{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "window_size = 8\n",
    "alpha = 0.025  # learning rate\n",
    "p = 100  # p = dimensions of document vectors (no. of features)\n",
    "m = len(vocab)  # number of words in the corpus \n",
    "\n",
    "D = np.random.rand(p, n)  # matrix of document embeddings\n",
    "U = scipy.stats.truncnorm.rvs(-2, 2, loc=0, scale=1, size=(m, p))  # matrix of softmax weights\n",
    "\n",
    "epochs = 1\n",
    "for epoch in range(epochs):\n",
    "    for i in range(n):\n",
    "        # Feed-forward\n",
    "        d = np.array(np.zeros(n), ndmin=2).T\n",
    "        d[i] = 1\n",
    "        \n",
    "        e = np.array(np.dot(D, d), ndmin=2)\n",
    "        k = np.array(np.dot(U, e), ndmin=2)\n",
    "        t_hat = softmax(k)\n",
    "        \n",
    "        doc_words = documents[i]\n",
    "        middle = randint(window_size, len(doc_words) - window_size - 1)\n",
    "\n",
    "        #window_words = []      \n",
    "        \n",
    "        errors_out = (np.array(np.zeros(m), ndmin=2).T)\n",
    "        errors_middle = (np.array(np.zeros(p), ndmin=2).T)\n",
    "        \n",
    "        for c in range(middle - window_size, middle + window_size):\n",
    "            #window_words.append(doc_words[c])\n",
    "            \n",
    "            t = (np.array(np.zeros(len(vocab)), ndmin=2).T)\n",
    "            t[vocab.index(doc_words[c])] = 1\n",
    "            errors_out += t_hat - t\n",
    "            errors_middle += np.dot(U.T, errors_out)\n",
    "            \n",
    "            if c == middle:\n",
    "                print(cross_entropy_loss(t, t_hat))\n",
    "\n",
    "        # Backprogation\n",
    "        U += - alpha * np.dot(errors_out, e.T)\n",
    "        D += - alpha * np.dot(errors_middle, d.T)"
   ]
  }
  ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [default]",
   "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": "3.6.3"
  },
  "toc": {
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": "block",
   "toc_window_display": false
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}