Parallel_Kernels.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Matplotlib is building the font cache; this may take a moment.\n"
     ]
    }
   ],
   "source": [
    "#Import potrzebnych bilbiotek\n",
    "import pynq\n",
    "import os\n",
    "import re\n",
    "import time\n",
    "import numpy as np\n",
    "import sys\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "class wait_handles:\n",
    "    \"\"\"\n",
    "    Create wait handles for multiple accelerators.\n",
    "    \"\"\"\n",
    "    def __init__(self, handles_list):\n",
    "        self.handles = handles_list\n",
    "\n",
    "    def wait(self):\n",
    "        for i in range(len(self.handles)):\n",
    "            self.handles[i].wait()\n",
    "            \n",
    "class MyOverlay(pynq.Overlay):\n",
    "    \"\"\"\n",
    "    Default init Overlay and run some functions for setup.\n",
    "    \"\"\"\n",
    "    #delete bitstream argument and change download to True in future use!\n",
    "    def __init__(self, xclbin, bitstream, download = False):\n",
    "        self.bitstream = bitstream\n",
    "        self.xclbin = xclbin\n",
    "        #Delete for future use!\n",
    "        self.reprogram()\n",
    "        #Delete for future use!\n",
    "        super().__init__(xclbin, download)\n",
    "        self.__getAcceleratorTypes()\n",
    "        self.__map_accels()\n",
    "        self.__create_accel_work_group()\n",
    "        # Other initialisation\n",
    "        \n",
    "    def reprogram(self):\n",
    "        \"\"\"\n",
    "        On edge devices XRT couldn't reprogram FPGA. \n",
    "        But we can stil do that manually.\n",
    "        \"\"\"\n",
    "        os.system(f\"fpgautil -b {self.bitstream}\")\n",
    "    \n",
    "    def __getAcceleratorTypes(self):\n",
    "        \"\"\"\n",
    "        Return name and number of accelerators:\n",
    "        If overlay have Vadd_1, Vadd_2, Vadd_3 in ip_dict()...\n",
    "        It should return\n",
    "        [[Vadd][3]]\n",
    "        \"\"\"\n",
    "        klucze = self.ip_dict.keys()\n",
    "        accelerators = list()\n",
    "        accelerator_types = list()\n",
    "        count = list()\n",
    "\n",
    "        for accelerator in klucze:\n",
    "            charind = accelerator.rfind('_')\n",
    "            accelerators.append(accelerator[:charind:])\n",
    "\n",
    "        accelerator_types = list(dict.fromkeys(accelerators))\n",
    "        for accelerator in accelerator_types:    \n",
    "            count.append(accelerators.count(accelerator))\n",
    "            \n",
    "        self.accel_container = [[accelerator_types, count]] \n",
    "        return [[accelerator_types, count]]\n",
    "    \n",
    "    def __get_accels(self, acc, cnt):\n",
    "        \"\"\"\n",
    "        Utility function that maps all accelerator to workgroups.\n",
    "        \"\"\"\n",
    "        func = getattr(self, f\"{acc}_{cnt}\")\n",
    "        return func\n",
    "\n",
    "    def __map_accels(self):\n",
    "        \"\"\"\n",
    "        Returns a tuple of mapped accelerators, so we can acces them like\n",
    "        accel_group[x], where x is index of accelerator.\n",
    "        \"\"\"\n",
    "        self.accel_group = list()\n",
    "        for i in range(len(self.accel_container)):\n",
    "            temp = tuple(map(self.__get_accels, self.accel_container[i][0]*self.accel_container[i][1][0], range(1,self.accel_container[i][1][0]+1)))\n",
    "            self.accel_group.append(temp)\n",
    "        return self.accel_group\n",
    "    \n",
    "    def __create_accel_work_group(self):\n",
    "        \"\"\"\n",
    "        Create new function based on accelerator groups found in overlay.\n",
    "        For example found three Vadd accelerators:\n",
    "        [[Vadd][3]] -> self.Vadd\n",
    "        \n",
    "        Where self is object that we create in main program.\n",
    "        \"\"\"\n",
    "        for i in range(len(self.accel_container)):\n",
    "            #Setattr powinien odpala&#263; wszystkie akclelatory\n",
    "            setattr(self, f\"{self.accel_container[i][0][0]}_group\", self.accel_group[i])\n",
    "    \n",
    "    def show_groups(self):\n",
    "        \"\"\"\n",
    "        Helper function that shows detected accelerator groups, thier count and imput arguments.\n",
    "        It is usefull for creating parallel run function.\n",
    "        \"\"\"\n",
    "        print(f\"Accelerator types: {len(self.accel_container)}\")\n",
    "        for i in range(len(self.accel_container)):\n",
    "            print(f\"Kernel: {self.accel_container[i][0][0]} Count: {self.accel_container[i][1][0]}\\nArguments: {self.accel_group[i][0].signature}\")\n",
    "            #nb of arguments len(elo[0][0].signature.parameters.keys())\n",
    "        \n",
    "    def split_buffer(self, buffer, accelerator_group):\n",
    "        \"\"\"\n",
    "        Utility function that splits input array into smaller arrays.\n",
    "        Arr1 = [10, 10, 10, 10, 10]\n",
    "        Arr2 = np.array_split(Arr1, 2)\n",
    "        Arr2 will become [[10, 10][10, 10][10]]\n",
    "        \"\"\"\n",
    "        acels = len(accelerator_group)\n",
    "        minibuffer = np.array_split(buffer, acels)\n",
    "        return minibuffer\n",
    "\n",
    "    def get_elements_size(self, array_to_split):\n",
    "        \"\"\"\n",
    "        Count elements of arrays in list.\n",
    "        [[10, 10][10, 10][10]] -> [2, 2, 1]\n",
    "        \"\"\"\n",
    "        size_split = list()\n",
    "        for element in array_to_split:\n",
    "            size_split.append(len(element))\n",
    "        return size_split    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Reprogram FPGA and xclbin\n",
    "xclbinary = \"/media/sd-mmcblk0p1/Vadd4/vadd.xclbin\"\n",
    "xcl_bitstream = \"/lib/firmware/base/vpl_gen_fixed.bit.bin\"\n",
    "t0 = time.time()\n",
    "ol = MyOverlay(xclbinary, xcl_bitstream)\n",
    "t1 = time.time()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#ol.show_groups()\n",
    "\"\"\"\n",
    "Parallel run function. This need to be created manually, because we can't detect dependancies,\n",
    "for example between arg[0][x] and arg[1][x] where arg[1][x] is size of arg[0][x]. \n",
    "\"\"\"\n",
    "def par_run(*args, accelerator_group=()):\n",
    "        acels = len(accelerator_group)\n",
    "        handle = list()\n",
    "\n",
    "        for i in range(acels):\n",
    "            #Basicly a template where you need to change args[0][i], args[1][i], args[2] only!\n",
    "            h = accelerator_group[i].start(args[0][i], args[1][i], args[2])\n",
    "            handle.append(h)\n",
    "\n",
    "        return wait_handles(handle)\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXgAAAEWCAYAAABsY4yMAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Il7ecAAAACXBIWXMAAAsTAAALEwEAmpwYAAAb1klEQVR4nO3de5hU9Z3n8feHBm0iBA30GBUVopGBDQiIQFZxCVFh1LAziS7emMdRh5nHQHDI6JioG8wsO25i1IyZjYIi3hlHY1QSbxnFW9RwEbwRUKFVXFTAoIggt+/+cU5r0XZXV3fX6eo+fF7PUw9VdS6/bx3g07/+nVO/o4jAzMzyp1OlCzAzs2w44M3McsoBb2aWUw54M7OccsCbmeWUA97MLKcc8LspSWdJeqrSdZhZdhzwOSapVtJmSR8VPH5R6bqKSWs+tq23zTtJp0paJmmTpNcljUrf7yMp6v0bubRguzmSttZbXpUuGyBpoaQ/pY/fSRpQr92hkp5It3tX0tS2/eS7t86VLsAy962I+F2li8gbSZ0jYnul6yiFpOOA/wNMAP4A7NfAansX+Tw/iYhLGnj//wEnA2+QdBa/C8wFBqXt9gIeBP4BuAvYA+jd8k9izeUevAEg6c8lPSLpfUnLJf2PgmVzJP1fSQ+kPbGnJX1Z0tVpz+2PkoYUrL+/pLslrZW0StL3CpZNl3SnpJslbZT0sqRh6bJbgIOA+9N2LpRULelWSeslbZC0QNK+DdT/uW3T98enbWyQNF9S/yLHICR9T9JKSesk/VRSp3TZWennvkrSemC6pB7p51gr6Q1Jl9Stn27zt2mveaOkVyQNLeH4DE97xR+mPd4r0/dLOg6NuAz4cUQ8GxE7I+LtiHi7xG0bFREbIqI2kq/DC9gBHFqwyjTgoYi4LSI+iYiNEbGste1aM0SEHzl9ALXAsY0sOwt4Kn2+F/AW8Dckv9UNAdYBA9Llc9LXRwDVwKPAKuCvgSrgfwGPpet2AhYB/5Okx/YVYCUwNl0+HdgCnJBu+y/As43VDPwdcD/whXT9I4AvlvJ5gcOATcBxQBfgQuA1YI9Gtg/gMeBLJD8sVgDnFhyv7cCU9Bh1BW4G7gW6A33S9c9J1z8FeBs4kiT8DgUOLuH4PANMTJ93A0Y2dRyAi4B5jXymKmBrus5rwGrgF0DXdHmf9HO/nS67EehVsP0c4P30sQj4TgNtbEiPzU7gkoL3HwV+DvweeC+t/6BK/7/YnR4VL8CPDP9yk8D7KP0PWPf423TZWXwW8BOAJ+ttex3wo/T5HGBWwbIpwLKC1wOBDenzEcCb9fb1A+DG9Pl04HcFywYAm+vVXBjSZ6cBMajEz1u47aXAnQWvO6VBNrqR7QMYV/D6POA/C47XmwXL6oJzQMF7fwfMT58/BExtoI2mjs8TJD3uXvXWKfk41Ntu//RzLSQZmukFPA3MSJd3A4aR/NDal2Qo5aGC7YcCPdPlJwAbgaMaaGev9HidWPDeivTf3JEkHYN/BZ6u9P+L3enhMfj8+8toegz+YGCEpA0F73UGbil4/W7B880NvO5WsK/96+2rCniy4PU7Bc8/BqqLjGnfAhwIzJW0N3ArcHFEbGviM0ESbm/UvYiInZLeAg4oss1bBc/fSPfR0LJeJL8VvFFv/bp9Hwi83sD+mzo+5wA/Bv4oaRVwWUTMo+XHYXP65zURsQYgHfa5JN3+I5LwB3hX0mRgjaTukQypLC7Y128l3QZ8m+SHxKciYpOka4G1kvpHxHtp2/dExIK03cuAdZJ6RMQHTdRtZeCAN0iC6/GIOK5M+1oVEV9t4fa7TG+aBthlwGWS+gC/BZYDNzS1LclJwIF1LySJJCSLjT8fCLycPj8o3UdD+18HbCMJ7FcK1q/b91vAIQ3sv+jxiYhXgdPSsfxvA3dJ6hkRmyj9OBTu70+SVtervdgUsnXLGjs/Vzfe3pBOJENIB5AMybzQjHYtAz7JagDzgMMkTZTUJX0cWeyEZBF/ADZK+idJXSVVSfqapCNL3P5dknFpACR9Q9JAJZfmfUgSqjtL2Ra4EzhR0jcldQG+D3xCMtTRmAsk7SPpQGAq8O8NrRQRO9L9z5DUXdLBJCcVb01XuR74R0lHKHFouk7R4yPpTEk1EbGTZHgDYGczj0N9NwJTJP2ZpH1IrmqZl7Y3QlI/SZ0k9SQZRplf18OWdLKkbuny44EzgfvSZcdJGpJ+hi8CVwJ/ApYVtPtXkganx/9SkmFB997biAM+/+quKql73FN/hYjYCBwPnErSY32H5LK6PZvbWBp8JwGDSU7EriMJux4l7uJfgEvSK0X+EfgyybjwhyTB8Ti7Dh01um1ELCcJpGvSOr5Fctno1iLt30tyMnEJ8BuK95CnkJzEXQk8BdwOzAaIiP8AZqTvbQR+DXyphOMzDnhZ0kckJyhPjYjNxY6DpB9KeqBInf8MLCAZE18GPJ/WBskPxAfTGl8i+QF4WsG2U0l+K9kA/JTkHM78dNnewB3AByTDUYeQnMPYkh6DR4EfkhzH90hONJ9epE4rM0X4tyYzSC6TBL4aEa9VuhazcnAP3swspxzwZmY55SEaM7Occg/ezCyn2tV18L169Yo+ffpUugwzsw5j0aJF6yKipqFl7Srg+/Tpw8KFC5te0czMAJD0RmPLMh2ikbS3pLuUzDa4TNLXs2zPzMw+k3UP/ufAgxFxsqQ9SL7GbGZmbSCzgJfUAziGZBY+0m8PFvsGoZmZlVGWPfi+wFrgRkmHk3z9e2o6adKnJE0CJgEcdNBBGZZjZu3dtm3bWL16NVu2bKl0Ke1OdXU1vXv3pkuXLiVvk9l18Eru0vMsydzRz0n6OfBhRFza2DbDhg0Ln2Q1232tWrWK7t2707NnT5LJPw2S+3asX7+ejRs30rdv312WSVoUEcMa2i7Lk6yrgdUR8Vz6+i6SmweYmTVoy5YtDvcGSKJnz57N/s0ms4CPiHeAtyT1S9/6Jp/Nm21m1iCHe8NaclyyvopmCnBbegXNSpJ7fpqZWRvINOAjYgnJ/R7NzJqtz0W/Kev+ai8/scl1qqqqGDhwINu3b6d///7cdNNNfOELrbvCu7a2lpNOOomXXnqJ+fPnc8UVVzBv3rxG1y9lnVK0q2+ytnfl/sfWmFL+EZpZNrp27cqSJUsAOOOMM7j22muZNm1ak9tt376dzp3bV6R6sjEzs0aMGjWK1157jfvvv58RI0YwZMgQjj32WN59N7nn/PTp05k4cSJHHXUUEydOpLa2llGjRjF06FCGDh3K739f7O6QsGnTJs4++2yGDx/OkCFDuPfee8taf/v6cWNm1k5s376dBx54gHHjxnH00Ufz7LPPIonrr7+en/zkJ/zsZz8D4JVXXuGpp56ia9eufPzxxzzyyCNUV1fz6quvctpppxWdX2vGjBmMGTOG2bNns2HDBoYPH86xxx5bts/ggDczK7B582YGDx4MJD34c845h+XLlzNhwgTWrFnD1q1bd7kWffz48XTt2hVIvqg1efJklixZQlVVFStWrCja1sMPP8x9993HFVdcASSXib755ptl+ywOeDOzAoVj8HWmTJnCtGnTGD9+PPPnz2f69OmfLttrr70+fX7VVVex7777snTpUnbu3El1dXXRtiKCu+++m379+u3yft0QUGt5DN7MrAkffPABBxxwAAA33XRT0fX2228/OnXqxC233MKOHTuK7nfs2LFcc8011M0o8Pzzz5evaNyDN7N2rL1cUTZ9+nROOeUU9tlnH8aMGcOqVasaXO+8887jO9/5DjfffDPjxo3bpXffkEsvvZTzzz+fQYMGsXPnTvr27dvqSyMLtat7srb3uWh8maRZtpYtW0b//v0rXUa71dDxqdRcNGZmVkEOeDOznHLAm5nllAPezCynHPBmZjnlgDczyylfB29m7df0HmXe3wdNrjJjxgxuv/12qqqq6NSpE9dddx2zZs1i2rRpDBgwoNlNFk4V3NYc8GZmqWeeeYZ58+axePFi9txzT9atW8fWrVu5/vrrK11ai3iIxswstWbNGnr16sWee+4JQK9evdh///0ZPXr0p7NCduvWjYsvvpjDDz+ckSNHfjpvzOuvv87IkSMZOHAgl1xyCd26dfvc/nfs2MEFF1zAkUceyaBBg7juuusy/TwOeDOz1PHHH89bb73FYYcdxnnnncfjjz/+uXU2bdrEyJEjWbp0KccccwyzZs0CYOrUqUydOpUXX3yR3r17N7j/G264gR49erBgwQIWLFjArFmzGp32oBwc8GZmqW7durFo0SJmzpxJTU0NEyZMYM6cObuss8cee3DSSScBcMQRR1BbWwskwzunnHIKAKeffnqD+3/44Ye5+eabGTx4MCNGjGD9+vW8+uqrmX0ej8GbmRWoqqpi9OjRjB49moEDB35u9sguXbog6dN1t2/fXvK+I4JrrrmGsWPHlrXmxrgHb2aWWr58+S496iVLlnDwwQeXtO3IkSO5++67AZg7d26D64wdO5Zf/vKXbNu2DYAVK1awadOmVlbdOPfgzaz9KuGyxnL66KOPmDJlChs2bKBz584ceuihzJw5k5NPPrnJba+++mrOPPNMZsyYwbhx4+jR4/OXeJ577rnU1tYydOhQIoKamhp+/etfZ/BJEp4uuBk8XbBZtjrydMEff/wxXbt2RRJz587ljjvuKPtNtJs7XbB78GZmZbBo0SImT55MRLD33nsze/bsSpfkgDczK4dRo0axdOnSSpexC59kNbN2pT0NG7cnLTkuDngzazeqq6tZv369Q76eiGD9+vVUV1c3a7tMh2gk1QIbgR3A9sZOBJiZAfTu3ZvVq1ezdu3aSpfS7lRXVzf6DdnGtMUY/DciYl0btGNmHVyXLl3o27dvpcvIDQ/RmJnlVNYBH8DDkhZJmtTQCpImSVooaaF/LTMzK5+sA/7oiBgK/AXwXUnH1F8hImZGxLCIGFZTU5NxOWZmu49MAz4i3k7/fA+4BxieZXtmZvaZzAJe0l6Sutc9B44H2v6eVWZmu6ksr6LZF7gnnVazM3B7RDyYYXtmZlYgs4CPiJXA4Vnt38zMivNlkmZmOeWANzPLKQe8mVlOOeDNzHLKAW9mllMOeDOznHLAm5nllAPezCynHPBmZjnlgDczyykHvJlZTjngzcxyygFvZpZTDngzs5xywJuZ5VSWN/xoW9N7tEEjt7dBG2Zm5eEevJlZTjngzcxyKj9DNFZebTLkBUz/oG3aaQ/a4pj6eGbQTsc9pu7Bm5nllAPezCynHPBmZjnlgDczyykHvJlZTjngzcxyygFvZpZTmQe8pCpJz0ual3VbZmb2mbbowU8FlrVBO2ZmViDTgJfUGzgRuD7LdszM7POy7sFfDVwI7GxsBUmTJC2UtHDt2rUZl2NmtvvILOAlnQS8FxGLiq0XETMjYlhEDKupqcmqHDOz3U6WPfijgPGSaoG5wBhJt2bYnpmZFcgs4CPiBxHROyL6AKcCj0bEmVm1Z2Zmu/J18GZmOdUm88FHxHxgflu0ZWZmCffgzcxyygFvZpZTDngzs5xywJuZ5VTRk6ySppWwj00RcV2Z6jEzszJpqgd/AdAN6F7k8f0sCzQzs5Zp6jLJWyLix8VWkLRXGesxM7MyKdqDj4gLm9pBKeuYmVnbK+kkq6Spkr6oxA2SFks6PuvizMys5Uq9iubsiPgQOB7YB5gIXJ5ZVWZm1mqlBrzSP08gGZd/ueA9MzNrh0oN+EWSHiYJ+IckdafITTzMzKzySp1s7BxgMLAyIj6W1BP4m8yqMjOzVisp4CNip6TtwDGSCrd5IZuyzMystUoKeEmzgUHAy3w2NBPArzKqy8zMWqnUIZqRETEg00rMzKysSj3J+owkB7yZWQdSag/+ZpKQfwf4hOQSyYiIQZlVZmZmrVJqwN9A8uWmF/HlkWZmHUKpAb82Iu7LtBIzMyurUgP+eUm3A/eTDNEAEBG+isbMrJ0qNeC7kgR74QRjvkzSzKwdK/WLTv7WqplZB1P0MklJk5raQSnrmJlZ22uqB3+RpHVFlguYCswsX0lmZlYOTQX848C3mljnkTLVYmZmZVQ04Fsz9i6pGngC2DNt566I+FFL92dmZs1T6lU0LfEJMCYiPpLUBXhK0gMR8WyGbZqZWSqzgI+IAD5KX3ZJH5FVe2Zmtqsse/BIqgIWAYcC/xYRz2XZnnU8fS76TeZt1F5+YuZtmLVHJc0mKWlfSTdIeiB9PUDSOU1tFxE7ImIw0BsYLulrDex7kqSFkhauXbu2meWbmVljSu3BzwFuBC5OX68A/p1kErImRcQGSY8B44CX6i2bSXqZ5bBhwzyEY2btSkf+LbPU+eB7RcSdpDNJRsR2YEexDSTVSNo7fd4VOA74Y8tLNTOz5ii1B78pvdF2AEgaCXzQxDb7ATel4/CdgDsjYl6LKzUzs2YpNeCnAfcBh0h6GqgBTi62QUS8AAxpXXlmZtZSpU42tljSfwP6kUxPsDwitmVamZk1W1uMF4OvTOooSgr4dJjlBKBPus3xkoiIKzOszczMWqHUIZr7gS34ln1mZh1GqQHf2zfYNjPrWEq9TPIBScc3vZqZmbUXpfbgnwXukdQJ2EZyojUi4ouZVWZmZq1SasBfCXwdeDGdRMzMzNq5Uodo3gJecribmXUcpfbgVwLz08nGPql705dJmpm1X6UG/Kr0sUf6MDOzdq7Ub7JelnUhZmZWXkUDXtIvImKypPtp4G5METE+s8rMzKxVmurB/zUwGbiiDWoxM7MyairgXweIiMfboBYzMyujpgK+RtK0xhb6Khozs/arqYCvArqRfHPVzMw6kKYCfk1E/LhNKjEzs7Jq6pus7rmbmXVQTQX8N9ukCjMzK7uiAR8R77dVIWZmVl6lTjZmZmYdjAPezCynHPBmZjnlgDczyykHvJlZTjngzcxyygFvZpZTmQW8pAMlPSbpFUkvS5qaVVtmZvZ5pd6yryW2A9+PiMWSugOLJD0SEa9k2KaZmaUy68FHxJqIWJw+3wgsAw7Iqj0zM9tVm4zBS+oDDAGea2DZJEkLJS1cu3ZtW5RjZrZbyDzgJXUD7gbOj4gP6y+PiJkRMSwihtXU1GRdjpnZbiPTgJfUhSTcb4uIX2XZlpmZ7SrLq2gE3AAs8639zMzaXpY9+KOAicAYSUvSxwkZtmdmZgUyu0wyIp7Cd4QyM6sYf5PVzCynHPBmZjnlgDczyykHvJlZTjngzcxyygFvZpZTDngzs5xywJuZ5ZQD3swspxzwZmY55YA3M8spB7yZWU454M3McsoBb2aWUw54M7OccsCbmeWUA97MLKcc8GZmOeWANzPLKQe8mVlOOeDNzHLKAW9mllMOeDOznHLAm5nllAPezCynHPBmZjmVWcBLmi3pPUkvZdWGmZk1Lsse/BxgXIb7NzOzIjIL+Ih4Ang/q/2bmVlxFR+DlzRJ0kJJC9euXVvpcszMcqPiAR8RMyNiWEQMq6mpqXQ5Zma5UfGANzOzbDjgzcxyKsvLJO8AngH6SVot6Zys2jIzs8/rnNWOI+K0rPZtZmZN8xCNmVlOOeDNzHLKAW9mllMOeDOznHLAm5nllAPezCynHPBmZjnlgDczyykHvJlZTjngzcxyygFvZpZTDngzs5xywJuZ5ZQD3swspxzwZmY55YA3M8spB7yZWU454M3McsoBb2aWUw54M7OccsCbmeWUA97MLKcc8GZmOeWANzPLKQe8mVlOOeDNzHLKAW9mllOZBrykcZKWS3pN0kVZtmVmZrvKLOAlVQH/BvwFMAA4TdKArNozM7NdZdmDHw68FhErI2IrMBf47xm2Z2ZmBRQR2exYOhkYFxHnpq8nAiMiYnK99SYBk9KX/YDlmRRUHr2AdZUuImd8TMvLx7P82vsxPTgiahpa0LmtK6kvImYCMytdRykkLYyIYZWuI098TMvLx7P8OvIxzXKI5m3gwILXvdP3zMysDWQZ8AuAr0rqK2kP4FTgvgzbMzOzApkN0UTEdkmTgYeAKmB2RLycVXttpEMMJXUwPqbl5eNZfh32mGZ2ktXMzCrL32Q1M8spB7yZWU454Esgabak9yS9VOla8kDSgZIek/SKpJclTa10TR2dpGpJf5C0ND2ml1W6pjyQVCXpeUnzKl1LSzjgSzMHGFfpInJkO/D9iBgAjAS+62ksWu0TYExEHA4MBsZJGlnZknJhKrCs0kW0lAO+BBHxBPB+pevIi4hYExGL0+cbSf4DHVDZqjq2SHyUvuySPnwFRStI6g2cCFxf6VpaygFvFSWpDzAEeK7CpXR46XDCEuA94JGI8DFtnauBC4GdFa6jxRzwVjGSugF3A+dHxIeVrqeji4gdETGY5FvjwyV9rcIldViSTgLei4hFla6lNRzwVhGSupCE+20R8atK15MnEbEBeAyfN2qNo4DxkmpJZsIdI+nWypbUfA54a3OSBNwALIuIKytdTx5IqpG0d/q8K3Ac8MeKFtWBRcQPIqJ3RPQhmWbl0Yg4s8JlNZsDvgSS7gCeAfpJWi3pnErX1MEdBUwk6RUtSR8nVLqoDm4/4DFJL5DMA/VIRHTIS/usfDxVgZlZTrkHb2aWUw54M7OccsCbmeWUA97MLKcc8GZmOeWAt3ZJ0pclzZX0uqRFkn4r6bAK1fLDFmxTK6lXwevRpcxIKOn36Z99JJ3e3HbNCjngrd1Jvwh1DzA/Ig6JiCOAHwD7VqikZgd8S0XEf02f9gEc8NYqDnhrj74BbIuIa+veiIilEfGkEj+V9JKkFyVNgE97yI9LulfSSkmXSzojnSP9RUmHpOvNkXStpIWSVqRzjiDpLEm/qGtP0rx0n5cDXdMvY92WLjsz3e8SSddJqmrOh5M0Pb3HwPy01u8VLKubEfJyYFTaxj9I+i8Fbb4g6astO7S2O3HAW3v0NaCxSZ6+TTLf+eHAscBPJe2XLjsc+HugP8k3ZQ+LiOEk071OKdhHH2A4yVSw10qqbqyQiLgI2BwRgyPiDEn9gQnAUenEXjuAM1rwGf8cGJvW8aN0bp5CFwFPpu1elX6un6dtDgNWt6BN2810rnQBZs10NHBHROwA3pX0OHAk8CGwICLWAEh6HXg43eZFkt8K6twZETuBVyWtJAnbUn0TOAJYkIwk0ZVket76GvqKeOF7v4mIT4BPJL1HMvxULLSfAS5O5yj/VUS82oyabTflHry1Ry+ThGhzfVLwfGfB653s2pmpH75Bcpepwv8PjfXqBdyU9qwHR0S/iJjewHrrgX0KXn8JWNdIrTtoorMVEbcD44HNwG8ljSm2vhk44K19ehTYU9KkujckDZI0CngSmJDe3KIGOAb4QzP3f4qkTum4/FeA5UAtMDh9/0CSoZM62wqGUP4TOFnSn6V1fUnSwQ20MZ9kmIh0jP5Mkil8S7UR6F73QtJXgJUR8a/AvcCgZuzLdlMeorF2JyJC0l8BV0v6J2ALSQCfDzwFfB1YStLzvjAi3pHUnGGWN0l+KHwR+PuI2CLpaWAV8ArJLQQXF6w/E3hB0uJ0HP4S4GFJnYBtwHeBN+q18c/ALyUtJen1Pwg0Zz7xF4Ad6fZzgD2BiZK2Ae8A/7sZ+7LdlGeTtN2KpDnAvIi4q9K1mGXNQzRmZjnlHryZWU65B29mllMOeDOznHLAm5nllAPezCynHPBmZjn1/wFHLVsUdwTYeQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEWCAYAAAB8LwAVAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Il7ecAAAACXBIWXMAAAsTAAALEwEAmpwYAAAgkklEQVR4nO3de7wVZd338c+XDbpJTA12JCJieUgTRUWkPIRHkAifSlNSy9LI2zB8KE2zu3aW5W2alnSLKNyAB8xX5gm14Cnx0K3FQRA84AFRUOJkoCIIG37PHzNbF9tZe68N68De+/t+vdZrzcx1zcxvhs36zVwzc40iAjMzs4baVToAMzPbNjlBmJlZJicIMzPL5ARhZmaZnCDMzCyTE4SZmWVygrCCSDpb0uOVjsPMyscJwt4naaGktZLeyfmMqnRcjUljPr7c87ZWkraXNFbSq5LeljRb0kkN6nxE0n9LWiFptaRHc8oukjQvnfcVSRc1mPfnkuZKqpNU20gc4ySFpL0yyvaWtE7SrUXYZGtE+0oHYNucL0bE/6t0EK2NpPYRUVfpOArQHlgEfB54DRgE3CmpV0QsTOuMSevtB7wJ9M6ZX8DXgaeBTwFTJC2KiDvS8peAi4Hz8gUg6ch03nx+D0xv1lbZFvEZhG0RSZ+WNFXSm5LmS/pqTtn49AjzofQs5O+SPiHpOkn/lvS8pINz6neTdJek5elR5/dyymol3SlpYnpU+oykPmnZLUAP4P50PRdLqpZ0q6SVklZJmi6pa0b8H5o3nT4kXccqSdMk7dfIPghJ35O0ID2a/rWkdmnZ2el2XytpJVAraad0O5anR+g/rq+fzvNtSc+l2/mspEMK2D99Jc2Q9JakpZJ+k04vaD80FBFrIqI2IhZGxKaImAy8Ahxa/+8ODAGGRcTyiNgYETNz5r8qImZFRF1EzAfuBY7IKZ8QEQ8Bb+fZp+2B64EL8pSfDqwC/trUtlgRRIQ//hARAAuB4/OUnQ08ng7vQHKU+U2SI8mDgRXA/mn5+HT8UKAa+BvJj8zXgSrgF8DDad12wEzgJ8B2wCeBBcCAtLwWWEdyJFsF/Ap4Ml/MwHeA+4GPpPUPBT5ayPYC+wBrgBOADiRHui8B2+WZP4CHgY+RJJsXgHNz9lcdyQ9de6AjMJHkB3NHoGda/5y0/qnA68BhJEfhewF7FLB/ngDOSoc7Af2a2g/AJcDkAv8muqb7/9Pp+NeBucC16b/xXOAreeYV8BRwXkbZrUBtxvSLgN/m7N+9cso+mu6z7unfxa2V/j/T2j8+g7CG7kmPOOs/386oMxhYGBH/E8mR4lPAXSQ/cvXujoiZEbEOuBtYFxETI2Ij8AeSpALJD2JNRFweEesjYgFwE3B6zrIej4gH03lvAQ5qJP4NQGeSH5aNaQxvFbjtpwEPRMTUiNgAXE3yw/65Rub5r4h4MyJeA64DhuaUvRER10fStLQ+3aZLI+LtSJprrgHOSuueC1wVEdMj8VJEvErT+2cDsJekLhHxTkQ82dR+iIgrI2JwUztDUgfgNmBCRDyfTu4OHACsBroBw4EJec60akkS3P80ta50fbuTJLaf5Knyc2BsRCwuZHm29XwNwhr6P9H0NYg9gMMlrcqZ1p7kx7ve0pzhtRnjnXKW1a3BsqqAx3LG/5Uz/C5Q3Uib/i3A7sAdknYmOVK9LP3Bb0o34NX6kYjYJGkRsFsj8yzKGX41XUZWWReSs5JXG9SvX/buwMsZy29q/5wDXA48L+kV4GeRNAttzX4gbfq6hSSxDc8pWkuSfH6R7v9HJD0MnAg8lzP/cJKzjaMi4r1C1kmSYC+PiNUZ8fQGjueDAwsrAycI2xKLgEci4oQiLeuViNh7C+ffrDvi9AfwZ8DPJPUEHgTmA2Obmhd4A+hVPyJJJD+yrzey/t2BZ9LhHukyspa/guSHdQ/g2Zz69cteRPaF2Ub3T0S8CAxNf9C/DPxRUueIWEPh+2Ez6XaPJWleGtQgqTydFUaD+b9F0ox1dDOP9o8DjpR0Vc60JySNAD5O0iz3WhIenYAqSftHxCHNWIc1g5uYbEtMBvaRdJakDunnsMYu6Dbin8Dbkn4oqaOkKkkHSDqswPmXkrTLAyDpGEm9JFUBb5H8KG8qZF7gTuALko5Lm1e+D7wH/G8j679I0i5p88gIkuazD0mbx+4ErpC0o6Q9gJEkR/YANwM/kHSoEnuldRrdP5LOlFQTEZtILt4CbGrmfmjoBpI7lL4YEWsblD1KcnfTpZLaSzoCOAb4SxrPGcAvgRPS5rDNpH8r1SS/Pe3Ti+lVafE+JM2HvfngzqgvkjRRjiFJoPVlo4EHgAEFbpNtiUpfBPFn2/mQXLRdC7yT87k7LTub9CJ1Or4vyX/Q5cBKkgvRvdOy8SRNEPV1zwWm5YzvBdTljHcDJpE0Jf0beJL04jENLkaSHEUG0D4dP5nkB2sV8AOSawDzSS42LwV+V183Y3s3mzed9iWSI/zVwCPAZxrZXwF8j+Si8UqSawpVWfsrnbYLSUJYTnJm8BOgXU75eWns7wDzgIML2D+3AsvSeZ4haSKksf0A/Ah4KM827ZFu17oGfwdn5NT5DMnF8TXpvvpSTtkrJMkod97ROeXj0+Xnfs5uZP/uladss78Lf0rzUbqzzayZJAWwd0S8VOlYzErBTUxmZpbJCcLMzDK5icnMzDL5DMLMzDK1qucgunTpEj179qx0GGZmLcbMmTNXRERNVlmrShA9e/ZkxowZlQ7DzKzFkPRqvjI3MZmZWSYnCDMzy1SyJiZJ40h6/VwWEQek0/5A8gQuwM7AqojonTHvQpL+4jeSPHHbp1RxmplZtlJegxgPjCLpAx+AiDitfljSNSTdGeRzTESsKFl0ZtbqbNiwgcWLF7Nu3bpKh7LNqa6upnv37nTo0KHgeUqWICLi0bQXyQ9Je4v8KnBsqdZvZm3P4sWL2XHHHenZsydpr69G0ufeypUrWbx4MXvuuWfB81XqGsRRwNJIuirOEiTvsp0paVhjC5I0LH3l4ozly5cXPVAzaznWrVtH586dnRwakETnzp2bfWZVqQQxlKR3ynyOjKSP95OA70o6Ol/FiBgTEX0iok9NTeatvGbWhjg5ZNuS/VL2BJG+lPzL5Ok3HyAiXk+/l5H0Bd+3PNGZmVm9SjwodzzwfOR505SkHUj6yH87HT6R5JWKZmbN0vOSB4q6vIVXfqHJOlVVVfTq1Yu6ujr2228/JkyYwEc+8pGtW+/ChQwePJh58+Yxbdo0rr76aiZPnpy3fiF1ClHK21wnAf2BLpIWAz+NiLEkL1uf1KBuN+DmiBhE8prDu9PTofbA7RHx51LFWU7F/mPNp5A/YjMrjY4dOzJ79mwAzjjjDEaPHs3IkSObnK+uro727betzi1KeRfT0DzTz86Y9gYwKB1eQPLaQTOzFu2oo47i6aef5v777+cXv/gF69evp3Pnztx222107dqV2tpaXn75ZRYsWECPHj341a9+xVlnncWaNWsAGDVqFJ/73OfyLn/NmjVccMEFzJs3jw0bNlBbW8vJJ59ctPi3rXRlZtZK1NXV8dBDDzFw4ECOPPJInnzySSRx8803c9VVV3HNNdcA8Oyzz/L444/TsWNH3n33XaZOnUp1dTUvvvgiQ4cObbR/uSuuuIJjjz2WcePGsWrVKvr27cvxxx9ftG1wgjAzK6K1a9fSu3dvIDmDOOecc5g/fz6nnXYaS5YsYf369Zs9izBkyBA6duwIJA/6DR8+nNmzZ1NVVcULL7zQ6LqmTJnCfffdx9VXXw0kt/m+9tprRdsWJwgzsyLKvQZR74ILLmDkyJEMGTKEadOmUVtb+37ZDjvs8P7wtddeS9euXZkzZw6bNm2iurq60XVFBHfddRf77rvvZtOXLl261dsB7qzPzKzkVq9ezW677QbAhAkTGq2366670q5dO2655RY2btzY6HIHDBjA9ddfT/2bQZ966qniBY3PIMysFdtW7uirra3l1FNPZZddduHYY4/llVdeyax3/vnn85WvfIWJEycycODAzc4usvznf/4nF154IQceeCCbNm1izz333OpbW3O1qndS9+nTJ7blFwb5Nlez0nruuefYb7/9Kh3GNitr/0iama/HbDcxmZlZJicIMzPL5ARhZmaZnCDMzCyTE4SZmWVygjAzs0x+DsLMWq/anYq8vNVNVrniiiu4/fbbqaqqol27dtx4443cdNNNjBw5kv3337/Zq8zt6rvcnCDMzIrkiSeeYPLkycyaNYvtt9+eFStWsH79em6++eZKh7ZF3MRkZlYkS5YsoUuXLmy//fYAdOnShW7dutG/f//3e2Xt1KkTl112GQcddBD9+vV7v9+kl19+mX79+tGrVy9+/OMf06lTpw8tf+PGjVx00UUcdthhHHjggdx4440l3R4nCDOzIjnxxBNZtGgR++yzD+effz6PPPLIh+qsWbOGfv36MWfOHI4++mhuuukmAEaMGMGIESOYO3cu3bt3z1z+2LFj2WmnnZg+fTrTp0/npptuytttRzE4QZiZFUmnTp2YOXMmY8aMoaamhtNOO43x48dvVme77bZj8ODBABx66KEsXLgQSJqnTj31VAC+9rWvZS5/ypQpTJw4kd69e3P44YezcuVKXnzxxZJtj69BmJkVUVVVFf3796d///706tXrQ723dujQgfSVylRVVVFXV1fwsiOC66+/ngEDBhQ15nx8BmFmViTz58/f7Ih+9uzZ7LHHHgXN269fP+666y4A7rjjjsw6AwYM4IYbbmDDhg0AvPDCC++/nrQUfAZhZq1XAbelFtM777zDBRdcwKpVq2jfvj177bUXY8aM4ZRTTmly3uuuu44zzzyTK664goEDB7LTTh++Rffcc89l4cKFHHLIIUQENTU13HPPPSXYkoS7+y4jd/dtVlotubvvd999l44dOyKJO+64g0mTJnHvvfcWdR3N7e67ZGcQksYBg4FlEXFAOq0W+DawPK32o4h4MGPegcBvgSrg5oi4slRxvq/YD9Rkur0M69hGlGV/UvYjxIoqxz5tS/vzjeK+fS2vbgcXVG3mzJkMHz6ciGDnnXdm3LhxJQ6saaVsYhoPjAImNph+bURcnW8mSVXA74ETgMXAdEn3RcSzpQrUzKzSjjrqKObMmVPpMDZTsovUEfEo8OYWzNoXeCkiFkTEeuAO4OSiBmdmrVZrajYvpi3ZL5W4i2m4pKcljZO0S0b5bsCinPHF6bRMkoZJmiFpxvLly/NVM7M2oLq6mpUrVzpJNBARrFy5kurq6mbNV+67mG4Afg5E+n0N8K2tWWBEjAHGQHKRemsDNLOWq3v37ixevJiCDhZXLSt9QACrnyvPeppQXV2d9wntfMqaICJiaf2wpJuAyRnVXgd2zxnvnk4zM2tUhw4d2HPPPQurXNuvtMG8v56We+G/rE1MknbNGf0SkNV/7XRgb0l7StoOOB24rxzxmZnZB0p5m+skoD/QRdJi4KdAf0m9SZqYFgLfSet2I7mddVBE1EkaDvyF5DbXcRHxTKniNDOzbCVLEBExNGPy2Dx13wAG5Yw/CHzo+QgzMysf98VkZmaZnCDMzCyTE4SZmWVygjAzs0xOEGZmlskJwszMMjlBmJlZJicIMzPL5ARhZmaZnCDMzCyTE4SZmWVygjAzs0xOEGZmlskJwszMMjlBmJlZJicIMzPL5ARhZmaZnCDMzCyTE4SZmWVygjAzs0xOEGZmlqlkCULSOEnLJM3LmfZrSc9LelrS3ZJ2zjPvQklzJc2WNKNUMZqZWX6lPIMYDwxsMG0qcEBEHAi8AFzayPzHRETviOhTovjMzKwRJUsQEfEo8GaDaVMioi4dfRLoXqr1m5nZ1qnkNYhvAQ/lKQtgiqSZkoY1thBJwyTNkDRj+fLlRQ/SzKytqkiCkHQZUAfclqfKkRFxCHAS8F1JR+dbVkSMiYg+EdGnpqamBNGambVNZU8Qks4GBgNnRERk1YmI19PvZcDdQN+yBWhmZkCZE4SkgcDFwJCIeDdPnR0k7Vg/DJwIzMuqa2ZmpVPK21wnAU8A+0paLOkcYBSwIzA1vYV1dFq3m6QH01m7Ao9LmgP8E3ggIv5cqjjNzCxb+1ItOCKGZkwem6fuG8CgdHgBcFCp4jIzs8L4SWozM8vkBGFmZpmcIMzMLJMThJmZZXKCMDOzTE4QZmaWyQnCzMwyOUGYmVkmJwgzM8vkBGFmZpmcIMzMLJMThJmZZXKCMDOzTI325ippZAHLWBMRNxYpHjMz20Y0dQZxEdCJ5B0O+T7fL2WAZmZWGU29D+KWiLi8sQrpW9/MzKyVafQMIiIubmoBhdQxM7OWp6CL1JJGSPqoEmMlzZJ0YqmDMzOzyin0LqZvRcRbwInALsBZwJUli8rMzCqu0ASh9HsQyXWJZ3KmmZlZK1RogpgpaQpJgviLpB2BTaULy8zMKq3QBHEOcAlwWES8C2wHfLOpmSSNk7RM0rycaR+TNFXSi+n3Lnnm/UZa50VJ3ygwTjMzK5KCEkREbALqgKMlfRn4PLBXAbOOBwY2mHYJ8NeI2Bv4azq+GUkfA34KHA70BX6aL5GYmVlpNPUcBJCcCQAHAs/wQdNSAH9qbL6IeFRSzwaTTwb6p8MTgGnADxvUGQBMjYg30/VPJUk0kwqJ18zMtl5BCQLoFxH7F2mdXSNiSTr8L6BrRp3dgEU544vTaR8iaRgwDKBHjx5FCtHMzAq9BvGEpGIliPdFRJCciWzNMsZERJ+I6FNTU1OkyMzMrNAziIkkSeJfwHskt7hGRBy4BetcKmnXiFgiaVdgWUad1/mgGQqgO0lTlJmZlUmhCWIsycNxc9n621vvA75B8qDdN4B7M+r8BfhlzoXpE4FLt3K9ZmbWDIUmiOURcV9zFy5pEsmZQBdJi0nuTLoSuFPSOcCrwFfTun2A8yLi3Ih4U9LPgenpoi6vv2BtZmblUWiCeErS7cD9JE1MAEREU3cxDc1TdFxG3RnAuTnj44BxBcZnZmZFVmiC6EiSGHI76GvyNlczM2u5CkoQEdHkU9NmZta6NHqba/qMQaMKqWNmZi1PU2cQl0ha0Ui5gBHAmOKFZGZm24KmEsQjwBebqDO1SLGYmdk2pNEE4WsPZmZtV6FdbZiZWRvjBGFmZpmcIMzMLFNBCUJSV0ljJT2Uju+fdpVhZmatVKFnEONJOtDrlo6/AFxYgnjMzGwbUWiC6BIRd5L25BoRdcDGkkVlZmYVV2iCWCOpM+nLfST1A1aXLCozM6u4QjvrG0nyHodPSfo7UAOcUrKozMys4grtrG+WpM8D+5J0rzE/IjaUNDKzAvS85IGSr2PhlV8o+TrMtkUFJQhJVcAgoGc6z4mSiIjflDA2MzOroEKbmO4H1lGcV46amVkLUGiC6B4RB5Y0EjMz26YUehfTQ5JObLqamZm1FoWeQTwJ3C2pHbCB5EJ1RMRHSxaZmZlVVKEJ4jfAZ4G5EREljMfMzLYRhTYxLQLmFSM5SNpX0uycz1uSLmxQp7+k1Tl1frK16zUzs+Yp9AxiATAt7azvvfqJW3Kba0TMB3rD+7fPvg7cnVH1sYgY3Nzlm5lZcRSaIF5JP9uln2I5Dng5Il4t4jLNzKwICn2S+mclWv/pwKQ8ZZ+VNAd4A/hBRDyTVUnSMGAYQI8ePUoSpJlZW9RogpA0KiKGS7qftKO+XBExZEtXLGk7YAhwaUbxLGCPiHhH0iDgHmDvrOVExBhgDECfPn18Ad3MrEiaOoP4OjAcuLoE6z4JmBURSxsWRMRbOcMPSvpvSV0iYkUJ4jAzswxNJYiXASLikRKseyh5mpckfQJYGhEhqS/J3VYrSxCDmZnl0VSCqJE0Ml/hlnbWJ2kH4ATgOznTzkuXOZqkK/H/kFQHrAVO9/MXZmbl1VSCqAI6kTw5XTQRsQbo3GDa6JzhUcCoYq7TzMyap6kEsSQiLi9LJGZmtk1p6knqop45mJlZy9FUgjiuLFGYmdk2p9EEERFvlisQMzPbthTaWZ+ZmbUxThBmZpbJCcLMzDI5QZiZWaZCu/s2szag5yUPlGU9C6/8QlnWY1vHZxBmZpbJCcLMzDI5QZiZWSYnCDMzy+QEYWZmmZwgzMwskxOEmZllcoIwM7NMflDOzKyEyvHwYakePPQZhJmZZXKCMDOzTE4QZmaWqWIJQtJCSXMlzZY0I6Nckn4n6SVJT0s6pBJxmpm1VZW+SH1MRKzIU3YSsHf6ORy4If02M7My2JabmE4GJkbiSWBnSbtWOigzs7aikgkigCmSZkoallG+G7AoZ3xxOs3MzMqgkk1MR0bE65I+DkyV9HxEPNrchaTJZRhAjx49ih2jmVmbVbEziIh4Pf1eBtwN9G1Q5XVg95zx7um0hssZExF9IqJPTU1NqcI1M2tzKpIgJO0gacf6YeBEYF6DavcBX0/vZuoHrI6IJWUO1cyszapUE1NX4G5J9THcHhF/lnQeQESMBh4EBgEvAe8C36xQrGZmbVJFEkRELAAOypg+Omc4gO+WMy4zM/vAtnybq5mZVZAThJmZZXKCMDOzTE4QZmaWyQnCzMwyOUGYmVkmJwgzM8vkBGFmZpmcIMzMLJMThJmZZXKCMDOzTE4QZmaWyQnCzMwyOUGYmVkmJwgzM8vkBGFmZpmcIMzMLJMThJmZZXKCMDOzTE4QZmaWyQnCzMwylT1BSNpd0sOSnpX0jKQRGXX6S1otaXb6+Um54zQza+vaV2CddcD3I2KWpB2BmZKmRsSzDeo9FhGDKxCfmZlRgTOIiFgSEbPS4beB54Ddyh2HmZk1rqLXICT1BA4G/pFR/FlJcyQ9JOkz5Y3MzMwq0cQEgKROwF3AhRHxVoPiWcAeEfGOpEHAPcDeeZYzDBgG0KNHj9IFbGbWxlTkDEJSB5LkcFtE/KlheUS8FRHvpMMPAh0kdclaVkSMiYg+EdGnpqampHGbmbUllbiLScBY4LmI+E2eOp9I6yGpL0mcK8sXpZmZVaKJ6QjgLGCupNnptB8BPQAiYjRwCvAfkuqAtcDpEREViNXMrM0qe4KIiMcBNVFnFDCqPBGZmVkWP0ltZmaZnCDMzCyTE4SZmWVygjAzs0xOEGZmlskJwszMMjlBmJlZJicIMzPL5ARhZmaZnCDMzCyTE4SZmWVygjAzs0xOEGZmlskJwszMMjlBmJlZJicIMzPL5ARhZmaZnCDMzCyTE4SZmWVygjAzs0xOEGZmlskJwszMMlUkQUgaKGm+pJckXZJRvr2kP6Tl/5DUswJhmpm1aWVPEJKqgN8DJwH7A0Ml7d+g2jnAvyNiL+Ba4L/KG6WZmVXiDKIv8FJELIiI9cAdwMkN6pwMTEiH/wgcJ0lljNHMrM1TRJR3hdIpwMCIODcdPws4PCKG59SZl9ZZnI6/nNZZkbG8YcCwdHRfYH6JN2FrdAE+tA22VbxPi8v7s/i29X26R0TUZBW0L3ckxRYRY4AxlY6jEJJmRESfSsfRmnifFpf3Z/G15H1aiSam14Hdc8a7p9My60hqD+wErCxLdGZmBlQmQUwH9pa0p6TtgNOB+xrUuQ/4Rjp8CvC3KHdbmJlZG1f2JqaIqJM0HPgLUAWMi4hnJF0OzIiI+4CxwC2SXgLeJEkirUGLaAprYbxPi8v7s/ha7D4t+0VqMzNrGfwktZmZZXKCMDOzTE4QZSBpnKRl6fMdtpUk7S7pYUnPSnpG0ohKx9TSSaqW9E9Jc9J9+rNKx9QaSKqS9JSkyZWOZUs4QZTHeGBgpYNoReqA70fE/kA/4LsZ3bVY87wHHBsRBwG9gYGS+lU2pFZhBPBcpYPYUk4QZRARj5LcjWVFEBFLImJWOvw2yX/A3SobVcsWiXfS0Q7px3ewbAVJ3YEvADdXOpYt5QRhLVra0+/BwD8qHEqLlzaHzAaWAVMjwvt061wHXAxsqnAcW8wJwlosSZ2Au4ALI+KtSsfT0kXExojoTdK7QV9JB1Q4pBZL0mBgWUTMrHQsW8MJwlokSR1IksNtEfGnSsfTmkTEKuBhfN1saxwBDJG0kKTH6mMl3VrZkJrPCcJanLTr97HAcxHxm0rH0xpIqpG0czrcETgBeL6iQbVgEXFpRHSPiJ4kPUH8LSLOrHBYzeYEUQaSJgFPAPtKWizpnErH1MIdAZxFclQ2O/0MqnRQLdyuwMOSnibpL21qRLTIWzOteNzVhpmZZfIZhJmZZXKCMDOzTE4QZmaWyQnCzMwyOUGYmVkmJwhrlSR9QtIdkl6WNFPSg5L2qVAsP9qCeRZK6pIz3r+QHkEl/W/63VPS15q7XrNcThDW6qQP0t0NTIuIT0XEocClQNcKhdTsBLGlIuJz6WBPwAnCtooThLVGxwAbImJ0/YSImBMRjynxa0nzJM2VdBq8f4T+iKR7JS2QdKWkM9J3JMyV9Km03nhJoyXNkPRC2ucOks6WNKp+fZImp8u8EuiYPsx3W1p2Zrrc2ZJulFTVnI2TVJu+Y2RaGuv3csrqe2S9EjgqXcf/lfSZnHU+LWnvLdu11pY4QVhrdACQr5O0L5O87+Ag4Hjg15J2TcsOAs4D9iN5UnufiOhL0l3zBTnL6An0JenKebSk6nyBRMQlwNqI6B0RZ0jaDzgNOCLtGG8jcMYWbOOngQFpHD9N+6bKdQnwWLrea9Pt+m26zj7A4i1Yp7Ux7SsdgFmZHQlMioiNwFJJjwCHAW8B0yNiCYCkl4Ep6TxzSc5K6t0ZEZuAFyUtIPmxLtRxwKHA9KQljI4k3Ws3lNXFQe60ByLiPeA9SctIms8a+9F/ArgsfUfBnyLixWbEbG2UzyCsNXqG5Ee4ud7LGd6UM76JzQ+mGv54B8lb7nL/P+U7qxAwIT2y7x0R+0ZEbUa9lcAuOeMfA1bkiXUjTRzsRcTtwBBgLfCgpGMbq28GThDWOv0N2F7SsPoJkg6UdBTwGHBa+nKcGuBo4J/NXP6pktql1yU+CcwHFgK90+m7kzT91NuQ0wT0V+AUSR9P4/qYpD0y1jGNpJmL9BrFmSRdcBfqbWDH+hFJnwQWRMTvgHuBA5uxLGuj3MRkrU5EhKQvAddJ+iGwjuQH/ELgceCzwBySI/+LI+JfkprTTPQaSVL5KHBeRKyT9HfgFeBZklegzsqpPwZ4WtKs9DrEj4EpktoBG4DvAq82WMfPgRskzSE56/gz0Jz3CTwNbEznHw9sD5wlaQPwL+CXzViWtVHuzdWsGSSNByZHxB8rHYtZqbmJyczMMvkMwszMMvkMwszMMjlBmJlZJicIMzPL5ARhZmaZnCDMzCzT/we8m5IJg+A7wgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def calc_run_time_par(cnt, size, increment_value):\n",
    "    workgroup = ol.vadd_group[0:cnt]  \n",
    "    a = pynq.allocate(size)   \n",
    "    \n",
    "    a_split = ol.split_buffer(a, workgroup)\n",
    "    a_elements_size = ol.get_elements_size(a_split)\n",
    "    #print(f\"A: {a}\")\n",
    "    t0 = time.time()\n",
    "    h = par_run(a_split, a_elements_size, increment_value, accelerator_group=workgroup)\n",
    "    h.wait()\n",
    "    t1 = time.time()\n",
    "    #print(f\"A: {a}\")\n",
    "    \n",
    "    execution_time = t1-t0\n",
    "    return execution_time\n",
    "\n",
    "def calc_run_time(size, increment_value):\n",
    "    workgroup = ol.vadd_group[0:1]\n",
    "    \n",
    "    a = pynq.allocate(size)\n",
    "    #print(f\"A: {a}\")\n",
    "    t0 = time.time()\n",
    "    handle = workgroup[0].start(a, size, increment_value)\n",
    "    handle.wait()\n",
    "    t1 = time.time()\n",
    "    #print(f\"A: {a}\")\n",
    "    \n",
    "    exec_time = t1-t0\n",
    "    return exec_time\n",
    "\n",
    "\n",
    "inc_val = 10\n",
    "parallel_elapsed_times = list()\n",
    "single_elapsed_times = list()\n",
    "power_indexes = list()\n",
    "cmp_unit_indexes = list()\n",
    "\n",
    "for i in range(14, 21):\n",
    "    power = 2**i\n",
    "    parallel_elapsed_times = list()\n",
    "    single_elapsed_times = list()\n",
    "    power_indexes = list()\n",
    "    cmp_unit_indexes = list()\n",
    "    for cmp_unit in range(1, 5):\n",
    "        parallel_elapsed_times.append(calc_run_time_par(cmp_unit, power, inc_val))\n",
    "        single_elapsed_times.append(calc_run_time(power, inc_val))\n",
    "        power_indexes.append(power)\n",
    "        cmp_unit_indexes.append(cmp_unit)\n",
    "\n",
    "\n",
    "    \n",
    "    x_axis_data = cmp_unit_indexes\n",
    "    \n",
    "    parallel_elapsed_axis_data = np.multiply(parallel_elapsed_times,1000.0)\n",
    "    single_elapsed_axis_data = np.multiply(single_elapsed_times,1000.0)\n",
    "\n",
    "    width = 0.3\n",
    "    move = np.ones(len(x_axis_data))*width\n",
    "    \n",
    "    fig, ax = plt.subplots()\n",
    "    \n",
    "    ax.bar(x_axis_data+move, parallel_elapsed_axis_data, width, label='Parallel')\n",
    "    ax.bar(x_axis_data, single_elapsed_axis_data, width, label='Single')\n",
    "\n",
    "    plt.xticks(x_axis_data+(move)*0.5 , range(1, 5))\n",
    "    ax.set_ylabel('Time [ms]')\n",
    "    ax.set_xlabel('Compute Units')\n",
    "    ax.set_title(f'Elements to process: {power}')\n",
    "    ax.legend()\n",
    "    \n",
    "    plt.savefig(f'/home/root/{power}.png')\n",
    "    plt.show()\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x_axis_data = cmp_unit_indexes\n",
    "\n",
    "parallel_elapsed_axis_data = parallel_elapsed_times\n",
    "single_elapsed_axis_data = single_elapsed_times\n",
    "\n",
    "width = 0.3\n",
    "move = np.ones(len(x_axis_data))*width\n",
    "               \n",
    "fig, ax = plt.subplots()\n",
    "#bar(x, height, width=0.8, bottom=None, *, align='center',\n",
    "#data=None, **kwargs)[source]\n",
    "ax.bar(x_axis_data, parallel_elapsed_axis_data, width, label='Parallel')\n",
    "ax.bar(x_axis_data+move, single_elapsed_axis_data, width, label='Single')\n",
    "\n",
    "plt.xticks(x_axis_data+(move)*0.5 , range(1, 5))\n",
    "ax.set_ylabel('Time')\n",
    "ax.set_xlabel('Compute Units')\n",
    "ax.set_title(f'Processing time {power}')\n",
    "ax.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Pass\n",
    "size = 2**20\n",
    "kkk = list()\n",
    "for elo in range(5): \n",
    "    kkk.append(pynq.allocate(size))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PynqBuffer([0, 0, 0, ..., 0, 0, 0], dtype=uint32)"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PynqBuffer([0, 0, 0, ..., 0, 0, 0], dtype=uint32)"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PynqBuffer([0, 0, 0, ..., 0, 0, 0], dtype=uint32)"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}