entropy_of_programming_languages.ipynb

{
  "cells": [
    {
      "metadata": {
        "_uuid": "1402cc0cdcedd669173186edb813eed2e6d2145c",
        "_cell_guid": "c9e1b183-f1e0-448e-bf64-9fff15d230d9"
      },
      "cell_type": "markdown",
      "source": "# Entropy rates of programming languages\n\nThe aim of this project was to estimate information entropy rates across programming languagues. Entropy rate estimation is a well-known research problem in computational linguistics: it was first posed by Shannon himself in his 1951 paper [\"Prediction and Entropy of Printend English\"](http://languagelog.ldc.upenn.edu/myl/Shannon1950.pdf), which estabilished that entropy rate average English corpus is approximately 1 bit per character. Similar estimates were later computed for other natural languagues. (For a modern review of this problem, see [Takahira et al., 2016, \"Entropy Rate Estimates for Natural Language—A New Extrapolation of Compressed Large-Scale Corpora\"](http://www.mdpi.com/1099-4300/18/10/364).) No research know to me, however, concerned the topic of entropy estimation for programming languages.\n\nIn this study, I compare entropy rates of Python, JavaScript and C++ using Kaggle's [GitHub Repos](https://www.kaggle.com/github/github-repos) dataset. I sample 10000 files of Python, JavaScript, and C++ code and train *n*-th order Markov models (better known as character-level *n*-gram language models in NLP community) with $n \\in [0, 5]$ range. Then, I compute their entropy rates and investigate their asymptotic behavior.\n\n## What is entropy rate\n\nLet us begin entropy, a measure of predictability of a dataset. Let $X$ be character sampled from some text; it can be any character from a distribution over an alpabet $A$ Then, its entropy $H(x)$ is expressed by the following equation:\n\n$$H(X) = \\sum_{x \\in A} P(X=x) \\log_2 (P(X = x))$$\n\nIntuitively, the distribution over an alpabet $A$ does depened on characters preceding $X$. This can be captured with the notion of conditional entropy $H(X|Y)$, where $Y$ is a character just before $X$:\n\n$$H(X|Y) = \\sum_{y \\in A} \\sum_{x \\in A} P(Y=y, X=X) \\log_2 (P(Y=y, X = x))$$\n\nFinally, entropy rate is entropy conditioned on a context of infinitely many preceding characers.\n\n$$h(X_n) = \\lim_{n \\to \\infty} H(X_n | X_1, ..., X_{n-1})$$\n\nAs mentioned above, we deal with this infinite limit with a rather crude esimate of $n \\in [0, 5]$."
    },
    {
      "metadata": {
        "collapsed": true,
        "_uuid": "a66e289ce63f335f90316f8b63584fb96b4e898c",
        "_cell_guid": "4ff92131-3070-43d4-839c-eca20aeeb479",
        "trusted": true
      },
      "cell_type": "code",
      "source": "from collections import defaultdict, deque, Counter\nimport random\nimport pandas as pd\nimport numpy as np\nimport seaborn as sns\nimport matplotlib.pyplot as plt\nfrom google.cloud import bigquery\nimport warnings\nwarnings.filterwarnings('ignore')\n\nsns.set()\n%matplotlib inline\nclient = bigquery.Client()",
      "execution_count": 1,
      "outputs": []
    },
    {
      "metadata": {
        "collapsed": true,
        "_uuid": "2f2fa14b1e82201f1a86a9053acd48c2932901bc",
        "_cell_guid": "5e74b138-c066-47b5-9bfc-e2cab0c43a25",
        "trusted": true
      },
      "cell_type": "code",
      "source": "def get_files(language_extension):\n    QUERY = ('''\n    SELECT SPLIT(content, '\\\\n') AS line\n    FROM `bigquery-public-data.github_repos.sample_contents`\n    WHERE sample_path LIKE \"%.{ext}\"\n    \n    '''.format(ext=language_extension))\n    query_job = client.query(QUERY)\n    bigquery_iterator = iter(query_job.result(timeout=120))\n    return bigquery_iterator",
      "execution_count": 2,
      "outputs": []
    },
    {
      "metadata": {
        "_uuid": "c27abdac84c3e9840692a9af7cf7d1127016c9ab",
        "_cell_guid": "963d5ba5-66ec-49e8-9c6d-71e3e49ade70"
      },
      "cell_type": "markdown",
      "source": "## An example data entry"
    },
    {
      "metadata": {
        "_uuid": "6e2970a6d0f61f2f2c242ba97b657d902f46f7d6",
        "_cell_guid": "d26cd9cd-afd9-4da3-aba9-a075c767adf1",
        "trusted": true
      },
      "cell_type": "code",
      "source": "bigquery_iterator = get_files('py')\nnext(bigquery_iterator)[0] ",
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 3,
          "data": {
            "text/plain": "['# -*- encoding: utf-8 -*-',\n '##############################################################################',\n '#',\n '#    Product - Simple Pricelist module for Odoo',\n '#    Copyright (C) 2015-Today GRAP (http://www.grap.coop)',\n '#    @author Sylvain LE GAL (https://twitter.com/legalsylvain)',\n '#',\n '#    This program is free software: you can redistribute it and/or modify',\n '#    it under the terms of the GNU Affero General Public License as',\n '#    published by the Free Software Foundation, either version 3 of the',\n '#    License, or (at your option) any later version.',\n '#',\n '#    This program is distributed in the hope that it will be useful,',\n '#    but WITHOUT ANY WARRANTY; without even the implied warranty of',\n '#    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the',\n '#    GNU Affero General Public License for more details.',\n '#',\n '#    You should have received a copy of the GNU Affero General Public License',\n '#    along with this program.  If not, see <http://www.gnu.org/licenses/>.',\n '#',\n '##############################################################################',\n '',\n '',\n 'from openerp.osv import fields',\n 'from openerp.osv.orm import Model',\n '',\n '',\n 'class res_partner(Model):',\n \"    _inherit = 'res.partner'\",\n '',\n '    # Function Fields Section',\n '    def _get_template_info(self, cr, uid, ids, name, args, context=None):',\n '        res = {}',\n \"        supplierinfo_obj = self.pool['product.supplierinfo']\",\n '        for partner in self.browse(cr, uid, ids, context=context):',\n '            supplierinfo_ids = supplierinfo_obj.search(cr, uid, [',\n \"                ('name', '=', partner.id)], context=context)\",\n '            supplierinfos = supplierinfo_obj.browse(',\n '                cr, uid, supplierinfo_ids, context=context)',\n '            template_ids = [x.product_id for x in supplierinfos]',\n '            res[partner.id] = {',\n \"                'template_ids': template_ids,\",\n \"                'template_count': len(template_ids)}\",\n '        return res',\n '',\n '    def _product_supplierinfo_2_res_partner(',\n '            self, cr, uid, ids, context=None):',\n '        \"\"\"Return \\'product.supplierinfo\\' Info',\n '        where somes \\'pricelist.partnerinfo\\' changes.\"\"\"',\n \"        supplierinfo_obj = self.pool['product.supplierinfo']\",\n '        supplierinfos = supplierinfo_obj.browse(cr, uid, ids, context=context)',\n '        res = [x.name.id for x in supplierinfos]',\n '        return res',\n '',\n '    # Column Section',\n '    _columns = {',\n \"        'template_ids': fields.function(\",\n \"            _get_template_info, type='one2many', relation='product.template',\",\n \"            string='Products', multi='template_info', readonly=True),\",\n \"        'template_count': fields.function(\",\n \"            _get_template_info, type='integer', string='Products Quantity',\",\n \"            multi='product_info', store={\",\n \"                'product.supplierinfo': (\",\n '                    _product_supplierinfo_2_res_partner, [',\n \"                        'name',\",\n '                    ], 10)',\n '                }),',\n '    }',\n '']"
          },
          "metadata": {}
        }
      ]
    },
    {
      "metadata": {
        "collapsed": true,
        "_uuid": "e46626b0fc7d29ff8b176319e02849710f45e711",
        "_cell_guid": "88127c77-f3f8-4024-8f73-6ff2a8307699",
        "trusted": true
      },
      "cell_type": "code",
      "source": "class MarkovModel(object):\n    def __init__(self, n):\n        self.n = n\n        self.model = defaultdict(Counter)\n        self.freqs = Counter()\n        self.buffer = deque(maxlen=self.n)\n    \n    def fit(self, stream):\n        for token in stream:\n            prefix = tuple(self.buffer)\n            self.buffer.append(token)\n            if len(prefix) == self.n:\n                self.freqs[prefix] += 1\n                self.model[prefix][token] += 1\n    \n    def entropy(self, prefix):\n        prefix_freqs = self.model[prefix].values()\n        normalization_factor = self.freqs[prefix]\n        return -np.sum(f/normalization_factor * np.log2(f/normalization_factor) \n                       for f in prefix_freqs)\n                \n    def entropy_rate(self):\n        normalization_factor = sum(self.freqs.values())\n        unnormalized_rate = np.sum(self.freqs[prefix] * self.entropy(prefix) for prefix in self.freqs)\n        try:\n            return unnormalized_rate/normalization_factor\n        except ZeroDivisionError:\n            return 0",
      "execution_count": 4,
      "outputs": []
    },
    {
      "metadata": {
        "collapsed": true,
        "_uuid": "1829ace69d7e9c7b4018ac27d806ed11058ccd85",
        "_cell_guid": "d5def369-086e-4483-a0e6-9a23a945576f",
        "trusted": true
      },
      "cell_type": "code",
      "source": "def tokenizer(text):\n    return text  # char-level tokens\n\ndef iterator(bigquery_iterator, items=10000):\n    counter = 0\n    while(counter < items):\n        counter += 1\n        file = next(bigquery_iterator)[0]\n        for line in file:\n            if line == '':\n                continue\n            for token in tokenizer(line.lower().strip()):\n                yield token",
      "execution_count": 5,
      "outputs": []
    },
    {
      "metadata": {
        "scrolled": false,
        "_uuid": "6c61399d4f1b6d8b2df51a620647459f023e75a9",
        "_cell_guid": "7d098fc0-2ebc-41fd-8064-01c0ccee195f",
        "trusted": true
      },
      "cell_type": "code",
      "source": "%%time\npy_files = list(iterator(get_files('py')))\ncpp_files = list(iterator(get_files('cpp')))\njs_files = list(iterator(get_files('js')))\n\nrates = pd.DataFrame(columns=['n', 'language', 'entropy_rate'])\nfor language, it in [('py', py_files), ('js', js_files), ('cpp', cpp_files)]:\n    for n in range(5):\n        model = MarkovModel(n=n)\n        model.fit(stream=it)\n        rate = model.entropy_rate()\n        print(n, language, rate)\n        rates.loc[len(rates)] = [n, language, rate]",
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": "0 py 4.97048782761\n1 py 4.17550438653\n2 py 3.35339423128\n3 py 2.550327241\n4 py 1.98113780103\n0 js 5.27707699389\n1 js 4.27686802426\n2 js 3.31937568224\n3 js 2.51101888077\n4 js 1.97103232939\n0 cpp 5.11329821155\n1 cpp 4.14952123542\n2 cpp 3.37304549766\n3 cpp 2.63247027467\n4 cpp 2.06613008248\nCPU times: user 44min 2s, sys: 7.5 s, total: 44min 9s\nWall time: 47min 59s\n",
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "_kg_hide-input": false,
        "_uuid": "0036a59036dea46d3c54cb31f70d866ec43c259b",
        "_cell_guid": "79b0cb02-8c72-43fe-9ae5-3f58d0a9d6a7",
        "trusted": true
      },
      "cell_type": "code",
      "source": "rates",
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 11,
          "data": {
            "text/plain": "    n language  entropy_rate\n0   0       py      4.970488\n1   1       py      4.175504\n2   2       py      3.353394\n3   3       py      2.550327\n4   4       py      1.981138\n5   0       js      5.277077\n6   1       js      4.276868\n7   2       js      3.319376\n8   3       js      2.511019\n9   4       js      1.971032\n10  0      cpp      5.113298\n11  1      cpp      4.149521\n12  2      cpp      3.373045\n13  3      cpp      2.632470\n14  4      cpp      2.066130",
            "text/html": "<div>\n<style>\n    .dataframe thead tr:only-child th {\n        text-align: right;\n    }\n\n    .dataframe thead th {\n        text-align: left;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>n</th>\n      <th>language</th>\n      <th>entropy_rate</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0</td>\n      <td>py</td>\n      <td>4.970488</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>1</td>\n      <td>py</td>\n      <td>4.175504</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>2</td>\n      <td>py</td>\n      <td>3.353394</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>3</td>\n      <td>py</td>\n      <td>2.550327</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>4</td>\n      <td>py</td>\n      <td>1.981138</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>0</td>\n      <td>js</td>\n      <td>5.277077</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>1</td>\n      <td>js</td>\n      <td>4.276868</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>2</td>\n      <td>js</td>\n      <td>3.319376</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>3</td>\n      <td>js</td>\n      <td>2.511019</td>\n    </tr>\n    <tr>\n      <th>9</th>\n      <td>4</td>\n      <td>js</td>\n      <td>1.971032</td>\n    </tr>\n    <tr>\n      <th>10</th>\n      <td>0</td>\n      <td>cpp</td>\n      <td>5.113298</td>\n    </tr>\n    <tr>\n      <th>11</th>\n      <td>1</td>\n      <td>cpp</td>\n      <td>4.149521</td>\n    </tr>\n    <tr>\n      <th>12</th>\n      <td>2</td>\n      <td>cpp</td>\n      <td>3.373045</td>\n    </tr>\n    <tr>\n      <th>13</th>\n      <td>3</td>\n      <td>cpp</td>\n      <td>2.632470</td>\n    </tr>\n    <tr>\n      <th>14</th>\n      <td>4</td>\n      <td>cpp</td>\n      <td>2.066130</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
          },
          "metadata": {}
        }
      ]
    },
    {
      "metadata": {
        "_uuid": "54f3ca864c27a493042fdc8a925a2c22758a698b",
        "_cell_guid": "4d3f3113-95f9-4264-b1c3-03dd6c4d1308"
      },
      "cell_type": "markdown",
      "source": "# Conclusions\n\nBelow are entropy rates for our three languages plotted against Markov model orders. I also fitted 2nd degree polynomials (plotted with confidence intervals) to show the trend as *n* approaches infinity. Although entropy rates seem to converge to 1, we cannot exclude them converging to 0 instead, given our confidence interval."
    },
    {
      "metadata": {
        "_uuid": "0ec79ce2d51bee5559d5e3e5d8c0503372568ed1",
        "_cell_guid": "437646ab-9f51-4a23-8ad3-0b2c8b2477fd",
        "trusted": true
      },
      "cell_type": "code",
      "source": "rates.n = np.float32(rates.n)\nf, (ax1, ax2, ax3) = plt.subplots(3, sharey=True, figsize=(15,15))\nax1.set_ylim(0, 8)\nsns.regplot(x='n', y='entropy_rate', data=rates[rates.language=='py'], ax=ax1, order=2).set_title(\"Python\")\nsns.regplot(x='n', y='entropy_rate', data=rates[rates.language=='cpp'], ax=ax2, order=2).set_title(\"C++\")\nsns.regplot(x='n', y='entropy_rate', data=rates[rates.language=='js'], ax=ax3, order=2).set_title(\"JavaScript\")\nplt.subplots_adjust(hspace=0.3)\n",
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": "<matplotlib.figure.Figure at 0x7f0be6bef6d8>",
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA28AAANsCAYAAADMW5BSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "metadata": {
        "trusted": true,
        "collapsed": true,
        "_uuid": "56eda5a35031c669f320670510989b528009c3d5"
      },
      "cell_type": "code",
      "source": "",
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "language_info": {
      "name": "python",
      "version": "3.6.4",
      "mimetype": "text/x-python",
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "pygments_lexer": "ipython3",
      "nbconvert_exporter": "python",
      "file_extension": ".py"
    },
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 1
}