ExampleMultiprocessingSimple.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "627d27c6",
   "metadata": {},
   "source": [
    "# Example Multiprocessing Simple\n",
    "\n",
    "This example consists of three files of a simple example of multiprocessing. The files are:\n",
    "<ol>\n",
    "    <li>\n",
    "        example_multiprocessing_simple\n",
    "    </li>\n",
    "    <li>\n",
    "        example_multiprocessing_simple_process_x\n",
    "    </li>\n",
    "    <li>\n",
    "        example_multiprocessing_simple_process_y\n",
    "    </li>\n",
    "</ol>\n",
    "The last two files contain the target functions for processes x and y, respectively.\n",
    "The target functions are put in separate files because of a feature of iPython and\n",
    "Jupyter notebooks.\n",
    "\n",
    "##  Processes and Streams in the Example\n",
    "\n",
    "The example consists of two processes: <b>process_x</b> and <b>process_y</b>, which\n",
    "have input streams called 'x' and 'y', respectively.\n",
    "When <b>process_x</b> gets a new item on the stream named 'x' the process prints the\n",
    "item and appends the item, incremented by 1, to the stream named 'y'.\n",
    "Likewise, when <b>process_y</b> gets a new item on the stream named 'y' the process prints the\n",
    "item and appends the item, incremented by 1, to the stream named 'x'.\n",
    "\n",
    "\n",
    "<b>process_y</b> starts by sending 1 on the stream called 'x'. So, <b>process_x</b> prints 1,\n",
    "then <b>process_y</b> prints 2, then <b>process_x</b> prints 3, and so forth. The order of\n",
    "printing is an arbitrary interleaving of these values because it depends on which process\n",
    "gets the printer first.\n",
    "\n",
    "\n",
    "<b>process_y</b> keeps track of the number of messages that it sends and halts execution after\n",
    "sending 3 messages.\n",
    "\n",
    "\n",
    "## Structure of the Code.\n",
    "\n",
    "<ol>\n",
    "    <li>\n",
    "        <b><i>Create the input queues for the processes</i></b>: \n",
    "        <br>\n",
    "        <b>q_x, q_y</b> are the input\n",
    "        queues for processes x and y. Putting <b>q_x, q_y</b> into a dict helps to ensure that\n",
    "        that both processes get identical values for <b>q_x, q_y</b>.\n",
    "    </li>\n",
    "    <li>\n",
    "        <b><i>Create the processes</i></b>\n",
    "    </li>\n",
    "    <li>\n",
    "        <b><i>Start, join, and terminate processes</i></b>\n",
    "    </li>\n",
    "</ol>\n",
    "\n",
    "## Serializing Data\n",
    "\n",
    "This code uses <i>JSON</i> for serializing data. You can also use <i>pickle</i> as illustrated\n",
    "in <b>example_multiprocessing_simple_pickle.py</b>.\n",
    "\n",
    "Certain types of Python objects cannot be serialized using <i>JSON</i>. So, use <i>pickle</i>\n",
    "when you want to pass complex types of classes or complex numbers between processes.\n",
    "The IoTPy library uses <i>JSON</i> as the default.\n",
    "\n",
    "We use <i>JSON</i> for distributed (message-passing) computing.\n",
    "<i>pickle</i> cannot be used with some message-passing libraries.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "390ad30b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "message received by process y:  2\n",
      "message received by process y:  4\n",
      "message received by process y:  6\n",
      "message received by process x:  1\n",
      "message received by process x:  3\n",
      "message received by process x:  5\n",
      "message received by process x:  7\n"
     ]
    }
   ],
   "source": [
    "import multiprocessing as mp\n",
    "\n",
    "# import files containing target functions of \n",
    "# processes x, y.\n",
    "# The target functions are in separate files because of\n",
    "# features of iPython and Jupyter notebooks.\n",
    "import example_multiprocessing_simple_process_x\n",
    "import example_multiprocessing_simple_process_y\n",
    "\n",
    "# q_x is the input queue for process x\n",
    "# q_y is the input queue for process y\n",
    "# put the queues into a dict that is passed to processes x, y.\n",
    "# Don't pass queues directly to external files.\n",
    "q_x, q_y = mp.Queue(), mp.Queue()\n",
    "dict_queues = {'x': q_x, 'y': q_y}\n",
    "\n",
    "# Create processes x and y\n",
    "process_x = mp.Process(\n",
    "    target=\\\n",
    "    example_multiprocessing_simple_process_x.process_target_x,\n",
    "    args=(dict_queues,),\n",
    "    name='process_x')\n",
    "\n",
    "process_y = mp.Process(\n",
    "    target=\\\n",
    "    example_multiprocessing_simple_process_y.process_target_y,\n",
    "    args=(dict_queues,),\n",
    "    name='process_y')\n",
    "\n",
    "# Start; join; terminate processes x and y\n",
    "process_x.daemon = True\n",
    "process_y.daemon = True\n",
    "\n",
    "process_x.start()\n",
    "process_y.start()\n",
    "\n",
    "process_x.join()\n",
    "process_y.join()\n",
    "    \n",
    "process_x.terminate()\n",
    "process_y.terminate()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}