{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Eliminating biased word embedding association\n",
    "\n",
    "In this experiment we show that replacing using a fair metric (EXPLORE) on the word embedding space eliminated most biases in word embeddings\n",
    "\n",
    "### Word embedding association test\n",
    "\n",
    "Developed by [Caliskan et al. (2017)](https://arxiv.org/pdf/1608.07187.pdf) to evaluate semantic biases in word embeddings, WEATs are inspired by implicit association tests (IAT) from the psychometric literature. Lt $X, Y$ be two sets of word embeddings of target words of equal size (e.g. African-American and European-American names respectively), and $A, B$ be two sets of attribute words (e.g. words with positive and negative sentiments respectively).\n",
    "\n",
    "For each word $x \\in X$, we measure its association with the attribute by:\n",
    "$$ s(x, A, B) \\triangleq \\frac{1}{|A|} \\sum_{a \\in A} \\frac{\\langle x, a \\rangle}{||x|| ||a||} - \\frac{1}{|B|} \\sum_{b \\in B} \\frac{\\langle x, b \\rangle}{||x|| ||b||} $$\n",
    "\n",
    "If $x$ tends to be associated with the attribute (e.g. it has positive or negative sentiment), then we expect $s(x, A, B)$ to be far from zero. To measure the association of $X$ with the attribute, we average the associations of the word in $X$:\n",
    "\n",
    "$$ s(X, A, B) \\triangleq \\frac{1}{|X|} \\sum_{x \\in X} s(x, A, B) $$\n",
    "\n",
    "\n",
    "#### Steps:\n",
    "1. Download the `sentiment_data` folder from the following GitHub repo into the `data` local folder: https://github.com/debarghya-mukherjee/Fair_metric_learning/tree/5333cbd35c3d8febcc9fc25a59a453b6c3cc0c88\n",
    "1. Download the `embeddings` folder from the following Github repo into the `embeddings` local folder: https://github.com/debarghya-mukherjee/Fair_metric_learning/tree/5333cbd35c3d8febcc9fc25a59a453b6c3cc0c88"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# uncomment and install dependencies before continuing\n",
    "#!pip install inFairness seaborn matplotlib"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import torch\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from inFairness.distances import EXPLOREDistance\n",
    "from inFairness.utils import datautils\n",
    "\n",
    "# import data loading script\n",
    "import data\n",
    "import utils"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "DATAPATH = \"./data/\"\n",
    "EMBEDPATH = \"./data/sentiment_glove.42B.300d.txt\"\n",
    "TEST_EMBEDPATH = \"./embeddings/\"\n",
    "\n",
    "# `ALL_TESTS` defines the words X, Y, A, and B used for association tests\n",
    "ALL_TESTS = [\n",
    "    ['flowers', 'insects', 'pleasant', 'unpleasant'],\n",
    "    ['instruments', 'weapons', 'pleasant', 'unpleasant'],\n",
    "    ['mental', 'physical', 'temporal', 'permanent'],\n",
    "    ['white', 'black', 'pleasant', 'unpleasant'],\n",
    "    ['white7', 'black7', 'pleasant', 'unpleasant'],\n",
    "    ['white7', 'black7', 'pleasant9', 'unpleasant9'],\n",
    "    ['male_names', 'female_names', 'career', 'family'],\n",
    "    ['math', 'arts', 'male_terms', 'female_terms'],\n",
    "    ['science10', 'arts10', 'male_terms10', 'female_terms10'],\n",
    "    ['young', 'old', 'pleasant9', 'unpleasant9'],\n",
    "]\n",
    "\n",
    "TEST_IDS = [\n",
    "    'FLvINS-PLvUPL', 'INSvWPN-PLvUPL', 'MENvPHY-TEMPvPERM',\n",
    "    'WHTvBLK-PLvUPL', 'WHT7vBLK7-PLvUPL', 'WHT7vBLK7-PL9vUPL9',\n",
    "    'MALEvFEM-CARvFAM', 'MATHvARTS-MALEvFEM', 'SCvART-MALEvFEM', \n",
    "    'YNGvOLD-PLvUPL'\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "_, X, y, vocab, all_names_embed, names = data.load_data(DATAPATH, EMBEDPATH, names_path=DATAPATH)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create comparable pairs\n",
    "\n",
    "n_pairs_comparable = 50000\n",
    "unique_names_idx = np.unique(names, return_index=True)[1]\n",
    "pairs_idx = datautils.generate_data_pairs(\n",
    "    n_pairs=n_pairs_comparable, datasamples_1=unique_names_idx\n",
    ")\n",
    "\n",
    "comparable_X1 = all_names_embed[unique_names_idx[pairs_idx[:, 0]]]\n",
    "comparable_X2 = all_names_embed[unique_names_idx[pairs_idx[:, 1]]]\n",
    "\n",
    "\n",
    "# Create incomparable pairs\n",
    "n_pairs_incomparable = 50000\n",
    "comparator_fn = lambda x1, x2: (x1 == 1) and (x2 == -1)\n",
    "pairs_idx = datautils.generate_data_pairs(\n",
    "    n_pairs_comparable, y, comparator=comparator_fn\n",
    ")\n",
    "\n",
    "incomparable_X1 = X[pairs_idx[:, 0]]\n",
    "incomparable_X2 = X[pairs_idx[:, 1]]\n",
    "\n",
    "\n",
    "# Join the two sets (comparable and incomparable) to create X and Y\n",
    "X1 = np.vstack((comparable_X1, incomparable_X1))\n",
    "X2 = np.vstack((comparable_X2, incomparable_X2))\n",
    "\n",
    "Y_pairs = np.zeros(n_pairs_comparable + n_pairs_incomparable)\n",
    "Y_pairs[:n_pairs_comparable] = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "X1 = torch.from_numpy(X1)\n",
    "X2 = torch.from_numpy(X2)\n",
    "Y_pairs = torch.from_numpy(Y_pairs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "metric = EXPLOREDistance()\n",
    "metric.fit(X1, X2, Y_pairs, iters=1000, batchsize=10000)\n",
    "sigma = metric.sigma.detach().numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "methods = {\n",
    "    'baseline': [None, None],\n",
    "    'explore': [sigma, None]\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Association test 0: target words are flowers and insects; attribute words are pleasant and unpleasant\n",
      "Association test 1: target words are instruments and weapons; attribute words are pleasant and unpleasant\n",
      "Association test 2: target words are mental and physical; attribute words are temporal and permanent\n",
      "Association test 3: target words are white and black; attribute words are pleasant and unpleasant\n",
      "Association test 4: target words are white7 and black7; attribute words are pleasant and unpleasant\n",
      "Association test 5: target words are white7 and black7; attribute words are pleasant9 and unpleasant9\n",
      "Association test 6: target words are male_names and female_names; attribute words are career and family\n",
      "Association test 7: target words are math and arts; attribute words are male_terms and female_terms\n",
      "Association test 8: target words are science10 and arts10; attribute words are male_terms10 and female_terms10\n",
      "Association test 9: target words are young and old; attribute words are pleasant9 and unpleasant9\n"
     ]
    }
   ],
   "source": [
    "result = []\n",
    "\n",
    "for idx, testset in enumerate(ALL_TESTS):\n",
    "    X_name, Y_name, A_name, B_name = testset\n",
    "    print('Association test %d: target words are %s and %s; attribute words are %s and %s' % (idx, X_name, Y_name, A_name, B_name))\n",
    "\n",
    "    X_embed = np.load(TEST_EMBEDPATH + X_name + '_embed.npy')\n",
    "    Y_embed = np.load(TEST_EMBEDPATH + Y_name + '_embed.npy')\n",
    "    A_embed = np.load(TEST_EMBEDPATH + A_name + '_embed.npy')\n",
    "    B_embed = np.load(TEST_EMBEDPATH + B_name + '_embed.npy')\n",
    "\n",
    "    for method_name, method_params in methods.items():\n",
    "        method_sigma, method_proj = method_params\n",
    "        p_val, effect = utils.run_test(\n",
    "            X_embed, Y_embed, A_embed, B_embed, Sigma=method_sigma, proj=method_proj\n",
    "        )\n",
    "\n",
    "        result.append(\n",
    "            [idx, TEST_IDS[idx], method_name, p_val, effect]\n",
    "        )\n",
    "\n",
    "result = pd.DataFrame(result, columns=['index', 'testname', 'method', 'p_val', 'effect_size'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9]),\n",
       " [Text(0, 0, 'FLvINS-PLvUPL'),\n",
       "  Text(1, 0, 'INSvWPN-PLvUPL'),\n",
       "  Text(2, 0, 'MENvPHY-TEMPvPERM'),\n",
       "  Text(3, 0, 'WHTvBLK-PLvUPL'),\n",
       "  Text(4, 0, 'WHT7vBLK7-PLvUPL'),\n",
       "  Text(5, 0, 'WHT7vBLK7-PL9vUPL9'),\n",
       "  Text(6, 0, 'MALEvFEM-CARvFAM'),\n",
       "  Text(7, 0, 'MATHvARTS-MALEvFEM'),\n",
       "  Text(8, 0, 'SCvART-MALEvFEM'),\n",
       "  Text(9, 0, 'YNGvOLD-PLvUPL')])"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1237.15x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set(font_scale=1.5)\n",
    "barplot = sns.catplot(x='testname', y='effect_size', hue='method', data=result, kind='bar', aspect=3.1)\n",
    "barplot.set_axis_labels(\"\", \"Effect Size\")\n",
    "plt.xticks(fontsize=14, rotation=45)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "1ff634e5a2199fb3fbd063cbcb535063e93d355be71aa4c5535f7c54643d696f"
  },
  "kernelspec": {
   "display_name": "Python 3.8.13 ('infairness_notebook')",
   "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": "3.8.13"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}