{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Coseg exeperiment\n",
    "\n",
    "Due to size limitation, we only include the super pixels version (which is in the supplemantry), also, we only display 1 frame, as the translation from super pixels to pixel is memory heavy (3MB per frame)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "using Images\n",
    "using Distributed\n",
    "using NPZ"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Data Loading and preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "436×1024 Array{Int64,2}:\n",
       "   0    0    0    0    0    0    0    0  …   32   32   32   32   32   32   32\n",
       "   0    0    0    0    0    0    0    0      32   32   32   32   32   32   32\n",
       "   0    0    0    0    0    0    0    0      32   32   32   32   32   32   32\n",
       "   0    0    0    0    0    0    0    0      32   32   32   32   32   32   32\n",
       "   0    0    0    0    0    0    0    0      32   32   32   32   32   32   32\n",
       "   0    0    0    0    0    0    0    0  …   32   32   32   32   32   32   32\n",
       "   0    0    0    0    0    0    0    0      32   32   32   32   32   32   32\n",
       "   0    0    0    0    0    0    0    0      32   32   32   32   32   32   32\n",
       "   0    0    0    0    0    0    0    0      32   32   32   32   32   32   32\n",
       "   0    0    0    0    0    0    0    0      32   32   32   32   32   32   32\n",
       "   0    0    0    0    0    0    0    0  …   32   32   32   32   32   32   32\n",
       "   0    0    0    0    0    0    0    0      32   32   32   32   32   32   32\n",
       "   0    0    0    0    0    0    0    0      32   32   32   32   32   32   32\n",
       "   ⋮                        ⋮            ⋱                   ⋮               \n",
       " 470  470  470  470  470  470  470  470     468  468  468  468  468  468  468\n",
       " 470  470  470  470  470  470  470  470  …  468  468  468  468  468  468  468\n",
       " 470  470  470  470  470  470  470  470     468  468  468  468  468  468  468\n",
       " 470  470  470  470  470  470  470  470     468  468  468  468  468  468  468\n",
       " 470  470  470  470  470  470  470  470     468  468  468  468  468  468  468\n",
       " 470  470  470  470  470  470  470  470     468  468  468  468  468  468  468\n",
       " 470  470  470  470  470  470  470  470  …  468  468  468  468  468  468  468\n",
       " 470  470  470  470  470  470  470  470     468  468  468  468  468  468  468\n",
       " 470  470  470  470  470  470  470  470     468  468  468  468  468  468  468\n",
       " 470  470  470  470  470  470  470  470     468  468  468  468  468  468  468\n",
       " 470  470  470  470  470  470  470  470     468  468  468  468  468  468  468\n",
       " 470  470  470  470  470  470  470  470  …  468  468  468  468  468  468  468"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = Dict()\n",
    "for i=1:50\n",
    "    group_data = copy(npzread(\"DATA/coseg/frame\" * string(i) * \".npy\")')\n",
    "    group_data[1:3,:] ./= 25\n",
    "    group_data[4:5,:] ./= 250\n",
    "    data[i] = group_data\n",
    "end\n",
    "\n",
    "pixels_to_super1 = npzread(\"DATA/coseg/pixels_to_super1.npy\")\n",
    "pixels_to_super2 = npzread(\"DATA/coseg/pixels_to_super2.npy\")\n",
    "pixels_to_super3 = npzread(\"DATA/coseg/pixels_to_super3.npy\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "┌ Info: Precompiling VersatileHDPMixtureModels [cb401d7d-c4f9-4284-af24-a1eca280e8e8]\n",
      "└ @ Base loading.jl:1273\n"
     ]
    },
    {
     "ename": "CompositeException",
     "evalue": "On worker 2:\nArgumentError: Package VersatileHDPMixtureModels [cb401d7d-c4f9-4284-af24-a1eca280e8e8] is required but does not seem to be installed:\n - Run `Pkg.instantiate()` to install all recorded dependencies.\n\n_require at ./loading.jl:993\nrequire at ./loading.jl:922\n#1 at /buildworker/worker/package_linux64/build/usr/share/julia/stdlib/v1.3/Distributed/src/Distributed.jl:78\n#105 at /buildworker/worker/package_linux64/build/usr/share/julia/stdlib/v1.3/Distributed/src/process_messages.jl:290\nrun_work_thunk at /buildworker/worker/package_linux64/build/usr/share/julia/stdlib/v1.3/Distributed/src/process_messages.jl:79\nrun_work_thunk at /buildworker/worker/package_linux64/build/usr/share/julia/stdlib/v1.3/Distributed/src/process_messages.jl:88\n#98 at ./task.jl:333\n\n...and 3 more exception(s).\n",
     "output_type": "error",
     "traceback": [
      "On worker 2:\nArgumentError: Package VersatileHDPMixtureModels [cb401d7d-c4f9-4284-af24-a1eca280e8e8] is required but does not seem to be installed:\n - Run `Pkg.instantiate()` to install all recorded dependencies.\n\n_require at ./loading.jl:993\nrequire at ./loading.jl:922\n#1 at /buildworker/worker/package_linux64/build/usr/share/julia/stdlib/v1.3/Distributed/src/Distributed.jl:78\n#105 at /buildworker/worker/package_linux64/build/usr/share/julia/stdlib/v1.3/Distributed/src/process_messages.jl:290\nrun_work_thunk at /buildworker/worker/package_linux64/build/usr/share/julia/stdlib/v1.3/Distributed/src/process_messages.jl:79\nrun_work_thunk at /buildworker/worker/package_linux64/build/usr/share/julia/stdlib/v1.3/Distributed/src/process_messages.jl:88\n#98 at ./task.jl:333\n\n...and 3 more exception(s).\n",
      "",
      "Stacktrace:",
      " [1] sync_end(::Array{Any,1}) at ./task.jl:300",
      " [2] _require_callback(::Base.PkgId) at ./task.jl:319",
      " [3] #invokelatest#1 at ./essentials.jl:709 [inlined]",
      " [4] invokelatest at ./essentials.jl:708 [inlined]",
      " [5] require(::Base.PkgId) at ./loading.jl:925",
      " [6] require(::Module, ::Symbol) at ./loading.jl:917",
      " [7] top-level scope at /buildworker/worker/package_linux64/build/usr/share/julia/stdlib/v1.3/Distributed/src/macros.jl:199",
      " [8] top-level scope at In[4]:3"
     ]
    }
   ],
   "source": [
    "using Distributed\n",
    "addprocs(4)\n",
    "@everywhere using VersatileHDPMixtureModels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "niw_hyperparams(1.0, [0.871428, 2.044336], 5.0, [0.2 0.0; 0.0 0.2])"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "global_hyper_params = niw_hyperparams(1.0,\n",
    "    [47.56,45.40,27.82] ./ 25,\n",
    "    9.0,\n",
    "    [[0.8662817 0.78323282 0.41225376];\n",
    "        [0.78323282 0.74170384 0.50340258];\n",
    "        [0.41225376 0.50340258 0.79185577]]*1.0)\n",
    "\n",
    "local_hyper_params = niw_hyperparams(1.0,\n",
    "    [217.857,511.084] ./ 250,\n",
    "    5.0,\n",
    "    Matrix{Float64}(I, 2, 2)*0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration: 1|| Global Counts: [50]|| iter time: 0.10778093338012695\n",
      "Iteration: 2|| Global Counts: [50]|| iter time: 0.08347201347351074\n",
      "Iteration: 3|| Global Counts: [50]|| iter time: 0.0806269645690918\n",
      "Iteration: 4|| Global Counts: [50]|| iter time: 0.0867609977722168\n",
      "Iteration: 5|| Global Counts: [51, 51]|| iter time: 0.07793903350830078\n",
      "Iteration: 6|| Global Counts: [51, 69]|| iter time: 0.2078390121459961\n",
      "Iteration: 7|| Global Counts: [53, 77]|| iter time: 0.12623095512390137\n",
      "Iteration: 8|| Global Counts: [54, 84]|| iter time: 0.11515498161315918\n",
      "Iteration: 9|| Global Counts: [54, 88]|| iter time: 0.14554595947265625\n",
      "Iteration: 10|| Global Counts: [55, 94, 55]|| iter time: 0.2730891704559326\n",
      "Iteration: 11|| Global Counts: [64, 97, 59]|| iter time: 0.1690070629119873\n",
      "Iteration: 12|| Global Counts: [67, 104, 63]|| iter time: 0.27566003799438477\n",
      "Iteration: 13|| Global Counts: [71, 104, 71]|| iter time: 0.15233278274536133\n",
      "Iteration: 14|| Global Counts: [71, 99, 81]|| iter time: 0.21695685386657715\n",
      "Iteration: 15|| Global Counts: [72, 100, 84, 72]|| iter time: 0.16602110862731934\n",
      "Iteration: 16|| Global Counts: [81, 101, 85, 79]|| iter time: 0.2005469799041748\n",
      "Iteration: 17|| Global Counts: [89, 105, 86, 77]|| iter time: 0.1858198642730713\n",
      "Iteration: 18|| Global Counts: [103, 104, 84, 78]|| iter time: 0.19151091575622559\n",
      "Iteration: 19|| Global Counts: [112, 105, 82, 80]|| iter time: 0.19164800643920898\n",
      "Iteration: 20|| Global Counts: [116, 103, 82, 81, 81]|| iter time: 0.20924782752990723\n",
      "Iteration: 21|| Global Counts: [121, 103, 80, 85, 81]|| iter time: 0.21548795700073242\n",
      "Iteration: 22|| Global Counts: [126, 105, 81, 87, 81, 126]|| iter time: 0.23274898529052734\n",
      "Iteration: 23|| Global Counts: [140, 107, 85, 92, 86, 115]|| iter time: 0.2566561698913574\n",
      "Iteration: 24|| Global Counts: [153, 110, 87, 98, 92, 116]|| iter time: 0.29587817192077637\n",
      "Iteration: 25|| Global Counts: [156, 111, 86, 102, 93, 119, 93]|| iter time: 0.27562499046325684\n",
      "Iteration: 26|| Global Counts: [159, 111, 84, 102, 97, 108, 93]|| iter time: 0.39360499382019043\n",
      "Iteration: 27|| Global Counts: [164, 113, 84, 106, 94, 107, 99]|| iter time: 0.3330957889556885\n",
      "Iteration: 28|| Global Counts: [162, 115, 85, 110, 98, 112, 108]|| iter time: 0.3002150058746338\n",
      "Iteration: 29|| Global Counts: [172, 115, 86, 112, 98, 118, 107]|| iter time: 0.3100910186767578\n",
      "Iteration: 30|| Global Counts: [170, 116, 88, 109, 103, 124, 119]|| iter time: 0.32840394973754883\n",
      "Iteration: 31|| Global Counts: [171, 112, 90, 108, 105, 123, 127, 105]|| iter time: 0.3411078453063965\n",
      "Iteration: 32|| Global Counts: [169, 112, 90, 115, 106, 119, 131, 112]|| iter time: 0.41132092475891113\n",
      "Iteration: 33|| Global Counts: [168, 108, 91, 121, 108, 121, 139, 123]|| iter time: 0.37489795684814453\n",
      "Iteration: 34|| Global Counts: [174, 109, 89, 125, 112, 120, 139, 132]|| iter time: 0.4627988338470459\n",
      "Iteration: 35|| Global Counts: [176, 111, 92, 120, 108, 121, 145, 135, 145]|| iter time: 0.4183499813079834\n",
      "Iteration: 36|| Global Counts: [178, 109, 93, 122, 116, 120, 126, 130, 176]|| iter time: 0.49109315872192383\n",
      "Iteration: 37|| Global Counts: [178, 110, 96, 123, 127, 113, 143, 123, 177]|| iter time: 0.4467480182647705\n",
      "Iteration: 38|| Global Counts: [181, 116, 90, 124, 126, 111, 150, 125, 193]|| iter time: 0.5052590370178223\n",
      "Iteration: 39|| Global Counts: [178, 118, 92, 125, 123, 117, 148, 131, 205]|| iter time: 0.5125830173492432\n",
      "Iteration: 40|| Global Counts: [176, 122, 94, 120, 131, 118, 134, 126, 211]|| iter time: 0.4916861057281494\n",
      "Iteration: 41|| Global Counts: [169, 119, 97, 118, 142, 116, 135, 126, 208]|| iter time: 0.476790189743042\n",
      "Iteration: 42|| Global Counts: [167, 125, 94, 114, 139, 117, 144, 127, 213]|| iter time: 0.553725004196167\n",
      "Iteration: 43|| Global Counts: [158, 124, 95, 114, 136, 111, 155, 128, 211, 155]|| iter time: 0.5287051200866699\n",
      "Iteration: 44|| Global Counts: [150, 132, 96, 111, 149, 114, 114, 130, 235, 158]|| iter time: 0.5126461982727051\n",
      "Iteration: 45|| Global Counts: [145, 131, 101, 119, 162, 123, 120, 127, 236, 165]|| iter time: 0.5309879779815674\n",
      "Iteration: 46|| Global Counts: [138, 126, 98, 118, 158, 116, 106, 132, 234, 175]|| iter time: 0.5517210960388184\n",
      "Iteration: 47|| Global Counts: [137, 129, 100, 114, 154, 119, 106, 130, 240, 177]|| iter time: 0.5154519081115723\n",
      "Iteration: 48|| Global Counts: [141, 126, 101, 116, 156, 116, 116, 137, 237, 177]|| iter time: 0.5026781558990479\n",
      "Iteration: 49|| Global Counts: [139, 133, 104, 111, 156, 108, 106, 139, 240, 186]|| iter time: 0.5472798347473145\n",
      "Iteration: 50|| Global Counts: [139, 134, 101, 109, 154, 105, 116, 142, 239, 200]|| iter time: 0.5156140327453613\n",
      "Iteration: 51|| Global Counts: [141, 129, 105, 105, 158, 117, 118, 133, 237, 204]|| iter time: 0.5359690189361572\n",
      "Iteration: 52|| Global Counts: [148, 129, 105, 107, 152, 126, 121, 140, 230, 211, 211]|| iter time: 0.5443282127380371\n",
      "Iteration: 53|| Global Counts: [166, 132, 104, 107, 161, 122, 121, 146, 221, 140, 266]|| iter time: 0.5775630474090576\n",
      "Iteration: 54|| Global Counts: [164, 131, 105, 101, 158, 127, 124, 143, 202, 133, 273]|| iter time: 0.576524019241333\n",
      "Iteration: 55|| Global Counts: [159, 125, 105, 95, 168, 123, 122, 148, 190, 134, 273]|| iter time: 0.5847108364105225\n",
      "Iteration: 56|| Global Counts: [163, 132, 106, 95, 173, 113, 116, 149, 176, 124, 285]|| iter time: 0.5636100769042969\n",
      "Iteration: 57|| Global Counts: [167, 130, 103, 97, 173, 117, 121, 149, 164, 123, 287, 123]|| iter time: 0.5881030559539795\n",
      "Iteration: 58|| Global Counts: [168, 134, 103, 95, 181, 120, 110, 157, 158, 86, 269, 123]|| iter time: 0.6195249557495117\n",
      "Iteration: 59|| Global Counts: [170, 124, 103, 98, 188, 116, 112, 159, 158, 86, 277, 120, 277]|| iter time: 0.6208369731903076\n",
      "Iteration: 60|| Global Counts: [174, 130, 106, 102, 185, 120, 106, 168, 163, 75, 316, 125, 194]|| iter time: 0.6573920249938965\n",
      "Iteration: 61|| Global Counts: [173, 132, 108, 98, 196, 115, 97, 176, 161, 73, 323, 127, 184]|| iter time: 0.682884931564331\n",
      "Iteration: 62|| Global Counts: [177, 127, 107, 85, 182, 116, 97, 179, 157, 68, 317, 129, 191]|| iter time: 0.8353350162506104\n",
      "Iteration: 63|| Global Counts: [171, 122, 103, 86, 181, 127, 98, 176, 156, 68, 311, 109, 187]|| iter time: 0.6869759559631348\n",
      "Iteration: 64|| Global Counts: [166, 130, 102, 89, 175, 125, 91, 175, 152, 65, 298, 100, 196]|| iter time: 0.7289021015167236\n",
      "Iteration: 65|| Global Counts: [163, 134, 103, 92, 169, 118, 85, 173, 152, 63, 296, 92, 201]|| iter time: 0.7682571411132812\n",
      "Iteration: 66|| Global Counts: [161, 127, 104, 92, 166, 116, 83, 169, 151, 63, 286, 96, 209]|| iter time: 0.815925121307373\n",
      "Iteration: 67|| Global Counts: [156, 130, 103, 81, 172, 111, 87, 164, 148, 63, 280, 88, 216]|| iter time: 0.7303910255432129\n",
      "Iteration: 68|| Global Counts: [151, 123, 102, 78, 166, 103, 79, 167, 141, 67, 291, 94, 210, 67]|| iter time: 0.6976690292358398\n",
      "Iteration: 69|| Global Counts: [143, 124, 102, 87, 166, 110, 74, 165, 147, 61, 287, 101, 212, 69]|| iter time: 0.6506190299987793\n",
      "Iteration: 70|| Global Counts: [143, 118, 107, 82, 154, 111, 79, 167, 146, 64, 283, 106, 201, 70]|| iter time: 0.719480037689209\n",
      "Iteration: 71|| Global Counts: [148, 124, 111, 83, 153, 107, 75, 179, 145, 65, 285, 92, 204, 77]|| iter time: 0.6475911140441895\n",
      "Iteration: 72|| Global Counts: [142, 119, 106, 80, 161, 111, 75, 172, 147, 72, 275, 93, 210, 78]|| iter time: 0.65647292137146\n",
      "Iteration: 73|| Global Counts: [138, 124, 103, 85, 158, 108, 73, 168, 147, 73, 285, 87, 225, 83, 73]|| iter time: 0.7394311428070068\n",
      "Iteration: 74|| Global Counts: [139, 119, 103, 84, 164, 103, 72, 173, 151, 65, 281, 93, 224, 83, 68]|| iter time: 0.7066409587860107\n",
      "Iteration: 75|| Global Counts: [133, 122, 104, 84, 150, 112, 73, 169, 153, 63, 279, 83, 234, 83, 68]|| iter time: 0.7807679176330566\n",
      "Iteration: 76|| Global Counts: [126, 120, 107, 80, 154, 119, 76, 153, 158, 61, 285, 73, 221, 82, 62]|| iter time: 0.9837119579315186\n",
      "Iteration: 77|| Global Counts: [125, 123, 106, 86, 149, 112, 78, 142, 157, 57, 293, 71, 227, 82, 62]|| iter time: 0.8511419296264648\n",
      "Iteration: 78|| Global Counts: [128, 121, 103, 86, 148, 117, 79, 151, 156, 59, 299, 67, 230, 80, 64]|| iter time: 0.7534670829772949\n",
      "Iteration: 79|| Global Counts: [126, 122, 106, 85, 145, 120, 78, 148, 150, 61, 307, 61, 235, 80, 63]|| iter time: 0.7948038578033447\n",
      "Iteration: 80|| Global Counts: [136, 121, 107, 87, 147, 118, 78, 140, 155, 60, 307, 64, 244, 82, 61]|| iter time: 1.0704200267791748\n",
      "Iteration: 81|| Global Counts: [147, 123, 107, 88, 151, 120, 78, 133, 147, 58, 308, 68, 241, 78, 63]|| iter time: 0.8805429935455322\n",
      "Iteration: 82|| Global Counts: [148, 127, 107, 92, 149, 121, 82, 129, 150, 62, 304, 66, 243, 85, 66]|| iter time: 0.8269190788269043\n",
      "Iteration: 83|| Global Counts: [140, 127, 105, 94, 148, 119, 79, 124, 153, 63, 308, 64, 241, 91, 61]|| iter time: 1.0011179447174072\n",
      "Iteration: 84|| Global Counts: [139, 122, 107, 92, 148, 114, 80, 121, 156, 60, 310, 59, 235, 93, 61]|| iter time: 0.7407159805297852\n",
      "Iteration: 85|| Global Counts: [134, 120, 107, 98, 146, 116, 78, 117, 154, 58, 298, 64, 238, 91, 61]|| iter time: 0.7711219787597656\n",
      "Iteration: 86|| Global Counts: [138, 119, 103, 95, 157, 118, 77, 116, 154, 59, 306, 60, 235, 98, 65]|| iter time: 0.7665741443634033\n",
      "Iteration: 87|| Global Counts: [139, 124, 103, 94, 148, 116, 75, 101, 154, 63, 304, 57, 236, 104, 64]|| iter time: 0.7455148696899414\n",
      "Iteration: 88|| Global Counts: [133, 118, 102, 102, 144, 113, 71, 107, 153, 61, 298, 56, 240, 103, 63]|| iter time: 0.7351949214935303\n",
      "Iteration: 89|| Global Counts: [136, 117, 102, 103, 140, 115, 70, 112, 153, 62, 307, 58, 246, 100, 69]|| iter time: 0.6651368141174316\n",
      "Iteration: 90|| Global Counts: [135, 118, 103, 103, 143, 125, 67, 124, 159, 71, 308, 63, 241, 99, 68]|| iter time: 0.6829671859741211\n",
      "Iteration: 91|| Global Counts: [132, 118, 104, 96, 149, 120, 71, 115, 161, 66, 306, 63, 253, 105, 68, 68]|| iter time: 0.7257449626922607\n",
      "Iteration: 92|| Global Counts: [133, 119, 102, 107, 151, 122, 76, 102, 157, 59, 291, 58, 255, 110, 68, 65]|| iter time: 0.7884271144866943\n",
      "Iteration: 93|| Global Counts: [127, 121, 105, 105, 170, 129, 71, 102, 161, 56, 298, 56, 243, 107, 70, 68]|| iter time: 0.8412179946899414\n",
      "Iteration: 94|| Global Counts: [131, 121, 107, 107, 167, 121, 75, 99, 159, 55, 299, 57, 239, 112, 69, 70]|| iter time: 0.7915999889373779\n",
      "Iteration: 95|| Global Counts: [131, 122, 111, 105, 167, 121, 81, 104, 162, 56, 307, 61, 240, 108, 69, 76]|| iter time: 0.7986149787902832\n",
      "Iteration: 96|| Global Counts: [133, 116, 108, 104, 165, 125, 80, 99, 157, 54, 302, 67, 253, 115, 66, 70]|| iter time: 0.7569680213928223\n",
      "Iteration: 97|| Global Counts: [129, 117, 108, 110, 171, 124, 80, 84, 152, 55, 296, 65, 271, 110, 68, 67]|| iter time: 0.7644290924072266\n",
      "Iteration: 98|| Global Counts: [131, 119, 109, 112, 175, 123, 79, 92, 157, 53, 302, 64, 265, 110, 69, 68]|| iter time: 0.731774091720581\n",
      "Iteration: 99|| Global Counts: [134, 120, 109, 107, 167, 123, 80, 87, 150, 55, 312, 62, 271, 104, 71, 72]|| iter time: 0.7772190570831299\n",
      "Iteration: 100|| Global Counts: [130, 118, 105, 102, 162, 122, 84, 81, 152, 54, 314, 61, 264, 109, 69, 72]|| iter time: 0.7253739833831787\n",
      "Iteration: 101|| Global Counts: [127, 116, 105, 101, 170, 118, 80, 74, 149, 54, 306, 58, 255, 107, 68, 77]|| iter time: 0.813727855682373\n",
      "Iteration: 102|| Global Counts: [130, 120, 104, 104, 162, 123, 80, 67, 149, 52, 306, 61, 262, 110, 69, 80]|| iter time: 0.7117459774017334\n",
      "Iteration: 103|| Global Counts: [129, 118, 103, 102, 165, 124, 87, 70, 147, 53, 304, 60, 259, 103, 70, 75]|| iter time: 0.7578458786010742\n",
      "Iteration: 104|| Global Counts: [124, 118, 103, 106, 155, 121, 80, 71, 150, 53, 311, 64, 266, 101, 69, 72]|| iter time: 0.7393341064453125\n",
      "Iteration: 105|| Global Counts: [129, 120, 101, 102, 168, 133, 84, 68, 149, 54, 311, 66, 259, 97, 70, 74]|| iter time: 0.7872180938720703\n",
      "Iteration: 106|| Global Counts: [118, 122, 101, 98, 153, 134, 81, 63, 152, 52, 317, 75, 250, 102, 69, 83]|| iter time: 0.7668509483337402\n",
      "Iteration: 107|| Global Counts: [125, 126, 101, 100, 151, 129, 81, 65, 153, 56, 309, 62, 248, 100, 75, 80]|| iter time: 0.6952660083770752\n",
      "Iteration: 108|| Global Counts: [129, 122, 100, 96, 153, 116, 83, 64, 159, 52, 307, 59, 248, 96, 69, 79]|| iter time: 0.6972389221191406\n",
      "Iteration: 109|| Global Counts: [130, 114, 100, 97, 151, 116, 86, 63, 160, 51, 306, 60, 249, 102, 71, 76]|| iter time: 0.8617889881134033\n",
      "Iteration: 110|| Global Counts: [129, 111, 104, 99, 152, 113, 84, 65, 161, 52, 314, 65, 257, 106, 75, 84]|| iter time: 0.658876895904541\n",
      "Iteration: 111|| Global Counts: [132, 113, 103, 93, 157, 119, 85, 64, 160, 54, 323, 66, 246, 105, 78, 83]|| iter time: 0.7206449508666992\n",
      "Iteration: 112|| Global Counts: [132, 112, 104, 92, 155, 122, 82, 60, 161, 54, 311, 67, 256, 114, 72, 79]|| iter time: 0.7343721389770508\n",
      "Iteration: 113|| Global Counts: [142, 111, 101, 90, 162, 117, 86, 61, 159, 53, 307, 63, 256, 108, 71, 80]|| iter time: 0.7347579002380371\n",
      "Iteration: 114|| Global Counts: [129, 114, 100, 95, 163, 117, 86, 63, 156, 52, 316, 59, 263, 111, 72, 77]|| iter time: 0.717426061630249\n",
      "Iteration: 115|| Global Counts: [131, 120, 102, 97, 168, 116, 84, 60, 156, 55, 307, 62, 270, 119, 70, 79]|| iter time: 0.801861047744751\n",
      "Iteration: 116|| Global Counts: [134, 115, 101, 95, 167, 125, 87, 62, 155, 52, 305, 63, 273, 110, 74, 73]|| iter time: 0.7939519882202148\n",
      "Iteration: 117|| Global Counts: [135, 109, 101, 93, 171, 123, 87, 60, 160, 55, 310, 61, 264, 105, 74, 75]|| iter time: 0.8832480907440186\n",
      "Iteration: 118|| Global Counts: [126, 109, 103, 91, 164, 124, 86, 55, 153, 54, 318, 54, 258, 112, 80, 70]|| iter time: 0.7690010070800781\n",
      "Iteration: 119|| Global Counts: [120, 112, 101, 94, 163, 122, 90, 51, 155, 55, 310, 50, 268, 105, 80, 77]|| iter time: 0.8248999118804932\n",
      "Iteration: 120|| Global Counts: [119, 111, 104, 89, 162, 115, 87, 55, 152, 54, 310, 54, 262, 112, 80, 79]|| iter time: 0.8788979053497314\n",
      "Iteration: 121|| Global Counts: [122, 109, 105, 89, 171, 110, 87, 57, 155, 54, 305, 58, 258, 113, 78, 80]|| iter time: 0.7652111053466797\n",
      "Iteration: 122|| Global Counts: [125, 112, 104, 93, 171, 115, 86, 54, 154, 53, 316, 57, 269, 114, 76, 75]|| iter time: 0.899583101272583\n",
      "Iteration: 123|| Global Counts: [127, 113, 106, 84, 164, 112, 87, 59, 156, 54, 316, 49, 274, 107, 76, 71]|| iter time: 0.8692469596862793\n",
      "Iteration: 124|| Global Counts: [124, 109, 108, 91, 160, 110, 84, 63, 156, 55, 316, 55, 266, 105, 80, 76]|| iter time: 1.0188908576965332\n",
      "Iteration: 125|| Global Counts: [127, 109, 108, 91, 158, 114, 84, 62, 155, 56, 314, 51, 270, 101, 76, 78]|| iter time: 0.7700638771057129\n",
      "Iteration: 126|| Global Counts: [124, 115, 109, 85, 152, 111, 87, 60, 158, 58, 304, 49, 276, 94, 80, 76]|| iter time: 0.6857879161834717\n",
      "Iteration: 127|| Global Counts: [125, 120, 109, 83, 148, 111, 89, 51, 158, 58, 312, 54, 280, 96, 81, 79]|| iter time: 0.6807539463043213\n",
      "Iteration: 128|| Global Counts: [124, 113, 107, 85, 151, 105, 91, 52, 162, 55, 311, 51, 287, 89, 77, 79]|| iter time: 0.689486026763916\n",
      "Iteration: 129|| Global Counts: [129, 112, 109, 87, 154, 115, 91, 47, 153, 57, 310, 50, 283, 92, 78, 76]|| iter time: 0.6824359893798828\n",
      "Iteration: 130|| Global Counts: [126, 114, 109, 90, 152, 114, 86, 50, 151, 58, 317, 53, 273, 86, 82, 83]|| iter time: 0.6679742336273193\n",
      "Iteration: 131|| Global Counts: [126, 114, 107, 84, 151, 116, 84, 51, 152, 56, 328, 49, 282, 94, 89, 83]|| iter time: 0.6871039867401123\n",
      "Iteration: 132|| Global Counts: [131, 112, 106, 82, 154, 126, 84, 50, 158, 56, 324, 49, 275, 92, 84, 85]|| iter time: 0.6917800903320312\n",
      "Iteration: 133|| Global Counts: [129, 114, 106, 86, 152, 117, 88, 50, 153, 55, 318, 50, 282, 97, 86, 77]|| iter time: 0.6923091411590576\n",
      "Iteration: 134|| Global Counts: [129, 115, 109, 82, 150, 121, 91, 51, 154, 60, 308, 52, 269, 92, 90, 79]|| iter time: 0.6472890377044678\n",
      "Iteration: 135|| Global Counts: [123, 118, 106, 84, 156, 115, 87, 53, 148, 59, 325, 49, 271, 92, 89, 79]|| iter time: 0.7219600677490234\n",
      "Iteration: 136|| Global Counts: [125, 112, 108, 84, 153, 118, 87, 52, 147, 58, 319, 50, 271, 97, 85, 75]|| iter time: 0.6798601150512695\n",
      "Iteration: 137|| Global Counts: [122, 116, 107, 88, 161, 120, 84, 50, 149, 61, 313, 54, 260, 100, 85, 75]|| iter time: 0.708198070526123\n",
      "Iteration: 138|| Global Counts: [128, 116, 107, 84, 152, 119, 85, 52, 149, 62, 307, 54, 270, 94, 85, 74]|| iter time: 0.7106540203094482\n",
      "Iteration: 139|| Global Counts: [122, 122, 107, 81, 147, 108, 81, 50, 153, 60, 303, 52, 278, 91, 84, 76]|| iter time: 0.6866228580474854\n",
      "Iteration: 140|| Global Counts: [128, 121, 105, 78, 144, 103, 83, 51, 157, 60, 304, 56, 265, 90, 80, 76]|| iter time: 0.6768529415130615\n",
      "Iteration: 141|| Global Counts: [129, 120, 106, 75, 146, 106, 83, 52, 152, 62, 307, 50, 269, 90, 80, 76]|| iter time: 0.6780869960784912\n",
      "Iteration: 142|| Global Counts: [125, 117, 105, 82, 148, 108, 86, 54, 149, 59, 306, 55, 267, 95, 77, 73]|| iter time: 0.6258788108825684\n",
      "Iteration: 143|| Global Counts: [123, 117, 106, 83, 157, 108, 87, 57, 147, 59, 315, 55, 266, 97, 82, 67]|| iter time: 0.684654951095581\n",
      "Iteration: 144|| Global Counts: [122, 121, 107, 80, 152, 109, 86, 61, 149, 62, 312, 47, 261, 97, 79, 72]|| iter time: 0.671638011932373\n",
      "Iteration: 145|| Global Counts: [123, 119, 110, 83, 155, 109, 85, 57, 152, 61, 316, 49, 268, 93, 76, 73]|| iter time: 0.6804401874542236\n",
      "Iteration: 146|| Global Counts: [130, 118, 106, 80, 157, 110, 85, 55, 152, 61, 316, 42, 263, 89, 83, 76]|| iter time: 0.6863150596618652\n",
      "Iteration: 147|| Global Counts: [127, 120, 107, 80, 162, 117, 90, 53, 152, 62, 310, 47, 265, 84, 80, 76]|| iter time: 0.6927080154418945\n",
      "Iteration: 148|| Global Counts: [130, 116, 111, 83, 150, 113, 86, 50, 155, 58, 319, 47, 278, 88, 80, 71]|| iter time: 0.6855320930480957\n",
      "Iteration: 149|| Global Counts: [130, 115, 111, 79, 154, 112, 84, 49, 158, 59, 328, 52, 264, 84, 79, 78]|| iter time: 0.6609241962432861\n",
      "Iteration: 150|| Global Counts: [122, 123, 111, 84, 144, 121, 84, 50, 158, 58, 328, 52, 267, 86, 86, 72]|| iter time: 0.6971859931945801\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "3×436×1024 Array{Float64,3}:\n",
       "[:, :, 1] =\n",
       " 0.129543  0.129543  0.129543  0.129543  …  0.129543  0.129543  0.129543\n",
       " 0.243669  0.243669  0.243669  0.243669     0.243669  0.243669  0.243669\n",
       " 0.261817  0.261817  0.261817  0.261817     0.261817  0.261817  0.261817\n",
       "\n",
       "[:, :, 2] =\n",
       " 0.129543  0.129543  0.129543  0.129543  …  0.129543  0.129543  0.129543\n",
       " 0.243669  0.243669  0.243669  0.243669     0.243669  0.243669  0.243669\n",
       " 0.261817  0.261817  0.261817  0.261817     0.261817  0.261817  0.261817\n",
       "\n",
       "[:, :, 3] =\n",
       " 0.129543  0.129543  0.129543  0.129543  …  0.129543  0.129543  0.129543\n",
       " 0.243669  0.243669  0.243669  0.243669     0.243669  0.243669  0.243669\n",
       " 0.261817  0.261817  0.261817  0.261817     0.261817  0.261817  0.261817\n",
       "\n",
       "...\n",
       "\n",
       "[:, :, 1022] =\n",
       " 0.0701149  0.0701149  0.0701149  …  0.0872286  0.0872286  0.0872286\n",
       " 0.17794    0.17794    0.17794       0.153544   0.153544   0.153544 \n",
       " 0.197234   0.197234   0.197234      0.161628   0.161628   0.161628 \n",
       "\n",
       "[:, :, 1023] =\n",
       " 0.0701149  0.0701149  0.0701149  …  0.0872286  0.0872286  0.0872286\n",
       " 0.17794    0.17794    0.17794       0.153544   0.153544   0.153544 \n",
       " 0.197234   0.197234   0.197234      0.161628   0.161628   0.161628 \n",
       "\n",
       "[:, :, 1024] =\n",
       " 0.0701149  0.0701149  0.0701149  …  0.0872286  0.0872286  0.0872286\n",
       " 0.17794    0.17794    0.17794       0.153544   0.153544   0.153544 \n",
       " 0.197234   0.197234   0.197234      0.161628   0.161628   0.161628 "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hdp, history = vhdp_fit(data,3,10.0,100.0,10.0,global_hyper_params,local_hyper_params,150)\n",
    "color_means = [(x.cluster_params.cluster_params.distribution.μ*25)/255 for x in hdp.global_clusters]\n",
    "glabels = create_global_labels(hdp.groups_dict[1])\n",
    "img = zeros(size(pixels_to_super1,1),size(pixels_to_super1,2),3)\n",
    "for i=1:size(pixels_to_super1,1)\n",
    "    for j=1:size(pixels_to_super1,2)\n",
    "        img[i,j,:] = color_means[glabels[(pixels_to_super1)[i,j]+1]]\n",
    "    end\n",
    "end\n",
    "image1 = img[:,:,[3,2,1]]\n",
    "image1 = permutedims(image1,[3,1,2])\n",
    "glabels = create_global_labels(hdp.groups_dict[2])\n",
    "img = zeros(size(pixels_to_super2,1),size(pixels_to_super2,2),3)\n",
    "for i=1:size(pixels_to_super2,1)\n",
    "    for j=1:size(pixels_to_super2,2)\n",
    "        img[i,j,:] = color_means[glabels[(pixels_to_super2)[i,j]+1]]\n",
    "    end\n",
    "end\n",
    "image2 = img[:,:,[3,2,1]]\n",
    "image2 = permutedims(image2,[3,1,2])\n",
    "\n",
    "glabels = create_global_labels(hdp.groups_dict[3])\n",
    "img = zeros(size(pixels_to_super3,1),size(pixels_to_super3,2),3)\n",
    "for i=1:size(pixels_to_super3,1)\n",
    "    for j=1:size(pixels_to_super3,2)\n",
    "        img[i,j,:] = color_means[glabels[(pixels_to_super3)[i,j]+1]]\n",
    "    end\n",
    "end\n",
    "image3 = img[:,:,[3,2,1]]\n",
    "image3 = permutedims(image3,[3,1,2])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "436×1024 reshape(reinterpret(RGB{Float64}, ::Array{Float64,3}), 436, 1024) with eltype RGB{Float64}:\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)   …  RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)      RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)      RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)      RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)      RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)   …  RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)      RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)      RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)      RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)      RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)   …  RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)      RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)      RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " ⋮                                          ⋱                                           \n",
       " RGB{Float64}(0.0872286,0.153544,0.161628)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.0872286,0.153544,0.161628)  …  RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.0872286,0.153544,0.161628)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.0872286,0.153544,0.161628)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.0872286,0.153544,0.161628)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.0872286,0.153544,0.161628)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.0872286,0.153544,0.161628)  …  RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.0872286,0.153544,0.161628)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.0872286,0.153544,0.161628)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.0872286,0.153544,0.161628)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.0872286,0.153544,0.161628)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.0872286,0.153544,0.161628)  …  RGB{Float64}(0.0872286,0.153544,0.161628)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "colorview(RGB,image1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "436×1024 reshape(reinterpret(RGB{Float64}, ::Array{Float64,3}), 436, 1024) with eltype RGB{Float64}:\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)  …  RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)  …  RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)  …  RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0533543,0.214307,0.24319) \n",
       " ⋮                                         ⋱                                           \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)  …  RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)  …  RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)  …  RGB{Float64}(0.0872286,0.153544,0.161628)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "colorview(RGB,image2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABAAAAAG0BAMAAABeUEh0AAAABGdBTUEAALGPC/xhBQAAACBjSFJNAAB6JgAAgIQAAPoAAACA6AAAdTAAAOpgAAA6mAAAF3CculE8AAAAJ1BMVEUhPkMONz4SLTJcMyMtMC84Pz8TExJBPDAYGBUSIyYWJykPGx3///+48TljAAAAAWJLR0QMgbNRYwAANkFJREFUeNrtnU1vG0e299sdyYq4SvQJGCG5F3d3ES9maxhNPsqsCIMtT3ZeULTjlXADh7OchWauMCvhQRJA8oaLBwJMf4I4X+5hv9epOqfqVHV1V1PiWdhiv1bV/1f/U/0eRfvwGF8PIU5OLSJ0i+1aZG02pmeH1r6IPQDdhaHNnoTWfg9Ax2FwgGEAYJMDQjforsUegEcehkYbBgA2OSB0g+5a7AF43BHvRArYA9BZmADYvePA0C26Y7EjAFiMAkO36I7FHoDHHnsAHnmYGi209GXsAegq9gA88jg15IAnoaUvgp8DQjfojkVsarU9AA86Sv01FrAH4EHH6a4AwB8EhG7S7rQax/4rtwdgZyI2DNbcYg/AjkRXtTNvdw/AEKIaq3m3gNi42aEAwB4Fhtaqk4g7q56Zq9DK7wGIRAB81y82bjS08lU8UADiMV+mTiq4MwBwLSCUkq7tz1pKBGDstwDxrgDAtYBQSroFU9C4yxrqi/AktO62MISS0iliJgCghpwVrGI3DIBLQCAp3SJmFThkFZ+EVtqagFBa2kRdSFZ/juUqmlfxF6F1tiegfzltoxl28QwgJABPQstMx8nOAlAXk6d/FLKOAwaAIiCQqhZRdeL6f8byewtgMxBKVmaM635f6z82rBIUgAEPAggCwujKjXgcCf6f/WEucVgAds4DwgjLiaxwp5nw2Y8YlHhsWC8kAAMnABwNbIH4OpS8xijVzk/+n8Z8SWMVgL4pD62xPmoT+Gb759cn4RQ2RKl0HGF9Ghc1zsBAABj3XPTQEpui6v+DTgG1jrj8qLD5JMZyXUdogU1xsu38X1dWEFpoKuJTc2wPEcA62aTw8g9+FJBl/uEPAjkAyOqybKKHGDwA4sFAaKGpYAIAKhFr5vUaofU1RuD20UZ91scq8lUxAMYh6hBa350FYGyvPKgHMgQYh6jHk9ACG+JEbriBhLP6jdKqCQSpSWiFdxKAmKm01gQGAsCT0BLrI3j7yNK3F7+qS4xP3gMwZAD8qF9UBtnWOESdQku8SwD40z/TeiAAPAmt8Q4B4E9/3AKCADBsCxgUAD71Px3vAdg1ALzqj14L2B8HDBkAz/rjEaRmoUXeA7AHYAcAiB8uAE9Cq7wHoI5xkLqFVnkPwN4C9gDsLWAPAL+Wz/LYA/AQARiz9fdIwJPQOg8AgOe7A8CzZ74JeEQAHCTShDTNp25jGADY6O+PgNA6hwDgMJN++086i6IkKQjIf4QFYGyj/7P/9gTAk9BC9whAkjwv9c8IyP89KABInudQhAXARMAXz7ogYLAAnHQBQG4CmfLb/p7/VwEwKaAIC4CV/t6SQGihewOgMvtC+TJK/ctf25nL5VJZcxAAKPr7IuBJaKX7BeA50D+dCwaQE7AF4LW8Zl8WMNYVX9XfUxJ4HABshW+SveoAiQhAMAvQAfDsWWcEhFaaitYACOO6pI7niPqNBcyeLjELCA/AFygAfpLAk9BKdwVAM65LjACUBMxyB1jKNhCHBuDZsw4BGKgFnLQFoFQU6g+GgAmIIgc8VQnYbiQwAJT+D3oUILaMS60O64H9gWgAIgDzRPaA86UQr4UNvRwmAA/4QEA0ABcAqrHeDBjA1gESnICJHoBeCAgGwAAJAPq7AFCLDB1AGPFncb6AY4AliO0xwVb+EqWQAHzRNQDDIwC2TAv9MwQkACaN/luVp8sGi/OlErWV9GABY7o+3Q4ChkeA1DKt9JcHe4IFVCovkO5fRX3eICQA3VvAsAg48QrAPKEsoJH5Au//IgDdWwANQOeDgGEhIOtvDcChHgDU7i8IAC76A4CuUC8ADOd8gEXLYAGGeSQAqtI4AMt6O+EA+KIfAJ6EFp7S3w6AAzMA06Vd9APAmKwSrb+/UeBwCDgZHgAFAi8fPABhCDjJXggthI03YgEO9HwBcNGDA5A10mSAZ9+P+S3DiAD6Nw1AGoC1A0AEVACs9c8s4PTb5MUAAfB7z3TfFqCojX84yh6AxDcAF+l/eCNgjE8mawQl/4sMgNeHikL1/jpOrJqGBkC0AB8ALF/+x/LbJPGif15MZwBWKwUAjxD0aQH8JrOqQqIAsEVgIvqA/RCgjMQHAU1JMS6wACngLwQAvjIBkOirDh3hxKnNuPpLAEAncAZguWifBMSyMgGQDAAA4NpNWAB8FXWWE2zkt6nac+yanzIYaAHANloxAErLs4AvQPenAfBDwBOof0cEtGk1XYArv7QFOBNQXj32pL89ACstAF7GAQIAyoQw3d8CAKA/kgjaDQPLUYA7AYo+pvm2AHghQDIA/xZgLX8LAOSxYOtxYGkBLzx0fwcAthngZzgI/B7bS0sSJP19W4CD/mwAEjYArhawaGMBJgAI2eQBoAEAuyZD4gnU37MFuOjvCwDxjJCbBVT6u1mAXNyYD8BfViUAPysnAv0TkAulEDF8ANQEoAHAzQKa1TsBAK9XAcDP5RAAAkDvbBw5h2QAni1gKABwLSBRAciM4EVvAHxRu/9fegJgqzjQ368FdAdAggU1BuADICy4qNbc/vGtvQnoARjHBgDy+LkXABQgdgGAAzMAogXwTV92gGlFwgvLasiayJWMUdF0AHzfutfwwqcFdAUArn8iiJ+6XA+Aiy7qVZOpAwEmANCKfdGcAlRSQF8AYBbgep2gKwBw/ctBQKG8y2EAtIvm1yJ/nsCSgLEGgDFVsUEAgIjt6gsdAXCgAYCYxde/pgWsubA/GIBljl0A+Jk3BPANgCI2TsXwANAFH4Bs4D9tTgNMawJe2FUEquwEAOs0kH8AJLG/QqYx46QLABIX/RlpYAEWX8grtrQACQCqmhIAAgNaAMbGVrOKJ6r+bgR0AYCb/I0HkGZArrio+GhTFTgjJiX7bxmAFSMDeH/5NKK/0yigCwBc9c+Fr/u16gcXJvOwv0FEkCVmV9MJAM/6C2p/1UwbBgDOBpAJKakKKLhI6RUNFvDtC0NdbLqscB4gIAAVAV+pkwID0EJ/Rda8a1+UAKQaAKbaA4FvSTJKmWMbAL7AAWjVaM4EfAUmPTQAir5dPSaI3VIOLSDBCcgmv9BWxqqWCgA/G4eAHb1i/WtwmTBysgDvABx4BmBSPTp6keoImNajxEb36o9vNQSMszLH7QBg6N/XRxYGAIBn/bPTx+elAZBnkbIzgc1xYqH5i29rGL6FYKi1ocDA42j1TDUAh0ZT9uIBE3sLcAEg1ZTAtwEUb4u6SMs3AxAAgCPFrK+XM14IABAWEOEGMFauBYld3gcAscRZ7MUo+gFg1iMA9CVkDIDqdFA1Jx//V38TysTUDEr/lX0GoABoCJB+9mYBTgBoLMC7/vAuEjMAi6yvN7O+bf7ELYB89SzUgtafMwTEAIAzYnI5Jw/go+AZAP8GYHiyVAEgtwB8MVznmKgmAOBIr78xA9AAjIW/vQCQEfBVbgasfHDiBMCsTwAEApo3CSblC+WmCgALOwDoEAnQJwBrAGLho/UwB5n1xW9UgQQAN+gCANICOgGgSQLVLSH5v4tCexmA7Iwwvp0XmjNCegCO/AKwVbwRHVqQSX0eJKIbdAMAhUAnADQesCwuBGeqT8nLRgvDxl5Y668xAFcAWMvBQEvWmgBnAGa9ATCZAw8A+d7uqmHpJy+ZFa0b+sgEwPfdA0Cg2ZqAHQAgSefCkQDs7Ej/N1nAhPd+IaYBrHgGwAaAEJezDB6mgwFnAJAccPD8AG/x1gSkBABYApiacgDzDWMWAHxvx5M9ADFjGUcC7PUnADhI6FuB5y0BEAzg3ATA0rg17ivm6qodmQBgbIwLAJ4DOMuQ4dsAKgBmMgCa7uvPAiS131oSME/YOeC0vitMp39+HsCGJg8A2DmAngAH/SsAZAvQadfWAhKi/7+9vr7+lw0AeUnYLxlkGEBGwPfmLUHNbAGIdRtrRcBJGwCABRzopWsLwAQfAGQAyAgYtjOXjwM0nyBiGEAGABslvEObAdDS1IoAF/1rAHILOHiu7f9V120LQImADMD1tUqAyUrmcBSg648cA9iOArwCgKnbGoDIpwFAAKLyG9AGAObtCcBeK38tE7AwnwVI8y2JjakXw6T/arUDADzxaACn4AtApst/qTcLQN4r/7YC4LoGwLiRbRLITiu8FJSJdbIZDYBFQGgAcAJO2gKQzg6Mze3RAhLZAhoA/mUBQBEvhbbUycbQf/U/xjaDmrUEYFswewb8GcCp4RMwnQJwgWaA2gKYFOUFEnpmTGvBMQCGBfABIG4eYyxlbQEnXgDQu/uEsxA7pksiA5QWwNpIWaT5C0EXWjWW/mYA+CmAvHdIAxSTgCcdAaATl0WJIvOCB8A1AGDB3X5OwCS/MKwHYDubZwDbHJBXEhOzDQDFTz/6Iy7gqL8KwNys/5yrTpJf4mUBIBrANgewAajOTb5oGjvGazpmGsBqVVYT0TJ2ByA2L9KGgBNvAJDy8tIEAsCy+F/WdEoaQDYIYO9gUhbICADXAGoAXqq91BqAsXExx/4vE+ARAEpfawCWlcyLZAEeFTUD8C++A0xKZl8Y2ppvAKv3qgUAtlwAiKkl3PUHBLjq7wLAnCn/tL7Dq7rSD2Sd0hnABoCkKpHQldCasg0AyQHVthEAYm3zjrVFagtAQ8CJRwAofSdsB6hu8lKDNAAJgGsLAKo7FEyi8PWvLOClopI1ANXS5FJjjs6a6AAAndeyAFjWN/spsYCLURng+pqvfx3P9aL8pwUAKwmAMdwwUIDRxmPTCep6684m4Ky/BQDME8HT3O+pV8RUBGwl9g2AYAFYRW30Ly2gBkAiC/RaTiOPeQC0OC/sEYC5CYC5QYdSaByAKglIF/3e+gDgudCSSlgZQJUDZP3LLQOZYve2VwFwywhPegVAr38jNAHAIjOBqzPpsu/QAFiJAKj9PVInuQsAt+QCQCsM+WcCJwwApoLQZCzOyqF+cwvQtQ8AUtCSUtjpX1qA3EVRAOJ2H74dw407WUAbBKJJNwDQIfd2zADcAJjRAFAG8EoPgCIHAsChCMB39jRIG3cFwNWKJAA08k5sMoAmrjnhAsCkAgDpDJTKWgCQ67Ryp91GM1ZI5aNHOwBihK7uPSDin+DxA8Bbjv5nTgCkZENAAzhqRKYAyAcBVC8DADRjhe9SJwLGsn79EhDxb/TaWkBiOAj0BICL/Ek9CEBqCbS9PFq9r8/2aQCYIW2sSCQoXgNg6wFjWOxx5BouBETMvp0vNs8e7RsoAJNqEGAygMvoqD7brwUAa2IZAFFwOI6yUwEU25kAZwAS1hnerI0NYR4EsjJAubUzCxLm+VOnHAPIbgxKGQBgBiADcJgKx4vqQNqGABkIt3AFwM8jPxwA+Ppnf12x95wm+WPH5nMAl1vZGgMgBwHEI/OSPvVmIADzqlQv2DrAzuueBZwBmLCSgDF8GMCVEwD5d8zTWYzUUTIAcOqDBGCmad9aHzHrf5eqBNh3yLYWYE1AJLSgBwJMFsAxgEL1M7vRQAUAVkdJf3juyw6AmAAgBfrXZuoMQJurxLHdjsQm9JAEfAFwZjkc3Io/b47ISAAusxZ6ZQbgfaprXAQAKUICYOcCwrMgXizAwxDwqtHfLgcwAIgUAPAc8B43APl68EABsCEggm3YmoDWABSSn13bErC1gAlxCC4ZAAQAt4D3Wns1AzBvCUA7/W2yAHwarHUSMIwBLBOA5SggmeMESAYQvXovaoVagLZlx50DMI7aBhsB6XFAw4kepgMsCCdIuACoxwQcAOYJngQkA4jS91AtCwAOYfc0APAiHADsLOD/iwAFAAgBC9CzyYAGYJMDiheQqRYg65quFMEkG7ik9M8P+XYCAC4BnQAgfRS01j/hWIArAEkJgGoB/6kA8F5RLE/8r0wGUC68EwDEIQFYqADk+nMs4ErCxAaAhAXAqxVuAanJAOpTvzMTAEWJXPX380mKkAAsk+UU059lAaoj8GJSAvCSBqBQdoUAkGbH/mwAKgK6OgjwAgDPAnwDUOq+kA2gmm8PwJXN7g2DgFzZo9UKyQGvVvVokNS/AaAkoKMM4OnLhGEAKG8GJAyANQxsCYCaA1QAEAt4XwGg6H+I9fiZDoCiMHy91bt6fRDgAMDUprHxWOQISEOARTO/WwA4g4CjFU5AAQDR7yUDKADwlQGwJ3t8EBBbA8B5LNv4stjFdLok9e9yEGAFADEQxPy/EhwsPKNHgGUDveDrj97Y74EAhgnYAjCZF8fb+qWmCaW/Qw6w0R9PAacAgPoHZgAIAFW/n0kAHHoCgHq2y4cHRLGJgQaAaf40B6P/p2kOgQGBabbFKRgA9gAAfhjQDgBc5FnqKQNonu70gwATgCnSW9EGnheV5J00RjYZYhRY54BL8U1R75EMoI4B0I4+izxlgG90D/j3CsCSC4BUT31Ml4qphADgtAUAmMSRxv+tjgG+Mb7iwwcEZgCmeL5uDwDyahjrHKABAJmF54AjDAAsAygE6JTWZQCGAfDe8uTFBhoIZEvIAADP8XF6GEh1UvsrFiAv4QuAs+srbISAWsC4Vt0aAGv9uQB8A/Xv1AIaAsayJUTKdTs2AKgFmIcGvgAg5k6wmwKaFwRFR9QgsL4SdNkSgKIcxv4vA9C1BaiPHhIAmF7RLtRUFXvSFwBn5Fzx5cFNG1b9PsoeDEKHACsUgM4M4FTWv3MLQL5RuM0GkXzS1jAMEAGYz7HZJgCsR4EOAKRKGzbGX1/1hQCsMAAOD10NwAEAigBfABCB3rihCT0ACWNo6BcA9TQBCoDwkrhKbAjAq2aBahhgL37KPwnwjQoAlQQGC0CCEDAAALIivjxNpde8gYM/3RCgBsCh91sYAAZA10mACYBeviYHzCciAKX5DwSAPEALHokAFPd/kABcuuvPNQAUAJKAXgEwyDdJq08/Juk8BdPz+7LNz5DaAnCt3wrNKGjBoyYD5KHcErCCFuAkv8UDASgA9Mue+gNgYQagbOFJKoz4JkIjzPVbOPMKgFw8CoCoOAhcvaKkeyXq79b9U4vTwDgAlAV0mAVsAShuvavauNJ6grSCNwCuNFuR502qOwNT+VWvhf4rWrsVCDf9510B0J0FWAMABK87e0sAtIlBA4Aya1J+RyrNLtjEYtsdrYibwRAL0C3GMADOdQDLFDA0AJTXhrYDQO8Lmq2obIgAwLiMDMq2B4BvAEMEYJFYACB/Q9ACAFXrfKqd/uTjw2n1eID6gPeRwdvbAlCXgaG/3VFAXwDYBGQeADA3rXuGKXxmpT8JwHYQUI5Q5JoakvurlmOAptYcA9h5ACaw2hO8KVgEXLUB4EolszxKpQBA+7akvoMFCJXm6I8DQEnf01GAOwATqi0YBFypkywAwM8E6wBAnwlAwk5/oQAsA+CfCRz3B4AFAXUrl2LbAoCcybM5BNABkEwIBziiCMD0tzQAcf8s/S1OBUfZVbt+ALD7VJNYc2sAVAEdAcDmF8WYUQDInRs1AL4FSNXlGYDFxaBMe+RSbicA8McBUu0nmhYhQtavcwAiVNsVGdncl+Lb3xjq828G/4ZrAD1fDXwrNyjCRHYHgWR+AAAmQ1A+qyHAXA/ABNVfvCfAqH4DwKmN/PzHgSwA6K73IwDITS7lhfrukUVaXAzABgFMAM7AvqwASJPaMbDZBY46AN6nmhdFC6OADIDvOInfVn8MAHLZTgFI4D1hcps2eQHeOZQBAN4sZ5sBcgnBD4sMkJ3t+xvNxwQfA7C+HFgvlFYAnPKkz4Lr/1lYnATo1AKgtm/zVhX6pjA2hABM8iu/2PVArv7JmaiujQFM8tcBbef+k4AtxSxgdWlWP4ogANJbgF/qMbeQ3wqATgmAvfut3PJLygIS5QYQWwOAYZUBirg2fORSMgATAPVy4nHgKSBAAEDZtU3vRwHQLt0dAgkQV2n6JWUB6ivlJpYGwAHgClt0QqsAAJhBA9ABcAkWVAAoxgEwGwDQbdVHADgxLN8tAKW6b5XONxUIABaAPCE8aaF/Ym8Amr0hg4AjAoD328EgbBFwnCDLgAHgoL4DAF0RUFdDAeBK7Pbi3xQArV433AEAYg6ovhODDfUvQYMciQagAPCdAoCT+ggA5jU6BiBZym/yBN1eJgADYNLiXbP4YeCVGwDSIAAXv+7mGgBeKioAAtz6PgqA0QBOOxoICK0mv8r3ShQ9EX/kvzsGIP+6ZDsAZmXv1+uPArCdnL1RVgWgIqBF33cGoBMEdCLkE6fiSeKpDoBWYwAVgOT6yggAxVszCNCN+1P8aKE0gFWKv3w8R6BV58cA4K7WKQByJl6KmksWUNwdigjjSgCS/6/AmQIMAMMg4GhlCORo4aia/p74/oSf+MbFALrwAF1HRK4UCkcFKSEM1SttAcjER0eBKcMCirkm/XOtZ1KLlAbw6n2Utv0wcAcAeLcAjRD/wi4VmwFwNIFmv2e1BWi7t+lMwCuO/NjHYfKvim0PDg9n/QFgsWbXACQEAIAADABzt2QCQD36jwIwp5bhyY9+Heho6wCraGsAbT8OrwtnADwToApxdYYDsCjmTysA5rQsrQC4ou/3QwCg9vV/uABgnwdavX91mf/RFwAnVqt2C8BVPRCQANiqvqgIsO2W1XsjNOYg5n49AJx70Nn649+Hqq4MdAjAqasBeCYAA6BUQAZguqy+BbEwqKICUN84QhKgAkAOAlJsX9UfixzTlgBU0SUApwMFoPloCwZA3vdRACY6AJLmxhGKAAmAfxIATCYSAGn1PGi+4aqMbP3JD4R1D0AzEDwZEABnzWd7xPQ/rQBYFr3MQAAyc153XhyA+gC0+Puff6Nu95LGALnyOQDZhusivxE0vtSdDnq/0rROtwDUBFiuNu4DgEQEIO9a4MqgNQFFD9UN3cHY//oaADARn0NGAKj3ulQJuIz05wMDAlC+JzpkBkDPAxR/TAEAyXRa2wGDgLlmHmECZ2Lav75OxRwwKbc4key/2Ne8AaspM1BXI//lUUAAnMKrAUSYDtWf8AjQCgCVgNRIwLbL133+DOaAapVJisW8nlwXefWm6f4GB9A2T2ixAwCQnDV/FpovpoXo0hkhDQDzBHmHoOLb6p5rwSe5BdQ8TOo1DADMpf7PAeByD4AmSsVzAqZsAFL9XJKApsfPMwu4AqvOSQcQQgTg79d/v94D0BKA6rhK/BgkA4A5Ot9IwJWw5PX1XNowZQBNnF9AAK4rAI5cAYhDq40TcJqh2Q8AyVRI+4ZBgKF/8x8hzGb/7Vq+49gMQG0AGQD/mwHwj4cIQBX9ACCe+jUcCJpGecz3SJRG8jfhrv9yk6YMcHHeHAQWB5VtARhODhirRekJAPGd71oLmJgBYH1voFpAsIByBVMGEE5blgBcmwHYkcMApCy9AZBIAFAWwACgWYY0AGwokU8yGMC5kAGWy7/zAdBaQBxaeFHsYACURrAwfFlowgBgYloApYQFALhw8RYA4H4YMBALGKOF6RmApHyRmGoBkxSc6VX1ld4pawJA8oBJeRxoGgKIFnANxgA7DwDhRz0DkDTfBCZ6bBkzabOTVCFgjm2/Xg6CNKlgUASfky5gAYB+EBA7yOU9xgSO46h1WAHQfBJ0gXfYSv/oQAZkLi1O9nyJgHzyPP8QzBYA5WySvOvl8hxkgIcGgFyangEoTgpMp5IFqPpDAGTFJ9QQsFmwFjZfcJ7My+mpAYDzZYoCYHUYAJs1tPhA59AAZAQUACwl2aD+cLuK5+uPEqHQ9U4w2uaK+1xAAKpBoNVhAHTW0OIDneUZ46htWAKQf2EW3haI6i9aQJ3CjQ5QqTuZEAAoYwAEgPP2AAzOAuiytCbAFoDsrNASWECKAiAQ0Eg5bybgG5flFJZLidk8AKwOA6RmjUPLXwGQYk+q9A5AgUAzDpRFUbfc9FdhJK+3AASACXomUJ52XqaA+g7DfzAsQAVgaBaQA3mIPqwaAICt9sKRgNIrqy3XFpDigSIwIReaoNcC5EPD6nSgFQDSKDA+lZ01tP46AMZRu3AAoD4dgMqrbpoAACVAlliYNccBgNPOy1MBCgDaHAD7kQJAHFr/onzpYACYLmkAZs2mD3BNtQSQGYDalAJA6gSA2Ixx1eIDAmBMAtA2B1hqX0ginA2SNVG3TQOA3jxcxwRZANuCPIUA4EgLQNWMWUsPEYDT8kNWBBu9ATApNBFOCEvy/nh7I1tAqos52cfnwnnBZvZc2YAOgKuz+iiAB0D+h9KsQwBgPAwAKuNuAJD1FwHICdAZAJIHwGQ1SbgDoMsB/yNcbh/HVZdrIrT6eXlyAF5ic3oGQEgCCwmArf63t+LWD8x3caXIHsy3CwmrowC8vS4ecfkHF4Cx7jpLaPHz8gwCgEklWX1zmNj0t7cyANGByQDwcd7cUAJxdXlKAcBV9mzB9p+mKEdaALTtGlp8PQDtCHABQEwCTfPf3iIARLYANIwxAUAuCNanKbcAPHcGQLCAOLT6p1UKIOb1BoDwIqiSgLr5K/3BIIDzEdY5tgtTCYRQEROvVDUARJYACM06AADG9BggAABCErjIf/94KwTcgREA1ALoEijDwES3xelrdwAECwgtf30Y+BKZ0S6SX10AmDcWALu/iwVgLj8nS6AeB2i3yAJgtQsARCgAY7PE+tgK5gCAbAE/tiFgjuyEBkC9XKAFYJl/E8UAwCXZ6EMBYFwCENNldAfAxgNEzUQLgATAXRgJwDTmA6DfYIHpaz0BRgDi08AR5QAgBfEBgIUFVI0/b9q29ADaAQwEzLBxCPOucQ5RRRmzYsxsARgPBoC8JPl1lk4A4FsAdO2l4AE/Et0fMpCgNxBhO5p7AqCkNLOAVAMArvEOADCmpT1WeiIFAN8DcACABeh2diAdPJTXDg+QPekAMCYBxQByAuh3h16OIzTPN607CADSyCYHHCNeTAKAmcB0CW/9qlo+P/su9i4wDtTuUoYmbabbhBGAhp6pCABpAZeUxsMGQCvssVENAABiAlXTLRQAEmkQIB4KancmrAIAOEjswjcA+S1BMdLwdcljlk49ANCgyNHVBgDZB+qmk24AzhsXywE/mvcoM4PcQerHAggAXnkHYNyPNyAA8GTNrYAPgECAoH/5buh5cQ4ml7569SvszxkAxxoEDkgAnC5JcABYNgCkNACXlMxGAHpKDgIAsY3+t+XfN4xFYSYA+otX/SalCcPFLqrZ2hwgKlIuL7DhE4C5Wo38PIrOAcrWHYtiN2UnhYlPe4i8CEV3iRn6HwtH5MeGxCwD8CsKQCVYnVwn1WILmJJ/1LAGvlJ6AR3AzgJcAJilGgsQdAaSjg0ARH0CIJSEkF0FwHxodktYgAxArhguh9jyM7Dxg+x6/AEUGRIF6fAGQFVSawAksRsAUJ3HPQEwlkpH9nuPAEzZAEA1lC4vXJE9EAG4kBf3qb8KQJmkLAEwDALGtDd0B0AUST8F1W8IVW0A+JWwAJABZAPIH/dTLKBaoiLgQM0BEbK4DwDmsBL19FccAGKk6YcEgEZJnIAb02pmC7hISQBKSapHNBoCDoQFGoWnQBORFzYB1gCkXQIQszRsFRz9/QGAWUCq1x/oMivFhCP7wwoIeCAIADhIeGHWv8oB0jEKmQNIAJqJuwEAPgowryYTkEEjZU82AIWtHzwHckaHdQevtjtRAJAt4K+bzf9zMwB4tSLddQDGESP8AXBbAPBUTqDMnjkrOvNzSbRZBUR9bkEZBMgWsMnCTf+ivBPFAIgUcGkGIKaE6V5/HgHA7dsBcJsDEMkZwEBAIcx8MkN7cw1AvdGFCoBEwAYjgK1/drvoVnygP3UU4AhAOb2H88EMAoDYLQEo1qktYHlxfsEGIMGv8VYANGNLxAGEta7yDKAQYCG/EuSVABYAmMjlwj24AAOAYxqAG3sA8lWUYaDB/Mvcmx5ko37FAeo/AQAzqTzFIvkzHRUANQGRdI/xzBYBJgDomSAKAGqe5wgCwFMNANJ3neueOaf8oZ5ReUCCWkBSfT6qygAFAdXsQwDAIVt6gwuQAOhzQI8AmC0AABB5ASAj4PVTAYBaRPWzfkYA0qlEwKKgRrKA+vNhIgCberYEgLUH8E4FxljDx7QoMVdG/wA0Yh0DsdsCUIL09HVUZoJUsADs2756ALJThdMFBKDyDVies+rzcVd/xQAQc8Asjew9gAUA82qAbl4vABwDeUWxW6cAMHBMz8+LsfQ8/6LQNQbARCQEHSPUAFRHAeV0YAFfNt+PFPTfrBELmEWHnQEQIw0/PACg6YPDgFu/AKTZ69fyNv/yGnziUwEAN4B8ZvXVgeYwoFxH3FXzDfFrHIDGAjJwZownz1gAXMrtwhIZzuuUAgQA6Png561fAA4rAER98EEA4QCJAECTAybFDMECri0B4Dx66AZAriluChQAnfoAJX+lrig2MaJzB6Bu5S+NANBjwHQqfXpoWa3X7Ocfwsb/vUHHAE0OmMGfvDh/s1ryAMANgeqVBS49AnAsSXWrCxcA4DpmAFINANP8oL9xgNICqhWf2hhAQ6PEAy+y/b5hDAIwP9Da8mnHKUAjXB8AHLYCoDghf44DsP3xtHh2T3i3lwTABilLZRx2BlACsLQFQD8GjIo3jAcC4CZC7gGi7ZwJgLRYCQDoo80gcK4FYFqeTl5CJMqMUMwtTjpRGQBaQOQOQJF7Vr+oLmAEgBGVVoSG48j9cGGs7AzkAK38hnv1j1kAFN0uAhZwVX1kdiJ+vZcFQNJYgHi++S1hANACisLMQMksDGC5/GW5XHYBQKwDoJHQDwAG0fkAEFtCltu2OUwBlQcITcx0gOZt00sHACLBAFwAWCoARF4AKBGoXjSnExDMcMkAJte3ASDiryMB0AzxWQAsTAAsSQDWcklmLgAsSQCOPBGQqxVrhaPW6g6AW2MB2ACoo8BMR6GJEyYAcJ4CwL83G60FiOFkABCAyy0Al34I2JqA8GjJ2IabKm5vWQDw9Wd4wDF7DeFcbXbNflrf2UEaQHVjHuoAUP8mB3QCwFLnAL4AAJJyAaiQqZpfyh7thgA8F2Aj8yWwf+FCUXYHjh4AZBBIAWDOATYAlJcyxD29kS3AMwBRea+QzdJC+5tGgDYJwMkENIuL47/S3cvuP8HPBJcALGUHWC7Ux08q6RUCaAAYg4AC0nMaAE8jgFYRAwH+rz4DWOvPGAgcM5cWDaAUsMz/fgAoFf83NweYAciUP5d2JAFwmW8qOxkxEAC2oau5vf4MAiLOwl+CA8ClCEDa3PWDDQJFAErhF7L+7zZkrMVi3hgAABNz7WUAtoOAX+AooLwDavFcrO6yHyTyh5JlwdbIgselPbsAAEQ9xvS/OUaXxQBIEADmk0QDQJn0wXpKbBgAVIWsIEAAAFeJz6mdKSeDykIWmz1IqnviXneufzEE/C8o153S9R18H8QN4EYtRtaiZrv4UjCAUtpyDJBQUXd0vfp6ADZIMxQMIAbQTJ2l9N6kNPBLaUuF/NkjjeVynRNQHgNAAu4E0xtt699WfrMr3DTNeyOzJ/74R33LLwAA1f4WeMABNAQLAD7e3iGdICsmagBRMSP745B2gCVmAEUOyPV/Wi3XNQBRcdgIAZCboGsAbgoA0DdKHMtTsAyA6l88aegDAPSaR1ZktP8LoQHgDWIAdaKKxPfL9BBS1e76BaDa+k1REqRsNADVUSCmf/Gwcan563YAoHFziPf/Jp7Su9MCIBS0FwAkAjYEADdR+8EA5gDiBDkDHFc7ruNABQA3gMIC2gNAE3DbAoD8WGBLwRuQAtSh6nAAuKn18A1AVQasZMfqomILkRngh9uSgB9EAHR6aMaAVA1+08sPn28ibODNGxGAxTABiOoXsHnUn3FyIJKeOxUAmIoAzCkAtlngh+XrbEH4njApfrr9fbm8I/WnHADqjxbfSEB5f0j1e6AAbPDlAgEgvPGLHAL8IO7maQUAlQEcAQAEzJwAWAIAmgxQlvQ+yCAQbQnv8vMAOMaWPUjk1y/oAaix0QHwkwMAIgQ4AE+NBAgALIQK5BPerZf9nB8+hnWyyoPdAlAtewOmbgX+6Xfx7RvoEFBa1QSAwxgAEIBXwAYA5ZLluy1AfegvtjSlv7YXdAbAMbHwtof/Lr59RWMA9arbyU9pAH663TgSoAfAbAHiGEACYBP1FKBCLr2gbwDEhkcAgPu5qQB4Tqmhq7ex8tpBIMMC3gjnASAA79oCkF3F0FzMRuXHRwCdWIAVALfKZHEEPqf7f71q9qC/DoA7B/FvwRgA/0xGNsDkAgAfX2BoZ9A/D+NWYIV6098SgBu5xL9hAPzwq2IA4o6oDNCi5lUxfqQqdZzlFw0E4lEgOBP0DhePK+uIuyCEwLEfdA/ArVpamYBM/V9V/QV2tADcOgFQFgAFVSgtTcDqzUr4JVhAvndSV1N6YC4Gt/yRzoRhAMBzANX8P9T/UHvSA4DTrwfgN0l/pF5FHUgAfgG/mkT213YAlBDw9c+R6VN/3omgY3XxY3EjYvsn1J5umADcYh1AD8C2AD/+Bj9deIPWwDASUAD4KyHzxgKAyNIAdBEKgEhdHADwm9IBWwFw56fqWPn5ADxvZF7TAPAsgA+AXv9Ag0BRcLVJbysD+NEbAD4PgiVe+QBEuhGcJQBrxlJ5GAzAAwIf/nAEQL4iKGyiPhLQI2ABADUMLtlo0wCsJFDo38gsCyjoxBKV4EJd3QxAKwI+/LkNVwAieIgtbCLlAVDvigFAl72AgUD5gQNSZj4AlIUIG1EX76r2f+bxhzsA4gfIFPkZBVAAaNT46XdxwTtDG3QHwOtt+V6rctAArLUNNsLXh3PX2HY7YOBDAYBkAVYAiCiU8Ruv9+dxIwGQTVtmF5Qk/+/UBrXDAKz7IjJvyDkmiYnNqNv13wEq/SULaAtAQQCzDMW6T6Hp/16eBebr3woAjQFIHZWWmQvASLsUt7/7AeDPJvwCcMvs/sLO0GFfHXchAHiHyCQItKZm6AcB+sWc9XepvqA/tABHAJzuS7mpVn5aZwA3AFrof0vrrwFA03WZAKzleVb9/88/P7cD4IMIALAAVwCOHdq+WfspbQAdA4AbAKaSqJAjADoL4Gr/qRStXfUhAH+WjvBHCwAcLOBGXB0agMhTtwAgvR/ci1wzAHsoDcBanKj08w2yaLk1O/1FAlzqDQn4o0kGzgAcW5cBrA4NoPrkGROAj876/07ZvyIonKoHAOEH7+flFBUwBgBNGnCqueQBdYQCQJh+U/0s/+8UgFuD/pk8mDYkACg9FAAOAdT67A7ArW8A7HMAsfZNPan80vVdYADw4EnqHwBoAnkm2FUAbvC1FV/pEgB1EMDWYa3v/NKCHgGQCPjsVnvvKaDdW4qO8cnMHNACgN+dASi6N28pvwBsJM3+GAYA1hbABuDODYA7MxpKBrh/Z6HDmqmodwA+SaLtKACi2VMAFNNNDULqvzEkyN9/cu//ubLc5coY+QFAUc2FgPBHAYLUx+jUevqdCwC6eXWYDgL1sWYu5q/vDxiAVm+qO0anKmJaAFAz8xHfdTHn7r7FEMAqRl3q/4c//dukgELFG4tcQKQPdZI9APq59exWBsCNtT/zx+TnASDf+9UBAM2LQ7gA3EjrghmtALhj4tELAH7jkxsAcp7oAoD6fVJcAIhhoErAnbZJPmr1V2aLM/vKAD5DFY2lv4RJJwDkKkZWwwEcgBspCVgCcKebLc683z0DcALgg+oUHzoBIJPxxm48WK2pyC9uRg+Ask1K/uzqEtjU/Q4aAJIDTG2MLdgRAOVbpSwAQKI6e+oKwB01c0sW3NL9DhqA/VHgB2zJblJAVL1Sro3+mUZyUnAG4KM4I5JH44MH4JMPALDh4ofOAFAd3d4ANsU1WEG8O10roRtQ5xeZRQdAaLkRqT9jVNgB8AED4M+uAXAj4G4jau0LgI14uDESZwxmCIAoXcqMLw3iD02LZjMxVroHIHI5N0gqq209IwDk9ZihOAAhNgXAJwMBgtJ/Nvd7iYv+2QsA9gRIFa0t4E7bfB+lrahLr8VSkQC8CwZA0/6f5amG5REExCl/ygAUYJAjAM8AWBMgV5RlAIoFGAAQssBQAGi69Gdlkn55BYAPotT5j1t1SVp/jwBYnRQmhP5ITCcWYwLQTL8fhv5FfK7/MQKgPxBopuX/Sd39D13/9wnAse0pIRoAY9u5AjAQAxAAQDTGlv0kaSqGVl9jBPR/EgBz20ELMAEwwgEIrL8VANACsAQwAAAcDgWVerLk30gWoM5mOcBQAeCMAhT7HwYA1gSoFf14x2o80QKQNQgA7gcEwCf5TIAeAM25gCEBYIuAe/vpNwLLhAPwLqj6qtgGAD6p2nmBwDcAxz0B8NEFgEGMAaGSbAA2GhH/aDEY2FUH2Oi2sYZlGqEABNIffdgLcsEBx1vsqgPojxjRoo02QzAASUlk6mdivd0AwI6ANu2oWx8H4H4QBgCEbMQ2A9CNBXgHwIqAVg2pu2SAlmwYGYAAwHAeQF5xwADYDAPataRu9brbEwC8CwcA6MnIZN6KAwbAwgK6a+V1NfYjDCAgACYH+Kxb2TsEHQBgYQEdNvN6VJNQxNOBAAAlFKZ/LudarD1QAOTv33wMAUAduAME0x9qaL/2TgBAfgArJABPh2EAnGTPWntnALgTSh8CgDXmAO/62HMnsRsAHO8B6Cx2AgCRALHwHwMAUOeAgQwB2sZuAEB9AS0kAE8fBgCec0BXABzjGn/cA9A2dgSACDUA2QJ6abF1WaKHMQTwnQM6A+AYB+BjMACe7gHoFYAIlzgYAIM5DdA6/OaA7gAgXvnXu/7lIODhAKDcUdKKiA4BiFCJBQu466nBHjQA5bTPG8fc0CUAxzgAH3s2gCIHAABCS9guxB7/mZg+CAAiuo/3qX8OwOsHMwbUX00aFgBeX5fZIoZzLdhLCAB8NswPDcBACJC/VvoudIFaBm0A2DJBARiFbqs8HpYBcG8eGgQAw7AACYDQxWkdtcfzlgoKwCh0Y+Vx/zABMCw2CACGQcDyYQGw4QHAIqBzAIaQBO73AAQEYBS6tXbg/ZAusTMADICABwnAdiRgWGAYY4BoAElguXxoOYARvIOAXgAYBW6K+0cJAE//XgAITYCs/0PJAYYYEACBk8AegOAABCVAyQD7HNA/AKOADbEHYAAAhCTgsQLAOwzoC4CASeCxAjAsBwhoAXsABgFAQAt4pAAM5lTwAC0gtDI9BWcU0B8A4Qh4tA7AsYAeAQiVBJAxwLvQyvQUAwNgFKYVlsvHSsDAUkAoD7jHCHgMWYCjf88AjEK0w7a73z9KAgYIQLCB4KMEgKN/NwCMiJf1hiPgcQ4DwjlAVQLNrH7jceYAThLoBoCqp6+HQsAjBYBhAl0BUBCAARCGgMcKgNEDOgMg371mVs/xOAcBG+kzxV0AsMYnj/K9EzMDtMMjPh1YmUA9wSsA5TYF4deRMNbHCRjZ1cBHPNpBwEZ9juiTLwBGYDe4tpw1e4lHD0DzuyMANqOsw0vTRGsIS8AjzgHlAKAObwBI+xlhO18LC2vW7SEerwPIjxH2CQBYeC2sO9r0HfePFwApfAEwEje6Jrq1AMCaXruPuH+0KUAJTwCIm6SdnUgB/SNwv9e/Cj8AjMRN0qqK0wVKCMPoMu73CaCMDgDY4FM30tHhWvgRIA3c7/XP41MHAKDnAdZSuliDX5tQF4cft/9vOgIAUbPRXwIg9CsjHr0FdAOAJOsaXXY9FAACFyFsdJICpFiDhUPXWIj9GADkgJ4AGHGK1Uvct9V/PaDKuEYfAGysFu4x2g8AB1QZ1/ABgHEnwwTAwwHAcCrjHH04QEbAZmD633s5/htKbVrEJw8AcAd26+Ho7yny6qx3u1I9ArBtq9C19RvW9R9ifPIAwIi7s90GANZ07dABBhl9ArDbTdVIvYb1D12wdvGpPQDcFhiNQle2VdTVXTvUfcDxqScA1js+BmxkXzd1zmu025mtSgJtABix9rPjAGC4PxAX6AmA3Y41rOrDSgOtHwyRN/gwmgUGTGH2KXDQ8ckvAA+mXUCMqqqNNob672S0AgB/AnAUuk7ugY3rGt3XcvVDF9dHtAOAeBfIKHStnGId6W9sVwDYzWqS1XMEYI1OD10t14bQttAeADYA7KYZyt0VmmILVK8daznkaAkAsQV206zDu0Uj7PZvteBroaoP7TCwqH4XMbLbf7jqryPhTM8aLclaqdJD0b6qTHgAWreoox1n2o6MBVcL6LSzgUZgANb2yKgVcH3ObI2UnFoMTHEu7ABjVwHgKMffu34zykjVtrBDjsAAuArYYpf6uvNW7FCQvmMHAViju/RUc14Byg2wqzngGAwA/MbEAWhtIeytrEGhd90NugGAu3fYlvYrOUFHboG3EXnl/rTqJAYDAL8hfe7UAQDX3Q40ugGA0Y55eAWAudOW5DpzN9AIC4C4e+4q63a71RfcYe8hVPMYDwqAViu7biGIbP6iGwDYu2+ac+ShyPy9tSg4wHy9BwCLEXv3jSQeimzarVl/m6OAAGJ1EYEBsOp7ZazddsuQn19yfg2HHp0AwN67IAq/SdcOu2WpvwegZwAcqaHKPCJX4Mr/GAH4/5cKEfbRanX6AAAAAElFTkSuQmCC",
      "text/plain": [
       "436×1024 reshape(reinterpret(RGB{Float64}, ::Array{Float64,3}), 436, 1024) with eltype RGB{Float64}:\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)  …  RGB{Float64}(0.0701149,0.17794,0.197234) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0701149,0.17794,0.197234) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0701149,0.17794,0.197234) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0701149,0.17794,0.197234) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0701149,0.17794,0.197234) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)  …  RGB{Float64}(0.0701149,0.17794,0.197234) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0701149,0.17794,0.197234) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0701149,0.17794,0.197234) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0701149,0.17794,0.197234) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0701149,0.17794,0.197234) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)  …  RGB{Float64}(0.0701149,0.17794,0.197234) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0701149,0.17794,0.197234) \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0701149,0.17794,0.197234) \n",
       " ⋮                                         ⋱                                           \n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)  …  RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)  …  RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)     RGB{Float64}(0.0872286,0.153544,0.161628)\n",
       " RGB{Float64}(0.129543,0.243669,0.261817)  …  RGB{Float64}(0.0872286,0.153544,0.161628)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "colorview(RGB,image3)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Julia 1.3.0",
   "language": "julia",
   "name": "julia-1.3"
  },
  "language_info": {
   "file_extension": ".jl",
   "mimetype": "application/julia",
   "name": "julia",
   "version": "1.3.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}