result_(a-p)の類似度ヒストグラム用.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# (a-p)間の距離のヒストグラムを確認する"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 必要なパーツ\n",
    "- 各モデルにおける埋め込みベクトル\n",
    "    - PATH/TO/MODEL/gallery.txt\n",
    "    - PATH/TO/MODEL/query.txt\n",
    "- 実際の画像のパス\n",
    "    - ./test_pairs.txt\n",
    "- それぞれはインデックスで整合性をとる   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import os\n",
    "import pickle\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# model = 'DualImpL2300a0.1b0.05'\n",
    "# model = 'Dual300a0.1'\n",
    "model = 'test'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "ALPHA = 0.1\n",
    "BETA = 0.05"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "vec_length = 500\n",
    "model_epoch = 99"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_PATH = './model/T_Shirt/Dual_exp/a{}b{}'.format(ALPHA,BETA)\n",
    "# model_PATH = './model/T_Shirt/tripletloss/a{}'.format(ALPHA)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2032, 500)\n"
     ]
    }
   ],
   "source": [
    "f = open(model_PATH+'/{}/gallery{}.txt'.format(vec_length,model_epoch),'rb')\n",
    "gallery = pickle.load(f)\n",
    "print(gallery.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2032, 500)\n"
     ]
    }
   ],
   "source": [
    "f = open(model_PATH+'/{}/query{}.txt'.format(vec_length,model_epoch),'rb')\n",
    "query = pickle.load(f)\n",
    "print(query.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# f = open('./model/T_Shirt/tripletloss/a0.1/300/gallery300.txt','rb')\n",
    "# gallery = pickle.load(f)\n",
    "# print(gallery.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# f = open('./model/T_Shirt/tripletloss/a0.1/300/query300.txt','rb')\n",
    "# query = pickle.load(f)\n",
    "# print(query.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## ユークリッド距離(の2乗)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def euclid_dist2(emb1, emb2):\n",
    "    return np.sum(np.square(emb1 - emb2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "ave_sim_negative = []\n",
    "for i,q in enumerate(query):\n",
    "    sum_ = 0\n",
    "    for j,g in enumerate(gallery):\n",
    "        if i == j: continue\n",
    "        else: sum_ += euclid_dist2(q,g)\n",
    "    ave_sim_negative.append(sum_/len(gallery))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "sim = []\n",
    "for q,g in zip(query,gallery):\n",
    "    sim.append(euclid_dist2(q,g))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "gap = np.array(ave_sim_negative)-np.array(sim)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plothist(sim1,sim2,model):\n",
    "# bins =　棒の数\n",
    "    #     plt.hist(sim,bins=40)\n",
    "    fig = plt.figure()\n",
    "    ax = fig.add_subplot(1,1,1)\n",
    "    ax.hist(sim1, bins=40,alpha=0.9, label='D(a,p)')\n",
    "    ax.hist(sim2, bins=40,alpha=0.9, label='D(a,n).Ave')\n",
    "    ax.legend(loc='upper right')\n",
    "    plt.grid('True')\n",
    "#     plt.title(\"Query-Negative Average Distance_{}\".format(model))\n",
    "    plt.xlabel(\"euclid dist^2\")\n",
    "    plt.ylabel(\"hist\")\n",
    "    plt.savefig('./histgram/D(p,q)vsAve(D(p,n))_{}.png'.format(model),dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/pip/koki_ishizuka/.conda/envs/py35-zukapy/lib/python3.5/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plothist(sim,ave_sim_negative,model=model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/pip/koki_ishizuka/.conda/envs/py35-zukapy/lib/python3.5/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'hist')"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(gap,bins=40)\n",
    "plt.grid('True')\n",
    "plt.title(\"Query-Positive Distance_{}\".format(model))\n",
    "plt.xlabel(\"euclid dist^2\")\n",
    "plt.ylabel(\"hist\")\n",
    "# plt.savefig('./histgram/avesimNegative_euclid_{}.png'.format(model),dpi=300)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## cos類似度"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cos_sim(v1, v2):\n",
    "    return np.dot(v1, v2) / (np.linalg.norm(v1) * np.linalg.norm(v2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "sim = []\n",
    "for q,g in zip(query,gallery):\n",
    "    sim.append(cos_sim(q,g))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([  9.,  18.,  44.,  44.,  40.,  37.,  37.,  56.,  59.,  67.,  57.,\n",
       "         68.,  78.,  61.,  71.,  85.,  93.,  84.,  82., 101.,  97., 110.,\n",
       "         98.,  83.,  91., 136., 116.,  87.,  18.,   5.]),\n",
       " array([-0.60164762, -0.57375712, -0.54586663, -0.51797613, -0.49008563,\n",
       "        -0.46219513, -0.43430463, -0.40641413, -0.37852363, -0.35063313,\n",
       "        -0.32274263, -0.29485214, -0.26696164, -0.23907114, -0.21118064,\n",
       "        -0.18329014, -0.15539964, -0.12750914, -0.09961864, -0.07172815,\n",
       "        -0.04383765, -0.01594715,  0.01194335,  0.03983385,  0.06772435,\n",
       "         0.09561485,  0.12350535,  0.15139585,  0.17928634,  0.20717684,\n",
       "         0.23506734]),\n",
       " <a list of 30 Patch objects>)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(sim,bins=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# print(sim)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "py35-zukapy",
   "language": "python",
   "name": "py35-zukapy"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}