main_SDs.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "using Revise, Plots, ArndtLabTheme, LightGraphs\n",
    "pyplot()\n",
    "theme(:arndtlab)\n",
    "\n",
    "includet(\"/Users/abdullae/julia_wd/object_oriented_SDs/modules_SDs.jl\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6791, 4)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = SDmoduleInit.Initialize_SD_object(\"../graphs_sticks/GRCh38GenomicSuperDup_sort.tab\");  # \"GRCh38_sort_notel.tab\"\n",
    "b = SDmoduleInit.Initialize_pyr_object(a.int_DF);\n",
    "c, ind_arr, ind_arr2 = SDmoduleInit.Initialize_SD_object(\"../graphs_sticks/GRCh38GenomicSuperDup_sort.tab\", comp3=\"filter\");\n",
    "size(b.int_DF)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Double edges = 3703 = 0.2308315671362673 of all edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Self loops = 269 = 0.040414663461538464 of all nodes (or 0.05940302896632848 without filtering singleton nodes)\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Tandem edges = 1383 = 0.08621119561151976 of all edges and 0.30815508021390375 of all intra\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Intra = 4488 and Inter = 11554 edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{6656, 16042} undirected Int64 metagraph with Float64 weights defined by :weight (default weight 1.0)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nex = deepcopy(b)\n",
    "graph_obj = NetworkAnalysis.Initialize_SD_network(b, a.int_DF);\n",
    "graph_obj.mg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Double edges = 2832 = 0.19540467812047196 of all edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Self loops = 254 = 0.038161057692307696 of all nodes (or 0.05940302896632848 without filtering singleton nodes)\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Tandem edges = 1305 = 0.09004346926102257 of all edges and 0.313475858755705 of all intra\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Intra = 4163 and Inter = 10330 edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{6656, 14493} undirected Int64 metagraph with Float64 weights defined by :weight (default weight 1.0)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "graph_obj2 = NetworkAnalysis.Initialize_SD_network(nex, c.int_DF)\n",
    "graph_obj2.mg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "assign_weights_edges! (generic function with 1 method)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "using GraphPlot, MetaGraphs\n",
    "using Distributions\n",
    "\n",
    "\n",
    "function assign_weights_edges!(graph; comp=\"normal\")\n",
    "    cc = connected_components(graph)\n",
    "    cl = map(length, cc)\n",
    "    \n",
    "    for i in vertices(graph)\n",
    "        neis = neighbors(graph, i)\n",
    "        for n in neis\n",
    "            neis2 = neighbors(graph, n)\n",
    "            mat_inds = findall(x -> x in neis, neis2)   # findall(x -> (x in neis) && (x != i), neis2)\n",
    "            #if length(mat_inds) > 0\n",
    "            if weights(graph)[i, n] == 1.0     # depends on specified weight\n",
    "                if comp == \"normal\"\n",
    "                    wei = 1.001 - (length(mat_inds) + 1)/length(neis)\n",
    "                elseif comp == \"normal_noise\"\n",
    "                    wei = 1.001 - (length(mat_inds) + 1)/length(neis) + rand(Uniform(-0.00005, 0.00005))\n",
    "                elseif comp == \"shuffle\"\n",
    "                    wei = 1.001 + rand()  \n",
    "                end\n",
    "                set_prop!(graph, i, n, :weight, wei)\n",
    "            else\n",
    "                if comp == \"normal\"\n",
    "                    wei = minimum([1.001 - (length(mat_inds) + 1)/length(neis), weights(graph)[i, n]])\n",
    "                elseif comp == \"normal_noise\"\n",
    "                    wei = minimum([1.001 - (length(mat_inds) + 1)/length(neis), weights(graph)[i, n]]) + rand(Uniform(-0.00005, 0.00005))\n",
    "                elseif comp == \"shuffle\"\n",
    "                    wei = minimum([1.001 + rand(), weights(graph)[i, n]])\n",
    "                end\n",
    "                set_prop!(graph, i, n, :weight, wei)\n",
    "            end\n",
    "        end\n",
    "    end\n",
    "end\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "brussels (generic function with 2 methods)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "function brussels(graph, N=100)\n",
    "    assign_weights_edges!(graph)\n",
    "    n_deg, ids_hui, edges = graph_nosec_from_graph(graph, comp=\"mst_krus\")\n",
    "    n_degs = deepcopy(n_deg)\n",
    "    for i in 1:(N-1)\n",
    "        assign_weights_edges!(graph, comp=\"normal_noise\")\n",
    "        n_deg, ids_hui, edg = graph_nosec_from_graph(graph, comp=\"mst_krus\")\n",
    "        n_degs = n_degs + n_deg\n",
    "        append!(edges, edg)\n",
    "    end\n",
    "    n_degs = n_degs./N\n",
    "    return n_degs, edges\n",
    "end\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "new_york (generic function with 1 method)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "function new_york(graph, N)\n",
    "    edg = brussels(graph, N)[2]\n",
    "    for e in edges(graph)\n",
    "        wei = 1.0 - sum(map(x ->(x.src == e.src) & (x.dst == e.dst) ? 1 : 0, edg))/N\n",
    "        set_prop!(graph, e.src, e.dst, :weight, wei)\n",
    "    end\n",
    "    n_deg, ids_hui, edg = graph_nosec_from_graph(graph, comp=\"mst_krus\")\n",
    "    return n_deg, edg\n",
    "end\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "stambul (generic function with 1 method)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# for deletion\n",
    "\n",
    "function tokyo(graph)\n",
    "    arr_out = Float64[]\n",
    "    vers = sort(collect(vertices(graph)))\n",
    "    for i in vers\n",
    "        neis = neighbors(graph, i)\n",
    "        gub = induced_subgraph(graph, [i, neis...])[1]\n",
    "        all_sum = sum(1 .- map(x -> get_prop(gub, x, :weight), edges(gub)))\n",
    "        gub = induced_subgraph(graph, neis)[1]\n",
    "        if length(edges(gub)) > 0\n",
    "            oth_sum = sum(1 .- map(x -> get_prop(gub, x, :weight), edges(gub)))\n",
    "            res = length(neis)*(all_sum - oth_sum)/all_sum\n",
    "        else\n",
    "            res = float(length(neis))\n",
    "        end\n",
    "        append!(arr_out, res)\n",
    "    end\n",
    "    return arr_out\n",
    "end\n",
    "    \n",
    "function stambul(graph)\n",
    "    gr = SimpleGraph(nv(graph), 0)\n",
    "    vers = sort(collect(vertices(graph)))\n",
    "    for i in vers\n",
    "        neis = neighbors(graph, i)\n",
    "        weis = collect(map(x -> x > i ? get_prop(graph, Edge(i, x), :weight) : get_prop(graph, Edge(x, i), :weight), neis))\n",
    "        node = neis[weis .== minimum(weis)][1]\n",
    "        if !(has_edge(gr, i, node))\n",
    "            add_edge!(gr, i, node)\n",
    "        end\n",
    "    end\n",
    "    arr_out = collect(map(x -> length(neighbors(gr, x)), vers))\n",
    "    return arr_out, gr\n",
    "end"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"done\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "using LightGraphs, Histograms\n",
    "\n",
    "\n",
    "function custom_cmp(x::String)\n",
    "   number_idx = findfirst(isdigit, x)\n",
    "   str, num = SubString(x, 1, number_idx-1), SubString(x, number_idx, length(x))\n",
    "   return str, parse(Int, num)\n",
    "end\n",
    "\n",
    "\n",
    "function graph_nosec_from_graph(graph; comp=\"mst_krus\", comp_vis=false)\n",
    "    if comp == \"mst_krus\"\n",
    "        edg_list = kruskal_mst(graph)\n",
    "        gr1 = induced_subgraph(graph, edg_list)[1]\n",
    "    elseif comp == \"mst_prim\"\n",
    "        ids, n_deg_int = [], []\n",
    "        ccs = connected_components(graph)\n",
    "        for cc in ccs\n",
    "            cc_gr = induced_subgraph(graph, cc)[1]\n",
    "            edg_list = prim_mst(cc_gr)\n",
    "            gr1 = induced_subgraph(cc_gr, edg_list)[1]\n",
    "            emap = vertices(gr1)\n",
    "            append!(ids, map(x -> get_prop(gr1, x, :id), emap))\n",
    "            append!(n_deg_int, map(x -> length(neighbors(gr1, x)), emap))\n",
    "        end\n",
    "        n_deg_int = n_deg_int[sortperm(ids, rev=false, by=custom_cmp)]\n",
    "        ids = sort(ids, by=custom_cmp)\n",
    "        return n_deg_int, ids                   # maybe problem!!! only 2 outputs\n",
    "    elseif comp == \"all\"\n",
    "        gr1 = graph\n",
    "        edg_list = edges(graph)\n",
    "    end\n",
    "    emap = vertices(gr1)\n",
    "    ids = map(x -> get_prop(gr1, x, :id), emap)\n",
    "    n_deg_int = map(x -> length(neighbors(gr1, x)), emap)\n",
    "    n_deg_int = n_deg_int[sortperm(ids, rev=false, by=custom_cmp)]\n",
    "    ids = sort(ids, by=custom_cmp)\n",
    "    if comp_vis\n",
    "        n_deg = convert.(UInt64, n_deg_int);\n",
    "        nd_h = Histograms.Histogram(n_deg, scale=:log);\n",
    "        p = plot(nd_h, yscale=:log10, xscale=:log10, line=false, markershape=:auto)\n",
    "        x = range(1, 30, length=10); y = 1.0*x.^(-3);\n",
    "        p = plot!(x, y, label=\"a = -3\")\n",
    "        gui(p)\n",
    "    end\n",
    "    return n_deg_int, ids, edg_list\n",
    "end\n",
    "\n",
    "assign_weights_edges!(graph_obj.mg);\n",
    "assign_weights_edges!(graph_obj2.mg) #, comp=\"normal_noise\");\n",
    "#assign_weights_edges!(graph_obj2.mg, \"shuffle\");\n",
    "n_deg, ids_hui = graph_nosec_from_graph(graph_obj2.mg, comp=\"mst_krus\")[1:2];\n",
    "display(\"done\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.00010178117048342816"
      ]
     },
     "execution_count": 148,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# to recognize the weights min difference\n",
    "\n",
    "weis = sort(collect(map(x -> get_prop(graph_obj.mg, x, :weight), edges(graph_obj.mg))))\n",
    "difs = sort(weis[2:end] - weis[1:end-1])\n",
    "min_dif = difs[difs .> 0][1]\n",
    "# 0.0001"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"These are top jumpers:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table class=\"data-frame\"><thead><tr><th></th><th>chr</th><th>coor_s</th><th>coor_e</th><th>ids</th></tr><tr><th></th><th>Int64</th><th>Int64</th><th>Int64</th><th>String</th></tr></thead><tbody><p>10 rows × 4 columns</p><tr><th>1</th><td>3</td><td>125690307</td><td>125935167</td><td>id1230</td></tr><tr><th>2</th><td>9</td><td>39588991</td><td>40873597</td><td>id3253</td></tr><tr><th>3</th><td>1</td><td>8893201</td><td>8894618</td><td>id36</td></tr><tr><th>4</th><td>1</td><td>96446864</td><td>96448625</td><td>id241</td></tr><tr><th>5</th><td>15</td><td>22358242</td><td>22768868</td><td>id4943</td></tr><tr><th>6</th><td>7</td><td>57497244</td><td>57823866</td><td>id2596</td></tr><tr><th>7</th><td>1</td><td>101786252</td><td>101787301</td><td>id247</td></tr><tr><th>8</th><td>2</td><td>131791143</td><td>132220279</td><td>id819</td></tr><tr><th>9</th><td>7</td><td>56787891</td><td>57137555</td><td>id2581</td></tr><tr><th>10</th><td>15</td><td>19987818</td><td>20428912</td><td>id4928</td></tr></tbody></table>"
      ],
      "text/latex": [
       "\\begin{tabular}{r|cccc}\n",
       "\t& chr & coor\\_s & coor\\_e & ids\\\\\n",
       "\t\\hline\n",
       "\t& Int64 & Int64 & Int64 & String\\\\\n",
       "\t\\hline\n",
       "\t1 & 3 & 125690307 & 125935167 & id1230 \\\\\n",
       "\t2 & 9 & 39588991 & 40873597 & id3253 \\\\\n",
       "\t3 & 1 & 8893201 & 8894618 & id36 \\\\\n",
       "\t4 & 1 & 96446864 & 96448625 & id241 \\\\\n",
       "\t5 & 15 & 22358242 & 22768868 & id4943 \\\\\n",
       "\t6 & 7 & 57497244 & 57823866 & id2596 \\\\\n",
       "\t7 & 1 & 101786252 & 101787301 & id247 \\\\\n",
       "\t8 & 2 & 131791143 & 132220279 & id819 \\\\\n",
       "\t9 & 7 & 56787891 & 57137555 & id2581 \\\\\n",
       "\t10 & 15 & 19987818 & 20428912 & id4928 \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/plain": [
       "10×4 DataFrames.DataFrame\n",
       "│ Row │ chr   │ coor_s    │ coor_e    │ ids    │\n",
       "│     │ \u001b[90mInt64\u001b[39m │ \u001b[90mInt64\u001b[39m     │ \u001b[90mInt64\u001b[39m     │ \u001b[90mString\u001b[39m │\n",
       "├─────┼───────┼───────────┼───────────┼────────┤\n",
       "│ 1   │ 3     │ 125690307 │ 125935167 │ id1230 │\n",
       "│ 2   │ 9     │ 39588991  │ 40873597  │ id3253 │\n",
       "│ 3   │ 1     │ 8893201   │ 8894618   │ id36   │\n",
       "│ 4   │ 1     │ 96446864  │ 96448625  │ id241  │\n",
       "│ 5   │ 15    │ 22358242  │ 22768868  │ id4943 │\n",
       "│ 6   │ 7     │ 57497244  │ 57823866  │ id2596 │\n",
       "│ 7   │ 1     │ 101786252 │ 101787301 │ id247  │\n",
       "│ 8   │ 2     │ 131791143 │ 132220279 │ id819  │\n",
       "│ 9   │ 7     │ 56787891  │ 57137555  │ id2581 │\n",
       "│ 10  │ 15    │ 19987818  │ 20428912  │ id4928 │"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#n_deg_tokyo = tokyo(graph_obj2.mg)\n",
    "#n_deg_stambul = stambul(graph_obj2.mg)[1]\n",
    "#vers = sort(collect(vertices(graph_obj2.mg)))\n",
    "#ids = map(x -> get_prop(graph_obj2.mg, x, :id), vers)\n",
    "\n",
    "zaz = ids_hui[sortperm(n_deg, rev=true)[1:10]]   # different ids and ndegs also possible\n",
    "display(\"These are top jumpers:\")\n",
    "display(b.int_DF[findfirst.(isequal.(zaz), (b.int_DF[!, :ids],)), :])\n",
    "b.int_DF[!, :jumps] = zeros(Int64, size(b.int_DF)[1])\n",
    "\n",
    "for (i, val) in enumerate(ids_hui)\n",
    "    b.int_DF[b.int_DF[!,:ids] .== val, :jumps] .= n_deg[i] \n",
    "end\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"../temp/df_ws_jumps.csv\""
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#using CSV\n",
    "\n",
    "#CSV.write(\"../temp/df_ws_jumps.csv\", b.int_DF)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#using GraphIO, ParserCombinator\n",
    "\n",
    "#edges(graph_obj.mg)\n",
    "#kruskal_mst(graph_obj.mg)\n",
    "#savegraph(\"./grav\", graph_obj.mg, \"graph\", GraphIO.EdgeList.EdgeListFormat())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Human:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Double edges = 6342 = 0.16994024491545862 of all edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Self loops = 640 = 0.050878448207329674 of all nodes (or 0.03175102169129205 without filtering singleton nodes)\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Tandem edges = 2203 = 0.05903159248640103 of all edges and 0.30776753283039954 of all intra\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Intra = 7158 and Inter = 30162 edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{12579, 37319} undirected Int64 metagraph with Float64 weights defined by :weight (default weight 1.0)"
      ]
     },
     "execution_count": 112,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a_hum = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Human_SDs_denovo.txt\", 2, \"both\"); \n",
    "b_hum = SDmoduleInit.Initialize_pyr_object(a_hum.int_DF);\n",
    "display(\"Human:\")\n",
    "graph_hum = NetworkAnalysis.Initialize_SD_network(b_hum, a_hum.int_DF);\n",
    "graph_hum.mg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [],
   "source": [
    "a_gor = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Gorilla_SDs_denovo.txt\", 2, \"both\");\n",
    "b_gor = SDmoduleInit.Initialize_pyr_object(a_gor.int_DF);\n",
    "a_chi = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Chicken_SDs_denovo.txt\", 2, \"both\");\n",
    "b_chi = SDmoduleInit.Initialize_pyr_object(a_chi.int_DF);\n",
    "a_cel = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Celegans_SDs_denovo.txt\", 2, \"both\");\n",
    "b_cel = SDmoduleInit.Initialize_pyr_object(a_cel.int_DF);\n",
    "#a_cmp = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Chimp_SDs_denovo.txt\", 2, \"both\");\n",
    "#b_cmp = SDmoduleInit.Initialize_pyr_object(a_cmp.int_DF);\n",
    "#a_dro = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Drosophila_SDs_denovo.txt\", 2, \"both\");\n",
    "#b_dro = SDmoduleInit.Initialize_pyr_object(a_dro.int_DF);\n",
    "#a_zfi = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Zebra_finch_SDs_denovo.txt\", 2, \"both\");\n",
    "#b_zfi = SDmoduleInit.Initialize_pyr_object(a_zfi.int_DF);\n",
    "a_zeb = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Zebrafish_SDs_denovo.txt\", 2, \"both\");\n",
    "b_zeb = SDmoduleInit.Initialize_pyr_object(a_zeb.int_DF);\n",
    "\n",
    "#a_yea = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Yeast_SDs_denovo.txt\", 2, \"both\");\n",
    "#b_yea = SDmoduleInit.Initialize_pyr_object(a_yea.int_DF);\n",
    "#a_pla = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Platypus_SDs_denovo.txt\", 2, \"both\");\n",
    "#b_pla = SDmoduleInit.Initialize_pyr_object(a_pla.int_DF);\n",
    "#a_cow = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Cow_SDs_denovo.txt\", 2, \"both\");\n",
    "#b_cow = SDmoduleInit.Initialize_pyr_object(a_cow.int_DF);\n",
    "a_sti = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Stickleback_SDs_denovo.txt\", 2, \"both\");\n",
    "b_sti = SDmoduleInit.Initialize_pyr_object(a_sti.int_DF);\n",
    "\n",
    "a_rat = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Rat_SDs_denovo.txt\", 2, \"both\");\n",
    "b_rat = SDmoduleInit.Initialize_pyr_object(a_rat.int_DF);\n",
    "a_mus = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Mouse_SDs_denovo.txt\", 2, \"both\");\n",
    "b_mus = SDmoduleInit.Initialize_pyr_object(a_mus.int_DF);\n",
    "a_gib = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Gibbon_SDs_denovo.txt\", 2, \"both\");\n",
    "b_gib = SDmoduleInit.Initialize_pyr_object(a_gib.int_DF);\n",
    "a_dog = SDmoduleInit.Initialize_SD_object(\"./oth_genomes/Dog_SDs_denovo.txt\", 2, \"both\");\n",
    "b_dog = SDmoduleInit.Initialize_pyr_object(a_dog.int_DF);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Gorilla:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Double edges = 4175 = 0.09790586966207819 of all edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Self loops = 1733 = 0.05602068854048812 of all nodes (or 0.012998712998712999 without filtering singleton nodes)\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Tandem edges = 10467 = 0.24545646413244845 of all edges and 0.5311041201542521 of all intra\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Intra = 19708 and Inter = 23308 edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{30935, 42643} undirected Int64 metagraph with Float64 weights defined by :weight (default weight 1.0)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Chicken:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Double edges = 3293 = 0.1925505788796632 of all edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Self loops = 438 = 0.13821394761754496 of all nodes (or 0.12190706095353047 without filtering singleton nodes)\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Tandem edges = 1449 = 0.08472693252251198 of all edges and 0.23247232472324722 of all intra\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Intra = 6233 and Inter = 10941 edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{3169, 17102} undirected Int64 metagraph with Float64 weights defined by :weight (default weight 1.0)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Celegans:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Double edges = 230 = 0.10459299681673488 of all edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Self loops = 186 = 0.1183206106870229 of all nodes (or 0.23529411764705882 without filtering singleton nodes)\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Tandem edges = 368 = 0.16734879490677582 of all edges and 0.4471445929526124 of all intra\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Intra = 823 and Inter = 1376 edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{1572, 2199} undirected Int64 metagraph with Float64 weights defined by :weight (default weight 1.0)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Zebrafish:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Double edges = 86885 = 0.14449790367028167 of all edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Self loops = 2531 = 0.07347946000870954 of all nodes (or 0.011679676206996241 without filtering singleton nodes)\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Tandem edges = 4745 = 0.007891380018593389 of all edges and 0.08422977242872866 of all intra\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Intra = 56334 and Inter = 544994 edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{34445, 601289} undirected Int64 metagraph with Float64 weights defined by :weight (default weight 1.0)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Stickleback:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Double edges = 3830 = 0.06510726549484921 of all edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Self loops = 429 = 0.07613132209405502 of all nodes (or 0.0698961937716263 without filtering singleton nodes)\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Tandem edges = 2091 = 0.03554550708870227 of all edges and 0.3869356032568468 of all intra\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Intra = 5404 and Inter = 53439 edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{5635, 58826} undirected Int64 metagraph with Float64 weights defined by :weight (default weight 1.0)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Rat:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Double edges = 44104 = 0.24019431646134912 of all edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Self loops = 2793 = 0.07775828948467385 of all nodes (or 0.011202307009760427 without filtering singleton nodes)\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Tandem edges = 18489 = 0.10069274254158089 of all edges and 0.41790606211292436 of all intra\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Intra = 44242 and Inter = 139569 edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{35919, 183618} undirected Int64 metagraph with Float64 weights defined by :weight (default weight 1.0)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Mouse:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Double edges = 7268 = 0.04374492160462247 of all edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Self loops = 628 = 0.0425301368007585 of all nodes (or 0.027094091610220642 without filtering singleton nodes)\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Tandem edges = 3793 = 0.022829456197899424 of all edges and 0.2703492516037063 of all intra\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Intra = 14030 and Inter = 152115 edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{14766, 166145} undirected Int64 metagraph with Float64 weights defined by :weight (default weight 1.0)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Gibbon:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Double edges = 25515 = 0.05747709026031952 of all edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Self loops = 334 = 0.011369825708061002 of all nodes (or 0.013685173266488262 without filtering singleton nodes)\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Tandem edges = 10320 = 0.023247641445678913 of all edges and 0.3398985574072854 of all intra\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Intra = 30362 and Inter = 413604 edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{29376, 443916} undirected Int64 metagraph with Float64 weights defined by :weight (default weight 1.0)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Dog:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Double edges = 2087 = 0.03349489632149965 of all edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Self loops = 281 = 0.01524026467078859 of all nodes (or 0.021740300274444384 without filtering singleton nodes)\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Tandem edges = 1788 = 0.02869615458689093 of all edges and 0.4079397672826831 of all intra\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Intra = 4383 and Inter = 57948 edges\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{18438, 62308} undirected Int64 metagraph with Float64 weights defined by :weight (default weight 1.0)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "display(\"Gorilla:\")\n",
    "graph_goril = NetworkAnalysis.Initialize_SD_network(b_gor, a_gor.int_DF);\n",
    "display(graph_goril.mg)\n",
    "display(\"Chicken:\")\n",
    "graph_chick = NetworkAnalysis.Initialize_SD_network(b_chi, a_chi.int_DF);\n",
    "display(graph_chick.mg)\n",
    "display(\"Celegans:\")\n",
    "graph_celeg = NetworkAnalysis.Initialize_SD_network(b_cel, a_cel.int_DF);\n",
    "display(graph_celeg.mg)\n",
    "#display(\"Chimp:\")\n",
    "#graph_chimp = NetworkAnalysis.Initialize_SD_network(b_cmp, a_cmp.int_DF);\n",
    "#display(\"Drosophila:\")\n",
    "#graph_drosop = NetworkAnalysis.Initialize_SD_network(b_dro, a_dro.int_DF);\n",
    "#display(\"Zebra finch:\")\n",
    "#graph_zfinch = NetworkAnalysis.Initialize_SD_network(b_zfi, a_zfi.int_DF);\n",
    "#display(graph_zfinch.mg)\n",
    "display(\"Zebrafish:\")\n",
    "graph_zebrafi = NetworkAnalysis.Initialize_SD_network(b_zeb, a_zeb.int_DF);\n",
    "display(graph_zebrafi.mg)\n",
    "\n",
    "#display(\"Yeast:\")\n",
    "#graph_yeast = NetworkAnalysis.Initialize_SD_network(b_yea, a_yea.int_DF);\n",
    "#display(\"Platypus:\")\n",
    "#graph_platy = NetworkAnalysis.Initialize_SD_network(b_pla, a_pla.int_DF);\n",
    "#display(\"Cow:\")\n",
    "#graph_cow = NetworkAnalysis.Initialize_SD_network(b_cow, a_cow.int_DF);\n",
    "display(\"Stickleback:\")\n",
    "graph_stick = NetworkAnalysis.Initialize_SD_network(b_sti, a_sti.int_DF);\n",
    "display(graph_stick.mg)\n",
    "display(\"Rat:\")\n",
    "graph_rat = NetworkAnalysis.Initialize_SD_network(b_rat, a_rat.int_DF);\n",
    "display(graph_rat.mg)\n",
    "display(\"Mouse:\")\n",
    "graph_mouse = NetworkAnalysis.Initialize_SD_network(b_mus, a_mus.int_DF);\n",
    "display(graph_mouse.mg)\n",
    "display(\"Gibbon:\")\n",
    "graph_gibbon = NetworkAnalysis.Initialize_SD_network(b_gib, a_gib.int_DF);\n",
    "display(graph_gibbon.mg)\n",
    "display(\"Dog:\")\n",
    "graph_dog = NetworkAnalysis.Initialize_SD_network(b_dog, a_dog.int_DF);\n",
    "display(graph_dog.mg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Human predicted f = 0.5298410636885584\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "4-element Array{Int64,1}:\n",
       " 1571\n",
       "  122\n",
       "  107\n",
       "  199"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "using LightGraphs, Optim\n",
    "\n",
    "function calc_inds_4_f(graph)\n",
    "    ccs = connected_components(graph)\n",
    "    lens = map(length, ccs)\n",
    "    ind_2_2, ind_3_2, ind_3_3 = 0, 0, 0\n",
    "    ind_2_2 = length(lens[lens .== 2])\n",
    "    \n",
    "    ind_more = length(lens[lens .> 3])\n",
    "    \n",
    "    for comps in ccs[lens .== 3]\n",
    "        neis = map(x -> length(neighbors(graph, x)), comps)\n",
    "        if sum(neis)/2 == 2\n",
    "           ind_3_2 += 1 \n",
    "        elseif sum(neis)/2 == 3\n",
    "           ind_3_3 += 1\n",
    "        end\n",
    "    end\n",
    "    \n",
    "    return [ind_2_2, ind_3_2, ind_3_3, ind_more]\n",
    "end\n",
    "    \n",
    "function predict_f(f, f_fin, ind_more, N)\n",
    "    for i in 1:ind_more\n",
    "        p = (0.6*f)/(0.6*f + 0.4*(1-f))\n",
    "        f = p*(f*N-1)/(N-1) + (1-p)*(f*N)/(N-1)\n",
    "        N -= 1\n",
    "    end\n",
    "    \n",
    "    return abs(f - f_fin)\n",
    "end\n",
    "\n",
    "function calc_f_species(graph, specie::String=\"Human\")\n",
    "    ind_2_2, ind_3_2, ind_3_3, ind_more = calc_inds_4_f(graph)\n",
    "    f_fin = ind_3_3/(ind_3_3 + ind_3_2); N = ind_3_2 + ind_3_3 + ind_more;\n",
    "    out = optimize(x -> predict_f(x, f_fin, ind_more, N), 0.0, 1.0)\n",
    "    display(\"$(specie) predicted f = $(out.minimizer)\")\n",
    "    return ind_2_2, ind_3_2, ind_3_3, ind_more\n",
    "end\n",
    "        \n",
    "ind_2_2, ind_3_2, ind_3_3, ind_more = calc_f_species(graph_obj.mg, \"Human\")\n",
    "display([ind_2_2, ind_3_2, ind_3_3, ind_more])\n",
    "\n",
    "#calc_f_species(graph_dog.mg, \"Dog\")\n",
    "#calc_f_species(graph_mouse.mg, \"Mouse\")\n",
    "#calc_f_species(graph_rhesus.mg, \"Rhesus\")\n",
    "#calc_f_species(graph_chick.mg, \"Chick\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Results of Nonlinear Solver Algorithm\n",
       " * Algorithm: Trust-region with dogleg and autoscaling\n",
       " * Starting Point: [0.0005, 0.5, 2000.0]\n",
       " * Zero: [9.77297e-5, 0.494198, 1876.71]\n",
       " * Inf-norm of residuals: 0.000000\n",
       " * Iterations: 7\n",
       " * Convergence: true\n",
       "   * |x - x'| < 0.0e+00: false\n",
       "   * |f(x)| < 1.0e-08: true\n",
       " * Function Calls (f): 8\n",
       " * Jacobian Calls (df/dx): 8"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "using NLsolve, QuadGK, Calculus\n",
    "\n",
    "function f!(y, incond)\n",
    "    Ns = [1571, 122, 107, 199]\n",
    "    N = sum(Ns)\n",
    "    delta, f, t = incond\n",
    "    pi = 1.0\n",
    "    N_2, N_3_2, N_3_3, N_more = Ns\n",
    "    y[1] = (pi/(2*delta))*(1 - exp(-2*delta*t)) - N_2\n",
    "    y[2] = (pi*(1-f)/(4*delta))*(1 - 2*exp(-2*delta*t) + exp(-4*delta*t)) - N_3_2\n",
    "    y[3] = (pi*f/(6*delta))*(1 - 1.5*exp(-2*delta*t) + 0.5*exp(-6*delta*t)) - N_3_3\n",
    "    qq(x) = pi*(1-f)*(1 - 2*exp(-2*delta*x) + exp(-4*delta*x)) + pi*f*(1 - 1.5*exp(-2*delta*x) + 0.5*exp(-6*delta*x))\n",
    "    #y[4] = quadgk(qq, 0, t)[1] - N_more\n",
    "    #y[4] = t - y[1] + y[2] + y[3] - N_2\n",
    "end\n",
    "\n",
    "nlsolve(f!, [0.0005, 0.5, 2000], ftol=1e-8)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "g! (generic function with 1 method)"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "function g!(G, incond)\n",
    "    delta, f, t = incond\n",
    "    pi = 1.0\n",
    "    Ns = [1571, 122, 107, 199]\n",
    "    weis = 1 ./ Ns\n",
    "    \n",
    "    y1 = (pi/(2*delta))*(1 - exp(-2*delta*t))\n",
    "    y2 = (pi*(1-f)/(4*delta))*(1 - 2*exp(-2*delta*t) + exp(-4*delta*t))\n",
    "    y3 = (pi*f/(6*delta))*(1 - 1.5*exp(-2*delta*t) + 0.5*exp(-6*delta*t))\n",
    "    c1 = (f/12 - 3/4)*pi/delta\n",
    "    y4 = c1 - f*pi*exp(-6*delta*t)/(12*delta) + f*pi*exp(-4*delta*t)/(4*delta) - \n",
    "    f*pi*exp(-2*delta*t)/(4*delta) - pi*exp(-4*delta*t)/(4*delta) + pi*exp(-2*delta*t)/delta + pi*t\n",
    "    ys = [y1, y2, y3, y4]\n",
    "    \n",
    "    K = zeros(4, 3)\n",
    "    \n",
    "    K[1,1] = -(pi*exp(-2*delta*t)*(-2*delta*t + exp(2*delta*t) - 1))/(2*delta^2)\n",
    "    K[1,2] = 0.0\n",
    "    K[1,3] = pi*exp(-2*delta*t)\n",
    "   \n",
    "    K[2,1] = ((f - 1)*pi*exp(-4*delta*t)*(exp(2*delta*t) - 1)*(-4*delta*t + exp(2*delta*t) - 1))/(4*delta^2)\n",
    "    K[2,2] = -(pi*exp(-4*delta*t)*(exp(2*delta*t) - 1)^2)/(4*delta)\n",
    "    K[2,3] = (1 - f)*pi*exp(-4*delta*t)*(exp(2*delta*t) - 1)\n",
    "    \n",
    "    K[3,1] = f*pi*exp(-6*delta*t)*(6*delta*t*(exp(4*delta*t) - 1) + 3*exp(4*delta*t) - 2*exp(6*delta*t) - 1)/(12*delta^2)\n",
    "    K[3,2] = pi*(exp(-6*delta*t) - 3*exp(-2*delta*t) + 2)/(12*delta)\n",
    "    K[3,3] = 0.5*f*pi*exp(-6*delta*t)*(exp(4*delta*t) - 1)\n",
    "    \n",
    "    coef2 = 0.5*exp(-6*delta*t)/(delta^2)\n",
    "    K[4,1] = coef2*(-0.5*f*pi*exp(2*delta*t) + 0.5*f*pi*exp(4*delta*t) - 2*f*delta*pi*t*exp(2*delta*t) +\n",
    "    f*delta*pi*t*exp(4*delta*t) + f*delta*pi*t + f*pi/6 + 0.5*pi*exp(2*delta*t) - 2*pi*exp(4*delta*t) + \n",
    "    2*delta*pi*t*exp(2*delta*t) - 4*delta*pi*t*exp(4*delta*t))\n",
    "    \n",
    "    coef1 = pi*exp(-6*delta*t)/delta\n",
    "    K[4,2] = coef1*(0.25*exp(2*delta*t) - 0.25*exp(4*delta*t) - 1/12)\n",
    "    K[4,3] = pi*(1 + (0.5*f - 2)*exp(-2*delta*t) + (1 - f)*exp(-4*delta*t) + 0.5*f*exp(-6*delta*t))\n",
    "    \n",
    "    G = map(x -> -sum(K[:,x].*weis.*sign.(ys - Ns)), 1:3)\n",
    "    display(G)\n",
    "end"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       " * Status: success\n",
       "\n",
       " * Candidate solution\n",
       "    Minimizer: [3.19e-04, 5.20e-01, 1.32e+03]\n",
       "    Minimum:   4.317710e-01\n",
       "\n",
       " * Found with\n",
       "    Algorithm:     Nelder-Mead\n",
       "    Initial Point: [5.00e-04, 5.00e-01, 2.00e+03]\n",
       "\n",
       " * Convergence measures\n",
       "    √(Σ(yᵢ-ȳ)²)/n ≤ 1.0e-08\n",
       "\n",
       " * Work counters\n",
       "    Seconds run:   0  (vs limit Inf)\n",
       "    Iterations:    135\n",
       "    f(x) calls:    264\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "using Optim, Distances, QuadGK\n",
    "\n",
    "\n",
    "function ff!(incond)\n",
    "    Ns = [1571, 122, 107, 199]\n",
    "    N = sum(Ns)\n",
    "    delta, f, t = incond\n",
    "    pi = 1.0\n",
    "    N_2, N_3_2, N_3_3, N_more = Ns\n",
    "    weis = 1 ./ Ns\n",
    "    y = Float64[0.0, 0.0, 0.0, 0.0]\n",
    "    y[1] = (pi/(2*delta))*(1 - exp(-2*delta*t))\n",
    "    y[2] = (pi*(1-f)/(4*delta))*(1 - 2*exp(-2*delta*t) + exp(-4*delta*t))\n",
    "    y[3] = (pi*f/(6*delta))*(1 - 1.5*exp(-2*delta*t) + 0.5*exp(-6*delta*t))\n",
    "    #qq(x) = pi*(1-f)*(1 - 2*exp(-2*delta*x) + exp(-4*delta*x)) + pi*f*(1 - 1.5*exp(-2*delta*x) + 0.5*exp(-6*delta*x))\n",
    "    #y[4] = quadgk(qq, 0, t)[1]\n",
    "    c1 = (f/12 - 3/4)*pi/delta\n",
    "    y[4] = c1 - f*pi*exp(-6*delta*t)/(12*delta) + f*pi*exp(-4*delta*t)/(4*delta) - \n",
    "    f*pi*exp(-2*delta*t)/(4*delta) - pi*exp(-4*delta*t)/(4*delta) + pi*exp(-2*delta*t)/delta + pi*t\n",
    "    return wcityblock(y, Ns, weis) # weighted cityblock distance \"wminkowski try\"\n",
    "end\n",
    "\n",
    "res = optimize(ff!, [0.0005, 0.5, 2000], NelderMead())\n",
    "display(res)\n",
    "\n",
    "#res = optimize(ff!, [0.0005, 0.5, 2000], Newton(), Optim.Options(iterations=10000, x_tol=1e-12, allow_f_increases=true))\n",
    "#display(res)\n",
    "#res = optimize(ff!, [0.0005, 0.5, 2000], GradientDescent(), Optim.Options(iterations=10000, x_tol=1e-12, allow_f_increases=true))\n",
    "#display(res)\n",
    "#res = optimize(ff!, [0.0005, 0.5, 2000], LBFGS(), Optim.Options(iterations=10000, x_tol=1e-12, allow_f_increases=true))\n",
    "#display(res)\n",
    "\n",
    "#Optim.abs_tol(res)\n",
    "#Candidate solution\n",
    "#    Minimizer: [3.19e-04, 5.20e-01, 1.32e+03]\n",
    "#    Minimum:   4.317710e-01"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 296,
   "metadata": {},
   "outputs": [],
   "source": [
    "#using BSON, LightGraphs\n",
    "\n",
    "#bson(\"sd_network_graph.bson\", Dict(\"sd_network\" => graph_obj.mg))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.4801903290282366"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "1.4169809598224496"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "PyPlot.Figure(PyObject <Figure size 600x400 with 1 Axes>)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "using Histograms\n",
    "\n",
    "grs = [graph_hum.mg, graph_goril.mg, graph_gibbon.mg, graph_rat.mg, graph_mouse.mg, graph_dog.mg, \n",
    "    graph_chick.mg, graph_zebrafi.mg, graph_celeg.mg];\n",
    "labels = [\"Human\", \"Gorilla\", \"Gibbon\", \"Rat\", \"Mouse\", \"Dog\", \"Chicken\", \"Zebrafish\", \"C. Elegans\"]\n",
    "inds = [1, 2]\n",
    "\n",
    "NetworkAnalysis.Node_degree_Comp_size_distr(grs[inds], labels[inds], nothing; \n",
    "    nd_bin=20, cc_bin=30, max_x=30000);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9-element Array{Any,1}:\n",
       " \"Human\"     \n",
       " \"Gorilla\"   \n",
       " \"Gibbon\"    \n",
       " \"Rat\"       \n",
       " \"Mouse\"     \n",
       " \"Dog\"       \n",
       " \"Chicken\"   \n",
       " \"Zebrafish\" \n",
       " \"C. Elegans\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"huyna.csv\""
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "using LightGraphs, Distances, DelimitedFiles\n",
    "using CSV, DataFrames\n",
    "\n",
    "function take_uniformly(arr, num)\n",
    "    inds = round.(Int, collect(range(1, length(arr), length=num)))\n",
    "    com_sizes = arr[inds]\n",
    "    return com_sizes\n",
    "end\n",
    "\n",
    "cols = [\"Human\", \"Gorilla\", \"Gibbon\", \"Rat\", \"Mouse\", \"Dog\", \"Chicken\", \"Zebrafish\", \"C. Elegans\"]\n",
    "\n",
    "cols_new = []\n",
    "lens_top = []\n",
    "com_num = 496\n",
    "\n",
    "for (ind, val) in enumerate(grs)\n",
    "    lens = sort(map(length, connected_components(val)))\n",
    "    if length(lens) >= com_num\n",
    "        push!(lens_top, lens[end-com_num+1:end])   #  lens[end-com_num+1:end] take_uniformly(lens, com_num)\n",
    "        push!(cols_new, cols[ind])\n",
    "    end\n",
    "end\n",
    "\n",
    "lens_top = hcat(lens_top...)\n",
    "Dist = pairwise(BrayCurtis(), dims=2, lens_top);\n",
    "display(cols_new)\n",
    "\n",
    "df = DataFrame(Dist)\n",
    "CSV.write(\"huyna.csv\", df, writeheader=false)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1×2 Array{Float64,2}:\n",
       " 0.129057  0.018383"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Fisher test p-value for self_loops in big/other components = 1.1281234254459717e-51\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "1×2 Array{Float64,2}:\n",
       " 0.244989  0.332652"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Fisher test p-value for intra_edges in big/other components = 3.0633321577303117e-19\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "1×2 Array{Float64,2}:\n",
       " 0.3416  0.0623821"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"Fisher test p-value for double_edges in big/other components = 2.236505818621229e-285\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"There are 0 simulations with bigger intra- fraction value than observed in middle size components out of 1000\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "NetworkAnalysis.Stat_tests_SD_net_features(graph_obj);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "gr_inter = NetworkAnalysis.Spliting_network_diff_ways(graph_obj, \"inter\");\n",
    "gr_intra = NetworkAnalysis.Spliting_network_diff_ways(graph_obj, \"intra\");\n",
    "gr_inter_nbig = NetworkAnalysis.Spliting_network_diff_ways(graph_obj, \"inter_nobig\");\n",
    "gr_intra_nbig = NetworkAnalysis.Spliting_network_diff_ways(graph_obj, \"intra_nobig\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6986.255783500655"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlgAAAGQCAYAAAByNR6YAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAAPYQAAD2EBqD+naQAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzs3X9clfX9//HHQX6DqCfUUhHwV6SJCuIPUDlZTS1/bKmpTdePWeZKMjcz/WzVqk+5bzPU2TZbTko3nanL3PzVRwFTNBAzW6W5VEQ0U1CQQBC4vn+cOIoHEA5HDsjzfrud203f1/u6rtd51+3Ws3O9r/fbZBiGgYiIiIg4jZurCxARERG52ShgiYiIiDiZu6sLaGq+//570tPT6datG76+vq4uR0REROrIMAwuXrxIu3btcHOz/nalgFXP0tPTiY2NdXUZIiIi4mSZmZl06NABUMCqd35+fgAkJyfTu3fvWp2bk5OD2Wy+EWXd9DR2jtPYOU5j5ziNXd1o/BznyNjl5eURFBRE8+bNbW0KWPXMZDIB4O/vT0BAQK3OLSkpqfU5YqWxc5zGznEaO8dp7OpG4+e4uoxd+X/jQZPcRURERJxOAUtERETEyRSwRERERJxMc7BEREScJDOngNVpJziRU0hHsw8TozoSZNaSPE2RApaIiIgTrE0/yZx1Byktu7ID3dLko8wfG864yA4urExcQY8IRURE6igzp8AuXAGUlBk8v+4gmTkFLqpMXEUBS0REpI5Wp52wC1flSsoMVqedqOeKxNUUsJwgKSkJHx8f+vbta/t89tlnri5LRETqyYmcwmqPZ17nuNx87ALW66+/zvjx4+nUqRMmk4mQkJBaXTApKQmTyVTtZ/fu3TXqX9uVzq8nNTWVuLg4YmJi8PPzw2QykZCQUO05q1atIjIyEh8fHwIDA5k0aRIZGRl2/Tp37sy+fftsn169ejm1dhERabg6mn2qPR50neNy87Gb5D5v3jzMZjMRERFcuHCh1he84447WLFihV17UVERTzzxBIGBgfTr18/u+BNPPMHgwYMrtDl7mf9Nmzbx1ltvERYWRu/evUlJSam2/5IlS5gxYwYxMTHEx8dz7tw5Fi5cyM6dO0lLS6Ndu3ZOrU9ERBqniVEdWZp8lJJKHhO6u5mYGNXRBVWJK9kFrG+++YZOnToBcOedd5Kfn1+rC7Zt25bJkyfbta9atYqysjJ+9rOf4eHhYXd84MCBlZ5XmW3bthESEkK3bt0qPZ6SkoKXlxeRkZEV2qdPn87s2bPx8/Nj7dq11Qas7Oxs5s6dS0REBElJSbi7W4dq+PDh9OvXjxdeeIF33nnH1j8jI4OBAwdiGAYjRoxg7ty5eHp61uj7iIhI4xZk9mX+2HCeX3ewQshydzPxu7HhWqqhCbILWOXhytnKw8jUqVOr7FNQUICbmxve3t5V9rl8+TLTp0+nsLCQxMREbr/99grHd+/ezYgRI+jZs2eFR5FgDX81tWHDBvLz84mLi7OFK4C+ffsyZMgQ1qxZwx//+Ec8PT2JjIwkKyuLgIAAzp8/z5QpU3jxxRd5/fXXa3w/ERFp3MZFdqB/qJnVaSfIzCkkSOtgNWn1Msn92LFjJCYmMmjQILtAVO6ZZ57Bz88PHx8fQkNDeeWVV7h8+bJdPw8PDzZv3ozJZMJisXDo0CHbsV27djF8+HCCgoJYv359nWpOTU0FIDo62u5YdHQ0Fy9etN27efPmto0hW7VqxbRp00hOTq7T/UVEpPEJMvsye1gYiyf1YfawMIWrJqxeAtZf//pXDMOo9NcrDw8PRo4cyfz58/nwww95++236dSpEy+88AKjRo2itLTU7pxu3bqRlJREs2bNsFgsfPXVV+zcuZMRI0YQHBxMYmJirX6tqkxWVhYAHTrYLw5X3nby5EkATp8+jWFYfxIuLi5mzZo19OnTp9rr5+fnk5eXZ/sUFRXVqV4RERFpOG74Su6lpaUkJCQQEBDA+PHj7Y7HxMSwcePGCm2PP/44U6dOZdmyZfzjH//goYcesjuva9euJCUlYbFYiI2NpaCggNDQUHbs2EHr1q3rXHdBgXVROC8vL7tj5Y8wy/ssX76cVatW4enpSXFxMdHR0cyfP7/a68fGxlb4+3PPPcecOXOqPef8+fM1rl8q0tg5TmPnOI2d4zR2daPxc5wjY5eXl2fXdsMD1tatWzl58iTTpk3D17fmP5X+5je/YdmyZfz73/+uNGABdOnShd/+9re2X8YWLVrklHAF2GotKirCx6fi67WFhYUV+sybN4958+bV6vrJyckVlqHw8vKqNMxdy9lvVjYlGjvHaewcp7FznMaubjR+jqvt2F09V7vcDX9EuGzZMqD6ye2VCQoKolmzZpw9e7bKPtu3b2fGjBn06tWLjh07MmHCBA4ePFinesu1b98euPIY8GrVPT6sKX9/fwICAmyfmoQrERERaRxuaMD67rvv2LhxI+Hh4fTt27dW5x49epTS0lJuvfXWSo9/9NFHjBo1irCwMHbs2EFSUhJ+fn4MHTqUAwcO1Ln2qKgogEqXckhJScHf35+wsLA630dERERuPnUKWKdPn+bQoUO2uUjXeu+997h8+XK1v159++23dm2lpaXMnTsXgNGjR9sd37ZtG6NHj6Z79+5s374ds9lMaGgoycnJBAQEcPfdd7N//34Hv5XVmDFj8PX1ZfHixZSUlNja9+3bx86dO3nwwQe1zpWIiIhUyu6h4YoVK2xbwZw9e5bi4mJeffVVAFq2bMnTTz9t6zt37lzeffddEhMTsVgsdhf/61//ire3d7ULiN53333ccsstDBo0iPbt23PmzBnef/99PvvsMx544AHGjh1boX/5Olg9e/Zk27ZttGzZ0nYsODiYpKQk7rrrLmbMmGG3DlZGRoZtlfkvvvgCgI0bN9oeA44ePZrw8HAAAgMDee2115g5cyYWi4UpU6Zw7tw54uPjadu2LS+//HKV30lERESaOOMasbGxBlDpJzg4uELfhx9+2ACMxMTEay9j7N692wCMhx56yO7Y1ebPn29ER0cbrVu3Ntzd3Y2AgAAjOjraWLp0qVFaWlrpOUeOHDEuXLhQ5TUzMzON06dP27UnJiZW+d0AY/ny5XbnrFy50ujTp4/h7e1tmM1mY8KECcbRo0er/U7VSU9PNwAjPT291udmZ2c7fN+mTmPnOI2d4zR2jtPY1Y3Gz3GOjF1ubq4BGLm5ubY2k2EY9hsnyQ2zf/9+IiMjSU9PJyIiolbn5uTk6K0QB2nsHKexc5zGznEau7rR+DnOkbHLy8ujRYsW5Obm2hYer5eFRkVERESaEgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwBIRERFxMgUsERERESdTwHKCpKQkfHx86Nu3r+3z2WefubosERERcRG7gPX6668zfvx4OnXqhMlkIiQkpNYXtVgsmEymSj8ffPCBXX/DMFiyZAk9evTA29ub2267jWnTppGdne3Ql6pKamoqcXFxxMTE4Ofnh8lkIiEhodpzVq1aRWRkJD4+PgQGBjJp0iQyMjLs+nXu3Jl9+/bZPr169XJq7SIiItJ4uF/bMG/ePMxmMxEREVy4cMHhCwcGBhIfH2/X3rdvX7u22bNns2DBAkaOHMnMmTM5duwYCxcuJCUlhb179+Ln5+dwHVfbtGkTb731FmFhYfTu3ZuUlJRq+y9ZsoQZM2YQExNDfHw8586dY+HChezcuZO0tDTatWvnlLpERETk5mIXsL755hs6deoEwJ133kl+fr5DF/bz82Py5MnX7ffVV18RHx/P6NGj2bBhg609MjKScePGER8fz69//esK52zbto2QkBC6detW6TVTUlLw8vIiMjKyQvv06dOZPXs2fn5+rF27ttqAlZ2dzdy5c4mIiCApKQl3d+tQDR8+nH79+vHCCy/wzjvv2PpnZGQwcOBADMNgxIgRzJ07F09Pz+t+fxEREbn52D0iLA9XzlBWVkZeXh5lZWVV9lm1ahVlZWXMmjWrQvvYsWMJCQlh5cqVFdovX77M9OnTsVgsHD582O56u3fvZvjw4cTFxdkda9u2bY1/DduwYQP5+fnExcXZwhVYf4EbMmQIa9asobi4GLCGwaysLPbs2cPmzZtJS0vjxRdfrNF9RERE5OZzwya5Z2Vl4e/vT4sWLfDz8+O+++5j3759dv1SU1Nxc3NjwIABdscGDhzI4cOHyc3NtbV5eHiwefNmTCYTFouFQ4cO2Y7t2rWL4cOHExQUxPr16+tUf2pqKgDR0dF2x6Kjo7l48aLt3s2bNycgIACAVq1aMW3aNJKTk+t0fxEREWm87B4ROkNISAjR0dH07NkTLy8vPv30UxYvXkxMTAybN29m6NChtr5ZWVkEBgbi5eVld50OHTrY+rRo0cLW3q1bN5KSkrjrrruwWCwkJiZy9uxZ7r//foKDg9mxYwdt2rSp03fIysqqUENldZ08eZLw8HBOnz7Nrbfeislkori4mDVr1tCnT59qr5+fn09eXp7t715eXpWOgYiIiDQ+NyRgXftm3gMPPMDkyZOJiIjgySef5Ouvv7YdKygoqDJYeHt72/pcq2vXriQlJWGxWIiNjaWgoIDQ0FB27NhB69at6/wdyu9ZWW3X1rV8+XJWrVqFp6cnxcXFREdHM3/+/GqvHxsbW+Hvzz33HHPmzKn2nPPnz9e4fqlIY+c4jZ3jNHaO09jVjcbPcY6M3dU/mJS7IQGrMrfffjsPPvggCQkJHDlyhK5duwLg6+vLd999V+k5hYWFtj6V6dKlC7/97W+ZOnUqAIsWLXJKuLr6nkVFRfj4+FRb17x585g3b16trp+cnEzv3r1tf6/pL1hms7lW95ErNHaO09g5TmPnOI1d3Wj8HFfbsbt6rna5el1otHxNrbNnz9ra2rdvz7lz5ygqKrLrX/6Yrn379pVeb/v27cyYMYNevXrRsWNHJkyYwMGDB51Sa/k9T548WWVdlT0+rCl/f38CAgJsHz0eFBERuXnUa8A6cuQIALfeequtLSoqirKyMvbu3WvXf8+ePXTr1q3C/KtyH330EaNGjSIsLIwdO3aQlJSEn58fQ4cO5cCBA3WuNSoqCqDSpRxSUlLw9/cnLCyszvcRERGRm0+dAtbp06c5dOhQhTlS58+fty1fcLV9+/axZs0aevToUWEpiIkTJ2IymXjzzTcr9F+/fj3Hjx+vdC2tbdu2MXr0aLp378727dsxm82EhoaSnJxMQEAAd999N/v376/LV2PMmDH4+vqyePFiSkpKKnyPnTt38uCDD2qdKxEREamU3UPDFStW2LaCOXv2LMXFxbz66qsAtGzZkqefftrWd+7cubz77rskJiZisVgA69yiadOmMX78eLp06YKXlxcHDhwgISEBDw8P/vKXv1S4X48ePXjmmWdYuHAho0aNYsyYMRw7doz4+HjuuOMOnn322Qr9y9fB6tmzJ9u2baNly5a2Y8HBwba3C2fMmMHu3bsrnJuRkcGKFSsA+OKLLwDYuHGj7THg6NGjCQ8PB6wr0b/22mvMnDkTi8XClClTOHfuHPHx8bRt25aXX365lkMtIiIiTYZxjdjYWAOo9BMcHFyh78MPP2wARmJioq3tyy+/NMaPH2907tzZ8Pf3Nzw8PIzg4GDjscceM77++utrb2cYhmGUlpYaixYtMsLCwgxPT0+jbdu2xuOPP26cPXu20v5HjhwxLly4UOkxwzCMzMxM4/Tp03btiYmJVX43wFi+fLndOStXrjT69OljeHt7G2az2ZgwYYJx9OjRKu99Penp6QZgpKen1/rc7Oxsh+/b1GnsHKexc5zGznEau7rR+DnOkbHLzc01ACM3N9fWZjIMw6j/WNd07d+/n8jISNLT04mIiKjVuTk5OXorxEEaO8dp7BynsXOcxq5uNH6Oc2Ts8vLyaNGiBbm5ubaFx+t1kruIiIhIU6CAJSIiIuJkClgiIiIiTqaAJSIiIuJkClgiIiIiTqaAJSIiIuJkClgiIiIiTqaAJSIiIuJkClgiIiIiTqaAJSIiIuJkClgiIiIiTqaAJSIiIuJkClgiIiIiTqaAJSIiIuJkClgiIiIiTqaAJSIiIuJkClgiIiIiTqaAJSIiIuJkClgiIiIiTqaAJSIiIuJkClgiIiIiTubu6gJERETEOTJzCliddoL/nr5Al9taMjGqI0FmX1eX1SQpYImIiNwE1qafZM66g5SWGQBsPZTN0uSjzB8bzrjIDi6urunRI0IREZFGLjOnoEK4KldSZvD8uoNk5hS4qLKmSwFLRESkkVuddsIuXJUrKTNYnXainisSBSwREZFG7kROYbXHM69zXJxPAUtERKSR62j2qfZ40HWOi/MpYImIiDRyE6M64u5mqvSYu5uJiVEd67kiUcASERFp5ILMvswfG24XstzdTPxubLiWanABLdMgIiJyExgX2YH+oWZWp53gm28v0PlWrYPlSgpYIiIiN4kgsy+zh4WRk5OD2Wx2dTlNmh4RioiIiDiZApaIiIiIkylgiYiIiDiZApaIiIiIkylgiYiIiDiZApYTJCUl4ePjQ9++fW2fzz77zNVliYiIiIvYBazXX3+d8ePH06lTJ0wmEyEhIbW64Pnz51m0aBE/+tGPCAoKwsfHh9tvv50nnniCzMxMu/5JSUmYTKZKP71793b4i1UmNTWVuLg4YmJi8PPzw2QykZCQUO05q1atIjIyEh8fHwIDA5k0aRIZGRl2/Tp37sy+fftsn169ejm1dhEREWk87NbBmjdvHmazmYiICC5cuFDrC37yySfMmjWLoUOH8tRTTxEYGMgXX3zB0qVLWbNmDSkpKXTv3t3uvCeeeILBgwdXaHP2Gh6bNm3irbfeIiwsjN69e5OSklJt/yVLljBjxgxiYmKIj4/n3LlzLFy4kJ07d5KWlka7du2cWp+IiIjcHOwC1jfffEOnTp0AuPPOO8nPz6/VBcPCwjh8+DBdunSp0H7//fdz77338uKLL/L+++/bnTdw4EAmT55co3ts27aNkJAQunXrVunxlJQUvLy8iIyMrNA+ffp0Zs+ejZ+fH2vXrq02YGVnZzN37lwiIiJISkrC3d06VMOHD6dfv3688MILvPPOO7b+GRkZDBw4EMMwGDFiBHPnzsXT07NG30dERERuLnaPCMvDlaNCQkLswhXAPffcg9ls5vPPP6/y3IKCAi5dulTt9S9fvsz06dOxWCwcPnzY7vju3bsZPnw4cXFxdsfatm2Ln59fDb4FbNiwgfz8fOLi4mzhCqBv374MGTKENWvWUFxcDEBkZCRZWVns2bOHzZs3k5aWxosvvlij+4iIiMjNp94muefm5nLx4kXatGlT6fFnnnkGPz8/fHx8CA0N5ZVXXuHy5ct2/Tw8PNi8eTMmkwmLxcKhQ4dsx3bt2sXw4cMJCgpi/fr1dao3NTUVgOjoaLtj0dHRXLx40Xbv5s2bExAQAECrVq2YNm0aycnJdbq/iIiINF71FrBeffVVLl++zMMPP1yh3cPDg5EjRzJ//nw+/PBD3n77bTp16sQLL7zAqFGjKC0ttbtWt27dSEpKolmzZlgsFr766it27tzJiBEjCA4OJjExkbZt29ap3qysLAA6dOhgd6y87eTJkwCcPn0awzAAKC4uZs2aNfTp06fa6+fn55OXl2f7FBUV1aleERERaTjqZbPnNWvWsGDBAu69914effTRCsdiYmLYuHFjhbbHH3+cqVOnsmzZMv7xj3/w0EMP2V2za9euJCUlYbFYiI2NpaCggNDQUHbs2EHr1q3rXHNBQQEAXl5edse8vb0r9Fm+fDmrVq3C09OT4uJioqOjmT9/frXXj42NrfD35557jjlz5lR7zvnz52tcv1SksXOcxs5xGjvHaezqRuPnOEfGLi8vz67thgesTZs2MWXKFPr06cP777+Pm1vNfjT7zW9+w7Jly/j3v/9dacAC6NKlC7/97W+ZOnUqAIsWLXJKuALw9fUFoKioCB8fnwrHCgsLK/SZN28e8+bNq9X1k5OTKyxD4eXlVWmYu5Z2R3ecxs5xGjvHaewcp7GrG42f42o7dlfP1S53Qx8RbtmyhQceeICwsDC2bdtGixYtanxuUFAQzZo14+zZs1X22b59OzNmzKBXr1507NiRCRMmcPDgQWeUTvv27YErjwGvVt3jw5ry9/cnICDA9qlJuBIREZHG4YYFrK1bt/KTn/yEbt26sX37dm655ZZanX/06FFKS0u59dZbKz3+0UcfMWrUKMLCwtixYwdJSUn4+fkxdOhQDhw4UOf6o6KiACpdyiElJQV/f3/CwsLqfB8RERG5+dQpYJ0+fZpDhw7Z5iKV27ZtGz/+8Y/p2rUrO3bsIDAwsMprfPvtt3ZtpaWlzJ07F4DRo0fbHd+2bRujR4+me/fubN++HbPZTGhoKMnJyQQEBHD33Xezf//+unw1xowZg6+vL4sXL6akpMTWvm/fPnbu3MmDDz6oda5ERESkUnYPDVesWGHbCubs2bMUFxfz6quvAtCyZUuefvppW9+5c+fy7rvvkpiYiMViAawBZMyYMRiGwWOPPcaWLVvsbnr1gqL33Xcft9xyC4MGDaJ9+/acOXOG999/n88++4wHHniAsWPHVji3fB2snj17sm3bNlq2bGk7FhwcTFJSEnfddRczZsxg9+7dFc7NyMhgxYoVAHzxxRcAbNy40fYYcPTo0YSHhwMQGBjIa6+9xsyZM7FYLEyZMoVz584RHx9P27Ztefnll2syviIiItIUGdeIjY01gEo/wcHBFfo+/PDDBmAkJiba2pYvX17l+eWfq82fP9+Ijo42Wrdubbi7uxsBAQFGdHS0sXTpUqO0tPTa8gzDMIwjR44YFy5cqPSYYRhGZmamcfr0abv2xMTEautavny53TkrV640+vTpY3h7extms9mYMGGCcfTo0SrvfT3p6ekGYKSnp9f63OzsbIfv29Rp7BynsXOcxs5xGru60fg5zpGxy83NNQAjNzfX1mYyjB8WcJJ6sX//fiIjI0lPTyciIqJW5+bk5OitEAdp7BynsXOcxs5xGru60fg5zpGxy8vLo0WLFuTm5toWHq+3hUZFREREmop6WWhUREREbm6ZOQWsTjvBiZxCOpp9mBjVkSCzr6vLchkFLBEREamTteknmbPuIKVlV2YdLU0+yvyx4YyLdHzNyMZMjwhFRETEYZk5BXbhCqCkzOD5dQfJzCmo4sybmwKWiIiIOGx12gm7cFWupMxgddqJeq6oYVDAEhEREYedyCms9njmdY7frBSwRERExGEdzT7VHg+6zvGblQKWiIiIOGxiVEfc3UyVHnN3MzExqmM9V9QwKGCJiIiIw4LMvswfG24XstzdTPxubHiTXapByzSIiIhInYyL7ED/UDOr006QmVNIkNbBUsASERGRugsy+zJ7WJiry2gw9IhQRERExMkUsEREREScTAFLRERExMkUsEREREScTAFLRERExMkUsEREREScTAFLRERExMm0DpaIiIi4XGZOAavTTnAip5CON8FCpQpYIiIi4lJr008yZ91BSssMW9vS5KPMHxvOuMgONb9QSTFggLuX84usJT0iFBEREZfJzCmwC1cAJWUGz687SGZOwfUvcvFb2PZr+H+h8M7dN6jS2lHAEhEREZdZnXbCLlyVKykzWJ12ouqTzx+Hf82CheGQ8gcozoeA9jem0FrSI0IRERFxmRM5hdUez6zs+NnD8PGb8Pn7YJRa24L6w+BfQdd7b0CVtaeAJSIiIi7T0exT7fGgq4+f+hQ+XgBf/Qv44VevzkNh8C8hOAZMphtXaC0pYImIiIjLTIzqyNLko5RU8pjQ3c3ExKiOcHy3NVh9s/3KwbCR1mDVPqIeq605BSwRERFxmSCzL/PHhvP8uoMVQpa7GyQMyiXogwfgxB5ro6kZ9BwHg56FNne4qOKaUcASERERlxoX2YH+oWZWp53gZHY+Q/mEETl/xzP1P9YOzTyhz2SIjgNzqGuLrSEFLBEREXG5oBYezG6TDl/HQ/YRa6OHH/R9FAY+DQG3ubbAWlLAEhEREdetpH65EPavgJTFkJtpbfNuCf2fhP7TwNd842u4ARSwREREmjinraReG5fyYN8y2PMWfH/W2ubXBgY+BVE/B6/mN+a+9UQBS0REpAm73krq/UPNzv0l6/ts+OTPkLoULuVa21p0hJg46zwrj+qXbWgsFLBERESasJqspD57WFjdb5R3ClKWQPpyuPzD9jeB3WDQLOubgc086n6PBkQBS0REpAlzaCX12sg5CrsXwYG/Q2mxte3WcBjyKwgbBW435659ClgiIiJNWK1WUq+NM1/Crnj4z1owyn64WbR1cdAudzeoVddvBAUsERGRJqxGK6nXxsl066rrh/99pa3LPT9sZxNdx2obDwUsERGRJqzqldRN/G5seM0muBsGHP/YGqyOJv3QaII7RlmDVbveN6T2hkwBywkSEhKYOXMmXbp0sbVt2rSJNm3auLAqERGRmrl6JfXMnEKCaroOlmHA11uswepkmrXN1AzCJ8CgmdD69htffANlN7Ps9ddfZ/z48XTq1AmTyURISIhDF05PT2f48OG0aNGC5s2bY7FY2LlzZ6V9i4uLefnll+ncuTNeXl4EBwczZ84cCgoKHLp3VT766COefPJJoqKi8PLywmQykZSUVGV/wzBYsmQJPXr0wNvbm9tuu41p06aRnZ1t13fQoEHs27fP9lG4EhGRxiTI7MvsYWEsntSH2cPCqg9XZaXw+Vr48yBYNdEarpp5QdRUiPsUfvKnJh2uoJJPQS4JAAAgAElEQVRfsObNm4fZbCYiIoILFy44dNG0tDRiY2Np06YNv/nNb/Dy8uLtt9/m7rvvZvPmzdxzzz0V+j/00EOsW7eOKVOmMGTIED777DMWLFjAvn37+Oijj3Bz0hsGf/vb3/j73//OnXfeSffu3Tlw4EC1/WfPns2CBQsYOXIkM2fO5NixYyxcuJCUlBT27t2Ln5+fU+oSERFpFEqK4eBq6+T1nKPWNk9/68KgA56C5m1dW19DYlzjm2++sf25R48eRnBw8LVdrmvAgAGGn5+fkZGRYWu7cOGC0b59e6Nr165GWVmZrX3Lli0GYMyYMaPCNX7/+98bgLFixQq76//jH/8wzpw5U+X9N27caBw9etSu/eTJk8alS5cMwzCMN954wwCMxMTESq/x5ZdfGm5ubsbo0aMrtK9du9YAjFdeecXWtnz5ciMwMNDo16+fMXDgQGPRokVV1paenm4ARnp6epV9qpKdnV3rc8RKY+c4jZ3jNHaO09jVjdPHr+h7w9jzJ8NYcIdhvBhg/cwPNozE+YZRkOPce7mYI2OXm5trAEZubq6tze6noU6dOtUpsB09epS9e/cyfvx4Ona88uZBixYtmDp1KkeOHOGTTz6xtf/tb38D4Je//GWF6/ziF7/Ax8eHlStXVmjPycnhiSeewGKxcObMGbv7r1u3jp/85Cf8+te/tjvWvn17vLy8avQ9Vq1aRVlZGbNmzarQPnbsWEJCQirUNXr0aE6cOMEnn3zC2rVrWbFiBUuXLq3RfURERBqswguw8/ew8E7YMgfyssD/VvjR/8LM/4BlDvi0cnWVDZLTV/dKTU0FIDra/lXM8rbyPuV/bteuHcHBwRX6+vj40Lt37wp9AcxmM5s3byYrKwuLxcK3335rO7Z27VomTpxIdHR0nQNOamoqbm5uDBgwwO7YwIEDOXz4MLm5ubaafHys64S0a9eOn/3sZyQnJ9fp/iIiIi6Tfxa2vwwLe8KOV6AgG1oGw8h4eOYziH4avPxdXWWD5vSAlZWVBUCHDvabQ5a3nTx5skL/yvqW9z9//rzdZPeBAweydetWTp06hcVi4fTp06xZs4ZJkyYRExPDpk2b8Pev2z/4rKwsAgMDK/3Fq7ze8u966tQp27Hvv/+eDz74gD59+lR7/fz8fPLy8myfoqKiOtUrIiJSZ7knYfMca7D6eAEU5UHrMHjgLzBjP/R9DDy8XV1lo+D0ZRrKw1BlwcTb27tCn/I/V/XY7ur+vr4V32YYMGAAW7duZdiwYfTv359Tp04xZMgQ/vWvf9n1dfR71KQugP/93//l448/xtPTk6KiIu6//36effbZaq8fGxtb4e/PPfccc+bMqfac8+fP17R8uYbGznEaO8dp7Bynsaub2o6f24VjeO/7M16H/omp7DIAJW3CKYx6isud7gGTG+Tm3YhSGxxH/t3Ly7MfG6cHrPJwU9kvMoWFhRX6lP+5ql9vKut/tQEDBhAXF8err75Ks2bN+POf/+yUcFV+z++++65Gdb311lu1vn5ycjK9e19ZeM3Ly6tG88PMZnOt7yVWGjvHaewcp7FznMaubmo0ft/+B3a9CV/888p2NiGDYfAvce9koflNvp1NVWr77567u32ccnrAat++PVDxMWC5yh4ftm/fvtK+5f1btWpVZWhasWIFr732GoMGDeI///kP9913H4mJiQQFBdX1a9C+fXu+/PJLioqK7IJP+fco/66O8Pf3JyAgoE41ioiIOCQzDT7+vXWR0HLdhsOgWdCxv+vquok4fQ5WVFQUACkpKXbHytvK+5T/+dSpU2RkZFToW1hYyIEDByr0vdp7773HI488wr333stHH33E//3f/5GTk0NsbKzdtRz9HmVlZezdu9fu2J49e+jWrRstWrSo831ERETqhWFYt7FJGAnL7vkhXJmgxwPw5C546B8KV05Up4B1+vRpDh06VGFOVefOnenXrx/vv/8+mZmZtva8vDyWLVtG586dK7yZ99BDDwGwYMGCCtf+05/+RGFhIZMnT7a7b0JCAo8++ijDhg1jw4YNeHt7ExkZyfbt28nNzSU2Npbjx4/X5asxceJETCYTb775ZoX29evXc/z48UrrEhERaXDKyuDQv+Gdu+G9MdY9A93coc9keHofjF8Ot/Z0dZU3HbtHhCtWrLD9AnT27FmKi4t59dVXAWjZsiVPP/20re/cuXN59913SUxMxGKx2NoXL16MxWJh8ODBxMXF4enpydKlSzl9+jSbNm3CdNUz3REjRvDjH/+YP/zhD+Tm5tpWcv/jH/+IxWLhpz/9aYX6cnJymDlzJvfddx9r166t8PiuT58+7Nixg3vuuYf/+Z//sa2xVe7gwYN8+OGHAOzevdv2fXft2gXAlClTbMtF9OjRg2eeeYaFCxcyatQoxowZw7Fjx4iPj+eOO+647iR2ERERlyotsc6t2vUmfPeltc3dByIfhoFPQ8u6T6eRaly7GmlsbKwBVPq5dlX3hx9+uMrV0FNTU417773XaN68ueHr62sMGTKkylXTL126ZLz44otGaGio4enpaQQFBRmzZ8828vPzK+1/8OBBo6ioqMoVVb/66ivj4sWLdu3Lly+v8rtV9j1KS0uNRYsWGWFhYYanp6fRtm1b4/HHHzfOnj1b5b2vRyu5u4bGznEaO8dp7BynsauDy5eM/OQlhrGw15VV11/rYBgfvWQYF79zdXUNnrNWcjcZhmHUf6xruvbv309kZCTp6elERETU6tycnBy9VeMgjZ3jNHaO09g5TmPngOLvIT0BUv4AF09b23xvgQHTIepx8Gnp0vIaC0f+3cvLy6NFixbk5ubaXmBz+luEIiIiUo8Kz0PqO7D3j1CYA0CZ3624DZ4JET8DTz8XF9g0KWCJiIg0RvnfwZ63IG0ZFF+0trUKhUHPciHoR5jb3Oba+po4BSwREZHG5EImpCyG/e9BySVrW5vuMPiX0P3H0MwdcnJcW6MoYImIiDQK547AroVwcDWUlVjb2veFIb+CrsPAzelLW0odKGCJiIg0ZKcPWjde/nID1pfegdBY6y9WoUOgiW5n09ApYImIiDQQmTkFrE47wYmcQga6f81Pvv8HPse3X+lw+33W7WyCKt/lRBoOBSwREZEGYG36Seas+4xoDvK0+wf0dzsEgIEbpp5jYdCz0LaHi6uUmlLAEhERcbHM7Hy2r1/GevcP6OV2FIBioxlrS4ewrGw0CXdNJMjs6+IqpTYUsERERFyl9DL8Zx0+W17nTx7HASg0PFlVOpS3S+7nW24BYHXaCWYPC3NhoVJbClgiIiL17fIlOPA32L0QLpwgEMgzfEko/REJJcPJIaBC98ycQtfUKQ5TwBIREakvRfmQvhxSlkD+t9Y230CSAx/k6a8juEjljwGDzD71WKQ4gwKWiIjIjVaQA6lvwyd/tm5tAxDQAWLioM8UOuVD4e+ToMx+e2B3NxMTozrWb71SZwpYIiIiN8rFM7BnCez7KxTnW9vMna1vBIZPAHdPAILMMH9sOM+vO0jJVSHL3c3E78aGa4J7I6SAJSIi4mznM37YzmYFlBZZ29r2hMGzoPsYcGtmd8q4yA70DzWzOu0EmTmFBJl9mBjVUeGqkVLAEhERcZazh2FXPBxcA0apta1Dvx+2s/nRdVddDzL76m3Bm4QCloiISF2d+hQ+fhO+2ohtO5tOd1m3swkZpO1smiAFLBEREUdlpMDO38M3V21nEzbS+iiwfaTr6hKXU8ASERGpDcOA//6fdQPmE3usbaZm0HOcdfJ6mztcW580CApYIiIiNVFWan0E+PEC+Pagta2ZJ/T+KcQ8A+ZQ19bXxF29UXbHBvCCgAKWiIhIdUovWyet74qH7CPWNg9f6PsYDHwaAm5zbX3yw0bZBym9aomLpclHmT82nHGRHVxSkwKWiIhIZS4XwqcrYfciyM20tnm3gP5PWj++ZtfWJ4D1l6trwxVASZnB8+sO0j/U7JJfshSwRERErnYpD/Ytgz1vwfdnrW1+bWDgU9ZfrbwDqj9f6tXqtBN24apcSZnhso2yFbBEREQAvs+2bmWTuhQu5VrbWgRZ51f1mQwe2g+wITpxnY2wXbVRtgKWiIg0bXmnrJsvpy+HywXWtlu6Wpda6Dkemnm4tj6pVsfrbITtqo2yFbBERBqYhvY21E0r5xjsXggH/g6lxda2W8Oti4PeMarS7Wyk4ZkY1ZGlyUcr7OFYzpUbZStgiYg0IA3xbaibzpkvrW8E/mctGGXWto4DYfCvoMvdWnW9kQky+zbIjbIVsEREGoiG+jbUTSMr3bqdzaF/XWnrco/1F6vgaNfVJXXWEDfKVsASEWkgGurbUI2aYcDxXfDx7+Fo0g+NJusjwMG/hHa9XVmdOFFD2yhbAUtEpIFoqG9DNUqGAV9vta66fjLV2mZqBuETYNBMaH27a+uTm54ClohIA9FQ34ZqVMpK4csPrI8Cz/zH2tbMCyKmQHQctAp2bX3SZChgiYg0EA31baj6VP4G5X9PX6DLbS1rPo+mpBgO/sM6eT3nG2ubpz9E/RwGPAXN297YwkWuoYAlItJANNS3oerLtW9Qbj2Uff03KIsLYP97kLIY8rKsbT6trFvZ9HtC29mIyyhgiYg0IA3xbaj6UOs3KC/lQto7sOePUHDO2uZ/K0Q/DZGPgpd/PVYvYk8BS0SkgWlob0PVhxq/Qfn9Odj7J0j9CxT9sJ1Ny2DrxPVeD4GHdz1WLVI1BSwREXG5671BefFMBmx5F9ITrmxn0zoMBs2CO8dCM/3nTBoW/RspIiIuV9UblMGmb3my2UYePPYxHC2xNrbrY111/fb7wM2tHqsUqTkFLCfKz8+nV69elJaWcvz4cVeXIyLSaFz7BuXtphP8wv1DRrrtoZnJAAMIHgRDfgmd7tJ2NtLg2UX/l156CZPJVOXHw+P6u4pbLJYqz//ggw/s+huGwZIlS+jRowfe3t7cdtttTJs2jezsbOd8yx+kpqYSFxdHTEwMfn5+mEwmEhISqj1n1apVREZG4uPjQ2BgIJMmTSIjI6PSvnFxcXTq1MmpNYuIYzJzCnhj6yFmrPqUN7YeIjOnwNUlSTXK36CMbPYNf/FYwFav5xnTLIVmJoPTbWLhsa3w6L+h81CFK2kU7H7BeuCBB+jSpYtdx4MHD/LGG28watSoGl04MDCQ+Ph4u/a+ffvatc2ePZsFCxYwcuRIZs6cybFjx1i4cCEpKSns3bsXPz+/Gt3zejZt2sRbb71FWFgYvXv3JiUlpdr+S5YsYcaMGcTExBAfH8+5c+dYuHAhO3fuJC0tjXbt2tn6/vOf/yQ/P5+f/vSnvPTSS06pV0Qcow2TGxnDgGPJjPvPAsZ57ASgDBNf33I3LX80h9tu7+fiAkVqzy5ghYeHEx4ebtdx2rRpAPz85z+v0YX9/PyYPHnydft99dVXxMfHM3r0aDZs2GBrj4yMZNy4ccTHx/PrX/+6wjnbtm0jJCSEbt26VXrNlJQUvLy8iIyMrNA+ffp0Zs+ejZ+fH2vXrq02YGVnZzN37lwiIiJISkrC3d06VMOHD6dfv3688MILvPPOOwCcPn2al156icTERD788MPrfmcRuXG0YXIjUlYGX2+xbmeTtc/a5uYOvSaSd+ejhHW2/x9ykcaiRrMDCwoKWL16Ne3bt2f48OE1vnhZWRl5eXmUlZVV2WfVqlWUlZUxa9asCu1jx44lJCSElStXVmi/fPky06dPx2KxcPjwYbvr7d69m+HDhxMXF2d3rG3btjX+NWzDhg3k5+cTFxdnC1dg/QVuyJAhrFmzhuLiYgAee+wx3njjDcxmLWgn4mo1ed1fXKy0BD5fC38eBKsnWcOVuzf0mwZxB2DMW5S10nQLadxqFLDWrFlDXl4ejz76KM2aNavRhbOysvD396dFixb4+flx3333sW/fPrt+qampuLm5MWDAALtjAwcO5PDhw+Tm5traPDw82Lx5MyaTCYvFwqFDh2zHdu3axfDhwwkKCmL9+vU1qrMqqanWzUGjo6PtjkVHR3Px4kXbvbds2cKwYcMwmUw8+uijZGRkYDKZuHDhQp1qEJHa04bJDVhJEaS/C0v6wrqfw3dfgGdzGPQszPwc7vt/0DLI1VWKOEWN3iJctmwZJpOJxx57rEYXDQkJITo6mp49e+Ll5cWnn37K4sWLiYmJYfPmzQwdOtTWNysri8DAQLy8vOyu06FDB1ufFi1a2Nq7detGUlISd911FxaLhcTERM6ePcv9999PcHAwO3bsoE2bNjWqtSpZWVkVaqisrpMnTxIeHo5hXPm/5YSEBF566aXrvkWYn59PXl6e7e9eXl6VjoGI1I42TG6Air+3BquUP8DFU9Y2HzMM/AVEPQ4+LV1bn8gNcN2AdfjwYXbt2sXdd99NaGhojS567Zt5DzzwAJMnTyYiIoInn3ySr7/+2nasoKCgymDh7e1t63Otrl27kpSUhMViITY2loKCAkJDQ9mxYwetW7euUZ3VKb9nZbVVV1dNxcbGVvj7c889x5w5c6o95/z58w7fr6nT2DmusY3d8K4BLE02Vblh8vCuAeTk5NRLLY1t7JzNVJSH18H38P50OW6XrGNe5teWwognKLpzInj4QmEZFNr/82jqY1dXGj/HOTJ2V/9gUu66AWvZsmUATJ06tdY3vNrtt9/Ogw8+SEJCAkeOHKFr164A+Pr68t1331V6TmFhoa1PZbp06cJvf/tbW22LFi1ySri6+p5FRUX4+FT8P97q6nrkkUd45JFHrnv95ORkevfubft7TX/B0jwvx2nsHNeYxs5shvljy6rcMLlnp3bVnH0j6mk8Y+c0+Wdh71uQ+g4UX7S2tQqBQc/i1msSfu5e1GQ2bJMcOyfS+DmutmN39VxtW1t1J5SUlPDee+9hNpv5yU9+UrvqKhESEgLA2bNnbQGrffv2fPnllxQVFdkFjPLHdO3bt6/0etu3b2fGjBn06tWL8+fPM2HCBLZv317pW5C1VX7PkydP2mq9tq7KHh/WlL+/PwEBAY4XKCJVaqobJrvchUzrY8D970LJJWtbm+4w+JfQ/cfazkaalGonuW/cuJEzZ84wZcoUp8wPOnLkCAC33nqrrS0qKoqysjL27t1r13/Pnj1069atwvyrch999BGjRo0iLCyMHTt2kJSUhJ+fH0OHDuXAgQN1rjUqKgqg0qUcUlJS8Pf3JyysaW3GKtKYlG+YvHhSH2YPC1O4upHO/Rc2PAWLe0PqUmu4ah8JE/8OT+6GnuMUrqTJqTZglT8erGrtq9OnT3Po0KEKc5HOnz9vW77gavv27WPNmjX06NGjwmrnEydOxGQy8eabb1bov379eo4fP17pWlrbtm1j9OjRdO/ene3bt2M2mwkNDSU5OZmAgADuvvtu9u/fX91Xu64xY8bg6+vL4sWLKSkpqfA9du7cyYMPPoinp2ed7iEi0qh9+zm8/4j1rcBPV0JZCYQOgZ9tgKnbIex+7RUoTVaV/0tx6tQptmzZQr9+/ejZs2elfebOncu7775LYmIiFosFsM4tmjZtGuPHj6dLly54eXlx4MABEhIS8PDw4C9/+UuFa/To0YNnnnmGhQsXMmrUKMaMGcOxY8eIj4/njjvu4Nlnn63Qv3wdrJ49e7Jt2zZatrzy9klwcLDt7cIZM2awe/fuCudmZGSwYsUKAL744gvA+ivdyZMnARg9erTt8WJgYCCvvfYaM2fOxGKxMGXKFM6dO0d8fDxt27bl5Zdfvu7giojclDJTYefv4cjWK23dRlgfBQZFua4ukQakyoCVkJBAaWlprSe333777cTGxrJlyxbOnDlDUVER7dq1Y/LkyTz//PN285kAFixYQGhoKH/605946qmnaNWqFZMnT+a1117D39+/Ql8PDw+2bt1K69atK3102LFjR5KTkyudcHbs2DF+85vfVGhbv369bc2sDh06VJi/9cwzzxAYGMiCBQuYOXMmvr6+3Hvvvbz++utVzgsTEbkpGQYcTYSP34TjH1vbTG7Q4ycwaBbceqdr6xNpYEzG1Ys4yQ23f/9+IiMjSU9PJyIiolbn5uTk6K0QB2nsHKexc9xNMXZlZXB4k3U7m1M/TL1w84DekyBmJtzS+Ybc9qYYOxfS+DnOkbHLy8ujRYsW5Obm2l5g06xDERGxV1oCX6y3/mJ19itrm7sPRD4C0U9DC22aLVIdBSwREbni8iVyUhJwS1lMyyLrkjRlns1x6/8EDPgF+AW6uECRxkEBS0REoCgf0hMoTF6IuegsAOeMAP5aMoK/F/+IX7ccwDiFK5EaU8ASEWnKCs/DJ2/DJ3+CwvP4AKcMM2+XjGR16V1cwroG4vPrDtI/1Kz1xERqSAFLRKQpunjGup1N2jIozgfgvHcHXr84gn+WDubyNf95KCkzWJ12gtnDtMCySE0oYImINCUXTsDuRbB/BZQWWdva9IDBs3jx8xA+PHimylMzcwrrqUjHZeYUsDrtBCdyCumoLZLEhRSwRESagrNfw654+HyNdcV1gA5RMPhX0G0YmEwEnToEVB2wgsw+VR5rCNamn2TOuoOUXrXJ99Lko8wfG864SL31KPVLAUtE5GZ26oB1DauvNgI/BI9OFuuq6yGDwWSydZ0Y1ZGlyUcpKbNfHtHdzcTEqI71UrIjMnMK7MIVWB9tav6YuII2iRIRuRllpMDKsfB2LHz1IWBA2EiYusO6V2DokArhCqwbZM8fG467W8V2dzcTvxsb3qADyuq0E3bhqlz5/DGR+qRfsEREbhaGAf/dDh//Hk7ssbaZ3ODOcTDoWWjb/bqXGBfZgf6hZlannSAzp5CgRjKP6cR15oc1hvljcnNRwBIRaezKyqy/Un28AL49aG1r5gm9JsGgmWDuVKvLBZl9G93bgh2vMz+soc8fk5uPApaISGNVehk+f986ef3c19Y2D1/o+xgMfAoC2rm2vnrUmOePyc1JAUtEpLG5XAifroTdiyH3h7lF3i2g3zTo/yT43eLa+lygfP7Y8+sOVghZjWH+mNycFLBERBqLoovWhUH3vAXff2dt82tt/bWq78/BO8C19blYY50/JjcnBSwRkYauIAc++bP1cynX2hbQAWKegYgp4KH5ReUa4/wxuTkpYImINFR5p2HPEti3HC5/b227pQsMmgU9x4O7p2vrE5EqKWCJiDQ0OccgZbF1nlVpsbXt1p7WxUHvGA1uzVxbn4hclwKWiEhD8d0h2PUmfL4WjFJrW1B/63Y2Xe+1WxhURBouBSwREVfL2m9dw+rQv660db4bBs+C4BgFK5FGSAFLRMQVDAMydluD1Tc7rrTfMco6x6p9hOtqE5E6U8ASEalPhgFHtlmDVeYn1jZTM+uk9UHPQhu9ASdyM1DAEhGpD2Wl8OUG+PhNOPO5ta2ZF/SZDDFx0CrEpeWJiHMpYImI3EglxfD5Gut2Ntn/tbZ5+EHUYzDwaWh+q2vrE5EbQgFLRORGuFyI12cJ8OkyyDtpbfNuCQOmQ78nwNfs0vJE5MZSwBIRcaZLudbtbPb+Eb/vz1rb/Ntaf63q+yh4NXdtfSJSLxSwRESc4fts+ORP8MnbUGTdzqY0oAPNBs+C3j8FD28XFygi9UkBS0SkLvJOQcoSSF8OlwusbYG3w+BZ5La7C3Prtq6tT0RcQgFLRMQR2d/A7kVw4O9Qdtnadltv63Y2YSPBzQ1yclxbo4i4jAKWiEhtnPnSup3Nf9aBUWZtC46xBqvOQ7XquogAClgiIjVzMt26OOjhf19p6/oj66rrwQNdV5eINEgKWCIiVTEMOLbTGqyOJf/QaILuY6z7BN7Wy6XlNVSZOQWsTjvBiZxCOpp9mBjVkSCzr6vLEqlXClgiItcyDPh6izVYnUyztrm5Q/gE63Y2gV1dW18Dtjb9JHPWHaS0zLC1LU0+yvyx4YyL7ODCykTqlwKWiEi5slL44p/W7Wy++8La5u4NET+D6BnQsqNr62vgMnMK7MIVQEmZwfPrDtI/1KxfsqTJUMASESkphoOrrdvZ5By1tnk2v7KdjX8b19bXSKxOO2EXrsqVlBmsTjvB7GHazFqaBgUsEWm6ir+H/e9Byh8gL8va5tMKBvwC+j1u/bPU2ImcwmqPZ17nuMjNRAFLRJqewguQ9g7s/SMUZFvbmt9mfQwY8TB4+bu2vkaqo9mn2uNB1zkucjNRwHKS/Px8evXqRWlpKcePH3d1OSJSmfyz1lCV9g4U5VnbWoVAzEzo/RC4e7m0vMZuYlRHliYfpaSSx4TubiYmRmkOmzQdbtc2mEymKj8XLlyo0UX/+9//Mm7cOG655RZ8fX3p378/69evr7SvYRgsWbKEHj164O3tzW233ca0adPIzs6u2ze7RmpqKnFxccTExODn54fJZCIhIaHac1atWkVkZCQ+Pj4EBgYyadIkMjIyKu0bFxdHp06dnFqziDhJ7knYPAcW9rQuElqUB63vgAfegafTrZswK1zVWZDZl/ljw3F3q7jYqrubid+NDdcEd2lSKv0Fa/DgwTzxxBN27X5+fte94PHjxxk4cCCGYfDss88SGBjIypUrGTt2LMuXL+eRRx6p0H/27NksWLCAkSNHMnPmTI4dO8bChQtJSUlh7969NbpnTWzatIm33nqLsLAwevfuTUpKSrX9lyxZwowZM4iJiSE+Pp5z586xcOFCdu7cSVpaGu3atbP1/ec//0l+fj4//elPeemll5xSr4g4QfY31onrn62+sp1NuwgY8ivoNsK6nY041bjIDvQPNbM67QSZOYUEaR0saaqMawDGww8/fG1zjU2cONEwmUxGWlqara24uNjo06eP0apVKyMvL8/W/uWXXxpubm7G6NGjK1xj7dq1BmC88sordtffunWrcfjw4Srvv3v3bmPfvn127d9++62Rn59vGIZhvP/++wZgLF++vNJrnDt3zvD39zciIiKMy5cv29rT0qukWikAACAASURBVNIMk8lk/PznP7e1nTp1yggPDzeys7ON5cuXG8HBwVXWZhiGkZ6ebgBGenp6tf0qk52dXetzxEpj57hGOXanPzeMNY8YxkstDePFAOtn+f2G8d8dhlFWVm9lNMqxayA0dnWj8XOcI2OXm5trAEZubq6trcr/fSsuLubixYu1Cmvff/89H3zwAbGxsfTt29fW7uHhQVxcHOfPn+ff/76yzcSqVasoKytj1qxZFa4zduxYQkJCWLlyZYX2y5cvM336dCwWC4cPH7a7/+7duxk+fDhxcXF2x9q2bVvjX8M2bNhAfn4+cXFxuLtf+ZGvb9++DBkyhDVr1lBcXAzAY489xhtvvIHZbK7RtUXkBspMhb9PgD/HwBfrrXsFdh0GP/8IHvkXdL5LewWKSL2oNGCtXbsWX19fAgICuOWWW5g6dSrffvvtdS/2+eefc+nSJaKjo+2Olbelpqba2lJTU3Fzc2PAgAF2/QcOHMjhw4fJzc21tXl4eLB582ZMJhMWi4VDhw7Zju3atYvhw4cTFBRU5XyvmiqvsarvcfHiRdu9t2zZwrBhwzCZTDz66KNkZGTUar6aiNSRYcA3iZAwEpbda12B3eQGd46FJ3fBT9dAUD9XVykiTYzdHKyoqCjGjRtH165dKSgoIDExkeXLl7Nt2zY++eQTbrvttiovlpVlXUemQwf77RDK206ePFmhf2BgIF5e9pNLy/tnZWXRokULW3u3bt1ISkrirrvuwmKxkJiYyNmzZ7n//vsJDg5mx44dtGlTt0UBa/o9wsPDMYwrb8skJCTw0ksv1egtwvz8fPLy8mx/9/LyqnQcRKQKZWXw9eb/396dx9d85Y8ff93sm0gjaaisKkQ1oYkMopqUMTra0JaqVqiWb1s/RVqDtigxtXRaYglG20FHO1QxNTpTu0RrqQiGoZbakiCILEQWWT6/Pz7u5bo3281NciPv5+Ph8eB8lnvu6SnvnOV91ONsLqaoZVa20GGQepxNs0dNfrWcpSeEqCmDAOveESaAwYMHExkZydChQ5k6dSqff/55uS/Lz88HMBooODg46N2j/X15QYWx+7UCAwNJTEwkKiqKyMhI8vPzCQgIYMeOHXh6epZbv6qq7vcwRWRkpN6fJ0yYwMSJEyt8Jjs7u0af2ZhJ25nO4tqurAS7Uz/gcGAJNlmnAFBsHChqP4jC0P+jrMmdDShZWSa9fuP/rjL9xzOU3pNpYGnSWaY804rox6v3w5vFtV0DIm1XM9J+pjOl7e4dMNGqUh6sIUOG8NFHH+mtnzLGyUn9Ca+oqMjgWkFBgd492t9fvXrV6LuM3X+v1q1bExcXx4gRIwCYP3++WYKrez+zqKgIR0f9xHgV1WvYsGEGuyTLk5SURMeOHXV/ruoIlqz1Mp20neksou1KiuDwP2D3PMg+r5bZu8Lv/g9N55E4uHjiUMOPSMvKZ/qms3rBFajHvPx501l6PF79kSyLaLsGqj7a7kEavZS+Z7rqtt2967V1ZVV92N/fn927d1d4T8uWLQH9aUAtY9NuLVu25Pjx4xQVFRkEF9r7te+83/bt2xk9ejQdOnQgOzubl19+me3btxMSElLVr1Sl7xEYGFjp9zCFi4sLrq6uNXqHEI3C7VuQskI9zubmZbXMqZl6nE34CHB0M9tHyVl6jdvalHSDw6qXJp1ldv8QBoTV7O980fhUKQmMoij89ttvNG/evML7goODsbe3N5pjSlsWHh6uKwsPD6esrIx9+/YZ3L93717atGmjt/5Ka+vWrURHRxMUFMSOHTtITEzE2dmZHj16cPjw4ap8pQpp61je93BxcSEoSP6SFaJWFWRD0l8g/nHY/KEaXDV5BJ6ZDbFH1VxWZgyuQM7Sa8zSsvINgitQA+v31x0hLatmy0JE46MXYF25csXoTQsXLiQ9PZ2+ffvqyjIzMzlx4oTeLj9nZ2f69etHUlISKSkpuvKSkhIWLlyIm5sbzz77rK580KBBaDQa5s6dq/d569ev5/z588TExBjUZcuWLfTt25fHHnuM7du34+7uTkBAAElJSbi6utKzZ08OHjxYzWbQ169fP5ycnFiwYAElJSW68gMHDrBr1y4GDhyInZ1djT5DCFGOvKuwdSrEB8POGVCQBe6toO9CGHsYuowEO/MkIL6fnKXXeFVl9FKI6tCbIpw1axbbtm3jueeew8/Pj4KCAhITE9m4cSOBgYF6WcoTEhKIi4szyM4+a9Ystm/fTu/evXWZ3FeuXMnBgwf58ssv9abF2rdvz9ixY5k3bx7R0dH069ePc+fOER8fT7t27Xj33Xf1KqvNgxUcHMyWLVtwc7v706ufn59ud+Ho0aMNpjMvXLjAypUrATh27BgAGzdu1E1n9u3bVze96OHhwcyZM4mNjSUqKoohQ4aQmZlJfHw8Xl5eTJ8+3dT2FkKUJycVdi+AQyuhpFAte/gx6D4OHnserGv/6FQ5S6/xktFLYW56f2P16NGDEydO8PXXX5OZmYlGo+HRRx9l0qRJjB8/3uh03f1atWrFnj17+OCDD/jss88oLCzk8ccf57vvvmPAgAEG98+ZM4eAgACWLFnCqFGjeOihh4iJiWHmzJm4uOifaG9ra8vmzZvx9PQ0WhdfX1+SkpKMLjY7d+4cU6ZM0Stbv369LmeWt7e33vqtsWPH4uHhwZw5c4iNjcXJyYlevXoxa9ascteFCSFMcO2UunD9yLdQdmfEuGUndQowsHedHmejPUvv/XVH9IIsOUvvwSejl8LcNMq9iZxErTt48CBhYWGkpKQQGhparWezsrJkV4iJpO1MV2ttd/m/ag6r4/8C7vw1FBCpjlgFPFWvGde1O8lqepae9DvT1XXbpWXl8/RnieWOXu78U1SDCrCl75nOlLa7ceMGTZs2JTc3VzdTV/tj7kIIca8Le9XA6retd8vaPgvd3wPvTuU/V4d83J1qtFtQG6D9djmH1i3cGvRW/8ZCRi+FuUmAJYSofYoCZ7bDrjmQemd3rvY4myffBa/2Bo801HxE92/133ziumz1byAGhHnTOcDdLKOXQkiAJYSoPWVlcGKjOmJ1+b9qmZUtdHwVuo0t9zibhpqPqLKt/p0D3OUfawtX09FLIbQkwBJCmF9pMRxdCz/PhUz1OBtsnSDsdeg6CpqWv1GkIQcpkqhUCKElAZYQwnyKC9U0C7sXQO6dvEH2TaHzm9D5bXD2qPQVDTlIka3+QggtCbCEEDVXdBMOLIO9iyDvTsJiJw91tCp8ODhUnuJFqyEHKbLVXwihJQGWEMJ0+Vnwy1L45a9QmKOWuXqr66ueiAG76k/lNeQgRRKVCiG06i6DnxDiwXEzAzZPUs8JTJqtBlfNWkO/RTDmkDolaEJwBWqQYmNlPAeWpQcp2q3+99dftvoL0fjICJYQouqyz8Pu+XDoGygtUsu8guGpcdCuL1hZ1/gjGno+onu3+p/JyOHR5pIHS4jGSAIsIUTlrp6An+Ph6HeglKplPp2h+58gsJfZs6439HxE2q3+kk1biMZLAiwhRPkuHcJl+2w4s+lu2aM91ONs/LrV6nE2ko9ICNGQSYAlhNCnKHBhD/z0GZzZgZ22vF00PPketKzeGZpCCNEYSYAlhFApCpzeqmZdT9unlmmsKWoTjX3PD+BhGU0SQoiqkgBLiMaurBR+/ZcaWGUcVcus7dQ0C93GcktxxV7WEQkhRLVIgCVEY1VaDEfWqIvXr59Wy2ydIfwN6PoONGmulmVl1V8dhRCigZIAS4jGprgADq6EPQsgN00tc3BTj7Lp/BY4yWiVEELUlARYQjQWhTfgwN/U42xuXVPLnB+GiHeg0xtg38SsH5eWlc/q5FRSswrwbWBpFoQQoqYkwBLiQXfrunqUzf6lUJirljX1hSfHQscYsHUw+0euTUln4rojeoc2L006y+z+IQwI8zb75wkhhKWRAEuIB9WNS+po1YFlUJyvlnm0UVMtBA8Aa9ta+di0rHyD4AqgpEzh/XVH6BzgLiNZQogHngRYQjxoss7B7nkoh/6Bpuw2ABnObbGJHI9Hp/5gVbtHkK5OTjUIrrRKyhRWJ6dKAlEhxANPAiwhHhRXjqs7Av+3FpQyNMAvZUEsLulHUmEINt9bMdv6Uq1P0aVmFVR4Pa2S60II8SCQAEuIhu5iCvw0F078oCtKLOvIouK+JCt3R4rqaorO192xwus+lVwXQogHQe3OFQghaoeiwLld8Pd+8EWPO8GVBh7rx1chf2fY7Ql6wZWWdoquNg0K98XGyvgZhTZWGgaF+9bq5wshhCWQAEuIhkRR4OQm+Nsf4KtoOJsIGmvo8CqM2g8D/86BoooDmNqeovNxd2J2/xCDIMvGSsMn/UNkgbsQolGQKUIhGoKyUjj+vToVeOV/apm1PYQOhYjR8JCf7lZLmKIbEOZN5wB3VienkpZVgI/kwRJCNDISYAlhyUpuw5Fv1cXrWWfUMjsXCB8OXUZBEy+DRwaF+7I06SwlRnby1eUUnY+7k+wWFEI0WhJgCWGJbufDwb+rx9ncuKiWOT4EnUdC5zfV35dDO0X3/rojekGWTNEJIUTdkQBLCEtSmAvJX8LexZCfqZa5NFenAcOGgb1LlV4jU3RCCFG/JMASwhLcyoR9S2D/F1B05zgbNz94MlZdwG7CcTYyRSeEEPVHAiwh6lPuRdizEFJWQMmd3X2eQepxNo/3B+v6+19Ue1jzb5dzaN3CTUbAhBCiGiTAEqI+XD8Du+fB4VVQVqyWPfIEdP8TtO1T68fZVOb+w5o3n7guhzULIUQ1SIAlRF26ckxNtXBsPShlapl/d+j+HrR6GjTGE3TWJTmsWQghak4CLCFqkXaazeriAfrdXE3r7J/uXgzsDd3HgW/nCp9NzSrAtw4XqcthzUIIUXMSYAlRS9YeSGPDP1fxttX3dLM+BkCZouFiy974RE+CFiHlP3vfFB1QZ1N0clizEELUnARYQphbWRnXDm6g9b/+zEpbNTlosWLN+tLu/LU0mrTzj7DTvjU+5Txe31N0lpAJXgghGjo5i1AIcyktgSPfwV+74fnDMDpanaFQsWV5SW8ii+KZWPIm55QWlR64XJUputokhzULIUTNyQiWEDVVUgT/XQU/z4PscwAUWDmz7HZPlpX8kes0NXikomm2+p6ik0zwQghRcxJgmUleXh4dOnSgtLSU8+fP13d1hJlUuND89i01f9WehXDzslrm1Ay6jOSLm1HM/elque+taJrNEqbo7s0EfyYjh0ebSx4sIYSoDr0pwlOnTvHRRx/RpUsXPD09adKkCR07dmTGjBncunWrSi+MiopCo9EY/fX9998b3K8oCgkJCbRv3x4HBwdatGjBW2+9xfXr183zDe/Yv38/Y8aMoVu3bjg7O6PRaFixYkWFz6xatYqwsDAcHR3x8PDglVde4cKFC0bvHTNmDK1atTJrnUX9WpuSTtRniSzaeYaN/73Eop1nePqzRDbsPQZJn0L847D5QzW4avIIPDMbYo/CU+N5oWt7k6fZLGWKTpsJfmZ0G8b3DpLgSgghqkEvwFq2bBlz584lICCAKVOm8Omnn9K2bVsmT55MREQEBQVVm5rw8PBg5cqVBr86depkcO/48eMZPXo0rVq1YuHChbz++uusXLmSqKioKgd1VfGf//yHRYsWkZOTQ8eOHSu9PyEhgVdffRVHR0fi4+OJjY1l69atREREcOnSJb17//nPf5KXl8fgwYPNVl9Rv4wtNPcgl3FW/6Dnpp6w82MoyIKHAiB6AYw9DF1Ggp0zcHea7f5AqSrTbDV5VgghhIVQ7pGcnKxkZ2cr95s0aZICKAkJCQbX7hcZGan4+flVep+iKMrx48cVKysrpW/fvnrla9euVQDlz3/+s8EzmzdvVk6ePFnuO3fv3q0cOHDAoDwjI0PJy8tTFEVRvvvuOwVQli9fbvQdmZmZiouLixIaGqoUFxfrypOTkxWNRqMMHz5cV3bp0iUlJCREuX79urJ8+fJKv3tKSooCKCkpKRXeZ8z169er/YxQVbft/rLpV8Vv4g+K38QflIiJy5XlkwYoBR81U5Sprooy1VW5+kmoohz5TlFKiit8T+r1W8pfNv2qjP7HQeUvm35VUq/fqnIdavKsOUm/M520nemk7WpG2s90prRdbm6uAii5ubm6Mr01WMZGmAAGDhzIjBkzOHr0aJUDt7KyMvLy8nBxccGqnGM/Vq1aRVlZGe+9955eef/+/fH39+frr79m8uTJuvLi4mJGjhxJQUEBO3fupG3btnrP7d69mz/+8Y8EBweze/duvWteXl5VrvuGDRvIy8tjzJgx2NjcbaJOnTrx1FNPsWbNGhYvXoydnR1vvPEGn376Ke7u7lV+v7B8qVkFtNJcYqT1v3jeeje2mlIADpW1JqGkHy5tnmV+cFil76nJgctyWLMQQjRcVUrTcPHiRQAefvjhKr304sWLuLi40LRpU5ydnenTpw8HDhwwuG///v1YWVnRpUsXg2tdu3bl5MmT5Obm6spsbW358ccf0Wg0REVFceLECd21n3/+mWeeeQYfHx/Wr19fpXqWZ//+/QBEREQYXIuIiODmzZu6z960aRO9e/dGo9Hw+uuvc+HCBTQaDTk5OTWqg6hHl//LqMw/s81uPC/Z7MJWU8rPpe155fYkXrgdx/ayMLybOdd3LYUQQliwSncRlpaWMn36dGxsbKq0xsjf35+IiAiCg4Oxt7fn0KFDLFiwgG7duvHjjz/So0cP3b0XL17Ew8MDe3t7g/d4e3vr7mna9O429zZt2pCYmMjTTz9NVFQUO3fu5Nq1azz77LP4+fmxY8eOKgeC5dEGlNo6GKtXeno6ISEhKMrdNTorVqxg2rRpVdpFmJeXx40bN3R/tre3N9oOog6l7oNdn8FvWwkC0MCW0jAWl/TjsNJad5vkghJCCFGZSgOsMWPGsG/fPj7++GODKTlj7t+Z9+KLLxITE0NoaChvv/02p06d0l3Lz88vN6hwcHDQ3XO/wMBAEhMTiYqKIjIykvz8fAICAtixYweenp6V1rEy2s80VreK6lUdkZGRen+eMGECEydOrPCZ7OzsGn1mY1Zu2ykKNqk/4Zi8CNtL6silorHidptoNjUdyLifrShR9HNBffTMozhTSFZWYV1Uvd5JvzOdtJ3ppO1qRtrPdKa03b0DJloVBliTJ09m8eLFjBgxgg8//LDaH6jVtm1bBg4cyIoVKzh9+jSBgYEAODk5cfWq8VxB2h2LTk7Gd0y1bt2auLg4RowYAcD8+fPNElzd+5lFRUU4OurnHKqoXsOGDWPYsGFV+oykpCS93YxVHcGStV6m02u7sjI48QP8NAcuH1bLrGyh46touo3Fvtmj9ANCu6h5sNKyCvCpwwOXLY30O9NJ25lO2q5mpP1MV922u3e9tq6svJunTZvGjBkzGDp0KEuXLkWjMZ6Xp6r8/f0BuHbtmi7AatmyJcePH6eoqMgguNBO07Vs2dLo+7Zv387o0aPp0KED2dnZvPzyy2zfvp2QkPIP0K0q7Wemp6fr6np/vYxNH1aHi4sLrq6uNXqHMEFpMfxvHfw0FzJPqmW2ThD2OnQdBU31+5ssNBdCCGEKo4vc4+LiiIuLIyYmhuXLl5e7C7A6Tp8+DUDz5s11ZeHh4ZSVlbFv3z6D+/fu3UubNm301l9pbd26lejoaIKCgtixYweJiYk4OzvTo0cPDh8+XOO6hoeHA7Bnzx6Da3v27MHFxYWgIPlHt0EpKYLkv8HCUPjnW2pwZd8UnhqvJgd9ZqZBcCWEEEKYyiBymj59OtOmTWPw4MGsWLGi3ODq8uXLnDhxQm8tUnZ2Nrdv3za498CBA6xZs4b27dvrZTsfNGgQGo2GuXPn6t2/fv16zp8/T0xMjMG7tmzZQt++fXnsscfYvn077u7uBAQEkJSUhKurKz179uTgwYNVbwEj+vXrh5OTEwsWLKCkpETve+zatYuBAwdiZ2dXo88QdaQoD3YvwG1Fd/j3e5CTCk4e0HMqvPs/6DEZnD3qu5ZCCCEeMHpThIsWLWLq1Kn4+vrSq1cvVq1apXezl5cXvXr1AuCDDz7gq6++YufOnURFRQHquqK33nqLl156idatW2Nvb8/hw4dZsWIFtra2fPHFF3rva9++PWPHjmXevHlER0fTr18/zp07R3x8PO3atePdd9/Vu1+bBys4OJgtW7bg5uamu+bn56fbXTh69GiDPFgXLlxg5cqVABw7dgyAjRs3kp6eDkDfvn1104seHh7MnDmT2NhYoqKiGDJkCJmZmcTHx+Pl5cX06dOr39KibuVnwf7PKd27BOuiHKyAG3ZelHYdzUPdhoNd41tHJYQQou7oBVjJyckApKamGl2sHRkZqQuwjGnbti2RkZFs2rSJK1euUFRUxCOPPEJMTAzvv/++wXomgDlz5hAQEMCSJUsYNWoUDz30EDExMcycORMXFxe9e21tbdm8eTOenp5Gpw59fX1JSkoyutjs3LlzTJkyRa9s/fr1upxZ3t7eeuu3xo4di4eHB3PmzCE2NhYnJyd69erFrFmzyl0XJizAzQzYuwgOLIPbeVgDZ8pa8NfSaL4vfBJlqy2zXbMYECYBlhBCiNqjUe5N5CRq3cGDBwkLCyMlJYXQ0NBqPZuVlSW7QsqTfQF2z4dDX0NpEQDHy/xYVNKPH8t+R9k9s+E2Vhp2/imqUe4GNIX0O9NJ25lO2q5mpP1MZ0rb3bhxg6ZNm5Kbm6vbwFZpHiwhLNq1k/BzPBxZA4p6nA0+nVnr/DJ/OuwFGO5+LSlTWJ2cKrsDhRBC1BoJsETDdOmQmsPq1x+AO4Owj/aA7uPArxtJqw8Dl8p9PC2roE6qKYQQonGSAEs0LBf2qMfZnNl+tyzoOej+HrS8e/iyr7ujkYfv8qnkuhBCCFETEmAJy6co8Ns2dcQqda9aprGG4AHw5LvwcDuDRwaF+7I06SwlZYZLDOUsQSGEELVNAixRq9Ky1KNmUrMK8K3uUTNlpfDrRjWwyjiillnbQcfB0G0suAeU+6iPuxOz+4fw/rojekGWjZWGT/qHyAJ3IYQQtUoCLFFr1qakM3HdEUrvCXCWJp1ldv8QBoRVcNRQaTEc/U49zua6egJAkcaBI81f5JE/jqelb6vyn73HgDBvOge4szo5lTMZOTza3K3RniUohBCibkmAJWpFWla+QXAF6g6+99cdoXOAu2GgU1ygplnYPR9y0wDIVZxZUdqb5SW9yTnXBJu/nmB2f7uKA7R7aM8SlC3LojGp0cixEMIsJMAStWJ1cqpBcKVlkCah8IaaGHTvIrh1FYBSJ08+u/F7Vpb0JA8nvWfLDdCEEKaPHAshzKrmpzgLYURqJWkQ0rIK1ONsds6EeY/DtqlqcNXUF/p8xoLH17KkJFovuNLSBmhCCH2VjRynZeWX86QQwtxkBEvUiorSJDxMNkNu/ADxG6D4llrYLFBNtRD8EljbcnbVoQrfL3mshDBUrZFjIUStkgBL1ApjaRJ8NFd42/oHXrJOwu5yiVrYPERNDtouGqysdfdKHishqq9KI8dCiDohAZaokKmLZe9Nk+CvpPH/bP5FX6s92GjK1Bt8u0L3P0HrnqAxPM5G8lgJUX3yg4kQlkMCLFGumi6WHdD8Cn3a/R2nMz/qygr8nsaxxwTwi6jwWcljJUT1yQ8mQlgOCbCEUSalWQA16/r5n9XkoGd33lmiroF2z0H3cTg+8kSV63BvHqu0rAJ8ZLu5EBWSH0yEsBwSYAmjqr1YVlHg9BY1sEr7RS3TWEPIQPU4G8+2JtVDm8dKCFE18oOJEJZBAixhVJUXy5aVwvENatb1K0fVMmt7CB0CEWPgIb9arqkQ4n7yg4kQ9U8CLGFUZYtl/dxs4OBK+Dkess6ohXYu0OkN6DoKmjSvg1oKIYQQlkkCLGFUeYtlHSjiFZtExh7bCr9culPoBl1Gwu/eBCc5jkYIIYSQAEsYdf9i2SbkM8R6K8NtfqSZ5gbkAS5eEDEawoaBfZP6rrIQQghhMSTAEuUaEOZNV68yUn+cS4fLa3Aqu5N13c0XusVCx8Fg61C/lRRCCCEskARYwrjci7A3gZYpK2hZfOf8Ms8gePI9eLw/WEvXEUIIIcoj/0oKfdfPwO55cHgVlBWrZS06qsfZBD0HVnI+uBBCCFEZCbCE6soxNdXCsfWg3DnOxu9J9QDmR3sYPc5GCCGEEMZJgNXYpR9Qk4Oe/M/dssA/qCNWvl3qr15CCCFEAyYBVmOkKHAuSQ2szu26U6iBx/qpgVWLkHqtnhBCCNHQSYDVCKRl5avHZly/RZQmhWdzVmF/5aB60coGQgbBk7HgEVi/FRVCCCEeEBJgPeDWpqQzcd0R/sge3rH5niCrNABKreyx7vSaepyNm08911IIIYR4sEiA9QBLy8pn4rojRPBfEuwWAnBTceTr0t+zoqwPa7s8j4+bHAArhBBCmJsEWA+w1cmplJYpnMKbTaXhHCvz46vSP3ADF911ORBWCCGEMD8JsB5gqVkFAFzBnbeL3zW4nnbnuhBCCCHMS7JGPsB83R0rvO5TyXUhhBBCmEYCrAfYoHBfbKyMJwi1sdIwKNy3jmskhBBCNA4SYD3AfNydmN0/xCDIsrHS8En/EHzcZYG7EEIIURtkDdYDbkCYN50D3NU8WFkF+Lg7MijcV4IrIYQQohbJCFYDUVRUxCeffEJRUVG1n/Vxd2J87yAWvPIE43sHNbrgqiZt19hJ25lO2s500nY1I+1nOnO2nUZRFMUMdRJVdPDgQcLCwkhJSSE0NLTKz924cYOmTZuSm5uLq6trLdbwwSNtZzppO9NJ25lO2q5mpP1MZ2rbGXtORrCEEEIIIcxMAiwhhBBCCDOTRe51rKBATe7566+/Vuu5vLw8AA4fPoyLi4vZ6/Ugk7YznbSd6aTtTCdtVzPSfqYzte20z+XnKjwLdwAACIJJREFU5+umCGUNVh375ptviImJqe9qCCGEEMLMkpKSeOqppwAJsOpcZmYmmzdvxt/fH0dHyaQuhBBCNHSKonDr1i3CwsJwdnYGJMASQgghhDA7WeQuhBBCCGFmEmAJIYQQQpiZBFhCCCGEEGYmAZaFW7VqFWFhYTg6OuLh4cErr7zChQsX6rtaDYJGoyn3V05OTn1XzyLMmjWLl156iVatWqHRaPD396/w/itXrvDGG2/g5eWFg4MDISEhfPHFF3VTWQtTnbabNm1auX0xNja27iptAU6dOsVHH31Ely5d8PT0pEmTJnTs2JEZM2Zw69Ytg/ulz91VnbaTPmfo5MmTDB48mHbt2tG0aVOcnZ1p164d48aNIyMjw+D+mvY9yYNlwRISEhg9ejTdunUjPj6ezMxM5s2bx65du0hOTuaRRx6p7ypavO7du/Pmm28alGt3eTR2H374Ie7u7oSGhlYadObk5PDkk09y8eJFYmNjCQgIYMOGDbz55ptcunSJqVOn1lGtLUN12k4rPj4eDw8PvbJ27drVRvUs1rJly0hISCA6OppXX30VOzs7du7cyeTJk1mzZg379u3T7bCWPqevOm2nJX3urvT0dDIyMnjhhRfw9vbGxsaGo0ePsnTpUlatWsWhQ4fw8vICzNT3FGGRMjMzFRcXFyU0NFQpLi7WlScnJysajUYZPnx4PdauYQCU1157rb6rYdHOnDmj+3379u0VPz+/cu99//33FUBZt26dXnl0dLRia2urnD17traqaZGq03ZTp05VAOXcuXO1XzELl5ycrGRnZxuUT5o0SQGUhIQEXZn0OX3VaTvpc1X37bffKoAyY8YMXZk5+p5MEVqoDRs2kJeXx5gxY7CxuTvQ2KlTJ5566inWrFnD7du367GGDcft27e5efNmfVfDIrVq1arK937zzTcEBATw4osv6pW/9957FBcX8+2335q7ehatOm13r5s3b1JcXGzm2jQcnTp1ws3NzaB84MCBABw9elRXJn1OX3Xa7l6Nvc9VJiAgAIDs7GxdmTn6ngRYFmr//v0AREREGFyLiIjg5s2bnDhxoq6r1eCsXbsWJycnXF1dadasGSNGjDA61y4qlpGRQVpaGl27djW41rVrVzQaja7PivJ16NABV1dXHBwc6NSpU6MLECpy8eJFAB5++GFA+lx13N9295I+Z6iwsJDMzEzS09PZtm0bI0eOBKBPnz6A+fqerMGyUNr/Yby9vQ2uacvS09MJCQmp03o1JOHh4QwYMIDAwEDy8/PZuXMny5cvZ8uWLfzyyy+0aNGivqvYYFTUH+3t7fHw8CA9Pb2uq9VguLm5MWLECLp160azZs04c+YMCQkJDBo0iNOnTzN58uT6rmK9Ki0tZfr06djY2DB48GBA+lxVGWs7kD5XkS+//JLRo0fr/uzj48NXX33F008/DZiv70mAZaHy8/MB9T/m/RwcHPTuEcbd/xPG4MGDiYyMZOjQoUydOpXPP/+8nmrW8FTUH0Htk9Ify2ds19Zbb71FeHg4cXFxDBkyBD8/v3qomWUYM2YM+/bt4+OPP6Zt27aA9LmqMtZ2IH2uIs8//zxBQUHk5eVx6NAhNm7cqDc9aK6+J1OEFsrJyQmAoqIig2sFBQV694iqGzJkCP7+/vz73/+u76o0KBX1R1D7pPTH6nF0dGT8+PGUlJSwZcuW+q5OvZk8eTKLFy9mxIgRfPjhh7py6XOVK6/tyiN9TuXt7c3vf/97nn/+eeLi4lixYgUTJkxg1qxZgPn6ngRYFqply5YARochKxq+FJXz9/fn2rVr9V2NBqWi/lhYWMj169elP5pAmzursfbHadOmMWPGDIYOHcrSpUvRaDS6a9LnKlZR21Wksfc5Y0JCQnjiiSdYvHgxYL6+JwGWhQoPDwdgz549Btf27NmDi4sLQUFBdV2tBk9RFH777TeaN29e31VpUJo3b463tzd79+41uLZv3z4URdH1WVF1p0+fBmiU/TEuLo64uDhiYmJYvnw5Vlb6/xxJnytfZW1Xkcbc5ypSUFBAVlYWYMa+Z9ZkEsJsrl27pjg5OZWbB+uNN96ox9pZvoyMDKPl8+fPVwBl1KhRdVwjy1dZLqcJEyaUmxfGxsZGLy9UY1NR2xUXFyuZmZkG5dnZ2UqrVq0UOzs7JT09vZZraFni4uIUQBk8eLBSUlJS7n3S5wxVpe2kzxl3+fJlo+U7duxQrKyslB49eujKzNH3NIqiKOaLAYU5zZ8/n9jYWLp168aQIUPIzMwkPj4eW1tbDhw4oBvGFIZiY2PZtm0bzz33HH5+fhQUFJCYmMjGjRsJDAxkz549BtmNG6OVK1fqjl5auHAht2/fZty4cYC6C+mdd97R3ZudnU2nTp3IyMjQy2z8ww8/MGXKFKZPn14v36G+VLXtcnJyaNGiBS+++CLBwcE0a9aMs2fPsmzZMq5evcq8efMYO3ZsvX2PurZo0SLeeecdfH19mT59OtbW1nrXvby86NWrFyB97n5VbTvpc8a98MILXL58mR49euDn50dhYSEpKSmsXr0aJycnEhMT6dixI2CmvleTaFDUvq+//lp54oknFAcHB8Xd3V15+eWXG132YlNs2LBB6d27t9KyZUvF3t5ecXBwUNq3b69MmjRJycnJqe/qWYzIyEgFMPrL2IjMpUuXlGHDhimenp6Kvb290r59e2XJkiV1X3ELUNW2KywsVIYPH64EBwcrbm5uio2NjeLp6ak899xzyrZt2+rvC9ST1157rdx2A5TIyEi9+6XP3VXVtpM+Z9y3336r9OnTR/H29tb9u9C2bVvlnXfeUS5cuGBwf037noxgCSGEEEKYmSxyF0IIIYQwMwmwhBBCCCHMTAIsIYQQQggzkwBLCCGEEMLMJMASQgghhDCz/w9dFMeEUCp2fQAAAABJRU5ErkJggg=="
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "PyPlot.Figure(PyObject <Figure size 600x400 with 1 Axes>)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "┌ Warning: implicit broadcasting in setindex! is deprecated; use `df[row_inds, col_ind] .= Ref(v)` broadcasting assignment to change the column in place\n",
      "│   caller = Pyr_length_node_degree!(::DataFrames.DataFrame, ::MetaGraph{Int64,Float64}) at modules_SDs.jl:529\n",
      "└ @ Main.PyrSDTypes /Users/abdullae/julia_wd/object_oriented_SDs/modules_SDs.jl:529\n"
     ]
    }
   ],
   "source": [
    "PyrSDTypes.Real_SD_length_distr(a.int_tree, b.int_tree, a.int_DF)\n",
    "PyrSDTypes.Pyr_length_node_degree!(b.int_DF, graph_obj.mg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Only long segments genome fraction:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.04709354583333333"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"All segments genome fraction:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.04914383993055556"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmgAAAGgCAYAAAAevJsNAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAAPYQAAD2EBqD+naQAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzs3XlYlXX+//HnYTmgsiiYEqFiGpKm4opSLpmW5ZapZeGUaa4o+qsGW2ZynBbNmikXVFQmy20sLZcsS8MwN0xwySE1HRc0V1AWF5Dl98f9lVFzYTlwc+D1uK7zh+ec+3O/z1yfodf1uT+LJS8vLw8RERERKTMczC5ARERERK6ngCYiIiJSxiigiYiIiJQxCmgiIiIiZYwCmoiIiEgZo4AmIiIiUsYooImIiIiUMQpoIiIiImWMApqIiIhIGaOAJiIiIlLGOJldQEFcuHCB+Ph4AgICqFy5stnliIiIiBRYXl4e6enp+Pr64uBQsLExuwho8fHxdOjQwewyRERERIosKSkJPz+/An3XLgJalSpVAIiNjSUoKKhIbaSkpODl5WXLssTOqU/ItdQf5EbqE3Kt4vSHtLQ0atWqhbu7e4GvsYuAZrFYAHBzc8PDw6NIbWRnZxf5Wimf1CfkWuoPciP1CbmWLfrD1TxTEFokICIiIlLG2MUImoiIiJRPmZmZ7Nu3j+zsbLNLua3U1FQ8PT1v+bmjoyN16tShatWqNrmfApqIiIiY4uDBgwQFBZGRkWF2KTYzZMgQZs2aVeDVmreigCYiIiKlLjc3l8GDB1O9enW+/fZbu99GKysriw0bNvDaa68BMHv27GK1p4AmIiIipe7EiRPExsayaNEiHnroIbPLsYk2bdoAMG7cOCZPnlysx50VfpFAckYmmw6cJTkj0+xSREREKowzZ84AUK9ePZMrsa327dsDcOTIkWK1U6FH0BbGHWHCykSycnKxOjowvmdDQoPrmF2WiIhIuZebmwuAk1PhokhObh7bDqVwOv0yNdxdaV3XC0eHgm9fUdKsVisAOTk5xWqnwga05IzM/HAGkJWTy4RViXRt5IO3m4vJ1YmIiMiN1uw5wYRViZxIvZz/3t2erozv0ZCuD9xdYvc9fPgwAwcOZMeOHdx3331s3769xO51VYV9xLn3ZHp+OLsqKzuXfSfTTapIREREbmXNnhOMWJBwXTgDOJl6mRELEliz50SJ3dvDw4N33nmHRYsWldg9blRhA1qgjztWx+t/vtXJgcC7tWu0iIhIWZKTm8eEVYnk3eSzq+9NWJVITu7NvlFwH3zwAcOGDcv/9/nz56levToADz30UP7Rk6WhwgY0bzcXxvdsiNXJ+J/A6uTA+B4N8apiNbkyERERuda2Qyl/GDm7Vh5wIvUy2w6lFOs+Q4YMYfny5aSmpgIQHR1Nr169TDmTtWLMQTsej/PxfdCiHzg6578dGlyHro182HcynQY+7pp7JiIiUgadTr91OCvK926latWq9OnTh3nz5hEeHs7MmTP54osvitVmUVWMgPb5C7inJsGWydAhAho/DY7GT/d2cyGkvoKZiIhIWVXD3dWm37ud8PBwnnzySerVq0fNmjVp1qxZsdssiorxiLPz38it5A3nDsHyERDZGnYtgdziLYEVERGRkte6rhd3e7pyq800LBirOVvXLf6jyMDAQPz9/RkxYgSjRo0qdntFVTECWuO+nB+4ATpPgEpekHIQvhoKM9rAL0sV1ERERMowRwcL43s0BPhDSLv67/E9GtpsP7QhQ4aQnZ1N3759AeNAdz8/P/r168fu3bvx8/Pj9ddft8m9bqViBDQA58rw0FgYuxseeQsqVYOz+2HZYJgZAnu+hNzcO7cjIiIipa7rA3czc0BzfDyvf4zp4+nKzAHNbboP2g8//MDIkSNxdjbmrbu4uHDs2DHOnDlDVlYWx44dY+LEiTa7381UjDlo13Jxh3avQKshEBcFW6bBmb2w9EWo8QF0fA0Ce0AxT6EXERER2+r6wN10aehTYicJ/P7773Tq1AkvLy/ef/99m7RZVBUvoF3l6gEd/gzBQ2HrTNgyA04nwufPQ80HoOPrENgNLGXn+AgREZGKztHBQtt63iXStq+vL3v37i2RtgtLw0Sunsao2dhd0D4CrO5wag8sCYWo9rDvW8gr3sZ3IiIiIoWhgHZVpWrQ6U1jjlq7V8DqBid3w+L+MOdh2P+9gpqIiIiUCgW0G1X2MhYRjNkND441Fhf8vgMW9YO5neHAOgU1ERERKVEKaLdSxRu6TDCCWshocKoEx7fDgj4Q/SgcXK+gJiIiIiVCAe1O3O6CR9+BMbugTRg4ucKxbTD/SfjkcTi0wewKRUREpJxRQCso95rQ9T0jqAUPB0cXOLoFPu0B87rD4U1mVygiIiIlICYmhuDgYBo2bMgDDzzAm2++SV4JP0VTQCssdx94/H0Ys9PYS83RCod/gnlPwKc94ehWsysUERERG6pWrRqLFy8mMTGR7du3Exsby+LFi0v0nhV3H7Ti8vCFbh8apxP89A9ImA+HYo1XvU7Q8Q2o1crsKkVERMq+vDy4crFk7+Fc+Y57m37wwQccOHCAqKgoAM6fP0/9+vXZv38/Xl7GOZ+urq4EBQXx3//+t0TLVUArLk8/6P4RPPT/YMOHsHMhHIwxXvW7wMOvwz0tzK5SRESk7LpyEd7zLdl7vPE7WKvc9itDhgyhQYMGTJ48GU9PT6Kjo+nVq1d+OAM4efIkS5cu5ZtvvinRcvWI01aq1oaeU2F0PDQbABZHOLAW5nSCRc8YW3WIiIhImVW1alX69OnDvHnzyMvLY+bMmYwaNSr/87S0NHr06EFERATNmzcv0Vo0gmZr1fyhVyQ89LIxorb737B/jfFq0M04teDuJmZXKSIiUnY4VzZGuEr6HgUQHh7Ok08+Sb169ahZsybNmjUDID09na5du9KzZ09efvnlkqwUKAMBLTMzk2HDhuHh4UFOTg6RkZFml2Qb3vWg90zjVIINk+GXL2DfauN1fw/jrM+ajcyuUkRExHwWyx0fP5aWwMBA/P39GTFiBJMnTwYgIyODrl278thjj/HXv/61VOqw6SPO8PBw/P39sVgs7Nmz57rPfvvtN0JCQggICKB169YkJiYC8OWXX9KhQwemTp1KtWrV2LJliy1LMl/1+vDUbBgZBw/0BSzw6yqYGQKfvwCnfzW7QhEREbnGkCFDyM7Opm/fvgBMmTKFbdu28dVXXxEUFERQUBDvvvtuidZg04DWt29fNm7cSJ06df7w2bBhwxg6dCj79+8nIiKCwYMHA3DkyBH8/f0BuPfeezly5IgtSyo77gqAvtEwcgs06m28l7gcZrSFpYPgzD5z6xMREREAfvjhB0aOHImzszMAb775JleuXGHnzp35rzfffLNEa7BpQGvfvj1+fn5/eP/06dMkJCQwYMAAAPr06cOhQ4c4fPgwtWrVyg9lhw8fpnbt2rdsPyMjg7S0tPxXZmamLcsvHTXuh37zYMRmuL8nkAd7lkFkMCwbAmcPmF2hiIhIhfT7778TGBjIzp07GTt2rKm1lMoctKSkJHx9fXFyMm5nsVioXbs2R48epU+fPgwfPpzdu3dz6dIlQkJC8q+LjIwkMjKSS5cuAdChQ4fr2o2IiGDcuHEFquHcuXM2+jU24nw3dJmCY9AwKsVNwfrf7+GXz8nbs5SsBk9yqfVocqv6m11luVbm+oSYSv1BbqQ+UbJSU1PNLuEPfH192bt3r03aSk1NJSUlBTBWfxZWqS0SsNywOdzVIxJcXV2ZN2/eTa8JCwsjLCyMhIQEWrRoQWxsLEFBQfmfu7i44OLiUuAart3HpMzweggaPAS/74QfJ2HZ/y0ue7/EZd8KaPostH8VvOqaXWW5VSb7hJhG/UFupD5Rcjw9Pc0uoUR5enrm95+rA1SFUSr7oNWqVYtjx46RnZ0NGOEsKSnpto8zb8bNzQ0PD4/8V2HCWZnnGwTP/RuGxMB9j0JeDuxcANNbwsrRcK6czs0TEZEKydHREYCsrCyTK7GtixeNExGuzl8rqlIZQatRowbNmjVjwYIFDBw4kGXLluHv75+/OECucU8LCP0Cjm2H9e/BwR8g4TPYudjYALfdK1C1ltlVioiIFIu/vz+urq78/e9/56233sJqtZpdUrFkZ2dz8OBBXn/9ddzd3QkICChWezYNaGFhYaxYsYKTJ0/SuXNn3NzcOHDAmPQeFRXFwIEDee+99/Dw8ODTTz+15a3LH7+W8Kcv4Wgc/Pge/PdHiP/EOEqq+fPGRrie95hdpYiISJF4enqyYsUKevXqxbfffmt2OTbTsWNHYmJiiv2Uz5J3dTJYGXZ1Dlp8fHyRj1ZISUmx77kERzYbI2qHfzL+7WiFFi8aZ4B63G1ubXbK7vuE2JT6g9xIfaJ0pKamcvjwYXJycswu5bZSU1NvO2/OwcGBGjVq4OPjg4PD9TPI0tLS8PT0JDU1FQ8PjwLdz/STBKSA6oTAwK/h0E/w40Q4sgm2RUHCp9ByEDw4Ftxrml2liIhIoXh6etK0aVOzy7ij0g7sOizd3tRtBwNXw/MroFYbyL4MW2fAlKbw3ZuQccbsCkVERKSYFNDskcUC93aEQWtgwJfg1wqyL8GW6TClCax9Cy4km12liIiIFJECmj2zWKD+IzB4LYQuBd/mcOUibJoCHzeGdRPgYorZVYqIiEghKaCVBxYL3NfF2EPt2SVwd1O4cgE2/hM+bgIx78Al7YgtIiJiLxTQyhOLBRp0haGx0H8R1GwMWemw4QMjqK2fCJfOm12liIiI3IECWnlksUBgNxi2AZ6eDzUaQWYaxE4y5qjFTobLhT8XTEREREqHAlp55uAADXvC8I3Qbx7cFQiXU2H9u8YctQ0fQma62VWKiIjIDRTQKgIHB2jUG0Zshj7RUD0ALp+HmLeNR58bP4LMjEI3m5yRyaYDZ0nOyCyBokVERCouBbSKxMERGveFkVvhqTngXR8upcC6vxmPPjdNhayLBWpqYdwR2k6MIXRuHG0nxrAwToe5i4iI2IoCWkXk4AhNnoaRcfDkLKhWFy4mw9q/GkFtSyRcuXTLy5MzMpmwMpGsnFwAsnJymbAqUSNpIiIiNqKAVpE5OkHQszBqO/SKhKp14MIZ+O4N42SCrbPgyuU/XLb3ZHp+OLsqKzuXfSc1n01ERMQWFNDECGrNBsDoeOgxFTxrQ8YpWDMOpgbBtjmQ/b/RsUAfd6yO13cdq5MDgXcX7ABYERERuT0FNPkfR2do8YIR1Lp/BB5+kH4CvnkVpjaDn6MhOwtvNxfG92yI1cnoPlYnB8b3aIhXFavJP0BERKR8cDK7ACmDnKzQchAEhULCZ/DTPyHtOKx+2Vjx2f5VQluG0rWRD/tOptPAxx1vNxezqxYRESk3NIImt+bkAq2HQPgOeHwyuPlAahKsGgPTWuC9/3NC6noqnImIiNiYAprcmbMrBA+DMTvhsYlQpQacPwIrR8H0VrBzEeRkm12liIhIuaGAJgXnXAnajoQxu+DRd6BydTh3CJaPgMjWsGsJ5OaYXaWIiIjdU0CTwrNWhpDRMHY3dJ4Albwg5SB8NRRmtIFfliqoiYiIFIMCmhSdtQo8NNYIao+8BZWqwdn9sGwwzAyBPV9Cbu6d2xEREZHrKKBJ8bm4Q7tXYMxuePgv4OoJZ/bC0hdh1oOQuEJBTUREpBAU0MR2XD2gw59h7C/Q8XVw8YTTifD58xDVHn79GvLyzK5SRESkzFNAE9tz9YSOr8HYXdA+AqzucOoXWBIKszvAvm8V1ERERG5DAU1KTqVq0OlNY45au1fA6gYndsHi/jDnYdj/vYKaiIjITSigScmr7GUsIhizGx4cC86V4fcdsKgfzO0MB9YpqImIiFxDAU1KTxVv6DLBCGoho8GpEhzfDgv6QPSjcHC9gpqIiAgKaGIGt7uMjW7H7II2YeDkCse2wfwn4ZPH4dAGsysUERExlQKamMe9JnR9zwhqwcPB0QWOboFPe8C87nB4k9kVioiImEIBTczn7gOPv2+c9dlqCDha4fBPMO8J+LQnHN1qdoUiIiKlSgFNyg4PX+j2IYTvgJaDwMEZDsXCvx6D+b0h6WezKxQRESkVCmhS9nj6QfePIDwBmr8ADk5wMAaiO8OCvnA83uwKRURESpQCmpRdVWtDz6kwOh6aDQCLIxxYC3M6waJnjK06biM5I5NNB86SnJFZSgWLiIjYhgKalH3V/KFXJIz6GZo+BxYH2L8GZneExc/Bid1/uGRh3BHaTowhdG4cbSfGsDDuSKmXLSIiUlQKaGI/vOtB75kQ9jM0ecYIavtWQ1Q7WDIATv0HMEbOJqxMJCvHOKA9KyeXCasSNZImIiJ2QwFN7E/1+vDUbBgZBw/0BSzw6yqYGQKfv8CRvfH54eyqrOxc9p1MN6deERGRQlJAE/t1VwD0jYaRW6FRb+O9xOU0+/oJplunU89yPP+rVicHAu/2MKlQERGRwlFAE/tXIxD6zYMRm+H+nljIo7vDZtZaI/jIOZIAp5OM79EQrypWsysVEREpEAU0KT9qNoJn5sPwjRDYHQdLHr0dN/Gd86uE/j4Rkg+aXaGIiEiBKKBJ+ePTGPovhKGxEPA4lrxc2LUYpreC5WGQcsjsCkVERG5LAU3KL98geO7fMCQG7nsU8nJg5wKY3hJWjsYh7ZjZFYqIiNyUApqUf/e0gNAv4KUfoN4jkJsNCZ/h+VknWDUWzieZXaGIiMh1FNCk4vBrCX/6EgZ9D/d2xJJ7BeI/gWnNYfUrkHr8jk2IiIiUBgU0qXhqB8PzK0jrswT820FOFvw8F6Y2g28iIO2E2RWKiEgFp4AmFVb2Pa1h4NfwwtdQ50HIyYRtUTA1CNa8DumnzC5RREQqKAU0kbrtYOBqeH4F1GoD2Zdh6wyY0hS+exMyzphdoYiIVDAKaCIAFgvc2xEGrYEBX4JfK8i+BFumw5QmsPYtuJBsdpUiIlJBKKCJXMtigfqPwOC1ELoUfJvDlYuwaQp83BjWTYCLKWZXKSIi5ZwCmsjNWCxwXxdjD7Vnl8DdTeHKBdj4T/i4CcS8A5fOmV2liIiUUwpoIrdjsUCDrsapBP0XQc3GkJUOGz4wgtr6iXDpvNlViohIOaOAJlIQFgsEdoNhG+Dp+VCjEWSmQewkY45a7GS4nFaoJpMzMtl04CzJGZklVLSIiNgrJ7MLELErDg7QsCcEdodfV8CPk+DMXlj/LmyJhJDREDwMXNxv28zCuCNMWJlIVk4uVkcHxvdsSGhwnVL6ESIiUtZpBE2kKBwcoFFvGLEZ+kRD9QC4fB5i3jYefW78CDIzbnppckZmfjgDyMrJZcKqRI2kiYhIPgU0keJwcITGfWHkVnhqDnjXh0spsO5vxqPPTVMh6+J1l+w9mZ4fzq7Kys5l38n0UixcRETKMgU0EVtwcIQmT8PIOHhyFlSrCxeTYe1fjaC2JRKuXAIg0Mcdq+P1/9ezOjkQeLeHGZWLiEgZpIAmYkuOThD0LIzaDr1mQNU6cOEMfPeGcTLB1ll4u+QxvmdDrE7G//2sTg6M79EQrypWk4sXEZGyQosEREqCoxM0CzVG1XYthtgPIPUorBkHmz4mtN0rdH31GfadzaKBjzvebi4FajY5I5O9J9MJLMQ1IiJifzSCJlKSHJ2h+fMwOh66fwwefpB+Ar55Fe9/tSHk3Aq8XS0Famph3BHaTowhdG4cbSfGsDDuSAkXLyIiZjE1oJ07d47BgwdTq1YtM8sQKXlOVmj5IoQnQLd/gLsvpB2H1S/DtOYQPw9yrtzycq38FBGpWAod0MLDw/H398disbBnz57rPvvtt98ICQkhICCA1q1bk5iYeNu2qlWrRnR0NA0aNChsGSL2yckFWr0E4Tvg8cng5gOpSbBqDExrAQnzbxrUtPJTRKRiKXRA69u3Lxs3bqROnT9uqjls2DCGDh3K/v37iYiIYPDgwQCcOnWKzp07X/eKiIgofvUi9srZ1djQdsxOeGwiVKkB54/AylEwvRXsXAQ52flf18pPEZGKpdCLBNq3b3/T90+fPk1CQgLff/89AH369GHUqFEcPnwYf39/1q1bV7xKgYyMDNLS/necjouLCy4umigtdsy5ErQdCS0GwvZo2PgxnDsEy0fAhg+hwzho3BdvNxfG92zIhFWJZGXnauWniEg5Z7NVnElJSfj6+uLkZDRpsVioXbs2R48exd/f/5bXDR8+nL179zJ8+HD+/Oc/U69evfzPIiMjiYyM5NIlY/+oDh06XHdtREQE48aNK1B9586dK+QvkvKuzPWJwFCo1xvX3Z/hGj8bh5SD8NVQcn58n0utw3n8vm60Gd6c385c5L67KlOtsjMpKSlmV11ulLn+IKZTn5BrFac/XDu4VFA23WbDYrl+NVpeXt4dr5k1a9YtPwsLCyMsLIyEhARatGhBbGwsQUFB+Z8XdgTNy8urwN+ViqHs9Qkv6PIGtB8N22bD5mk4njuI23djIGEmXh3GUa/pk8ZRU2JzZa8/iNnUJ+RaRe0PVwevCsNmf+Vr1arFsWPHyM425s3k5eWRlJRE7dq1bXUL3Nzc8PDwyH/p8aaUWy7u0O4VGLMbHv4LuHoah7IvfRFmPQiJKyA3987tiIiIXbJZQKtRowbNmjVjwYIFACxbtgx/f//bPt4UkTtw9YAOf4axv0DH18HFE04nwufPQ1R7+PVrKMBItYiI2JdCB7SwsDD8/Pw4duwYnTt3pn79+vmfRUVFERUVRUBAAJMmTSI6OtqmxYpUWK6e0PE1GLsL2keA1R1O/QJLQmF2B9j3rYKaiEg5YskryEQxk12dgxYfH0/z5s2L1EZKSormEsh17LpPXEyBLdMhLgqyMoz3fJtBxzfgvi5gKdjpBPI/dt0fpESoT8i1itMf0tLS8PT0JDU1FQ+Pgm2PpJnGIvaoshc88pYxR+3BseBcGX7fAYv6wdzOcGCdRtREROyYApqIPaviDV0mGEEtZDQ4VYLj22FBH/jXY3BwfZGDWnJGJpsOnNVxUiIiJlBAEykP3O6CR9+BMbugTRg4uUJSHMx/Ej55HA5tKFRzOphdRMRcCmgi5Yl7Tej6nhHUgoeDowsc3QKf9oB53eHwpjs2oYPZRUTMp4AmUh65+8Dj7xtnfbYaAo5WOPwTzHsCPu0JR7fe8lIdzC4iYj4FNJHyzMMXun0I4Tug5SBwcIZDscb8tPm9IennP1yig9lFRMyngCZSEXj6QfePIDwBmr8ADk5wMAaiO8OCvnA8Pv+rVw9mtzoZfx50MLuISOmz6VmcIlLGVa0NPadCu5dhwwewczEcWGu8Aroam+H6NiM0uA5dG/mw72Q6DXzc8XbTsWoiIqVJI2giFVE1f+gVCaO3Q9PnwOIA+9fA7I6w+Dk4sRtvNxdC6lcvVDjT1hwiIrahgCZSkXndC71nQtjP0OQZI6jtWw1R7WDJADj1nwI3pa05RERsRwFNRKB6fXhqNoyMgwf6Ahb4dRXMDIHPX4DTv972cm3NISJiWwpoIvI/dwVA32gYuRUa9TbeS1wOM9rC0kFwZt9NL9PWHCIitqWAJiJ/VCMQ+s2DEZvh/p5AHuxZBpHBsGwInD1w3de1NYeIiG0poInIrdVsBM/Mh+EbIbA7kAe/fA6RreCr4ZB8ELCfrTm0iEFE7IW22RCRO/NpDP0Xwu874cdJsP9b2LUYdn8OTZ+F9q8SGly3TG/NsTDuSP48OaujA+N7NiQ0uI7ZZYmI3JRG0ESk4HyD4Ll/w5AYuO9RyMuBnQtgektYORrvKycLvTVHadAiBhGxNwpoIlJ497SA0C/gpR+g3iOQmw0Jn8G0FrBqLJxPMrvC62gRg4jYGwU0ESk6v5bwpy9h0Pdwb0fIvQLxn8C05rD6FUg9bnaFgBYxiIj9UUATkeKrHQzPr4AXvwX/dpCTBT/PhanN4JsISDthann2sohBROQqLRIQEdupEwIDv4ZDP8GPE+HIJtgWBQmfQstB8OBYcK9Z7NskZ2Sy92Q6gYVYjKDzRUXEniigiYjt1W0H/g/BoVhYPxGStsLWGbD9E2g12AhqbncVqenirMY0zhdVMBORsk+POEWkZFgsxry0QWtgwJfg1wqyL8GW6TClCax9Cy4kF6pJrcYUkYpCAU1ESpbFAvUfgcFrIXQp+DaHKxdh0xQjqK2bABdTCtSUVmOKSEWhgCYipcNigfu6GHuoPbsE7m4KWRmw8Z/wcROIeQcunbttE1qNKSIVhQKaiJQuiwUadIWhsdB/EdRsDFnpsOEDI6itnwiXzt/0Uq3GFJGKQosERMQcFgsEdoOAx2Hv18YRUqf/A7GTIG4mtB0FwcPB9frRMa3GFJGKQCNoImIuBwdo2NM4kL3fPLjrfricCuvfhY8bw4YPIfP6OWbGasyyd6SUiIitKKCJSNng4ACNesOIzdD3X1A9AC6fh5i3jUefGz+CzAyzqxQRKRUKaCJStjg4wAN9YORWeGoueNeHSymw7m/Gqs9NUyHrotlVioiUKAU0ESmbHByhST8YGQe9o6BaXbiYDGv/agS1LZFw5ZLZVYqIlAgFNBEp2xydoGl/GLUdes2AqnXgwhn47g2Y0hS2zoIrl82uUkTEphTQRMQ+ODpBs1AYHQ89p4Fnbcg4BWvGwdQg2DYHsnWigIiUDwpoImJfHJ2h+fNGUOv+MXj4QfoJ+OZVmNoMfo6G7CyzqxQRKRYFNBGxT05WaPkihCdAt3+Auy+kHYfVL8O05hA/D3KumF2liEiRKKCJiH1zcoFWL0H4Dnh8Mrj5QGoSrBoD01pAwny7CGrJGZlsOnBWB7+LCKCAJiLlhbMrBA+DMTvhsYlQpQacPwIrR8H0VrBzEeRkm13lTS2MO0LbiTGEzo2j7cQYFsYdMbskETGZApqIlC/OlaDtSBizCx59BypXh3OHYPkIiGwNu5ZAbo7ZVeZLzshkwspEsnJyAcjKyWXCqkSNpIlUcApoIlI+WStDyGgYuxs6T4BKXpByEL4aCjPawC9Ly0RQ23syPT+cXZWVncu+k+m3uEJEKgIFNBEp36xV4KGxRlB75C2oVA3O7odlg2FmCOz5EnJz79zTqjcIAAAgAElEQVROCQn0ccfqeP2fYquTA4F3e9ziChGpCBTQRKRicHGHdq/AmN3w8F/A1RPO7IWlL8KsB3E+8K0pQc3bzYXxPRtidTL+HFudHBjfoyFeVaylXouIlB1OZhcgIlKqXD2gw58heChsnQlbZsDpRNy/GQnxM6HjaxDYDSyWUispNLgOXRv5sO9kOg183PF2cym1e4tI2aQRNBGpmFw9jTA2dhe0jyDP2Q1O/QJLQmF2B9j3LeTllVo53m4uhNSvrnAmIoACmohUdJWqQac3Of/iT8YjUKsbnNgFi/vDnIdh//elGtREREABTUQEgDzXqsYigjG74cGx4FwZft8Bi/rB3M5wYJ2CmoiUGgU0EZFrVfGGLhOMoBYyGpwqwfHtsKAP/OsxOLheQU1ESpwCmojIzbjdZWx0O2YXtAkDJ1dIioP5T8InT8Chn8yuUETKMQU0EZHbca8JXd8zglrwcHB0gaOb4dPuMK87HN5kdoUiUg4poImIFIS7Dzz+vnHWZ+uh4GiFwz/BvCfg055wdKvZFYpIOaKAJiJSGB6+8MQHEL4DWg4CB2c4FGvMT5vfG5J+NrtCESkHFNBERIrC0w+6fwThCdD8BXBwgoMxEN0ZFvSF4/FmVygidkwBTUSkOKrWhp5TYXQ8NBsAFkc4sBbmdIJFzxhbdYiIFJICmoiILVTzh16RMHo7NH0OLA6wfw3M7giLn4MTu82uUETsiAKaiMhtJGdksunAWZIzMgt2gde90HsmhP0MTZ4xgtq+1RDVDpYMgFP/KdmCRaRcUEATEbmFhXFHaDsxhtC5cbSdGMPCuCMFv7h6fXhqNoyMgwf6Ahb4dRXMDIHPX4DTv5ZY3SJi/xTQRERuIjkjkwkrE8nKyQUgKyeXCasSCz6SdtVdAdA3GkZuhUa9jfcSl8OMtrB0EJzZZ+PKRaQ8UEATEbmJvSfT88PZVVnZuew7mV60BmsEQr95MGIz3N8TyIM9yyAyGJYNgbMHil2ziJQfCmgiIjcR6OOO1fH6P5FWJwcC7/YoXsM1G8Ez82H4RgjsDuTBL59DZCv4ajgkHyxe+yJSLpga0GJjYxk8eDChoaGMGjXKzFJERK7j7ebC+J4NsToZfyatTg6M79EQrypW29zApzH0XwhDYyHgccjLhV2LYXorWB4GKYdscx8RsUuFDmjh4eH4+/tjsVjYs2fPdZ/99ttvhISEEBAQQOvWrUlMTLxtWx06dCA6OpqFCxdy9OhR0tOL+OhARKQEhAbXYctrnVj0UjBbXutEaHCdAl1XqJWfvkHw3L9hSAzc9yjk5cDOBTC9JawcDecKsTBBRMqNQge0vn37snHjRurU+eMfqmHDhjF06FD2799PREQEgwcPBuDUqVN07tz5uldERET+dd988w33338/7u7uxfgpIiK25+3mQkj96ni7uRTo+0Ve+XlPCwj9Al76Aeo9ArnZkPAZTGsBq8bC+aRi/AoRsTdOhb2gffv2N33/9OnTJCQk8P333wPQp08fRo0axeHDh/H392fdunU3vW7OnDkcP36c999//473zsjIIC0tLf/fLi4uuLgU7I+miEhJu9XKz66NfAoc8PBrCX/6Eo7GwY/vwX9/hPhPYMcCaPECPPQyeN5Tcj9CRMqEQge0W0lKSsLX1xcnJ6NJi8VC7dq1OXr0KP7+/je9ZtmyZfz973+nW7duDB8+nLfffpu77ror//PIyEgiIyO5dOkSYDwSvVZERATjxo0rUH3nzp0rwq+S8kx9Qq5li/6w7UjqTVd+/vzb77Su41m4xtzug+6f4HR8G5XiPsb52Bb4eS558Z+R2fhZLrUYQZ5bzWLXLLemvxFyreL0h2sHlwrKZgENjFB2rby8vNt+v0+fPvTp0+eWn4eFhREWFkZCQgItWrQgNjaWoKCg/M8LO4Lm5eVV4O9KxaA+Idcqbn9oba2C1XHvdSHN6uRA64B7ir64wKsrNO4Kh36CHydiObIJ112f4vqfJdByEDw4FtwLF9SSMzLZezKdQB/3go/sVVD6GyHXKmp/uDp4VRg2W8VZq1Ytjh07RnZ2NmCEs6SkJGrXrm2rW+Dm5oaHh0f+S483RaQsKdGVn3XbwcDV8PwKqNUGsi/D1hkwpSl89yZknClQM8U6HaGQCn1Mlojks9kIWo0aNWjWrBkLFixg4MCBLFu2DH9//1s+3hQRKY9Cg+vQtZEP+06m08DWI1QWC9zbEep2gP+uh/XvwbGfYct02P4vaD0EQsZAFe+bXm6TOXIFtDDuSP69rI4OjO/ZsMCrYEWkCCNoYWFh+Pn5cezYMTp37kz9+vXzP4uKiiIqKoqAgAAmTZpEdHS0TYsVEbEHhV35WWgWC9TrBIPXQugy8G0OVy7CpinwcWNYNwEupvzhMpufjnALNjsmS6QCK/QI2tWJ+zfToEEDtmzZUuyiRESkACwWuK8z1H8Efvse1r8LJ3bBxn/CtjnQZji0DYNK1YD/nY5w4xy5Yp+OcIPbBcGQ+pqaIlIQOupJRMTeWSwQ8JhxKkH/xcYpBVnpsOED+LgJrJ8Il86X/OkI/6fEjskSqUBsuopTRERMZLFA4BMQ0BX2rTaC2en/QOwkiJsJbUcRGjycro06lcwcuf9zNQhOWJVIVnZuiQVBkfJMAU1EpLxxcID7e0CDbvDrSvhxEpz51XgEuiUS75DRhAQPgxJcCV+iiyVEKgA94hQRKa8cHKDRkzBiM/T9F1QPgMvnIeZt49Hnxo8gM6PEbl/iiyVEyjEFNBGR8s7BAR7oAyO3wlNzwbs+XEqBdX+DKU1g01TIumB2lSJyDQU0EZGKwsERmvSDkXHQOwqq1YWLybD2r8aGt5unQ9ZFs6sUERTQRETsWpF263d0gqb9YdR26DUDqtaBC2fg+zdhahBsnQlXLpdc0SJyRwpoIiJ2qtjHNjk6QbNQGB0PPaeBZ23IOAVrXjOC2rY5kG0EPx3bJFK6tIpTRMQO2fTYJkdnaP48NOkPOxfChg8h7Rh88yps/IhttV7kxZ2BXMhxKJVjm3SYu4hG0ERE7FKJHNvkZIWWL0J4AnT7B7j7QtpxWv/nHb5zGkt/xxhyc7JK9Nim0jzMXaQsU0ATEbFDJbpbv5MLtHoJwnfw35ZvcSqvKn6Ws0xynst66yv0yoth/+9/POuzuHSGp8j/KKCJiNihUjm2ydkVz46j6Jw9lb9f+RNn8jyp5XCGD5xn03p1V9i5CHKybXa70jrMXcQeaA6aiIidKo3d+r3dXHitZxATVllZlNmJgc4/MLbSalxTD8PyEcZ8tQ7joHFfYxuPYiitw9xF7IECmoiIHTN26y/ZifTXB8FuuFonGys8N0+FlIPw1VDjYPaOr0Gj3kUOajrDU+R/FNBEROSOrg+CLvDQWGOe2rbZRlBL/g2WDYbYyUZQa/ikcYJBIekMTxGD5qCJiEjRuLhBu5dhzG7o9Bdw9YSz+2DpizDrQUhcAbm5d27nBjrDU0QBTUREisvVA9r/Gcb+Ah3fABdPOJ0Inz8PUe3g11WQl2d2lSJ2RQFNRERsw9UTOo6DsbuNhQNWdzi1B5YMgKj2sPcbBTWRAlJAExER26pUFR5+wwhq7V4Fqxuc3A3/fhbmPAz7v1NQE7kDBTQRESkZlb3gkb8ac9Qe+n/gXAV+3wGLnoa5j8Bv6xTURG5BAU1EREpWFW/o/DdjRC0kHJwqwfF4WNgHoh+FgzEKaiI3UEATEZHSUaU6PPq2EdTajgInVzi2Deb3hk8eh0MbzK5QpMxQQBMRkdLlVgMeexfG7ILgEeDoAke3wKc94JNucHij2RWKmE4BTUREzOHuA49PgjE7ofVQcLTCkY0wr5sR1o5sMbtCEdMooImIiLk8fOGJDyB8B7QcBA7OxuPOT7rCZ09C0jazKxQpdQpoIiJSNnj6QfePIDwBmr8ADk7w3/UQ3QUW9IFj8WZXKFJqFNBERKRsqVobek6F0fHQbABYHOHAOpjbCRY+bWzVIVLOKaCJiEjZVM0fekXC6O3Q9DmwOMBv38HsjrD4WTixy+wKRUqMApqIiJRtXvdC75kwajs06W8EtX3fGMdH/TsUTu4xu0IRm1NAExER++BdD56KgrBt0LgfYIG9X8OsB42D2U8lml2hiM0ooImIiH2pfh/0mQsjt0KjpwALJK6AmSHwxYtwZp/ZFYoUmwKaiIjYpxqB0O8TGLEZGvYC8uA/X0JkMCx7Cc7+ZnaFIkWmgCYiIvatZkN4+jMYvhECuwN58MsXENkavhwGyQfNrlCk0BTQRESkfPBpDP0XwrAN0OAJyMuF3f+G6a1g+UhIOWR2hSIFpoAmIiLly91N4dnFMGQ93PcY5OXAzoUwvSWsGAXnjphdocgdKaCJiEj5dE9zCP0cXvoB6neG3GzYMR+mNYdVY+B8ktkVitySApqIiJRvfi1hwDIYvBbufdgIavHzYGozWP0KpB43u0KRP1BAExGRiqFWa3h+Oby4Buq2h9wr8PNcmBoE30RA2gmzK7xOckYmmw6cJTkj0+xSxAROZhcgIiJSquq0hRdWwaGf4MeJcGQTbIuChE+h5SB4cCy41zS1xIVxR5iwMpGsnFysjg6M79mQ0OA6ptYkpUsjaCIiUjHVbQcDV8PzK6FWG8i+DFtnwJSm8N2bWC6eNaWs5IzM/HAGkJWTy4RViRpJq2AU0EREpOKyWODeDjBoDfzpK/BrBdmXYMt0qs5rD2vfggvJpVrS3pPp+eHsqqzsXPadTC/VOsRcCmgiImL3ij1fy2KBep2MhQShy8C3OZbsS7BpCnzcGNZNgIspti36FgJ93LE6Xv+fZ6uTA4F3e5TK/aVsUEATERG7tjDuCG0nxhA6N462E2NYGFeMfc4sFrivMwyJIb1HtLGn2pULsPGf8HETiHkHLp2zXfE34e3mwvieDbE6Gf+Jtjo5ML5HQ7yqWEv0vlK2aJGAiIjYrVvN1+rayAdvN5eiN2yxcKVuJ2jeB/Z9Cz++Byd/gQ0fQFwUtBkJbUZApao2+iXXCw2uQ9dGPuw7mU4DH/fi/RaxSxpBExERu1Xi87UsFgh8Aob9BM8sgJoPQGYaxE6CKU0gdjJcTrPNvW7g7eZCSP3qCmcVlAKaiIjYrVKbr2WxwP09SB6wjr3tppPtHQiXU2H9u8YctQ0fQqYm8YvtKKCJiIjdKs35WgvjjtB20o90XetFo5NvsTFoMlRvAJfPQ8zbxhy1jR9BZobN7y0Vj+agiYiIXSuN+Vo3znXLzIFB22uzJeJHvA+vNh55Jh+AdX+DzdOMzW5bvQTWyjavRSoGjaCJiIjdK+n5Wrec63b6IjTpByPjoHcUeN0LF5Nh7V+NOWpbIuHKpRKpSco3BTQREZE7uONcN0cnaNofwn6GXjOgmj9cOAPfvWGcTLB1Fly5XPqFi91SQBMRkQqpMJvbFnium6MTNAuFUduh5zSoWhsyTsGaccah7NvmQLaObJI70xw0ERGpcIpyGHmh5ro5OkPz56FJf9i5kJzYD3BMPw7fvGosJGj3CjT7EziZv/lsckYme0+mE6j91soUjaCJiEiFUpzDyAs9183JysKcTjRJmcRfrrzIyTwvSDsOq1+Gac0hfh7kXCnGrykem57CIDalgCYiIhVKaR5GfjUMXshxZEFOFzpk/pO/5wwkt0pNSE2CVWOMoJYwv9SDWnGCqpQ8BTQREalQSvMw8hvDYCZW/nXlUbb1iIGuk6BKDTh/FFaOguktYeciyMm2eR0FqQ1KLqhK4SmgiYhIhVKam9veKgwG1KphnOU5Zhc8+i5Urg7nDsPyERDZGnYtgdwcm9dTkNpKIqhK4SmgiYhIhRMaXIctr3Vi0UvBbHmt0x0XCBTVHcOgtTKEjIKxu6HL36GyN6QchK+GQmQw/LK0UEGtRFamiikseXl5eWbdfM+ePUyfPp2MjAwcHR2ZN28eFovlD99LSEigRYsWxMfH07x58yLdKyUlBS8vr+KWLOWI+oRcS/1BbmTLPpGckVmw1Z+ZGbBtNmyeCpfOGe9VbwAdX4OGT4LDrcdVirIytVC1VXDF6Q9paWl4enqSmpqKh0fBRigLPYIWHh6Ov78/FouFPXv2XPfZb7/9RkhICAEBAbRu3ZrExMTbtvXAAw8wa9YsFixYwMWLF7lw4UJhyxERESnzCrz608UN2r0MY3ZDp7+Aqyec3QdLX4RZD0LiCsjN/cNlpboyVUpFoQNa37592bhxI3Xq/DGVDxs2jKFDh7J//34iIiIYPHgwAKdOnaJz587XvSIiIgBYu3Ytzz33HNWrV6dyZZ1ZJiIigqsHtP8zjP0FOr4BLp5wOhE+fx6i2sGvX8M1D8A04b/8KfRGte3bt7/p+6dPnyYhIYHvv/8egD59+jBq1CgOHz6Mv78/69atu+l1Xbp0oUuXLoSFhbFz587bPsLMyMggLS0t/98uLi64uCjxi4hIOeXqCR3HQfAw2DoDts6EU3tgSSj4NIGH34CArvkT/q8NaZrwb99sdpJAUlISvr6+ODkZTVosFmrXrs3Ro0fx9/e/6TVr165l+fLlZGcbS4obN2583eeRkZFERkZy6ZJx0GyHDh2u+zwiIoJx48YVqL5z584V5udIBaA+IddSf5Ablbk+0XQ4lgb9cd0xF9ed87Cc3A2L+5NdownWNmN5tdP9fBhzmKycPKyOFl7t5A+ZGaRoWzObKE5/uHZwqaBsetTTjRP877T+4Oro2a2EhYURFhaWv0ggNjaWoKCg/M8LO4KmCcByI/UJuZb6g9yo7PUJL/B9Dzq+AlumQdxsnE7vxn3lIIbe05L+/V9lj2srGtztoTllJaCo/eHq4FVh2GybjVq1anHs2LH80bC8vDySkpKoXbu2rW6Bm5sbHh4e+S893hQRkQqpijd0/puxPUdIODhXhuPb8VjWn5DY5/A+tfm6OWpif2wW0GrUqEGzZs1YsGABAMuWLcPf3/+WjzdFRESkmKpUh0ffNja8bTsKnFzh2DaY/yR88jgc2mB2hVJEhQ5oYWFh+Pn5cezYMTp37kz9+vXzP4uKiiIqKoqAgAAmTZpEdHS0TYsVERGRm3CrAY+9awS14BHg6AJHt8CnPWBedzi8yewKpZBM3ai2oLRRrZQE9Qm5lvqD3Miu+0Ta77DxI4ifBzlZxnt1OxirPmu3MbU0e1XmN6oVERGRMs7DF574AMJ3QMtB4OAMh2LhX4/B/N6Q9LPZFcodKKCJiIiUV55+0P0jCE+A5i+AgxMcjIHozrCgLxyPN7tCuQUFNBERkfKuam3oORVGx0OzAWBxhANrYU4nWPQM/L7D7ArlBgpoIiIiFUU1f+gVCaO3Q9PnwOIA+9fA7I6w+Dk4sdvsCuX/KKCJiIhUNF73Qu+ZEPYzNHnGCGr7VhvnfC4ZAKf+Y3aFFZ4CmoiISEVVvT48NRtGxsEDfQEL/LoKZobA5y/A6V/NrrDCUkATERGp6O4KgL7RMHIrNOptvJe4HGa0haWD4Mw+c+urgBTQRERExFAjEPrNgxGb4f6eQB7sWQaRwbBsCJw9YHaFFYYCmoiIiFyvZiN4Zj4M3wiB3YE8+OVziGwFXw2H5INmV1juKaCJiIjIzfk0hv4LYWgsBDwOebmwazFMbwXLwyDlkNkVllsKaCIiInJ7vkHw3L9hSAzc9yjk5cDOBTC9JawcDeeOmF1huaOAJiIiIgVzTwsI/QJe+gHqPQK52ZDwGUxrAavGwvkksyssNxTQREREpHD8WsKfvoRB38O9HSH3CsR/AtOaw+pXIPW42RXaPQU0ERERKZrawfD8CnjxW/BvBzlZ8PNcmNoMvomAtBNmV2i3FNBERESkeOqEwMCv4YWvoc6DkJMJ26JgahCseR3ST5ldod1RQBMRERHbqNsOBq42RtVqtYHsy7B1BkxpCt+9CRlnzK7QbiigiYiIiO1YLMa8tEFrYMCX4NcKsi/BlukwpQmsfQsuJJtdZZmngCYiIiK2Z7FA/Udg8FoIXQq+zeHKRdg0BT5uDOsmwMUUs6sssxTQREREpORYLHBfF2MPtWeXwN1N4coF2PhP+LgJxLwDl86ZXWWZo4AmIiIiJc9igQZdjVMJ+i+Cmo0hKx02fGAEtfUT4dJ5s6ssMxTQREREpPRYLBDYDYZtgKfnQ41GkJkGsZOMOWqxk+FymtlVmk4BTUREREqfgwM07GkcyN5vHtwVCJdTYf27xhy1DR9CZrrZVZpGAU1ERETM4+AAjXrDiM3QJxqqB8Dl8xDztvHoc+NHkJlhdpWlTgFNREREzOfgCI37wsit8NQc8K4Pl1Jg3d+MR5+bpkLWBbOrLDUKaCIiIlJ2ODhCk6dhZBw8OQuq1YWLybD2r8aGt5unQ9ZFs6sscQpoIiIiUvY4OkHQszBqO/SaAVXrwIUz8P2bxhFSW2fClctmV1liFNBERESk7HJ0gmahMDoeek4Dz9qQcQrWvGYEtW1zIDvT7CptTgFNREREyj5HZ2j+vBHUun8EHn6QfgK+eRWmNoOfoyE7y+wqbUYBTUREROyHkxVaDoLwBOj2D3D3hbTjsPplmNYc4udBzhWzqyw2BTQRERGxP04u0OolCN8Bj08GNx9ITYJVY4ygljDfroOaApqIiIjYL2dXCB4GY3bCYxOhSg04fxRWjoLpLWHnIsjJNrvKQlNAExEREfvnXAnajoQxu+DRd6BydTh3GJaPgMjWsGsJ5OaYXWWBKaCJiIhI+WGtDCGjYexu6DwBKnlBykH4aihEBsMvS+0iqCmgiYiISPljrQIPjTWC2iNvQaVqkPwbLBsMM9rCni8hN9fsKm9JAU1ERETKLxd3aPcKjNkND/8FXD3h7D5Y+iLMehASV5TJoKaAJiIiIuWfqwd0+DOM/QU6vg4unnA6ET5/HqLawa9fQ16e2VXmU0ATERGRisPVEzq+BmN3QfsIsLrDqT2wJBSi2sO+b8tEUFNAExERkUJLzshk04GzJGfY6TFLlapBpzeNOWrtXgGrG5zcDYv7w5yHYf/3pgY1BTQREREplIVxR2g7MYbQuXG0nRjDwrgjZpdUdJW9jEUEY3bDg2PBuTL8vgMW9YO5neHAOlOCmgKaiIiIFFhyRiYTViaSlWNMrM/KyWXCqkT7HUm7qoo3dJlgBLWQ0eBUCY5vhwV9IPpRnI5tLdVyFNBERESkwPaeTM8PZ1dlZeey72S6SRXZmNtdxka3Y3ZBmzBw+v/t3V1IXGcex/HfxMFJiZpoto5NfDk2iSZuCzGCCW5RLMK2lOSi411SCLioZIrtRTvpRSFpy4ZaSktfhEqRFjF709qFbCk0BIqQXbMBpZTBGl1w4njRSJoQNVWD6dkLcTrjW5xxdJ4Zvx84mMwzzznP0b8PP58zM2e7NHZdmf88Jd0Z2bRhODftSAAAIOkdzMtUetq2iJCW7tymg09kJXBUGyDTLT13QfpLi/SfT/TgdkCuHX/atMOzggYAANZsd4ZL506UKd05HyHSndt07niZcnakJ3hkGyQzT/rr33X/+U/mP1Ntk7CCBgAAonLyaJGe+3OebvwyqdK8TO3OcCV6SCmHgAYAAKK2O8Olqv0Es43CJU4AAADDENAAAAAMQ0ADAAAwDAENAADAMAQ0AAAAwxDQAAAADENAAwAAMAwBDQAAwDAENAAAAMMQ0AAAAAxDQAMAADAMAQ0AAMAwRgS0N998U3V1dYkeBgAAgBGiDmgtLS2yLEsOh0N+vz+ibXh4WFVVVSopKVFlZaUGBgYeub+vvvpKR44ciXYYAAAAKSvqgFZfX6+rV6+qqKhoSVtTU5MaGxs1NDQkn8+nhoYGSdKtW7dUV1cXsfl8Pg0PD6uvr08vvvji+s8EAAAgRTij7VBdXb3s4+Pj4+rv79fly5clSR6PRy+//LICgYAsy9KVK1eW9Pnoo4907949vfrqqxoaGtKlS5d04sSJFY89NTWliYmJ0P9dLpdcLle0pwAAAGC0qAPaSoLBoPbs2SOnc36XDodDhYWFGh0dlWVZy/Z55ZVXQv/2+/1LwllbW5va2to0PT0tSaqpqYlo9/l8Onv27JrGd/fu3bWeCrYIagLhqAcsRk0g3HrqIXxxaa3iFtCk+VAWzrbtNfddboXN6/XK6/Wqv79fFRUV6unp0eHDh0Pt0a6g5eTkrPm52BqoCYSjHrAYNYFwsdbDwuJVVH1iOtIyCgoKNDY2prm5OTmdTtm2rWAwqMLCwngdQhkZGcrKyorb/gAAAEwUt4/ZyM3NVXl5ubq6uiRJ3d3dsixrxcubAAAAWF7UAc3r9So/P19jY2Oqq6vT/v37Q23t7e1qb29XSUmJ3n33XXV0dMR1sAAAAFtB1Jc4F164v5zS0lL19vaue1AAAABbmRF3EgAAAMAfCGgAAACGIaABAAAYhoAGAABgGAIaAACAYQhoAAAAhiGgAQAAGIaABgAAYBgCGgAAgGEIaAAAAIYhoAEAABiGgAYAAGAYAhoAAIBhCGgAAACGIaABAAAYhoAGAABgGAIaAACAYQhoAAAAhiGgAQAAGIaABgAAYBgCGgAAgGEIaAAAAIYhoAEAABiGgAYAAGAYAhoAAIBhCGgAAACGIaABAAAYhoAGAABgGAIaAACAYQhoAAAAhiGgAQAAGIaABgAAYBgCGgAAgGEIaAAAAIYhoAEAABiGgAYAAGAYAhoAAIBhCGgAAACGIaABAAAYhoAGAABgGAIaAACAYQhoAAAAhiGgAQAAGIaABgAAYBgCGgAAgGEIaAAAAIYhoAEAABiGgAYAAGAYAhoAAIBhCGgAAACGIaABAAAYhoAGAABgGAIaAADYFL9Ozerf/7utX6dmEz0U4zkTPQAAAJD6Lv73pt66NKAHD39Xeto2nTtRppNHixI9LGOxggYAADbUr1OzoXAmSQ8e/q63/jXAStoqCGgAAGBDDf4yGQpnCx7M/a4bv0wmaETmI6ABAIANddG2guEAAAavSURBVDAvU+lpkZEj3blNB5/IStCIzJfQgBYIBHTo0CE1Nzfr/PnziRwKAADYILszXDp3okzpzvnYke7cpnPHy5SzIz3BIzNX1AGtpaVFlmXJ4XDI7/dHtA0PD6uqqkolJSWqrKzUwMDAI/e3c+dOzczMqLi4ONqhAACAJHHyaJF633hW//jbUfW+8SxvEHiEqN/FWV9fL5/Pp2eeeWZJW1NTkxobG3X69Gl9/fXXamhoUG9vr27duqWTJ09GPPfIkSNqbW3VtWvXZNu26uvrVVNTI8uyYj4ZAABgrt0ZLlXtdyV6GEkh6hW06upq5efnL3l8fHxc/f39OnXqlCTJ4/FoZGREgUBAbrdbV65cidjee+89ORwOSZLD4ZDb7dbk5OovFpyamtLExERom51d27s/Zmdn1drauubnI/VREwhHPWAxagLhElEPDtu27Vg6Wpalb7/9Vk899ZQkqa+vTy+99FLEZc3Kykq9//77qq6uXnYfP/zwgzo7OyVJu3bt0ocffhjR3tbWpra2Nk1PTysQCCzp7/P5dPbs2UeOdWJiQsXFxRoZGVFWFi9IBDWBSNQDFqMmEG699bDQ/969e2vuH9cPql1YEVvwqOxXW1ur2traFdu9Xq+8Xq/6+/tVUVGhnp4eHT58ONTucrnkcj16qdTpnD/NnJwcftEgiZpAJOoBi1ETCLfeeljoH1WfqHusoKCgQGNjY5qbm5PT6ZRt2woGgyosLIzXIZSRkcEvCgAASHlxC2i5ubkqLy9XV1eXTp8+re7ublmWFZcX/U9PT0uSfv7555j6T01NSZJ+/PFHZWRkrHs8SH7UBMJRD1iMmkC49dbDQv/ffvtt7QtNdpTOnDlj7927105LS7Pdbre9b9++UNvg4KB97Ngx+8CBA3ZFRYXt9/uj3f2yurq6bElsbGxsbGxsbEm79fT0rDn7xPwmgc10+/Ztff/997IsS4899lhM+/B4POru7o7zyACkCuYIAKtZzxxh27bu37+viooK7dixY019kiKgxUNZWdmaPjgXwNbEHAFgNZs9R2yZe3F6vd5EDwGAwZgjAKxms+eILbOCBgAAkCy2zAoaAABAsiCgAQAAGCaudxIw3ezsrJqampSVlaWHDx+qra1tTW1IXav93AOBgJ5//nnV1NQoLy9P58+fT9xAsSnu3r2r1157TZcvX1YwGIxoY47YmlarCeaIraenp0ednZ2amZlRdna2Pv3001BbvOeIpF1Ba2lpkWVZcjgc8vv9EW3Dw8OqqqpSSUmJKisrQ++6+Oabb1RTU6OPP/5Y2dnZ6u3tDfVZrQ3JId41IUk7d+7UzMyMiouLN+08EB+x1EN2drY6OjpUWlq6ZH/MEckv3jUhMUcks1jqoaamRh0dHbp48aJGR0c1OTkZ6hPvOSJpA1p9fb2uXr2qoqKiJW1NTU1qbGzU0NCQfD6fGhoaJEk3b94M3dngySef1M2bN0N9VmtDcoh3TRQVFenatWv64osvdOnSJQUCgc04DcRJLPWwGuaI5BfvmmCOSG7rqYfvvvtOhw4dUmZmZuixeM8RSRvQqqurlZ+fv+Tx8fFx9ff369SpU5LmP1huZGREgUBABQUFoW9YIBCIuE/oam1IDvGuCYfDEfrqdrsj/lKC+WKph9UwRyS/eNcEc0Ryi7UePv/8c12/fl2tra0R/eI9R6Tca9CCwaD27NkTunO8w+FQYWGhRkdH5fF41NzcrJ9++knT09OqqqrSjRs31N7ergsXLixpQ2qItSaOHz+uzs5OSdKuXbv09NNPJ/I0ECer1YNlWWpubtbg4KCam5v1+uuva25ujjkixcVaE8wRqWm1eujr69Pbb7+tF154Qc3NzXrnnXd0586dDZkjUi6gSX/8VbNg4aPetm/fri+//DKirbS0VB988IEkLWlD6oi1JmprazdlfNhcK9WDJH322WdLns8ckfpirQnmiNS0Uj14PB55PJ6Itscff3xD5oikvcS5koKCAo2NjWlubk7S/Dc1GAxyOWILoyYQjnrAYtQEwplSDykX0HJzc1VeXq6uri5JUnd3tyzLCr1wD1sPNYFw1AMWoyYQzph6sJPUmTNn7L1799ppaWm22+229+3bF2obHBy0jx07Zh84cMCuqKiw/X5/AkeKzUJNIBz1gMWoCYQzvR64FycAAIBhUu4SJwAAQLIjoAEAABiGgAYAAGAYAhoAAIBhCGgAAACGIaABAAAYhoAGAABgGAIaAACAYQhoAAAAhiGgAQAAGIaABgAAYBgCGgAAgGH+D+Z4inYpmzyAAAAAAElFTkSuQmCC",
      "text/plain": [
       "PyPlot.Figure(PyObject <Figure size 600x400 with 1 Axes>)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "segs_int = SegmentBlocks.plot_segments_distribution(a, b; len_thr=1000);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "mean_covs, covers = SegmentBlocks.pyramid_coverage_calc(b, a);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5368147802897596"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "StatsModels.TableRegressionModel{GLM.LinearModel{GLM.LmResp{Array{Float64,1}},GLM.DensePredChol{Float64,LinearAlgebra.Cholesky{Float64,Array{Float64,2}}}},Array{Float64,2}}\n",
       "\n",
       "length ~ 1 + self + double + degree + comp_s + mean_cov + breaks + max_cov + intra_frac\n",
       "\n",
       "Coefficients:\n",
       "─────────────────────────────────────────────────────────────────────────────────\n",
       "               Estimate  Std. Error     t value  Pr(>|t|)   Lower 95%   Upper 95%\n",
       "─────────────────────────────────────────────────────────────────────────────────\n",
       "(Intercept)   3980.64    1227.98       3.24161     0.0012   1573.41     6387.87  \n",
       "self          -129.047     74.5044    -1.73207     0.0833   -275.099      17.0051\n",
       "double        1318.35      57.6762    22.8578      <1e-99   1205.29     1431.42  \n",
       "degree        3515.83     129.574     27.1338      <1e-99   3261.82     3769.83  \n",
       "comp_s          14.579      1.22775   11.8746      <1e-31     12.1723     16.9858\n",
       "mean_cov     -4841.96     246.403    -19.6506      <1e-82  -5324.99    -4358.94  \n",
       "breaks         -45.1433    85.044     -0.530822    0.5956   -211.856     121.57  \n",
       "max_cov        190.458    213.177      0.893427    0.3717   -227.436     608.352 \n",
       "intra_frac    8468.25    1383.89       6.11916     <1e-9    5755.39    11181.1   \n",
       "─────────────────────────────────────────────────────────────────────────────────"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "┌ Warning: implicit broadcasting in setindex! is deprecated; use `df[row_inds, col_ind] .= Ref(v)` broadcasting assignment to change the column in place\n",
      "│   caller = (::getfield(Main.PyrSDTypes, Symbol(\"##10#16\")){Main.SDmoduleInit.Interval_object,Int64})(::String) at modules_SDs.jl:567\n",
      "└ @ Main.PyrSDTypes /Users/abdullae/julia_wd/object_oriented_SDs/modules_SDs.jl:567\n"
     ]
    }
   ],
   "source": [
    "forest_df = PyrSDTypes.Modify_DF_for_RandForset!(a, b, segs_int, graph_obj, mean_covs);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6663821914339847"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "8-element Array{Float64,1}:\n",
       "  0.045041618720603194 \n",
       "  0.5007084515405196   \n",
       "  0.2511178489641151   \n",
       "  0.02564355752445835  \n",
       "  0.5616874225186914   \n",
       " -0.00638904891287706  \n",
       "  0.08566583890037205  \n",
       "  0.0028992556264587632"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"For feature self the p-value:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"For feature double the p-value:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"For feature degree the p-value:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"For feature comp_s the p-value:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"For feature mean_cov the p-value:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"For feature breaks the p-value:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "1.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"For feature max_cov the p-value:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\"For feature intra_frac the p-value:\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "using ScikitLearn, Random, Statistics\n",
    "#@sk_import ensemble : RandomForestRegressor\n",
    "using DecisionTree\n",
    "\n",
    "function RF_cross_valid(features, labels, comp, folds=10, col_ind=-1)\n",
    "    cv_inds = CrossValidation.KFold(size(features)[1], n_folds=folds)\n",
    "    r_squares = []\n",
    "    for i in cv_inds\n",
    "        rf = build_forest(labels[i[1]], features[i[1], :], -1, 100, 0.7, -1)\n",
    "        if comp == \"real\"\n",
    "            labels_pred = apply_forest(rf, features[i[2], :])\n",
    "        elseif comp == \"perm\"\n",
    "            features_p = copy(features)\n",
    "            features_p[:, col_ind] = shuffle(features[:, col_ind])\n",
    "            labels_pred = apply_forest(rf, features_p[i[2], :])\n",
    "        end\n",
    "        mse = mean((labels[i[2]] .- labels_pred).^2)\n",
    "        var_tot = var(labels[i[2]])\n",
    "        push!(r_squares, 1 - mse/var_tot)\n",
    "    end\n",
    "    return mean(r_squares)\n",
    "end\n",
    "\n",
    "function RF_perm_importance(features, labels, comp)\n",
    "    #rf = RandomForestRegressor(n_estimators=100, random_state=42, oob_score=false)\n",
    "    #fit!(rf, features, labels)\n",
    "    #r2_r = rf.oob_score_\n",
    "    r2_r = RF_cross_valid(features, labels, \"real\", 10)\n",
    "    r2_ps = Float64[]\n",
    "    for i in 1:size(features)[2]\n",
    "        r2_p = RF_cross_valid(features, labels, \"perm\", 10, i)\n",
    "        push!(r2_ps, r2_p)\n",
    "    end\n",
    "    if comp == \"verbose\"\n",
    "        display(r2_r)\n",
    "        display(r2_r .- r2_ps)\n",
    "    end\n",
    "    return r2_r .- r2_ps\n",
    "end\n",
    "\n",
    "labels = float(forest_df[!, 1])\n",
    "features = Matrix(forest_df[!, collect(2:9)])\n",
    "imp_perms = reshape(Float64[], 0, size(features)[2])\n",
    "perm_num = 1                                                   # better much more)\n",
    "\n",
    "imp_r = RF_perm_importance(features, labels, \"verbose\")\n",
    "\n",
    "for i in 1:perm_num\n",
    "    labels_p = shuffle(labels)\n",
    "    imp_p = RF_perm_importance(features, labels_p, \"noverbose\")\n",
    "    imp_perms = vcat(imp_perms, transpose(imp_p))\n",
    "    display(i)\n",
    "end\n",
    "\n",
    "for i in 1:size(imp_perms)[2]\n",
    "    arr = imp_perms[:,i]\n",
    "    display(\"For feature $(names(forest_df[:, collect(2:9)])[i]) the p-value:\")\n",
    "    display(length(arr[arr .>= imp_r[i]])/size(imp_perms)[1])\n",
    "end"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#=\n",
    "Explained r^2 = 0.67\n",
    "400\n",
    "\"For feature self the p-value:\"\n",
    "0.0\n",
    "\"For feature double the p-value:\"\n",
    "0.0\n",
    "\"For feature degree the p-value:\"\n",
    "0.0\n",
    "\"For feature comp_s the p-value:\"\n",
    "0.0625\n",
    "\"For feature mean_cov the p-value:\"\n",
    "0.0\n",
    "\"For feature breaks the p-value:\"\n",
    "0.73\n",
    "\"For feature max_cov the p-value:\"\n",
    "0.0\n",
    "\"For feature intra_frac the p-value:\"\n",
    "0.25"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2131, 4)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(224, 12)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(235, 12)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(1385, 4)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(203, 12)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(218, 12)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(683, 4)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(231, 12)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(178, 12)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(8540, 4)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(827, 12)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(892, 12)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "using StatsPlots, StatsBase, Plots\n",
    "\n",
    "filenames = [\"../CNV_pop_cancer/CNVs_black_popul_15_hg38.txt\", \"../CNV_pop_cancer/CNVs_black_popul_all_hg38.txt\", \"../CNV_pop_cancer/CNVs_only_popul_hg38.txt\"]\n",
    "filename = filenames[3]\n",
    "\n",
    "CNVspart.find_gene_overlap!(b, \"./genes_GRCh38_noprom.bed\")\n",
    "\n",
    "function main_CNV_overlap(b, MAF_from::Int64, MAF_to::Int64, gene::String, samplesize::Int64, perm::Int64=100)\n",
    "    CNV_int = CNVspart.read_CNVs_to_DF(filename, [MAF_from, MAF_to], samplesize)\n",
    "    CNVspart.overlap_CNVs_pyrs!(CNV_int, b, \"real\")\n",
    "    \n",
    "    display(size(CNV_int.int_DF))\n",
    "    display(size(b.int_DF[(b.int_DF[!, :cnv_over] .> 0.0) .& (b.int_DF[!, :gene_over] .== 1.0), :]))\n",
    "    display(size(b.int_DF[(b.int_DF[!, :cnv_over] .> 0.0) .& (b.int_DF[!, :gene_over] .== 0.0), :]))\n",
    "\n",
    "    shuff_df = CNVspart.CNV_rare_fixed_overlap(b, CNV_int, \"shuff\", gene, perm)\n",
    "    norm_df = CNVspart.CNV_rare_fixed_overlap(b, CNV_int, \"norm\", gene)\n",
    "    \n",
    "    return CNV_int, norm_df, shuff_df\n",
    "end\n",
    "\n",
    "comp_value = \"both\"   # \"gene\", \"nogene\"\n",
    "samplesize = 5008   # 30, 122, 5008\n",
    "\n",
    "MAF_from, MAF_to = 2, 3    # 2, 3\n",
    "CNV_int, norm_df_rare, shuff_df_rare = main_CNV_overlap(b, MAF_from, MAF_to, comp_value, samplesize);\n",
    "\n",
    "MAF_from, MAF_to = 4, 15        # 4, 15\n",
    "CNV_int, norm_df_mid, shuff_df_mid = main_CNV_overlap(b, MAF_from, MAF_to, comp_value, samplesize); \n",
    "\n",
    "MAF_from, MAF_to = 16, 2504         # 16, 6000\n",
    "CNV_int, norm_df_fix, shuff_df_fix = main_CNV_overlap(b, MAF_from, MAF_to, comp_value, samplesize);\n",
    "\n",
    "MAF_from, MAF_to = 1, 2504\n",
    "CNV_int = main_CNV_overlap(b, MAF_from, MAF_to, comp_value, samplesize)[1];\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "using GLM, Statistics, Histograms, IntervalTrees\n",
    "\n",
    "function bangkok(from::Int, to::Int, comp::String)\n",
    "    acs = Int64[]\n",
    "    if (from > 0) && (to > 0)\n",
    "        if comp == \"both\"\n",
    "            piece_df = b.int_DF[from .<= b.int_DF[!, :degree] .<= to, :]\n",
    "        elseif comp == \"gene\"\n",
    "            piece_df = b.int_DF[(from .<= b.int_DF[!, :degree] .<= to) .& (b.int_DF[!, :gene_over] .== 1.0), :]\n",
    "        elseif comp == \"nogene\"\n",
    "            piece_df = b.int_DF[(from .<= b.int_DF[!, :degree] .<= to) .& (b.int_DF[!, :gene_over] .== 0.0), :]\n",
    "        end\n",
    "        piece = SDmoduleInit.Interval_object(piece_df, SegmentBlocks.itree_from_df(piece_df, \"nothing\"))\n",
    "        for row in eachrow(CNV_int.int_DF)\n",
    "            piece.int_tree[row[:chr]] = get(piece.int_tree, row[:chr], IntervalTree{Int, IntervalValue{Int, Int}}())\n",
    "            over = collect(intersect(piece.int_tree[row[:chr]], (row[:coor_s], row[:coor_e])))\n",
    "            if length(over) > 0\n",
    "                push!(acs, row[:ac])\n",
    "            end\n",
    "        end\n",
    "    elseif (from == 0) && (to == 0)\n",
    "       for row in eachrow(CNV_int.int_DF)\n",
    "            b.int_tree[row[:chr]] = get(b.int_tree, row[:chr], IntervalTree{Int, IntervalValue{Int, Int}}())\n",
    "            over = collect(intersect(b.int_tree[row[:chr]], (row[:coor_s], row[:coor_e])))\n",
    "            if length(over) == 0\n",
    "                push!(acs, row[:ac])\n",
    "            end\n",
    "        end\n",
    "    end\n",
    "    return acs\n",
    "end\n",
    "\n",
    "acs = bangkok(21, 100, \"both\")\n",
    "#acs = map(x -> x <= 2504 ? x : 5008 - x, acs)\n",
    "cc_h = Histograms.Histogram(acs, N=30, scale=:log10)\n",
    "cc_p = plot(cc_h, xscale=:log10, yscale=:log10, markershape=:circle, line=false)\n",
    "\n",
    "x = range(1, 5000, length=10); y = 0.5*x.^(-1.5);\n",
    "cc_p = plot!(x, y)\n",
    "x = range(50, 5000, length=10); y = 0.15*x.^(-1.2);\n",
    "cc_p = plot!(x, y)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "6.634896601021213"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "3.841458820694126"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "91.71704148692778"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "538"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#[NA19700, NA19818, NA19705, NA19712, NA19902, NA20322, NA20334, NA20339, NA20341, NA19919, NA19921, NA20276, NA19914, NA20346, NA20288]\n",
    "using Distributions, DelimitedFiles\n",
    "\n",
    "#10746 + 8663 + 7053 +6273\n",
    "val = -2*(244.2 - 244.7)\n",
    "display(val)\n",
    "display(cquantile(Chisq(1), 0.01))\n",
    "display(cquantile(Chisq(1), 0.05))\n",
    "display(cquantile(Chisq(1), 1e-21))\n",
    "\n",
    "acs = bangkok(2, 5, \"both\")   # 0 0; [1 1; 2 5; 6 30; 31 200]\n",
    "#acs = map(x -> x <= 15 ? x : 30 - x, acs)\n",
    "\n",
    "arr = map(x -> length(acs[acs .== x]), 1:122)\n",
    "arr = convert(Array{Float64,1}, arr)\n",
    "arr =  arr .+ 0.01\n",
    "writedlm(\"/Users/abdullae/Downloads/prfreq_archive/sfs_all.txt\", arr, \"\\n\")\n",
    "length(acs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5×3 Array{Float64,2}:\n",
       " 0.810446  0.145393  0.0441608\n",
       " 0.651387  0.19421   0.154403 \n",
       " 0.552045  0.217472  0.230483 \n",
       " 0.48062   0.258398  0.260982 \n",
       " 0.305439  0.317992  0.376569 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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"
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "using StatsBase\n",
    "\n",
    "bins_nodes = [0 0; 1 1; 2 5; 6 30; 31 200]\n",
    "bins_cnvs = [1 3; 4 15; 16 2504]\n",
    "\n",
    "cools = reshape(Float64[], 0, 3)\n",
    "for i in eachrow(bins_nodes)\n",
    "    acs = bangkok(i[1], i[2], \"both\")\n",
    "    acs = acs[acs .>= bins_cnvs[1,1]]\n",
    "    cool = map(x -> length(acs[x[1] .<= acs .<= x[2]])/length(acs), eachrow(bins_cnvs))\n",
    "    cools = vcat(cools, transpose(cool))\n",
    "end\n",
    "display(cools)\n",
    "#labs = [\"no SDs\", \"degree=1\", \"2<=degree<=5\", \"6<=degree<=30\", \"31<=degree<=140\"]\n",
    "labs = [\"\", \" \", \"  \", \"   \", \"    \"]\n",
    "\n",
    "groupedbar(labs, cools, grid=false, bar_position = :stack, legend=:bottomleft, bar_width=0.6, \n",
    "    lab=[\"rare\" \"middle\" \"high\"], legendfont=font(9), lw=0)\n",
    "\n",
    "#savefig(\"./pics/CNV_fractions_freqs.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# -11.2\n",
    "plot([0, 1, 2, 3], [-3, -1.3, -2.3, 4.8], ylims=[-12, 5.5], xlims=[-0.12, 3.12], \n",
    "    color=1, label=\"both\", grid=false, lw=2.3, markershape=:xcross, markersize=9)\n",
    "#xticks!(0:4, [\"no SDs\", \"degree=1\", \"2<=degree<=5\", \"6<=dergee<=30\", \"31<=dergee<=140\"])\n",
    "xticks!(0:3, [\"\", \"\", \"\", \"\"])\n",
    "plot!([0, 1, 2, 3], [-2.7, -3, -2, 0.3], color=2, label=\"nogene\", lw=2.3, markershape=:xcross, markersize=9)\n",
    "plot!([0, 1, 2, 3], [ -3, -0.17, -1.8, 4.8], color=3, label=\"gene\", lw=2.3, markershape=:xcross, markersize=9)\n",
    "plot!([-0.2, 3.2], [-11.2, -11.2], label=\"rest\", color=4, lw=2.3)\n",
    "plot!([-0.2, 3.2], [0, 0], label=\"\", color=\"grey\", linestyle=:dash, lw=1)\n",
    "\n",
    "#savefig(\"./pics/CNV_selection_plots.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "TCGA_cnvs = CNVs_TCGA.read_CNVs_TCGA(\"../TCGA_CNV_data/CNVs_BRCA_all.txt\", 0.3);\n",
    "#TCGA_shuff_df = CNVspart.CNV_rare_fixed_overlap(b, TCGA_cnvs, \"shuff\", 10);\n",
    "TCGA_norm_df = CNVspart.CNV_rare_fixed_overlap(b, TCGA_cnvs, \"norm\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "CNVspart.overlap_CNVs_pyrs!(TCGA_cnvs, b, \"real\")\n",
    "TCGA_shuff = CNVspart.shuffle_DF_to_Intobject(TCGA_cnvs.int_DF)\n",
    "CNVspart.overlap_CNVs_pyrs!(TCGA_shuff, b, \"shuffle\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "degs = sort(unique(b.int_DF[!, :degree]))\n",
    "overs_cnv = map(x -> mean(b.int_DF[b.int_DF[!, :degree] .== x,:cnv_over]), degs)\n",
    "overs_shuff = map(x -> mean(b.int_DF[b.int_DF[!, :degree] .== x,:shuff_over]), degs)\n",
    "\n",
    "plot(degs, overs_cnv, markershape=:auto, line=false)\n",
    "plot!(degs, overs_shuff, markershape=:auto, line=false)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "PyPlot.Figure(PyObject <Figure size 600x400 with 1 Axes>)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "using DataFrames, Statistics\n",
    "\n",
    "rectangle(x, y, w, h) = Shape(x .+ [0,w,w,0,0], y .+ [0,0,h,h,0])\n",
    "wisker(x, y0, y1, w) = [ Shape([x,x], [y0,y1]), Shape([x-w/2,x+w/2], [y1,y1]) ]\n",
    "\n",
    "colors = ArndtLabTheme.arndtlab_palette\n",
    "\n",
    "function add_box!(plt, x, yw0, y0, ym, y1, yw1; width=1.0, fillcolor=:blue, labl=\"\") \n",
    "    plot!(plt,[rectangle(x-width/2, y0, width, y1-y0), wisker(x,y0,yw0,width/2)..., wisker(x,y1,yw1,width/2)...],\n",
    "        fillcolor= fillcolor, label=labl)\n",
    "    plot!(Shape([x-width/2, x+width/2], [ym, ym]), linewidth=2, label=\"\")\n",
    "end\n",
    "\n",
    "function plot_boxplot(p, shuff_df, step, comp=\"real\", labs=[\"\", \"\", \"\"])\n",
    "    if comp == \"shift\"\n",
    "        shift = [mean(shuff_df[!, 1]), mean(shuff_df[!, 2]), mean(shuff_df[!, 3])]\n",
    "    elseif comp == \"real\"\n",
    "        shift = [0, 0, 0]\n",
    "    end\n",
    "    \n",
    "    add_box!(p, step, minimum(shuff_df[!, 1]) - shift[1], quantile(shuff_df[!, 1], 0.25) - shift[1], mean(shuff_df[!, 1]) - shift[1], \n",
    "        quantile(shuff_df[!, 1], 0.75) - shift[1], maximum(shuff_df[!, 1]) - shift[1], width=0.7, fillcolor=colors[1], labl=labs[1])\n",
    "    add_box!(p, step+4, minimum(shuff_df[!, 2])- shift[2], quantile(shuff_df[!, 2], 0.25) - shift[2], mean(shuff_df[!, 2]) - shift[2], \n",
    "        quantile(shuff_df[!, 2], 0.75) - shift[2], maximum(shuff_df[!, 2]) - shift[2], width=0.7, fillcolor=colors[2], labl=labs[2])\n",
    "    add_box!(p, step+8, minimum(shuff_df[!, 3]) - shift[3], quantile(shuff_df[!, 3], 0.25) - shift[3], mean(shuff_df[!, 3]) - shift[3], \n",
    "        quantile(shuff_df[!, 3], 0.75) - shift[3], maximum(shuff_df[!, 3]) - shift[3], width=0.7, fillcolor=colors[3], labl=labs[3])\n",
    "    \n",
    "    return shift\n",
    "end\n",
    "\n",
    "function plot_everything(p, norm_df, shuff_df, positions, step, mod=\"shift\")\n",
    "     if mod == \"real\"\n",
    "        shift = plot_boxplot(p, shuff_df, step, \"real\");\n",
    "        p = plot!(positions, collect(norm_df[1,:]) .- shift, markersize = 5.0, label=\"\", line=false, markershape=:circle, color=colors[4]);\n",
    "    elseif mod == \"shift\"\n",
    "        shift = plot_boxplot(p, shuff_df, step, \"shift\");\n",
    "        p = plot!(positions, collect(norm_df[1,:]) .- shift, markersize = 5.0, label=\"\", line=false, markershape=:circle, color=colors[4]);\n",
    "    elseif mod == \"scale\"\n",
    "        shift = plot_boxplot(p, shuff_df, step, \"shift\");\n",
    "        sh_std = [std(col) for col = eachcol(shuff_df)];\n",
    "        p = plot!(positions, (collect(norm_df[1,:]) .- shift)./sh_std, markersize = 5.0, label=\"\", line=false, markershape=:circle, color=colors[4]);\n",
    "    end   \n",
    "end\n",
    "\n",
    "p = plot(xlim=(0,12), grid=false, xticks=nothing, legend=:topleft)   # ylim=(-0.02,0.12)\n",
    "plot_everything(p, norm_df_rare, shuff_df_rare, [1,5,9], 1, \"scale\")\n",
    "plot_everything(p, norm_df_mid, shuff_df_mid, [2,6,10], 2, \"scale\")\n",
    "plot_everything(p, norm_df_fix, shuff_df_fix, [3,7,11], 3, \"scale\")\n",
    "gui(p)\n",
    "\n",
    "#savefig(\"./pics/CNV_nogene_fix_boxplot.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "PyPlot.Figure(PyObject <Figure size 600x400 with 1 Axes>)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "using StatsBase\n",
    "\n",
    "max_deg = 30\n",
    "mlen = map(x -> mean(b.int_DF[b.int_DF[!, :degree] .== x, :length]), 1:max_deg)\n",
    "\n",
    "p = Plots.plot(log10.(b.int_DF[!, :degree]), log10.(b.int_DF[!, :length]), markershape=:auto, line=false,\n",
    "    legend=false, grid=false, xtickfont=Plots.font(13), ytickfont=Plots.font(13), \n",
    "    xticks=[0,1,2], yticks=[3,4,5,6]) #, size=(600,600))\n",
    "\n",
    "p = plot!(log10.(collect(1:max_deg)), log10.(mlen), markershape=:auto, markersize=6, line=false, color=\"red\")\n",
    "p = plot!(log10.(collect(1:max_deg)), log10.(collect(1:max_deg)*4700), color=\"red\")\n",
    "gui(p)\n",
    "\n",
    "#savefig(\"./pics/degree_length_nomean.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{1325, 12436} undirected Int64 metagraph with Float64 weights defined by :weight (default weight 1.0)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "using StatsBase\n",
    "\n",
    "function copy_model_bigcomp(N::Int64, frac::Float64)\n",
    "    edg_copy::Array{LightGraphs.SimpleGraphs.SimpleEdge{Int64},1} = []\n",
    "    gr = MetaGraph(complete_graph(2), 1.0)\n",
    "    append!(edg_copy, [Edge(1, 2)])\n",
    "    for i in 1:N-2\n",
    "        add_vertex!(gr)\n",
    "        weis = Weights(map(x -> neighbors(gr, x) |> length, vertices(gr)))\n",
    "        m_node = StatsBase.sample(vertices(gr), weis)\n",
    "        for nei in neighbors(gr, m_node)\n",
    "           if rand() < frac\n",
    "              add_edge!(gr, nei, i+2)  \n",
    "           end\n",
    "        end\n",
    "        add_edge!(gr, m_node, i+2)\n",
    "        append!(edg_copy, [Edge(m_node, i+2)])\n",
    "    end\n",
    "    return gr, edg_copy\n",
    "end\n",
    "\n",
    "copy_gr, edg_copy_real = copy_model_bigcomp(1325, 0.3);  # to make the number of nodes and edges proportional to what we have in the SD giant component\n",
    "copy_gr\n",
    "#copy_gr2, edg_copy = copy_model_bigcomp(1325, 0.47)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 404,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{1325, 8622} undirected Int64 metagraph with Float64 weights defined by :weight (default weight 1.0)"
      ]
     },
     "execution_count": 404,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# for deletion\n",
    "\n",
    "f_r = 0.2\n",
    "for e in edges(copy_gr)\n",
    "    if (length(neighbors(copy_gr, e.src)) > 1) & (length(neighbors(copy_gr, e.dst)) > 1)\n",
    "        if rand() < f_r\n",
    "           rem_edge!(copy_gr, e) \n",
    "        end\n",
    "    end\n",
    "end\n",
    "\n",
    "copy_gr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 443,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.07"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.48"
      ]
     },
     "execution_count": 443,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# for deletion\n",
    "\n",
    "outs = []\n",
    "for i in 1:300\n",
    "    edg_copy_real = copy_model_bigcomp(1325, 0.3)[2]\n",
    "    ndgs = samarkand(edg_copy_real)\n",
    "    v1, v2 = ndgs[1:2]\n",
    "    append!(outs, length(ndgs[ndgs .> maximum([v1, v2])]))\n",
    "end\n",
    "\n",
    "display(mean(outs))\n",
    "length(outs[outs .== 0])/length(outs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for deletion\n",
    "\n",
    "for i in 1:nv(copy_gr)\n",
    "    id = \"id$i\"\n",
    "    set_prop!(copy_gr, i, :id, id)\n",
    "end"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "ename": "MethodError",
     "evalue": "MethodError: no method matching assign_weights_edges!(::MetaGraph{Int64,Float64}, ::String)\nClosest candidates are:\n  assign_weights_edges!(::Any; comp) at In[5]:6",
     "output_type": "error",
     "traceback": [
      "MethodError: no method matching assign_weights_edges!(::MetaGraph{Int64,Float64}, ::String)\nClosest candidates are:\n  assign_weights_edges!(::Any; comp) at In[5]:6",
      "",
      "Stacktrace:",
      " [1] top-level scope at In[20]:12"
     ]
    }
   ],
   "source": [
    "using Statistics, StatsBase, LightGraphs\n",
    "\n",
    "function samarkand(arr_in)\n",
    "    arr_out = Int64[]\n",
    "    for i in arr_in\n",
    "        append!(arr_out, i.src)\n",
    "        append!(arr_out, i.dst)\n",
    "    end\n",
    "    sort!(arr_out)\n",
    "    return counts(arr_out)\n",
    "end\n",
    "\n",
    "assign_weights_edges!(copy_gr, \"shuffle\")       # \"shuffle\"\n",
    "edg_copy_mst = kruskal_mst(copy_gr)  # both can be used\n",
    "#edg_copy_mst = prim_mst(copy_gr)    # both can be used\n",
    "edg_copy_mst = map(x -> x.src < x.dst ? x : Edge(x.dst, x.src), edg_copy_mst)\n",
    "edg_stambul = edges(stambul(copy_gr)[2])\n",
    "\n",
    "brr = collect(map(x -> x in edg_copy_mst ? 1 : 0, edg_copy_real))\n",
    "println(\"Fraction of edges match (MST):\", sum(brr)/length(brr))\n",
    "#brr2 = collect(map(x -> x in edg_stambul ? 1 : 0, edg_copy_real))\n",
    "#println(\"Fraction of edges match (Stambul):\", sum(brr2)/length(brr2))\n",
    "\n",
    "n_deg_mst = samarkand(edg_copy_mst)\n",
    "n_deg_simple = samarkand(edges(copy_gr))\n",
    "n_deg_real = samarkand(edg_copy_real)\n",
    "n_deg_mult1 = brussels(copy_gr, 50)[1]\n",
    "n_deg_mult2, edg_mult_mst = new_york(copy_gr, 50)\n",
    "\n",
    "brr = collect(map(x -> x in edg_mult_mst ? 1 : 0, edg_copy_real))\n",
    "println(\"Fraction of edges match (Multy):\", sum(brr)/length(brr))\n",
    "#n_deg_tokyo = tokyo(copy_gr)\n",
    "#n_deg_stambul = stambul(copy_gr)[1]\n",
    "\n",
    "println(\"Correlation node degrees MST:\", cor(n_deg_real, n_deg_mst));\n",
    "#println(\"Correlation node degrees Stambul:\", cor(n_deg_real, n_deg_stambul));\n",
    "#println(\"Correlation node degrees Tokyo:\", cor(n_deg_real, n_deg_tokyo));\n",
    "println(\"Correlation node degrees Simple:\", cor(n_deg_real, n_deg_simple));\n",
    "println(\"Correlation node degrees Multiple I:\", cor(n_deg_real, n_deg_mult1));\n",
    "println(\"Correlation node degrees Multiple II:\", cor(n_deg_real, n_deg_mult2));\n",
    "\n",
    "n_deg_mst = n_deg_mst .+ rand(length(n_deg_mst)) .- 0.5\n",
    "#n_deg_stambul = n_deg_stambul .+ rand(length(n_deg_stambul)) .- 0.5\n",
    "n_deg_real = n_deg_real .+ rand(length(n_deg_real)) .- 0.5\n",
    "\n",
    "#p = plot(n_deg_real, n_deg_mst, line=false, markershape=:auto, markersize=2.0, aspect_ratio=:equal);\n",
    "#p = plot(n_deg_real, n_deg_stambul, line=false, markershape=:auto, markersize=2.0, aspect_ratio=:equal);\n",
    "#p = plot(n_deg_real, n_deg_tokyo, line=false, markershape=:auto, markersize=2.0, aspect_ratio=:equal);\n",
    "p = plot(n_deg_real, n_deg_mst, line=false, markershape=:auto, markersize=2.0, aspect_ratio=:equal);\n",
    "gui(p)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 339,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"done\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#ccs = connected_components(graph_obj2.mg)\n",
    "#ccl = sort(map(length, ccs))\n",
    "\n",
    "ccl = collect(3:5:500)\n",
    "#append!(ccl, collect(53:50:1500))\n",
    "f, n = 0.47, 30\n",
    "cors, match_fracs = [], []\n",
    "\n",
    "for cc in ccl\n",
    "    cors_n, match_fracs_n = [], []\n",
    "    for j in 1:n\n",
    "        c_gr, edg_cr = copy_model_bigcomp(cc, f)        \n",
    "        assign_weights_edges!(c_gr)\n",
    "        edg_cm = kruskal_mst(c_gr)   # both can be used\n",
    "        # edg_cm = prim_mst(c_gr)    # both can be used\n",
    "        # edg_cm = map(x -> x.src < x.dst ? x : Edge(x.dst, x.src), edg_cm)\n",
    "        brr = collect(map(x -> x in edg_cm ? 1 : 0, edg_cr))\n",
    "        append!(match_fracs_n, sum(brr)/length(brr))\n",
    "        n_deg_mst = samarkand(edg_cm)\n",
    "        n_deg_real = samarkand(edg_cr)\n",
    "        append!(cors_n, cor(n_deg_real, n_deg_mst))\n",
    "    end\n",
    "    append!(match_fracs, mean(match_fracs_n))\n",
    "    append!(cors, mean(cors_n))\n",
    "end\n",
    "\n",
    "display(\"done\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 340,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "execution_count": 340,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot(ccl, match_fracs, line=false, markershape=:circle, markersize=4.0, \n",
    "    label=\"edges match\", grid=false, legendfontsize=12, legend=:bottomright, xticks=[0, 200, 400, 600, 800, 1000, 1200, 1400])\n",
    "plot!(ccl, cors, line=false, markershape=:circle, markersize=4.0, color=:red, label=\"correlation\")\n",
    "#savefig(\"./pics/correl_edges_frac2.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 713,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5710837141132856"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.011120779303465693"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.03153544103272776"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.17939481268011528"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.07918394218937369"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "5-element Array{Float64,1}:\n",
       " 0.17124788304644983\n",
       " 0.2310635749603144 \n",
       " 0.25098290556886355\n",
       " 0.33332630932163687\n",
       " 0.3423590766991681 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "5-element Array{Float64,1}:\n",
       " 0.004363322020578603\n",
       " 0.004396923635081896\n",
       " 0.004402259328162291\n",
       " 0.004935749062396907\n",
       " 0.005463572697809613"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "5-element Array{Float64,1}:\n",
       " 0.0425316220227364  \n",
       " 0.042805786561712106\n",
       " 0.043189646630420436\n",
       " 0.04453129790708444 \n",
       " 0.06338667774213956 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "5-element Array{Float64,1}:\n",
       " 0.05506183255698014\n",
       " 0.07086471718990695\n",
       " 0.07756302241815664\n",
       " 0.14121812664777708\n",
       " 0.23935464011363144"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "5-element Array{Float64,1}:\n",
       " 0.027010101666968697\n",
       " 0.02802332254545506 \n",
       " 0.03066465965556037 \n",
       " 0.03348903839510442 \n",
       " 0.04012053629958151 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "using LightGraphs, GraphPlot\n",
    "\n",
    "cc = connected_components(graph_obj.mg)\n",
    "cl = map(length, cc)\n",
    "\n",
    "big_c = induced_subgraph(graph_obj.mg, cc[cl .== maximum(cl)][1])[1]\n",
    "rand_gr = SimpleGraph(nv(big_c), ne(big_c))\n",
    "barab_gr = barabasi_albert(nv(big_c), 7)            # took 7 to make it comparable to the SD network;\n",
    "same_degree = random_configuration_model(nv(big_c), degree(big_c))\n",
    "while length(connected_components(same_degree)) > 1\n",
    "    same_degree = random_configuration_model(nv(big_c), degree(big_c))\n",
    "end\n",
    "\n",
    "\n",
    "display(global_clustering_coefficient(big_c))\n",
    "display(global_clustering_coefficient(rand_gr))\n",
    "display(global_clustering_coefficient(barab_gr))\n",
    "display(global_clustering_coefficient(copy_gr))\n",
    "display(global_clustering_coefficient(same_degree))\n",
    "\n",
    "big_bet = sort(betweenness_centrality(big_c))\n",
    "display(big_bet[end-4:end])\n",
    "rand_bet = sort(betweenness_centrality(rand_gr))\n",
    "display(rand_bet[end-4:end])\n",
    "barab_bet = sort(betweenness_centrality(barab_gr))\n",
    "display(barab_bet[end-4:end])\n",
    "copy_bet = sort(betweenness_centrality(copy_gr))\n",
    "display(copy_bet[end-4:end])\n",
    "copy_bet = sort(betweenness_centrality(same_degree))\n",
    "display(copy_bet[end-4:end])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.2241509433962264"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.03018867924528302"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "6.477339787936634"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "2.942859260103745"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "2.818996750840791"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "5.349074463160479"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "3.018758479165479"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "using DataFrames\n",
    "\n",
    "bc = biconnected_components(big_c)\n",
    "\n",
    "bc_sizes = Int64[]\n",
    "for i in bc\n",
    "    nodes = Int64[]\n",
    "    for j in i\n",
    "        append!(nodes, [j.src, j.dst])\n",
    "    end\n",
    "    push!(bc_sizes, length(unique(nodes)))\n",
    "end\n",
    "\n",
    "display((nv(big_c) - maximum(bc_sizes))/nv(big_c))\n",
    "\n",
    "ba = articulation(big_c)\n",
    "\n",
    "splits = Int64[]\n",
    "\n",
    "for i in ba\n",
    "    temp_gr = copy(big_c)\n",
    "    rem_vertex!(temp_gr, i)\n",
    "    temp_cc = connected_components(temp_gr)\n",
    "    push!(splits, maximum(map(length, temp_cc)))\n",
    "end\n",
    "\n",
    "display((nv(big_c) - minimum(splits))/nv(big_c))\n",
    "\n",
    "m_path = mean(map(x -> sum(dijkstra_shortest_paths(big_c, x).dists)/(nv(big_c) - 1), vertices(big_c)))\n",
    "display(m_path)\n",
    "\n",
    "m_path = mean(map(x -> sum(dijkstra_shortest_paths(rand_gr, x).dists)/(nv(rand_gr) - 1), vertices(rand_gr)))\n",
    "display(m_path)\n",
    "\n",
    "m_path = mean(map(x -> sum(dijkstra_shortest_paths(barab_gr, x).dists)/(nv(barab_gr) - 1), vertices(barab_gr)))\n",
    "display(m_path)\n",
    "\n",
    "m_path = mean(map(x -> sum(dijkstra_shortest_paths(copy_gr, x).dists)/(nv(copy_gr) - 1), vertices(copy_gr)))\n",
    "display(m_path)\n",
    "\n",
    "m_path = mean(map(x -> sum(dijkstra_shortest_paths(same_degree, x).dists)/(nv(same_degree) - 1), vertices(same_degree)))\n",
    "display(m_path)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8-element Array{Int64,1}:\n",
       "  1\n",
       "  2\n",
       "  5\n",
       "  6\n",
       "  8\n",
       " 12\n",
       " 16\n",
       " 23"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "939"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "using StatsBase, HypothesisTests, Distances\n",
    "\n",
    "big_part = label_propagation(big_c);\n",
    "big_mods = sort(collect(values(countmap(big_part[1]))), rev=true);\n",
    "\n",
    "cli = big_part[1]\n",
    "b_cli = sort(collect(filter(x -> countmap(cli)[x] > 40, keys(countmap(cli)))))\n",
    "display(b_cli)\n",
    "display(sum(map(x -> length(cli[cli .== x]), b_cli)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 430,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 430,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "using GraphIO, ParserCombinator\n",
    "\n",
    "savegraph(\"./graphs/BC_gr\", big_c, \"graph\", GraphIO.EdgeList.EdgeListFormat())\n",
    "savegraph(\"./graphs/rand_gr\", rand_gr, \"graph\", GraphIO.EdgeList.EdgeListFormat())\n",
    "savegraph(\"./graphs/barab_gr\", barab_gr, \"graph\", GraphIO.EdgeList.EdgeListFormat())\n",
    "savegraph(\"./graphs/copy_gr\", copy_gr, \"graph\", GraphIO.EdgeList.EdgeListFormat())\n",
    "savegraph(\"./graphs/same_degree\", same_degree, \"graph\", GraphIO.EdgeList.EdgeListFormat())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "write_graph_with_inds (generic function with 1 method)"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "function write_graph_with_inds(graph, filename)\n",
    "    rm(filename, force=true)\n",
    "    io = open(filename, \"a\")\n",
    "\n",
    "    cc = connected_components(graph)\n",
    "    cl = map(length, cc)\n",
    "    for i in cc[argmax(cl)]\n",
    "        neis = neighbors(graph, i)\n",
    "        for n in neis\n",
    "            if i < n\n",
    "                id1 = graph[i, :id]\n",
    "                id2 = graph[n, :id]\n",
    "                write(io, id1*\"\\t\"*id2*\"\\n\");\n",
    "            end\n",
    "        end\n",
    "    end\n",
    "    close(io)\n",
    "end\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 432,
   "metadata": {},
   "outputs": [],
   "source": [
    "write_graph_with_inds(graph_obj.mg, \"./graphs/BC_with_inds\");\n",
    "write_graph_with_inds(graph_obj2.mg, \"./graphs/BC_nosec_with_inds\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "PyPlot.Figure(PyObject <Figure size 600x400 with 1 Axes>)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "gr_use = MetaGraph(graph_obj2.mg, 1.0)\n",
    "assign_weights_edges!(gr_use);\n",
    "p = graph_nosec_from_graph(gr_use)[1];\n",
    "gui(p)"
   ]
  },
  {
   "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": [
    "# trash\n",
    "using DataFrames\n",
    "\n",
    "df = DataFrames.DataFrame(d1=Float64[], d2=Float64[], d3=Float64[])\n",
    "\n",
    "for k in keys(dic)\n",
    "    push!(df, dic[k])\n",
    "end\n",
    "\n",
    "display(cor(df[!, :d1], df[!, :d2]))\n",
    "display(cor(df[!, :d1], df[!, :d3]))\n",
    "display(cor(df[!, :d2], df[!, :d3]))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Julia 1.1.0",
   "language": "julia",
   "name": "julia-1.1"
  },
  "language_info": {
   "file_extension": ".jl",
   "mimetype": "application/julia",
   "name": "julia",
   "version": "1.1.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}