Simple Python3 parallelism

simple_python3_parallelism.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Blocking parallelism in Python 3\n",
    "Dispatch parallel commands and block until the last of them completes à la GNU Parallel.\n",
    "- Uses available logical CPUs\n",
    "- Python 3.2+\n",
    "- Functions and arguments must be pickleable"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1) Parallelising functions\n",
    "### Just run the functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1\n",
      "3\n",
      "2\n"
     ]
    }
   ],
   "source": [
    "import concurrent.futures\n",
    "\n",
    "def run(function, iterable):\n",
    "    with concurrent.futures.ProcessPoolExecutor() as x:\n",
    "        result = x.map(function, iterable)\n",
    "        \n",
    "run(print, [1,2,3]) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Return list of function return values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[1, 2, 3]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import concurrent.futures\n",
    "\n",
    "def run(function, iterable):\n",
    "    with concurrent.futures.ProcessPoolExecutor() as x:\n",
    "        return [r for r in x.map(function, iterable)]\n",
    "    \n",
    "run(abs, [-1,-2,-3])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Return dict associating input and return values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{-3: 3, -2: 2, -1: 1}"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import concurrent.futures\n",
    "\n",
    "def run(function, iterable):\n",
    "    with concurrent.futures.ProcessPoolExecutor() as x:\n",
    "        return {iterable:r for iterable, r in zip(iterable, x.map(function, iterable))}\n",
    "    \n",
    "run(abs, [-1,-2,-3])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2) Parallelising subprocesses\n",
    "Returns a list of `CompletedProcess` instances with input arguments, stdout, stderr and return code."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import functools\n",
    "import subprocess\n",
    "import concurrent.futures\n",
    "\n",
    "def run_cmds(cmds):\n",
    "    run = functools.partial(subprocess.run,\n",
    "                            shell=True,\n",
    "                            check=True, # Raise CalledProcessError with non zero exit code\n",
    "                            universal_newlines=True,\n",
    "                            stdout=subprocess.PIPE, # Capture stdout\n",
    "                            stderr=subprocess.PIPE) # Capture stderr \n",
    "    with concurrent.futures.ProcessPoolExecutor() as x:\n",
    "        return [r for r in x.map(run, cmds)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[CompletedProcess(args='sleep 4; echo sleep4', returncode=0, stdout='sleep4\\n', stderr=''),\n",
       " CompletedProcess(args='sleep 5; echo sleep5', returncode=0, stdout='sleep5\\n', stderr=''),\n",
       " CompletedProcess(args='sleep 3; echo sleep3', returncode=0, stdout='sleep3\\n', stderr=''),\n",
       " CompletedProcess(args='sleep 2; echo sleep2', returncode=0, stdout='sleep2\\n', stderr='')]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cmds = ['sleep 4; echo sleep4',\n",
    "        'sleep 5; echo sleep5',\n",
    "        'sleep 3; echo sleep3',\n",
    "        'sleep 2; echo sleep2']\n",
    "\n",
    "run_cmds(cmds)"
   ]
  }
 ],
 "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": "3.6.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}