T-SNE-Visualization-train_word_embeeding_vs_general-text.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "##Visualizing Word Embeddings using t-SNE (train wordembeeding on the COCO-text vs general word embedding)\n",
    "## original code from https://github.com/sismetanin/word2vec-tsne"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "import gensim\n",
    "from gensim.models import KeyedVectors\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.cm as cm\n",
    "#% matplotlib inline\n",
    "%matplotlib inline\n",
    "from sklearn.manifold import TSNE\n",
    "import numpy as np\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "OUTPUT_FILE_PATH = 'w2v_model_w2.bin'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = KeyedVectors.load(OUTPUT_FILE_PATH)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "keys = ['st','stop','kia','tennis','airways','12', 'car','parking', 'shop']\n",
    "\n",
    "embedding_clusters = []\n",
    "word_clusters = []\n",
    "for word in keys:\n",
    "    embeddings = []\n",
    "    words = []\n",
    "    #for similar_word, _ in model_gn.most_similar(word, topn=30):\n",
    "    for similar_word, _ in model.most_similar(word, topn=5):\n",
    "        words.append(similar_word)\n",
    "        embeddings.append(model[similar_word])\n",
    "    embedding_clusters.append(embeddings)\n",
    "    word_clusters.append(words)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "embedding_clusters = np.array(embedding_clusters)\n",
    "n, m, k = embedding_clusters.shape\n",
    "tsne_model_en_2d = TSNE(perplexity=13, n_components=2, init='pca', n_iter=3500, random_state=32)\n",
    "embeddings_en_2d = np.array(tsne_model_en_2d.fit_transform(embedding_clusters.reshape(n * m, k))).reshape(n, m, 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def tsne_plot_similar_words(title, labels, embedding_clusters, word_clusters, a, filename=None):\n",
    "    plt.figure(figsize=(16, 9))\n",
    "    colors = cm.rainbow(np.linspace(0, 1, len(labels)))\n",
    "    for label, embeddings, words, color in zip(labels, embedding_clusters, word_clusters, colors):\n",
    "        x = embeddings[:, 0]\n",
    "        y = embeddings[:, 1]\n",
    "        plt.scatter(x, y, c=color, alpha=a, label=label)\n",
    "        for i, word in enumerate(words):\n",
    "            plt.annotate(word, alpha=0.5, xy=(x[i], y[i]), xytext=(5, 2),\n",
    "                         textcoords='offset points', ha='left', va='bottom', size=14)\n",
    "    plt.legend(loc=4)\n",
    "    plt.title(title)\n",
    "    plt.grid(True)\n",
    "    #plt.grid(False)\n",
    "    \n",
    "    if filename:\n",
    "        # try to make it better in term of quality. \n",
    "        plt.savefig(filename, format='pdf', dpi=1000, bbox_inches='tight')\n",
    "        # original size\n",
    "        #plt.savefig(filename, format='pdf', dpi=150, bbox_inches='tight')\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "tsne_plot_similar_words('', keys, embeddings_en_2d, word_clusters, 0.7,\n",
    "                        'similar_words.pdf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "# General word embedding "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "from gensim.models import KeyedVectors\n",
    "from gensim.scripts.glove2word2vec import glove2word2vec\n",
    "from gensim.models.keyedvectors import KeyedVectors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "OUTPUT_FILE_PATH_2 = 'glove.840B.300d.txt'\n",
    "#from gensim.keyedvectors import KeyedVectors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2196017, 300)"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from gensim.scripts.glove2word2vec import glove2word2vec\n",
    "glove2word2vec(glove_input_file=OUTPUT_FILE_PATH_2, word2vec_output_file=\"gensim_glove_vectors.txt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "glove_model = KeyedVectors.load_word2vec_format(\"gensim_glove_vectors.txt\", binary=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "keys = ['st','stop','kia','tennis','airways','12', 'car','parking', 'shop']\n",
    "\n",
    "embedding_clusters = []\n",
    "word_clusters = []\n",
    "for word in keys:\n",
    "    embeddings = []\n",
    "    words = []\n",
    "    #for similar_word, _ in model_gn.most_similar(word, topn=30):\n",
    "    for similar_word, _ in glove_model.most_similar(word, topn=5):\n",
    "        words.append(similar_word)\n",
    "        embeddings.append(glove_model[similar_word])\n",
    "    embedding_clusters.append(embeddings)\n",
    "    word_clusters.append(words)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "embedding_clusters = np.array(embedding_clusters)\n",
    "n, m, k = embedding_clusters.shape\n",
    "tsne_model_en_2d = TSNE(perplexity=13, n_components=2, init='pca', n_iter=3500, random_state=32)\n",
    "#tsne_model_en_2d = TSNE(perplexity=7, n_components=2, init='pca', n_iter=3500, random_state=7)\n",
    "embeddings_en_2d = np.array(tsne_model_en_2d.fit_transform(embedding_clusters.reshape(n * m, k))).reshape(n, m, 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n",
      "*c* argument looks like a single numeric RGB or RGBA sequence, which should be avoided as value-mapping will have precedence in case its length matches with *x* & *y*.  Please use the *color* keyword-argument or provide a 2-D array with a single row if you intend to specify the same RGB or RGBA value for all points.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def tsne_plot_similar_words(title, labels, embedding_clusters, word_clusters, a, filename=None):\n",
    "    plt.figure(figsize=(16, 9))\n",
    "    colors = cm.rainbow(np.linspace(0, 1, len(labels)))\n",
    "    for label, embeddings, words, color in zip(labels, embedding_clusters, word_clusters, colors):\n",
    "        x = embeddings[:, 0]\n",
    "        y = embeddings[:, 1]\n",
    "        plt.scatter(x, y, c=color, alpha=a, label=label)\n",
    "        for i, word in enumerate(words):\n",
    "            plt.annotate(word, alpha=0.5, xy=(x[i], y[i]), xytext=(5, 2),\n",
    "                         textcoords='offset points', ha='left', va='bottom', size=14)\n",
    "    plt.legend(loc=4)\n",
    "    plt.title(title)\n",
    "    plt.grid(True)\n",
    "    #plt.grid(False)\n",
    "    \n",
    "    if filename:\n",
    "        # try to make it better in term of quality. \n",
    "        plt.savefig(filename, format='pdf', dpi=1000, bbox_inches='tight')\n",
    "        # original size\n",
    "        #plt.savefig(filename, format='pdf', dpi=150, bbox_inches='tight')\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "tsne_plot_similar_words('', keys, embeddings_en_2d, word_clusters, 0.7,\n",
    "                        'similar_words.pdf')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}