Arduino RPC transfer benchmark

Transfer benchmark.ipynb

{
 "metadata": {
  "name": "",
  "signature": "sha256:6760c4239f4ddabe62c7b561cd70b206bcae136a1286a55be99cdda4117e03ce"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from serial import Serial\n",
      "from arduino_rpc.node import Proxy\n",
      "import pandas as pd\n",
      "from si_prefix import si_format"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 40
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "-------------------------------------------------------------------"
     ]
    },
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Transfer rate benchmark test: one-way versus round-trip"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "d = Serial('/dev/ttyUSB1', baudrate=115200)\n",
      "p = Proxy(d)\n",
      "p.str_echo(map(ord, 'hello, world!')).tostring()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 32,
       "text": [
        "'hello, world!'"
       ]
      }
     ],
     "prompt_number": 32
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Test one-way (i.e., `array_length`) and round-trip (i.e., `str_echo`)\n",
      "# for byte counts from 1-100.\n",
      "frames = []\n",
      "for n in range(1, 102, 10):\n",
      "    data = map(ord, n * 'h')\n",
      "    \n",
      "    echo_timeit = %timeit -qo p.str_echo(data)\n",
      "    length_timeit = %timeit -qo p.array_length(data)\n",
      "    frame = pd.DataFrame([[n, echo_timeit.best, length_timeit.best]],\n",
      "                         columns=['byte_count', 'echo_seconds',\n",
      "                                  'length_seconds'])\n",
      "    frames.append(frame)\n",
      "times = pd.concat(frames)\n",
      "# __NB__ Support for returning arrays currently **copies** the result array\n",
      "# to the output buffer.  This copy is not strictly necessary from a technical\n",
      "# perspective, but it an implementation detail.  The impact of this is the\n",
      "# added runtime overhead associated with the `memcpy`, though this should be\n",
      "# insignificant compared to the communication time."
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 35
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%matplotlib inline"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 36
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Compute minimum serial communication time for one-way transfer of\n",
      "# each byte.\n",
      "min_seconds_per_byte = (8. / d.baudrate)\n",
      "\n",
      "times['baud_seconds'] = (times['length_seconds'].min() +\n",
      "                         min_seconds_per_byte * times['byte_count'])\n",
      "\n",
      "axis = times.plot(kind='scatter', x=['byte_count', 'byte_count'],\n",
      "                  y=['echo_seconds', 'length_seconds'], xlim=0,\n",
      "                  marker='x', s=8 ** 2)\n",
      "times.plot(x='byte_count', y='baud_seconds', xlim=0, ax=axis)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 42,
       "text": [
        "<matplotlib.axes.AxesSubplot at 0x7f7886b42590>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAZUAAAEUCAYAAADjt6tGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nREREREREREREREQklfx/9bfQj4g+hBYAAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f7886b4c910>"
       ]
      }
     ],
     "prompt_number": 42
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "diff = times.iloc[-1] - times.iloc[0]\n",
      "slope = diff[['echo_seconds', 'length_seconds']] / diff.byte_count\n",
      "slope.name = 'time_per_byte'\n",
      "slope['baud_seconds'] = min_seconds_per_byte\n",
      "pd.set_eng_float_format(accuracy=2, use_eng_prefix=True)\n",
      "axis = slope.plot(kind='bar')\n",
      "axis.set_yticklabels(['%ss' % si_format(t, 0) for t in axis.get_yticks()])\n",
      "pass"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAFECAYAAAAuvukdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkiQp+H9Jw8EKrQh1uwAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f7886e61f90>"
       ]
      }
     ],
     "prompt_number": 46
    }
   ],
   "metadata": {}
  }
 ]
}