.08258633999999998),\n",
       " (0.0, 0.08299805000000005),\n",
       " (0.0, 0.08464198),\n",
       " (0.0, 0.08729047000000001),\n",
       " (0.0, 0.08744105999999996),\n",
       " (0.0, 0.08878103999999998),\n",
       " (0.0, 0.08969459000000002),\n",
       " (0.0, 0.09321352000000005),\n",
       " (0.0, 0.09341871000000002),\n",
       " (0.0, 0.09479172999999996),\n",
       " (0.0, 0.09541854999999999),\n",
       " (0.0, 0.10237912999999998),\n",
       " (0.0, 0.10285434000000004),\n",
       " (0.0, 0.10715870999999999),\n",
       " (0.0, 0.10830088999999998),\n",
       " (0.0, 0.10954903000000005),\n",
       " (0.0, 0.11366507000000003),\n",
       " (0.0, 0.11958183),\n",
       " (0.0, 0.14547953000000002),\n",
       " (0.0, 0.18964298999999996),\n",
       " (0.0, 0.39963213),\n",
       " (0.0, 0.40540978000000005),\n",
       " (0.0, 0.42515959999999997),\n",
       " (0.0, 0.50723615),\n",
       " (0.0, 0.66869636),\n",
       " (0.0, inf)]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rips_simplex_tree_protein0.persistence_intervals_in_dimension(0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The last bars (0.0, inf) die at infinity."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally we can plot the points (birth, death) in the so-called **persistence diagram**:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3Xl8VPX1+P/XmWyQkAUIoIYtrahBUNEQ910q4rfqp3VBq8atbmGR0qoftVFRf1U+1X60UpVaNfZTq5ZWixbFahVXGoLsQSgQxACyZ4Psc35/3JlxEiYb5GZmkvPMg0dm7ty598wAc+Z93ssVVcUYY4wB8IQ7AGOMMZHDkoIxxpgASwrGGGMCLCkYY4wJsKRgjDEmwJKCMcaYAEsKJuqJSJWIfC/ccXQGEXlARP7Pd3uo77XFhDsu03NYUjCuEZGNIlLt+2DbJiIvikifzj6PqvZR1Q1txHKWiJR29rndpKqbfK+tMdyxmJ7DkoJx2w9VtQ9wPDAWuK+jBxCR2E6Pqoez1odpiSUF0yVUdTPwDjAKQERSReQPIrJVRDaLyMP+DyoRuU5EPhOR34jIbuABETlcRBaISLmI7BSR1/zHFhEVkcN9tyeISLGIVPqO+3MRSfKd+zBfq6VKRA4TEY+I3C0i60Vkl4i8LiL9fMcZ7jturohs8p3z3qBzxojIPb7nVorIYhEZ4nvsKBH5p4jsFpE1InJ5S++LiGT6XleliPwTSA96zB9DrO/+9SKy2rfvBhG5pdmx7vS9n1tE5KZm78tLIvKMiMwTkb3A2SJyoYgsEZEKEflGRB4Ice7rfY/tEZFbRWSsiCwXkTIRefpA/i2YCKeq9sf+uPIH2Aic57s9BFgFPOS7/ybwHJAEDAQKgVt8j10HNACTgVigN/Bn4F6cLzK9gNOCzqPA4b7bW4HTfbf7Asf7bp8FlDaL7w5gITAYSPDF82ffY8N9x/297/zHArVAlu/xXwArgCMB8T3e3/d6vgGu98V+PLATOLqF9+gL4Anf+c8AKoH/axZDrO/+hcD3fec7E9gX9PrGA98CRwOJwB+bvS8vAeXAqUHv4VnAaN/9Y4BtwCXNzv2sb98fADW+v7eBQAawHTgz3P/O7E8n/78NdwD2p/v+8SWFKqAM+Br4ne8DdpDvA7Z30L5XAh/6bl8HbGp2rJeB2cDgEOcJ/vDbBNwCpDTbJ1RSWA2cG3T/UKDe92Hu/1AcHPR4ITDRd3sNcHGIWK4APmm27Tng/hD7DsVJfklB215pKSmEeP6bwFTf7ReAXwU9dniIpPByG39f/wv8ptm5M4Ie3wVcEXT/r8Ad4f53Zn8694+Vj4zbLlHVNFUdpqq3q2o1MAyIA7b6yhBlOB+cA4Oe902z49yJ8w25UERWicgNLZzvx8AE4GtfWebkVmIbBrwRFMNqoBEnafl9G3R7H+DvKB8CrG/hmCf6j+k77k+AQ0LsexiwR1X3Bm37uqVgReQCEVnoK0uV+V6nv9x0GE3fs+bv337bROREEflQRHaISDlwa9Dx/LYF3a4Ocb/TBw6Y8LKkYMLhG5yWQrovYaSpaoqqHh20T5Ple1X1W1X9qaoehtMS+J2/Xt5sv0WqejFOgnkTeD3U8YLiuCAohjRV7aVO/0d7XsP3W9i+oNkx+6jqbSH23Qr09fV5+A0NdTIRScD5Zv5rYJCqpgHzcBKl/1iDg54yJMRhmr8HrwBzgSGqmopTKpL9nmV6FEsKpsup6lbgPeBxEUnxdfh+X0TObOk5InKZiPg/9PbgfMA1NtsnXkR+IiKpqloPVATtsw3oLyKpQU95FnhERIb5nj9ARC5u58t4HnhIREaI4xgR6Q+8DRwhIteISJzvz1gRyQrxPnwNFAEP+mI/DfhhC+eLx+l32AE0iMgFOHV+v9eB60UkS0QSgfx2vIZkYLeq1ohIDnBVO1+76cYsKZhwuRbng64Y50N+Dk5NvyVjgX+LSBXOt9upqloSYr9rgI0iUoFTDrkaQFW/wums3uAr6xwGPOk71nsiUonT6XxiO+N/AueD+D2c5PMHnD6SSpwP64nAFpzy02M4H+ihXOU7527gfpy+k/34jjvFd849vufNDXr8HeAp4ENgHU4HNjgtspbcDszwvfZ8vmtVmR5MVO0iO8Z0N76WyUogQVUbwh2PiR7WUjCmmxCR//KVofritE7esoRgOsqSgjHdxy04fQ7rcfpSQnVuG9MqV8tHIjIep24bAzyvqo82e3woUACk+fa5W1XnuRaQMcaYVrmWFMRZsmAtMA4oBRYBV6pqcdA+s4ElqvqMiIwE5qnqcFcCMsYY0yY3FxrLAdapb/VKEXkVuBhntImfAim+26k4ozValZ6ersOHD+/cSI0xpptbvHjxTlUd0NZ+biaFDJrOoCxl/+F+D+AMB5yMs2bMeaEOJCI3AzcDDB06lKKiok4P1hhjujMRaXG2fDA3O5pDzYxsXqu6EnhJVQfjTNn/o4jsF5OqzlbVbFXNHjCgzURnjDHmALmZFEppOtV+MPuXh27EN2FGVb/AWY2x+dorxhhjuoibSWERMMK3Xnw8zgzPuc322QScC4HJNr1whtQZY4wJA9eSgm/SzCRgPs7qk6+r6ioRmSEiF/l2mw78VESW4SxBcJ3aFGtjjAkbVy9z6JtzMK/Ztvyg28U4F/0wxhgTAWxGszHGmABLCsYYYwIsKRhjjAmwpGCMMSbAkoIxxpgASwrGGGMCLCkYY4wJcHWegptWrlzJ/PnzOfzwwznqqKN46623GDZsGJdddlnnnKCwEAoKoKQEMjMhNxdycjrn2MYYE6GitqXw1VdfMX36dHr16sX777/PtGnTGDJkCFu3bj34gxcWQn4+7NwJGRnO7/x8Z7sxxnRjUZsU/KthJCUlUVtbi9frRVXplFUyCgogNRXS0sDjcX6npjrbjTEmyrz33nvt3tfVy3G6YcCAAfqjH/2I3bt3U1paSp8+fejfvz8bN24kKSmJww8//OBP8s47kJS0//a9e+GCCw7++MYY0wWqqqr4/PPP2bVrF5WVlYtVNbut50Rdn8KwYcN47rnn3D1JXp5TMkpL+25bWRmkp8OsWe6e2xhjDlJtbS2PP/44TzzxBFOmTOHOO++kd+/e7Xpu1JaPXJWbC+XlTiLwep3f5eXOdmOMiWDz589n9OjR/Pvf/2bRokXk5+fTq1evdj8/6loKXSInB2bMaDr6aPp0G31kjIlYmzZtYtq0aSxbtownn3ySCy+88ICOY0mhJTk5lgSMMREvuFQ0depU/vSnP3WoZdCcJQVjjIlS7777LlOmTGHkyJEsWrSIzMzMgz6mq0lBRMYDTwIxwPOq+mizx38DnO27mwgMVNU0jDHGtOjrr79m2rRpLF++nKeeeooJEyZ02rFd62gWkRhgFnABMBK4UkRGBu+jqtNU9ThVPQ74LfA3t+IxxphoV1tbyyOPPMIJJ5zA8ccfz8qVKzs1IYC7LYUcYJ2qbgAQkVeBi4HiFva/ErjfxXiMMSZqvfvuu0yePJlRo0ZRVFTE8OHDXTmPm0khA/gm6H4pcGKoHUVkGJAJ/KuFx28GbgYYOnRo50ZpjDERbOPGjUybNo2VK1fy1FNPcYHLE2jdnKcgIba1NH16IjBHVRtDPaiqs1U1W1WzBwwY0GkBGmNMpKqpqeHhhx8mOzubE044gRUrVrieEMDdlkIpMCTo/mBgSwv7TgTyXIzFGGOixjvvvMOUKVNcLxWF4mZSWASMEJFMYDPOB/9VzXcSkSOBvsAXLsZijDERr6tLRaG4Vj5S1QZgEjAfWA28rqqrRGSGiFwUtOuVwKsabSvzGWNMJwkuFWVnZ3dZqSgUV+cpqOo8YF6zbfnN7j/gZgwHzC6yY4zpAu+88w6TJ0/mmGOOYfHixQwbNiys8diCeKHYRXaMMS7buHEjl1xyCVOmTOHpp5/mb3/7W9gTAlhSCM0usmOMcUlNTQ0PPfQQ2dnZjB07lpUrVzJ+/PhwhxVgax+FUlLitBCCpaQ4240x5gDNmzePKVOmREypKBRLCqFkZu5/kZ2KCme7McZ00MaNG7njjjsoLi7m6aefjqiWQXNWPgrFLrJjjOkENTU1zJgxg+zsbHJyclixYkVEJwSwpBCa/yI76emwebPze8YMG31kjGm3f/zjHxx99NEsW7aMxYsXc88995CQkBDusNpk5aOW2EV2jDEHoKSkhKlTp/LVV1/xu9/9jvPPPz/cIXWItRSMMaYTVFdX8+CDDzJ27FhOPvlkVqxYEXUJAaylYIwxB+3tt99m6tSpjBkzhi+//DKqV3O2pGCMMQdow4YN3HHHHaxZs4ZnnnmGH/zgB+EO6aBZ+cgYYzrIXyrKycnh5JNPZvny5d0iIYC1FIwxpkPefvttpkyZwvHHHx/1paJQLCkYY0w7bNiwgalTp7J27VqeffbZbtMyaM7KR8YY04rq6moeeOABcnJyOPXUU7tVqSgUaykYY0wL3nrrLaZOncoJJ5zQLUtFoVhSMMaYZvylov/85z8899xzjBs3LtwhdRlXy0ciMl5E1ojIOhG5u4V9LheRYhFZJSKvuBmPMca0JrhUdNppp7F8+fIelRDAxZaCiMQAs4BxQCmwSETmqmpx0D4jgP8GTlXVPSIy0K14jDGmNf5SUXZ2NkuWLGHIkCHhDiks3Cwf5QDrVHUDgIi8ClwMFAft81NglqruAVDV7S7GY4wx+1m/fj1Tp05l3bp1zJ49m/POOy/cIYWVm+WjDOCboPulvm3BjgCOEJHPRGShiIRcU1ZEbhaRIhEp2rFjh0vhGmN6kurqau6//35OPPFEzjjjDJYvX97jEwK421KQENs0xPlHAGcBg4FPRGSUqpY1eZLqbGA2QHZ2dvNjGGNMu6lqoFSUk5PD0qVLGTx4cLjDihhuJoVSILgoNxjYEmKfhapaD5SIyBqcJLHIxbiMMT1AIYUUUEAJJWSSSS659F/fnylTprBhwwZ+//vfW8sgBDfLR4uAESKSKSLxwERgbrN93gTOBhCRdJxy0gYXYzLG9ACFFJJPPjvZSQYZfLvvW67Mv5ITTjyBM888k2XLlllCaIFrLQVVbRCRScB8IAZ4QVVXicgMoEhV5/oe+4GIFAONwC9UdZdbMRljupdQrYEcciiggFRSSdVUvp77NV/c8QVpOWlcvPRi7hx8Z7jDjmiiGl0l+uzsbC0qKgp3GMaYMPO3BlJJJYUUKqignHJmMIMHeIA+6/qwcOpCKjdUcurTp3LouYeymc3MY164Qw8LEVmsqtlt7Wczmo0xUamAAhpoYCUrqaSSZJI5lEN5ft/z7Hh0Bx/87gPG3DWGUW+MIiY+hjLKyCQz3GFHPEsKxpio0LxU9DEfs4c9JJBAEklUazVL5i7h0zs+5fQTT+fkpSczaPAgBKGMMsopZzrTw/0yIp4lBWNMxAsuFWWQwU52UkIJiSSSTDJ16+rYPmU7tRtryXg+g3+e+8/9ksh0ppNDTrhfSsSzpGCMiXj+juM00gBII40EEti7by/1v6qn8plKUu5KIXVqKofFHwZAju/HdIwlBWNMxCuhhIygBRFUlcS/J1I2rYzYE2MZuHQgfQf35VAO5UiODGOk0c+SgjEmIrQ0vBQgk0x2spM00ihfV87nUz5n98bd9P9Df84+5+wmo49yyQ3zK4luduU1Y0zYBU82iyee+cznQi7kUi6lkEJyyWX3vt188stPePOkN+l7Tl9OXXoqj53zGOmks5nNpJPODGZYyeggWUvBGNNl2ppsto1tFFJIPfUoyru8S6mW8uO//5hVd6wi6eQkTll2ClkZWYHnXs/14X5Z3YolBWNMlwg1giiffGYwgxJKiCeeQgqppZYYYgDY9599fDnlS4q/LubvL/6ds88+O8yvovuz8pHpFgopJI88JjCBPPIopDDcIZlmgkcQefCQRhqppFJAAZlkspKV1FGHBw/sA+99XvRkxXOeh7RlaZYQuoi1FEzUa+0bqNWX3dFap3BLgkcQbWc7y1jGdrbTSCMncRI72Ymqom8q3mle5BSh97LekAESciV+4wZLCibqhRrD7t9uSaHzNU/Ca1jDpVzKYAYzhjEtJgj/CKI66ljAAiqoQFFiiGE5y/H+xwuTwfuNl5gXY+h1di8AvHgZy9iufpk9lpWPTNQroYQUUppsSyGFEkrCFFH3FpyEd7KTNaxBUcooC7TSmpfvCilkG9v4F//in/yTMpzraHnwELM3hop7K6g6uYrEcYmkLU0j8exEvL6fTDK5E1vZtKtYUjBRL5NMKqhosq2CClv8zCXBSXgta4knniSS2MUuVrKSZSzjBm4IJIYXeZFLuZTP+MxZo4hqFAWFmL/FUDuyFkqg1/JejJ4+mnPjzqUvfRnAACYwgRd50Vp8XcjKRy0pLISCAigpgcxMyM2FHPuHGYlyySWffIAmk5hs8TN3BE8kq6SSJJKopJK97KUPfUgllV3sIp98ruAK7ud+FCWVVOqoQxBkreCd4qX+m3oSCxLxnOWhgQbGMIZZzAr3S+zRrKUQSmEh5OfDzp2QkeH8zs93tpuIk0MOM5hhk5i6SC65lFNOGWX0oQ9VVFFGGamkEk88ddTRj36kksrjPE499fShD4IQtzcOz70e9BSFH0Dvpb2Rs4Q66kgm2WYjRwBXWwoiMh54EufKa8+r6qPNHr8O+B9gs2/T06r6vJsxtUtBAaSmQprTYRn4XVBgrYUIZYufdR1/Ei6ggDTSqKKKBBJIJplaaqmnnmM5lhRS2MY2BjCAGq2h7o06tk3bRvxp8XiXe+l1mNOR3EgjqaTyKI/a32EEcC0piEgMMAsYB5QCi0RkrqoWN9v1NVWd5FYcB6SkxGkhBEtJcbYbY5ok4UIKuYEb2MUu+tGPYzmWgQykjDIGMYiUtSl8OflLvJu9HFpwKHqWUk89p3Ea+9jX7iGtpmu42VLIAdap6gYAEXkVuBhonhQiT2amUzLytxAAKiqc7caYJnLI4QVeaHJpzDLK2LV3F8MfGc57s99j+D3DiZkcQ1lcGXHE8St+ZctTRCg3+xQygG+C7pf6tjX3YxFZLiJzRGRIqAOJyM0iUiQiRTt27HAj1qZyc6G8HMrKwOt1fpeXO9uNMfsJ7tcp1VLK/1rOqpGrSP06lTeWv8G4n40jMy6TH/Ej5jDHEkIEE1V158AilwHnq+pNvvvXADmqOjlon/5AlarWisitwOWqek5rx83OztaioiJXYm7CRh8Z02Fr165l8uTJbN68mVmzZnHmmWeGOyTjIyKLVTW7rf3cLB+VAsHf/AcDW4J3UNVdQXd/DzzmYjwdk5NjScCYdtq7dy+PPPIIs2fP5t5772XSpEnExcWFOyxzANwsHy0CRohIpojEAxOBucE7iMihQXcvAla7GI8xppOpKnPmzCErK4tNmzaxYsUKpk2bZgkhirnWUlDVBhGZBMzHGZL6gqquEpEZQJGqzgWmiMhFQAOwG7jOrXiMMZ3rq6++YvLkyXz77bf88Y9/tFJRN+Fan4JbuqxPwZhu7kBWOgWoqqri4Ycf5vnnn+e+++4jLy/PWgZRoL19Cjaj2ZgeKPjyl8HLjbd2HQpV5S9/+QsjR45k8+bNrFixgjvuuMMSQjdjax8Z0wO1tNz4TGYyiEH7tR6CS0X/93//xxlnnBHO8I2LrKVgTA8UarnxWmpZwIImrYd7qu4h9+5cTj/9dC688EK+/PJLSwjdnLUUjOmBglc69VvJykDrQVXZPWc3n//sc4acNYTly5dz6KGHtnJE011YUjCmB2ppufFTOIWyr8r4bNJnVG+v5pw/nYP3DC+HYgmhp7DykTE9UKjlxk+tOpU1d61h7ulzGfrDofzoyx+RdEaSXayoh7GWgjE9lH+lU/+oosnTJxN7dizjVoxj0CGD7GJFPZQlBWN6sNWrVzN58mS2b9/OnFfmkHB6QpO5C9OZbkta9zCWFIyJQAc6say9qqqqeOihh3jhhRcCE9BiY52PA0sCPZv1KRgTYQ5kYll7qSqvv/46WVlZbN26lRUrVjB16tRAQjDGkkJLCgshLw8mTHB+2/WZTRcJnljmwUMaaaSSSgEFTfZ7kRcZxSgGMIBRjOJFXmz1sdWrV3Peeefx8MMP88orr/Dyyy9zyCGHdPXLMxHO1j4KpbAQ8vOd6zSnpDhXXSsvhxkzbDlt46pCCrmMy6immgYaiCOOdNIZwQjqqGMe8wDnQ/8X/ILe9CaJJPayl2qquY3b+IzP+JRPiSee/vSnobKB3Q/tJvbFWB755SPcfvvt1jLogdq79pElhVDy8va/HGdZGaSnw6xZ7p7b9Fj+stFKVrKLXcQQg6KkkUYNNaSRxpEcSSaZvMM71FNPLLGUUUYNNdRTj6LEEIMHD3EaR/3r9TRMbyDh3AQGPzaYNYesCffLNGESCRfZiV4lJZDR7MqhKSnOdmM6qL2dxv6yURxxCIIHD4pSTjn11OPBw0Y2soQlbGMbaaRRRx2K0kADjTQC4MVLTHEMDZMb0J1K4quJZJyWwS527XdOY5qzPoVQMjOdklGwigpnuzEd0JFOY/96RI00MoABxBBDAw3sYx+NNLKb3ZRTThppxBDDHvYEEoIgzkEqgV9A45mNcAn0WdwHPU3Zy14GMahrX7yJSu1uKYjIKcDw4Oeo6ssuxBR+ubkwdSosXgy1tZCQAAMGwHSbxGM6pqXVSAso2K+1kEgiH/Ihe9iDBw9JJFFLLR7fj7/V4O8r2Ma2QEsBBV4Dfg6e8zzoSkUHKV68AFRTzQxmdOVLN1GqXUlBRP4IfB9YCr42qvPPsNWkICLjgSdxrrz2vKo+2sJ+lwJ/AcaqamRcQaeyEnbs+C4p9OoV7ohMFCqhhAyaliJTSKGE70qRhRQyk5m8z/vUUANAHXVUUBFoATTSiCDsYx9f8zWJJJJAAg004C32wiSI2R1DzGsxJJyaQAIJlFNOLbUMZjC/5Jdcz/Vd98JN1GpvSyEbGKkd6JUWkRhgFjAOKAUWichcVS1utl8yMAX4d3uP7bqZM52O5vR0iI+Hujrn/syZMGdOuKM7YG5PiDL7C7UaaQUVgfWE/OWldawjkUSqqaaRxkDpSFESSKCWWqdF4FNNNfGV8STMSKD+pXoS8xPpc1sf6mLriCOOOuq4iIu4kzvt79h0SHv7FFYCHR3QnAOsU9UNqloHvApcHGK/h4CZ4PuKFAkWLYLERCchgPM7MdHZHqXcnBBlWpZLLuWUU0YZXryUUUY55eSSC3xXXqqjjmqqSSCBRBLx4CGW2EDZqAkFeVWoyaoheUcyP1n5E0ZOHklabBqHcziXcAn/4B/MYY4lBNNhrbYUROQtnDJRMlAsIoVArf9xVb2oladnAN8E3S8FTmx2/DHAEFV9W0R+3kocNwM3AwwdOrS1kE0LOlLbNp3HvxppS+sJ+ctLySSzk50kkNCk89iLlzrqvjvgKmAyNO5upM9rfTj+1OObTFoz5mC1VT769UEcW0JsC3zlEREP8BvgurYOpKqzgdngzFM4iJjaZ+xY+PRT57a/fLRvH5x2muundkt7atvGHf7VSEPxl5eO4Ai+5uvAXAN/68BfOqISeBAogN7390ZuFTyxHlvW2nS6VstHqrpAVRcAE/y3g7e1cexSYEjQ/cHAlqD7ycAo4CMR2QicBMwVkTYnV7juzjth+HDweGDvXuf38OHO9iiVSSYVNB1mG1zbNuHhLy9VUkkyyVRTzT72EUMMscQiKnj+7IEsYBckrEwgZlIMGqsIEihDGdNZ2tunMC7EtgvaeM4iYISIZIpIPDARmOt/UFXLVTVdVYer6nBgIXBRRIw+ysmBp56C88+H0aOd3089FdVLXLRV2zbhkUMOV3AFa1kLwFCGEk88McSQsioFzgFmQsLrCcS/GE/soFgUpS99OY/zrPRnOl1bfQq3AbcD3xOR5UEPJQOft/ZcVW0QkUnAfJwhqS+o6ioRmQEUqerc1p4fdjk5UZ0Emmurtm3Cp4giTubkQD/PgsoFbH5wM3sK9nDSAyeRfms6C2MW0oc+nM3ZgYvf3En0tlxN5Gp17SMRSQX6Ar8C7g56qFJVd7scW0hdsvaRMV1oAhPIIANRYf2r6/n855/jOd9D0qNJXDzwYiqoYBObyCCDfeyz4cTmgHTK2keqWg6UA1f6DjoQ6AX0EZE+qrqpM4I1pifLJJP1q9azPG85deV1nD/nfKpOruJbvmUzm61VZ7pUe2c0/xB4AjgM2A4MA1YDR7sXmjHdX0VFBWUPlvHhHz9k1AOjOO6W46iKqaKGGl7gBUsEpsu1t6P5YZzRQWtVNRM4F/jMtagM4Ew4yyOPCUwgjzybaNaNqCqvvPIKWVlZ9Crrxd9X/p2Tbj+JrTFbSSedGcywhGDCor3LXNSr6i4R8YiIR1U/FJHHXI2sh/PPQE4ltckMZPuwiH4rV65k0qRJVFRUMGfOHE4++WQAxjM+zJEZ0/6WQpmI9AE+Af4kIk8CDe6FZdp7SUbTudxsnVVUVPCzn/2Mc845h8svv5xFixYFEoIxkaK9SeFiYB9wB/AusB74oVtBme/W1g9mM5Dd5db6UKrKn/70J7KysigvL2fVqlXcfvvtxMTEdFLkxnSedpWPVHWviAwDRqhqgYgk4sw9MC5pa3VN0/ncWB9q5cqV5OXlUVlZ2aRUZEykaldLQUR+CswBnvNtygDedCsoYzOQw6EzW2fBpaKJEydaqchEjfaWj/KAU8FZPEdV/wMMdCso890M5HTS2cxmG5HSBTpjfajgUlFFRQWrVq3itttus1KRiRrtHX1Uq6p1Is7CpyISC80XeTedrbXVNbuLSLrwTy655JMPOC0E/3IS02nfZVhXrFjBpEmTqKqq4q9//SsnnXSSm+Ea44r2thQWiMg9QG8RGYdz6cy33AvL9ASRduGfA22dlZeXM23aNM4991wmTpxIYWGhJQQTtdrbUrgbuBFYAdwCzAOedyso0zNE4oV/OtI685eK7rrrLi644AJWrVrFgAEDXI7QGHe1d/SRV0TeBN5U1R0ux2R8Iqm04oZovvDPihUryMvLY+/evVYqMt1Kq+UjcTwgIjuBr4A1IrJDRPK7Jrye62BKK9GyPEY0XviehpxzAAAdyElEQVQnuFR01VVXWanIdDtt9SncgTPqaKyq9lfVfjjXWT5VRKa5Hl0PdqAzmiOtTt+aaBp2q6r88Y9/JCsri6qqKoqLi7n11lttVJHpdtoqH10LjFPVnf4NqrpBRK4G3sO5xrJxgb+0sp3trGUtlVTShz5NJrOFEol1+pZEy4V/li9fTl5eHtXV1bzxxhuceOKJ4Q7JGNe0lRTighOCn6ruEJG4tg4uIuOBJ3FmPz+vqo82e/xWnDkQjUAVcLOqFrc3+O4sk0zW+H7iiSeJJCqppIoqCils8YMz2ur0kTzstry8nPz8fP785z/z0EMPcdNNN1nLwHR7bZWP6g7wMUQkBpiFcy3nkcCVIjKy2W6vqOpoVT0OmIlzzQaDU1rxX7c3jjhqqQXgCI5otYQUjXX6SKOqvPzyy2RlZVFdXU1xcTG33HKLJQTTI7TVUjhWRCpCbBecK7C1JgdYp6obAETkVZyF9QItAVUNPnYSNiEuIIccBjOYMsqooopkkjmWY0knvdVv/Qc7Aau7ONCRW8uWLWPSpEnU1NTw5ptvktONrtNtTHu02lJQ1RhVTQnxJ1lV2yofZQDfBN0v9W1rQkTyRGQ9TkthSqgDicjNIlIkIkU7dvScEbFjGMNoRnMBF3AapzGQgW1+67flMQ6ss72srIypU6cybtw4rr76ahYuXGgJwfRI7Z3RfCAkxLb9WgKqOktVvw/cBdwX6kCqOltVs1U1uydNDjrQ0Tk55DCLWcxjHrOY1aMSAnRs5JaVioxpqr0zmg9EKTAk6P5gYEsr+78KPONiPFEnWkbnRJr2drYvW7aMvLw8amtr+fvf/24tA2NwNyksAkaISCawGZgIXBW8g4iM8K24CnAh8B9ME5E8OidStXUtirKyMvLz83nttdd46KGHuPHGG61lYIyPa+UjVW0AJgHzgdXA66q6SkRmiMhFvt0micgqEVkK/AwicNaSiTotld2u8V5DQUEBWVlZ1NbWUlxczM0332wJwZggohpdA36ys7O1qKgo3GGYCNd89NHJy07m2bxnqa2t5Xe/+x1jx44Nd4jGdCkRWayq2W3t52ZHszFh4+9sf6XsFWKmxDD9B9O59tprWbhwoSUEY1phSSFCRcuidpHK6/VaqciYA+BmR7M5QP5x9qmkNhln39PmGxyopUuXkpeXR319PXPnzrWWgTEdYEkhArW2qJ3/d3e5xkJnXjOirKyMX/7yl7z++us8/PDD3HjjjXg81hg2piPsf0wEKqGEFFKabEshhSUsIZ981rCGjWzkb/yNS7mUF3kxTJEenM5a5tvr9fLSSy+RlZVFfX09xcXF/PSnP7WEYMwBsJZCBGppnH0FFSSSGFg5NY00qqjifu7naI6OuhZDZyzzvWTJEvLy8mhoaOCtt94iO7vNwRXGmFbYV6kI1NI4+xRS2MIW4n0/AEkkUU99mxffiUQttYjas8z3nj17mDRpEuPHj+eGG25g4cKFlhCM6QSWFCJQS4vajWEMe9gTSAgAddTRl74Re72E1rS2zHdLo6+8Xi8vvvgiI0eOpLGxkeLiYm666SYrFRnTSax8FKFaWt7iLd6iiiqSSKKOOuqp53t8Lyqvl9DSMt/jGR9y9NVPlvyEZ/KeobGx0UpFxrjEvl5FmNbmJ+SQw4M8iCCUU04venEkRxJLbERe17gtLbWIiihqsspp7z29WT9pPbeMv4UbbriBL774whKCMS6xZS4iyIu8yP3cTz319KUvh3EYscTuNz+hM4dxdrX2xD6BCWSQgXiFtQVrKfzvQob91zAyHsng/X7vhylyY6Jbe5e5sPJRhCikkPu5H0VJI41aalnDGo7kyP1G43TVyqmdnXzaOykvk0z+8+V/WJa3DPUq498eT1x2HOmkd8bLMsa0wspHEaKAAuqpJ4kkgMAIoy1sCUsncmfNIQjWnovf7Nmzh21521hwwQKG3jSUi764iLjsuHZdXMgYc/AsKUSIEkroS1/qqAtsiyeePewJSydyR65e1l6tDUH1er288MILZGVlMVAH8vbqt8m5MYctni098pKixoSLlY8iRCaZ1FHHGtYATkLYy17iiAvLN+T2Xr2sI1qalNfnyz6cmncqqso//vEPTjjhBADGMe6Az2WMOTCWFCKEf3jmkRzJFrawhz3EEceDPEgOOa3W993oeG7r6mXNhYoBmq7TlE02r/Ea4CSYnXt2svi+xez9615mPjKT66+/3uYbGBNmro4+EpHxwJNADPC8qj7a7PGfATcBDcAO4AZV/bq1Y3bn0UctfbgHd9AGj+efwQyAFh/rrE7hto4bat9NbEJRhjGsyfOv4AoWeRfxyUufsP6e9Uz40QRmPzybfv36HdR7Z4xpXXtHH7mWFEQkBlgLjANKca7ZfKWqFgftczbwb1XdJyK3AWep6hWtHbc7J4WW5JHHTnZSRx1rWUsllcQTz3EcxyAG7feNvowy0klnFrMO6rztbZ1sYhOHcAjf5/uB537ABwCcy7lN4tIvlU23bwJg1qxZgVKRMcZdkTAkNQdYp6obfAG9ClwMBJKCqn4YtP9C4GoX44laJZQQTzyLWUw88XjwBEYlxRBDBhkcx3EMZCDQeu2/I6Wm4KGv/uc9wAMkkkgppQxjGBlksIQllFFGMsmBGGqpbXKsmt01rLhvBev/tp5n/r9nuO6666xUZEwEcvN/ZQbwTdD9Ut+2ltwIvBPqARG5WUSKRKRox44dnRhiKwoLIS8PJkxwfheG78pnmWSykpXEE48XL9vYRi21eHx/fVvZyud8zna2Ay3X/lsbZtraTOrmz1vKUjaykTrq8OChH/0QhLWsDTwnwfejXuWrP3zFX0b+hTqp4+riq7nhhhssIRgTodxsKUiIbSFrVSJyNZANnBnqcVWdDcwGp3zUWQG2qLAQ8vMhNRUyMmDnTuf+jBmQ07XDIgspZBvb2MQmetELRWmgAUFIIIFGGhGEvewNLKldTjnTmb7fsWYyk3Wso446kknmCI4glVRmMpMqqlqcVNZ8ies66kgkkbWsZSADOYIj+IzPKKGECipIIIHe9IbF8Ne8vxIrsZw27zQ8x3u4hVu69P0zxnSMm1/XSoEhQfcHA1ua7yQi5wH3Ahepam3zx8OioMBJCGlp4PE4v1NTne1dyP8NPYYYDuMwvHjZxz4UJY64wIJ4ddRRTTXrWU8RRVRQQQEF+33bX8ACvHhJIokaaiiiiFpqWcSiwIf+TnaykpUsYxk3cAOFFO43vyCZZAAqqQxsa6SROOKc27sb2X7bdjZduIljbzmW4z87nhHHj7C5BsZEATdbCouAESKSCWwGJgJXBe8gImOA54DxqrrdxVg6pqTEaSEES0lxtrskVK0/+Bv6GMZQhNPBXkNNk0lu4muU+VsMAxjQ5Ns+wA3cwF72Ukst/elPL3oBsJKVzssjhe1sp4gi4oknlVR2sYt88ulDHyqoCLQUjuAIvuAL+tAHL15WsII44jjdezp7XtjDovsWkXFpBv9v9f/jD33/4Np7ZozpfK4lBVVtEJFJwHycIakvqOoqEZkBFKnqXOB/gD7AX0QEYJOqXuRWTO2WmemUjNK+G9FDRYWz3QUtrQlUQQVHczQAAxlINtksZSlf83WTRACgKDHEkEgi61jHaZwGOCWjzWxmAxsCLY1aajmUQ/HgoYIKRjGKD/mQbWwL9BHUUUc/+pFKKo00Uk454CSPeOLJJJMMMtjMZuqp58jFR/JF3heIR7jgnQvoN6Yfm9nsyvtljHGPq5PXVHUeMK/Ztvyg2+e5ef4Dlpvr9CGA00KoqIDycpi+f52+M7R0WcpSSpt8Qx/IQHLIoZpqetObr/kaL97AcRpoIJ74QFknhRTe4i0A4oijwfdTRx2llHIIhzCKUVRTTRVV1FGHFy9VVJFAAmMZSwopbGYzM5jRpCUznenkkMPu3bs5494zKHyjkBN/dSJH5B6BeIQyyqLyGg/G9HQ2BCSUnBynUzk9HTZvdn672Mnc0ppAKaSEvCznaZzGAAYgCDHEEOvL7f55DGWUsZ3tLGMZ230/e9lLNdWB4zfSyE52Uk45wxjGkRyJ1/cTQwyCsIENgSSQQw655JJJJiWU8JL3Je59/l6ysrLIisnilNWnMOj6QahHA3HaAnbGRB+7nkIE8E9OCzUBzd+3sIQlVFAR6DfYyMbAvtpsUFccccQTTw01NNJIDDE00thkH0FIJ51KKvkBP2Ad6wIf5oLgxRsYajqHOcB3M6dri2r5JO8TvDFenpr1FNeMuSaqr/FgTE8QCZPXTDu1dFlKf4lmFat4i7cop5x97AOcRODB06R8JAiKEkss1VQHkkNwC8GvF73oRz+qqGIVq6ijjhRSiCWWPexBEJJJJo00csghjzx67erFintXsPHNjeQ8msPAawey0LOQa7imy67xYIxxl5WPIkCoy1JewRUUUMApnMJkJlNJJXvZ26RVEJwQYoklhphAIvDiDbQO/J3Rweqoo4oqBjGIcsqJJ55aammkMTBstY46Zyis18vHv/+Y+SPn44nzcPnqyznyuiNJ9aSG5VoPxhj3WEshQjRfUsJfqimjjAYaqKUWL17iiNuvFAROJ7Off7azotRRRwIJ1FMfeF4MMSjKTnYymtGcwAnsZjef8RkNNJBAArHEUkUVXxV9xTF5x7Ardhenzz+dzOO+6zxubdVUY0x0spZCS8K4zEXwaKQqqgIjh/yzmUMlhWA11ARaCv7nBGukEQ8eUkhhG9u4kAuJJ5500gNXfivbVYbnVg/rf7ieQbcP4o1P3sBznGe/Tm/rTDame7GkEIp/mYudO5suc9FFiWEJS1jBCt7hHcooo4aawGPN+xBaE0MMXrxNEkkMMYE1k5JIYjCDKaKIVFKdmdPew0j9fSo1I2uoja/lstWXkZCbwEmek/YrcdkMZWO6HysfhRK8zAV897ugwPW1jwoppJRSFKUPfdjBjv2+6fs1H3XUnBcv8cQ3mf3sn+Q2gAHEEMMYxgSusha7KJaNeRvxxHsYMn8I3uO81FIbKBFZZ7Ix3Z+1FEIpKXEmrQVzeZkLvwIKOIIjAKdvIPgDHQisL9Qe/j4FcFoVguDBQyONTco/h+w8hH/d/C+2XLSFxEmJHPLJIXiO8wQW17MSkTE9hyWFUDIznVnMwVxc5iLYEpawhS3UUcdOdjYpF8UR12SyWkcpGjheHXWc0XgGS55bwptHv0lD7wYuWH0BZ1x7Bh7xUEklx3GclYiM6WGsfBRKFy9z4RdcOupNb3azu8nj9b6fA+Gf1xCY21AIq/NWsy1hGx++9yG1x9YGJp+dz/k2+cyYHsqSQig5OXDFFfD447BtGwwa5CQEl/oT/LOB3+ZtGmmkkkr2sa/NUUYdpSjenV6893jhLbj7sbu55ppr8C1GaEnAGGPlo5AKC+G11+Doo+GSS5zfr73myuij4Kua+Us7lVS22YncYY2gzyk6UunVuxcnrD6Ba6+9NpAQjDEGrKUQWkEBNDTAypVQWQnJyXDooa6MPgqek5BCCqWUBtYe6jSFwO1Ab8j8ZyaZx2ZyJEd23vGNMd2GJYVQliyBFStg3z7wep2rr23a5Nzv7FP5LnpfRRWxxO63lMVB2Qn8N/APkMeE2KtjyZZsG1FkjGmRlY9CWbcOqqqchADO76oqZ3sn8ncsV1JJEkmBtYcOWiPwLDASJEmIXR2LXCPES7xNOjPGtMrVpCAi40VkjYisE5G7Qzx+hoh8KSINInKpm7F0yI4dHdt+gJrPSdjL3oM/6L+BE4E/Ae9D7P/GkpyazPEcTy65zGKWJQRjTItcKx+JSAwwCxgHlAKLRGSuqhYH7bYJuA74uVtxRDL/tQeSSWYZywJXTDsgQaUiHgOuht7Sm5/wk8BS3FYyMsa0xc2WQg6wTlU3qGod8CpwcfAOqrpRVZdDZ/aqRg//VcyWsYytbD2wvoSgUhFJwGrgGhARDuEQW6fIGNMhbnY0ZwDfBN0vxSlsdJiI3AzcDDB06NCDjywCFFJIMcV8xmeBiWUdTgr/xhlVlAi8Dxzz3UMDGMCrvGqJwBjTIW62FEINgD+gYTWqOltVs1U1e8CAAQcZVvj55yZ8wzekkx5Yo6itVU8DdgA3Af8FTAM+JpAQPHhIIolHedQSgjGmw9xMCqXAkKD7g4EtLp4vavjnJtRRRyqpJJOMBw8xvp8WNQLPAEcDyTiloqtpkn7TSee3/Jbrud7Nl2CM6abcLB8tAkaISCawGZgIXOXi+aKGf6nqZJKpoYY00qikstUrq7EQyMPpN/gAGN304RhiOJ3TeYzHrIVgjDlgrrUUVLUBmATMx/lO+7qqrhKRGSJyEYCIjBWRUuAy4DkRWeVWPJEkk0wqqOAIjqCKKrazHa/vp5bapjvvAG4EfgT8DFhAICHEEksccZzFWXzO53zIh5YQjDEHxdUZzao6D5jXbFt+0O1FOGWlHiWXXPLJp4EG6qijmur9O5kbgeeAB3BKRF8BKc51EWKJJZtsxjDGVjM1xnQqW+aiCxVSyExmsohFVPp+Ql5VzV8q6sN+paL+9OdMzmQOc7omaGNMj2JJoYsUUsiN3Mh61lNHXeh+gx3A3cC7wEycHhhfJ3IvejGKUfSnP3dyZ5fFbYzpWSwpdJGZzGQ962n0/TTRvFS0Gkhxhpf2pjcDGchRHEUmmVYuMsa4ypJCFyikkPnMp4aa/fsOvsApFSUD/wJGOZs9eEgggdGM5kmetERgjOkStkqqywopZCpT9+9M3g7cAFyKs/LTR8AoJxmAM8T0RE60hGCM6VLWUnDZTGZSQgkJJFBDDd5Gr7NW0YPANQRKRQADGUgvejGWsdzJnZYMjDFdzpKCyxaxiEQSSSCBrV9sxZvndZJAs1LRsRzLl3wZzlCNMcaSQldo3N5I+d3leOd76fU/vai5sgbEmXyWSCKDGcyzPBvuMI0xxpKCmxobG+n3bD+WPbCM5GuTyVydSWNKI+WUk0qqjSgyxkQcSwou+fzzz8nLy8OT6mHUh6PYN2ofNdSQQAIjGMFTPGWJwBgTcSwpdLLt27dz11138d577/HrX/+aiRMnskgWUUBB4Epr1jIwxkQqSwqdpKGhgWeffZYHH3yQ3NxcVq9eTUqKM6wox/djjDGRzpJCJ/jss8/Iy8ujb9++fPTRRxx99NHhDskYYw6IJYWDsG3bNu666y7ef/99fv3rX3PFFVcg0s6rpxljTASyGc0HoKGhgd/+9reMGjWKAQMGsHr1aiZOnGgJwRgT9ayl0EH+UlG/fv1YsGABI0eODHdIxhjTaVxtKYjIeBFZIyLrROTuEI8niMhrvsf/LSLD3YznYGwDrgOuuOIK7r77bj744ANLCMaYbse1pCAiMcAs4AJgJHCliDT/FL0R2KOqhwO/AR5zK54Oue++wM0G4Lc4K1IMOOkkKxUZY7o1N1sKOcA6Vd2gqnXAq8DFzfa5GCjw3Z4DnCuR8Gn70ENw3318mpzMCcAbsbEsuOUW/ueLL0hOTg53dMYY4xo3k0IG8E3Q/VLftpD7qGoDUA70b34gEblZRIpEpGjHjh0uhfudbdu2kbtpExNTUrjn1Vf5oK6Okc/a2kTGmO7PzaQQ6hu/HsA+qOpsVc1W1ewBAwZ0SnChNDQ08NRTTzFq1CgGDRrE6tWrbZipMaZHcXP0USkwJOj+YGBLC/uUikgskArsdjGmFn366afk5eWRnp5uo4qMMT2Wmy2FRcAIEckUkXhgIjC32T5zgVzf7UuBf6nqfi0FN23bto3c3FyuvPJK7rnnHt5//31LCMaYHsu1pODrI5gEzMe5vtjrqrpKRGaIyEW+3f4A9BeRdcDPgP2GrbqleamouLjYSkXGmB7P1clrqjoPmNdsW37Q7RrgMjdjCOWTTz5h0qRJpKen8/HHH5OVldXVIRhjTETqUTOav/32W37xi1/w0Ucf8fjjj3PZZZdZy8AYY4L0iLWPGhoa+N///V9GjRrFYYcdxurVq7n88sstIRhjTDPdvqXw8ccfM2nSJAYOHMgnn3xipSJjjGlFt00KW7du5c477+Sjjz7iiSee4NJLL7WWgTHGtKHblY/8paLRo0eTkZHB6tWrre/AGGPaqVu1FD7++GPy8vI45JBD+PTTTznqqKPCHZIxxkSVbpEUtm7dyi9+8QsWLFhgpSJjjDkIUV0+qq+v5ze/+Q2jR49m8ODBVioyxpiDFLUtBSsVGWNM54u6pFBfX8/VV1/Nxx9/zBNPPMGPf/xjaxkYY0wnibqksGrVKiZMmEBxcTF9+vQJdzjGGNOtSBcvSnrQRGQH8HUXnjId2NmF5ztQ0RBnNMQIFmdni4Y4oyFGOLg4h6lqmxekibqk0NVEpEhVs8MdR1uiIc5oiBEszs4WDXFGQ4zQNXFG9egjY4wxncuSgjHGmABLCm2bHe4A2ika4oyGGMHi7GzREGc0xAhdEKf1KRhjjAmwloIxxpgASwrGGGMCLCn4iMh4EVkjIutE5O4QjyeIyGu+x/8tIsMjMMYzRORLEWkQkUu7Or6gONqK82ciUiwiy0XkAxEZFqFx3ioiK0RkqYh8KiIjIzHOoP0uFREVkS4fWtmO9/I6Ednhey+XishNXR1je+L07XO579/nKhF5patj9MXQ1vv5m6D3cq2IlHXayVW1x/8BYoD1wPeAeGAZMLLZPrcDz/puTwRei8AYhwPHAC8Dl0bwe3k2kOi7fVtXv5cdiDMl6PZFwLuRGKdvv2TgY2AhkB1pMQLXAU+H499kB+McASwB+vruD4zEOJvtPxl4obPOby0FRw6wTlU3qGod8CpwcbN9LgYKfLfnAOdK1y661GaMqrpRVZcD3i6Mq7n2xPmhqu7z3V0IDO7iGKF9cVYE3U0CwjEqoz3/NgEeAmYCNV0ZnE97Ywy39sT5U2CWqu4BUNXtXRwjdPz9vBL4c2ed3JKCIwP4Juh+qW9byH1UtQEoB/p3SXTNzu8TKsZI0NE4bwTecTWi0NoVp4jkich6nA/cKV0UW7A24xSRMcAQVX27KwML0t6/8x/7SoZzRGRI14TWRHviPAI4QkQ+E5GFIjK+y6L7Trv/D/lKr5nAvzrr5JYUHKG+8Tf/VtiefdwU7vO3V7vjFJGrgWzgf1yNKLR2xamqs1T1+8BdwH2uR7W/VuMUEQ/wG2B6l0W0v/a8l28Bw1X1GOB9vmt1d6X2xBmLU0I6C+cb+PMikuZyXM115P/6RGCOqjZ21sktKThKgeBvLoOBLS3tIyKxQCqwu0uia3Z+n1AxRoJ2xSki5wH3Ahepam0XxRaso+/nq8AlrkYUWltxJgOjgI9EZCNwEjC3izub23wvVXVX0N/z74ETuii2YO39f/53Va1X1RJgDU6S6Eod+bc5kU4sHQHW0ezrqIkFNuA0w/wdO0c32yePph3Nr0dajEH7vkT4Oprb816OwelIGxHhf+cjgm7/ECiKxDib7f8RXd/R3J738tCg2/8FLIzE9xIYDxT4bqfjlHH6R1qcvv2OBDbim4Tcaefv6r+YSP0DTADW+j6s7vVtm4HzTRagF/AXYB1QCHwvAmMci/MtYy+wC1gVoe/l+8A2YKnvz9wIjfNJYJUvxg9b+zAOZ5zN9u3ypNDO9/JXvvdyme+9PCoS30uc0s0TQDGwApgYiXH67j8APNrZ57ZlLowxxgRYn4IxxpgASwrGGGMCLCkYY4wJsKRgjDEmwJKCMcaYAEsKxrRCRBp9K1Eu861Ae4pv+2EiMqeF5wwXkauC7l8nIk93VczGHAxLCsa0rlpVj1PVY4H/xhlvj6puUdX9lif3zXYfDlzV/DFjokFsuAMwJoqkAHvAaQ0Ab6vqKBG5DrgQZ4JjEpAIZInIUpw1fvYAh4nIu8D3gTdU9c4uj96YdrCkYEzrevs+3HsBhwLntLDfycAxqrpbRM4Cfq6q/w+c8hFwHM7yHrXAGhH5rap+08KxjAkbKx8Z0zp/+egonHVxXm7hOhr/VNXWFkj8QFXLVbUGZwmFsFxtzpi2WFIwpp1U9QucRdIGhHh4bxtPD14JthFrpZsIZUnBmHYSkaNwLpW4q41dK3GWtDYm6ti3FWNa5+9TAGcFzVxVbWzjSqzLgQYRWfb/t3PHNADAMBDEyp9sKXSodAiSzUaQ7aQMf/6M+d09EeZYSQUg3kcARBQAiCgAEFEAIKIAQEQBgIgCAHmqbZ8vybhIjQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "gd.plot_persistence_diagram(BarCodes_Rips0);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this representation, 0-dimensional features are the red points (holes of dimension 0, namely connect components). The last connected component dies at infinity in the filtration (red point at the top). 1-dim features are represented in green. \n",
    "\n",
    "The most persistent topological features are those points that are far from the diagonal. Further in the tutorial we give statistical methods to identify topological features that are significant.\n",
    "\n",
    "Note that this representation does not say which points are \"at the origin\" of a given feature. Moreover, a given topological feature (namely a homology class that corresponds to a hole of dimension $d$) is by definition a class of cycles defined on the point cloud and thus it can be represented by several cycles ."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Bottleneck distance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To exploit the topological information and topological features inferred from persistent homology, one needs to be able to compare persistence diagrams.\n",
    "\n",
    "We see a persistence diagram as the union of its points and of the diagonal, where the point of the diagonal are counted with infinite multiplicity. \n",
    "\n",
    "A matching  between two diagrams $\\mathrm{dgm}_1$ and $\\mathrm{dgm}_2$ is a subset $m \\subseteq \\mathrm{dgm}_1 \\times \\mathrm{dgm}_2$ such that every points in $\\mathrm{dgm}_1 \\setminus \\Delta$ and $\\mathrm{dgm}_2 \\setminus \\Delta$ appears exactly once in $m$. \n",
    "\n",
    "<img src=\"Images/MatchingDiag.png\"  style=\"width: 400px;\"/>\n",
    "\n",
    "The **Bottleneck distance** between $\\mathrm{dgm}_1$ and $\\mathrm{dgm}_2$ is then defined by\n",
    "\n",
    "$$\n",
    "\\mathrm{d}_b (\\mathrm{dgm}1, \\mathrm{dgm}_2) =\\inf_{\\scriptsize{\\mbox{matching}} \\ m} \\max_{(p,q) \\in m} \\| p-q \\|_\\infty.\n",
    "$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let us compute the Rips complex filtration for another configuration of protein"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "mat_dist1 = dist_list[1]\n",
    "\n",
    "skeleton_protein1 = gd.RipsComplex(distance_matrix=mat_dist1.values,\n",
    "                                  max_edge_length=0.8) \n",
    "\n",
    "Rips_simplex_tree_protein1 = skeleton_protein1.create_simplex_tree(max_dimension=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "and the barcode for this filtration:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "BarCodes_Rips1 = Rips_simplex_tree_protein1.persistence()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The bottleneck distance between the two persistence diagrams can be computed using the <code>bottleneck_distance()</code> function. The bottleneck distance is computed per dimension (for dimension 1 in the example below). We can give in argument of the function the persistence_intervals for a given dimension, which can be computed using the `persistence_intervals_in_dimension()` function. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.05052142999999998"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "I0 = Rips_simplex_tree_protein0.persistence_intervals_in_dimension(1)\n",
    "I1 = Rips_simplex_tree_protein1.persistence_intervals_in_dimension(1)\n",
    "\n",
    "gd.bottleneck_distance(I0,I1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "By default, the function uses an expensive algorithm to compute the exact bottleneck distance. It is also possible to compute an approximated bottleneck distance (additive approximation error given in argument), which is usually a lot faster to compute. See the [documention](http://gudhi.gforge.inria.fr/python/latest/bottleneck_distance_user.html#gudhi.bottleneck_distance) for more details."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.05061900451758505"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gd.bottleneck_distance(I0,I1,0.01)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### MDS on the bottlneck distances"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We simple approach for comparing the MPB configurations is using the bottleneck distances between configurations.\n",
    "\n",
    "We first compute the Rips complex filtration for each MPB:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "1\n",
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n",
      "10\n",
      "11\n",
      "12\n",
      "13\n"
     ]
    }
   ],
   "source": [
    "persistence_list0 = []\n",
    "persistence_list1 = []\n",
    "\n",
    "i=0\n",
    "for d in dist_list:\n",
    "    print(i)\n",
    "    rips_complex = gd.RipsComplex(distance_matrix=d.values,max_edge_length=0.8)\n",
    "    simplex_tree = rips_complex.create_simplex_tree(max_dimension=2)\n",
    "    diag = simplex_tree.persistence()\n",
    "    persistence_list0.append(simplex_tree.persistence_intervals_in_dimension(0))\n",
    "    persistence_list1.append(simplex_tree.persistence_intervals_in_dimension(1))\n",
    "    i = i +1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we compute the matrix of bottleneck distances for dimensions 0 and 1:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "l = len(files_list)\n",
    "B0= np.zeros((l,l))\n",
    "B1 =np.zeros((l,l))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(l):\n",
    "    for j in range(i):\n",
    "        B0[i,j] = gd.bottleneck_distance(persistence_list0[i], persistence_list0[j])\n",
    "        B1[i,j] = gd.bottleneck_distance(persistence_list1[i], persistence_list1[j])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.        , 0.27012237, 0.20270489, 0.23304443, 0.10767524,\n",
       "        0.12799122, 0.30432059, 0.20270489, 0.20270489, 0.11054764,\n",
       "        0.2939635 , 0.132072  , 0.08415318, 0.15367433],\n",
       "       [0.27012237, 0.        , 0.08062446, 0.05610207, 0.25558757,\n",
       "        0.26251234, 0.06350129, 0.12608031, 0.11633436, 0.29995159,\n",
       "        0.0742677 , 0.25168449, 0.22630345, 0.23520893],\n",
       "       [0.20270489, 0.08062446, 0.        , 0.04354652, 0.19339696,\n",
       "        0.23391601, 0.11482268, 0.04545585, 0.03982919, 0.21932713,\n",
       "        0.10446559, 0.17106003, 0.20410059, 0.18580079],\n",
       "       [0.23304443, 0.05610207, 0.04354652, 0.        , 0.21850963,\n",
       "        0.23027366, 0.07127616, 0.08900237, 0.07925642, 0.26287365,\n",
       "        0.06345506, 0.21460655, 0.19464639, 0.19813099],\n",
       "       [0.10767524, 0.25558757, 0.19339696, 0.21850963, 0.        ,\n",
       "        0.07830397, 0.28978579, 0.23364638, 0.20730722, 0.18338388,\n",
       "        0.2794287 , 0.09956064, 0.11198855, 0.11894431],\n",
       "       [0.12799122, 0.26251234, 0.23391601, 0.23027366, 0.07830397,\n",
       "        0.        , 0.29671056, 0.27416543, 0.24782627, 0.22390293,\n",
       "        0.29372872, 0.14007969, 0.13932216, 0.15946336],\n",
       "       [0.30432059, 0.06350129, 0.11482268, 0.07127616, 0.28978579,\n",
       "        0.29671056, 0.        , 0.16027853, 0.15053258, 0.33414981,\n",
       "        0.02031487, 0.28588271, 0.26050167, 0.26940715],\n",
       "       [0.20270489, 0.12608031, 0.04545585, 0.08900237, 0.23364638,\n",
       "        0.27416543, 0.16027853, 0.        , 0.03422325, 0.18236719,\n",
       "        0.14992144, 0.15171769, 0.20410059, 0.18686864],\n",
       "       [0.20270489, 0.11633436, 0.03982919, 0.07925642, 0.20730722,\n",
       "        0.24782627, 0.15053258, 0.03422325, 0.        , 0.18361723,\n",
       "        0.14017549, 0.13535013, 0.20410059, 0.17705646],\n",
       "       [0.11054764, 0.29995159, 0.21932713, 0.26287365, 0.18338388,\n",
       "        0.22390293, 0.33414981, 0.18236719, 0.18361723, 0.        ,\n",
       "        0.32379272, 0.08382324, 0.08458077, 0.09512993],\n",
       "       [0.2939635 , 0.0742677 , 0.10446559, 0.06345506, 0.2794287 ,\n",
       "        0.29372872, 0.02031487, 0.14992144, 0.14017549, 0.32379272,\n",
       "        0.        , 0.27552562, 0.25014458, 0.25905006],\n",
       "       [0.132072  , 0.25168449, 0.17106003, 0.21460655, 0.09956064,\n",
       "        0.14007969, 0.28588271, 0.15171769, 0.13535013, 0.08382324,\n",
       "        0.27552562, 0.        , 0.0868246 , 0.10404144],\n",
       "       [0.08415318, 0.22630345, 0.20410059, 0.19464639, 0.11198855,\n",
       "        0.13932216, 0.26050167, 0.20410059, 0.20410059, 0.08458077,\n",
       "        0.25014458, 0.0868246 , 0.        , 0.16471239],\n",
       "       [0.15367433, 0.23520893, 0.18580079, 0.19813099, 0.11894431,\n",
       "        0.15946336, 0.26940715, 0.18686864, 0.17705646, 0.09512993,\n",
       "        0.25905006, 0.10404144, 0.16471239, 0.        ]])"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "B0 = B0 + B0.transpose()\n",
    "B1 = B1 + B1.transpose()\n",
    "B0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We apply a Multidimensional Scaling method (from the [<code>scikit-learn</code>](http://scikit-learn.org/stable/) library)  to visualize the bottleneck distances:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "mds = manifold.MDS(n_components=2, max_iter=3000, eps=1e-9,\n",
    "                   dissimilarity=\"precomputed\", n_jobs=1)\n",
    "pos = mds.fit(B0).embedding_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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": [
    "plt.scatter(pos[0:7,0], pos[0:7, 1], color='red', label=\"closed\")\n",
    "plt.scatter(pos[7:l,0], pos[7:l, 1], color='blue', label=\"red\")\n",
    "plt.legend( loc=3, borderaxespad=1);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "mds = manifold.MDS(n_components=2, max_iter=3000, eps=1e-9,\n",
    "                   dissimilarity=\"precomputed\", n_jobs=1)\n",
    "pos = mds.fit(B1).embedding_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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": [
    "plt.scatter(pos[0:7,0], pos[0:7, 1], color='red', label=\"closed\")\n",
    "plt.scatter(pos[7:l,0], pos[7:l, 1], color='blue', label=\"red\")\n",
    "plt.legend( loc=2, borderaxespad=1);"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "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": "3.6.8"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}