Minimal character-level language model in Python. Always picks the same prediction from equally probably predictions.

Minimal Character Model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "First Citizen:\n",
      "We have power in me as in a kinsman.\n",
      "\n",
      "YORK:\n",
      "I am thine.\n",
      "\n",
      "EXETER:\n",
      "They seek revenge and therefore for assurance\n",
      "Let's each one send unto his wife;\n",
      "And he whose wife is most obedient subjects, follow him\n",
      "To his new kingdom of perpetual peace\n",
      "By this one bloody trial of sharp war.\n",
      "\n",
      "OXFORD:\n",
      "Every man's conscience and my soul aches\n",
      "To know, when maidens sue,\n",
      "Men give like gods; but when they shall be proud, our garments are not spotted with our company is coming hither.\n",
      "\n",
      "PETRUCHIO:\n",
      "Why, therefore hence amain.\n",
      "3 KING HENRY VI:\n",
      "And I, with grief and shame to utter:\n",
      "He would not stay at your petition with more successful to the queen: please your honour will command me as my sovereign liege, I do beseech you, sir, a word:\n",
      "and as I told you, I, he was a frantic fool,\n",
      "Hiding his bitter jest or two!\n",
      "\n",
      "BIANCA:\n",
      "Am I your bird? I mean to shift my bush;\n",
      "And I, the hapless male to one sweet birds, O, how they sing!\n",
      "Doth set my pugging tooth on edge;\n",
      "For a quart of ale is a dish for a king.\n",
      "The lark, t\n"
     ]
    }
   ],
   "source": [
    "from collections import defaultdict\n",
    "\n",
    "order = 12\n",
    "filename = 'shakespeare-tiny.txt'\n",
    "seed = 'First Citizen:'\n",
    "output_length = 1000\n",
    "\n",
    "text = open(filename).read()\n",
    "frequency = defaultdict(lambda: defaultdict(int))\n",
    "\n",
    "for i in xrange(len(text)-(order+2)):\n",
    "    context = text[i:i+order]\n",
    "    prediction = text[i+order]\n",
    "    frequency[context][prediction] += 1\n",
    "\n",
    "for i in range(output_length):\n",
    "    predictions = frequency[seed[-order:]]\n",
    "    best = max(predictions, key=lambda key: predictions[key])\n",
    "    seed += best\n",
    "    \n",
    "print seed"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}