barplots.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {},
   "outputs": [],
   "source": [
    "dfmini = pd.DataFrame([['MAML code','5-way 1-shot',47.6],\n",
    "                   ['MAML code','5-way 5-shot',62.3],\n",
    "                   ['MAML code','2-way 1-shot',78.8],\n",
    "                   ['MAML code','2-way 5-shot',82.6],\n",
    "                   ['MAML paper','5-way 1-shot',48.7],\n",
    "                   ['MAML paper','5-way 5-shot',63.1],\n",
    "                   ['MAML paper','2-way 1-shot',74.9],\n",
    "                   ['MAML paper','2-way 5-shot',84.4],\n",
    "                   ['our R2D2*','5-way 1-shot',45.7],\n",
    "                   ['our R2D2*','5-way 5-shot',63.7],\n",
    "                   ['our R2D2*','2-way 1-shot',74.7],\n",
    "                   ['our R2D2*','2-way 5-shot',83.0],\n",
    "                   ['our R2D2','5-way 1-shot',51.7],\n",
    "                   ['our R2D2','5-way 5-shot',63.3],\n",
    "                   ['our R2D2','2-way 1-shot',74.6],\n",
    "                   ['our R2D2','2-way 5-shot',86.8],\n",
    "                   ['R2D2 paper','5-way 1-shot',51.5],\n",
    "                   ['R2D2 paper','5-way 5-shot',68.8],\n",
    "                   ['R2D2 paper','2-way 1-shot',76.7],\n",
    "                   ['R2D2 paper','2-way 5-shot',86.8]],columns=['code','wayshot','val'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>code</th>\n",
       "      <th>MAML code</th>\n",
       "      <th>MAML paper</th>\n",
       "      <th>our R2D2*</th>\n",
       "      <th>our R2D2</th>\n",
       "      <th>R2D2 paper</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>wayshot</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5-way 1-shot</th>\n",
       "      <td>47.6</td>\n",
       "      <td>48.7</td>\n",
       "      <td>45.7</td>\n",
       "      <td>51.7</td>\n",
       "      <td>51.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5-way 5-shot</th>\n",
       "      <td>62.3</td>\n",
       "      <td>63.1</td>\n",
       "      <td>63.7</td>\n",
       "      <td>63.3</td>\n",
       "      <td>68.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2-way 1-shot</th>\n",
       "      <td>78.8</td>\n",
       "      <td>74.9</td>\n",
       "      <td>74.7</td>\n",
       "      <td>74.6</td>\n",
       "      <td>76.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2-way 5-shot</th>\n",
       "      <td>82.6</td>\n",
       "      <td>84.4</td>\n",
       "      <td>83.0</td>\n",
       "      <td>86.8</td>\n",
       "      <td>86.8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "code          MAML code  MAML paper  our R2D2*  our R2D2  R2D2 paper\n",
       "wayshot                                                             \n",
       "5-way 1-shot       47.6        48.7       45.7      51.7        51.5\n",
       "5-way 5-shot       62.3        63.1       63.7      63.3        68.8\n",
       "2-way 1-shot       78.8        74.9       74.7      74.6        76.7\n",
       "2-way 5-shot       82.6        84.4       83.0      86.8        86.8"
      ]
     },
     "execution_count": 197,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfp = dfmini.pivot(\"wayshot\", \"code\", \"val\")\n",
    "dfplot = dfp[['MAML code', 'MAML paper', 'our R2D2*','our R2D2','R2D2 paper']]\n",
    "dfplotmini = dfplot.reindex(['5-way 1-shot', '5-way 5-shot', '2-way 1-shot','2-way 5-shot'])\n",
    "dfplotmini"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plotz = dfplotmini.plot(kind='bar', rot=0,ylim=(40,100), grid=True, width=0.7, figsize=(15,4), color=tuple([\"palegreen\", \"green\",\"mistyrose\",\"coral\",\"red\"]))\n",
    "plotz.set_xlabel(\"miniImageNet few-shot results\")\n",
    "plotz.set_ylabel(\"accuracy (%)\")\n",
    "for p in plotz.patches:\n",
    "    plotz.annotate(str(p.get_height()), (p.get_x() * 1.001, p.get_height() * 1.01))\n",
    "\n",
    "plt.savefig('plots/miniImageNet.pdf', bbox_inches = 'tight', pad_inches = 0)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {},
   "outputs": [],
   "source": [
    "dfcifar = pd.DataFrame([['MAML code','5-way 1-shot',56.8],\n",
    "                   ['MAML code','5-way 5-shot',70.8],\n",
    "                   ['MAML code','2-way 1-shot',83.1],\n",
    "                   ['MAML code','2-way 5-shot',88.5],\n",
    "                   ['MAML paper','5-way 1-shot',58.9],\n",
    "                   ['MAML paper','5-way 5-shot',71.5],\n",
    "                   ['MAML paper','2-way 1-shot',82.8],\n",
    "                   ['MAML paper','2-way 5-shot',88.3],\n",
    "                   ['our R2D2*','5-way 1-shot',54.3],\n",
    "                   ['our R2D2*','5-way 5-shot',69.7],\n",
    "                   ['our R2D2*','2-way 1-shot',78.3],\n",
    "                   ['our R2D2*','2-way 5-shot',87.7],\n",
    "                   ['our R2D2','5-way 1-shot',60.2],\n",
    "                   ['our R2D2','5-way 5-shot',70.9],\n",
    "                   ['our R2D2','2-way 1-shot',83.6],\n",
    "                   ['our R2D2','2-way 5-shot',89.0],\n",
    "                   ['R2D2 paper','5-way 1-shot',65.3],\n",
    "                   ['R2D2 paper','5-way 5-shot',79.4],\n",
    "                   ['R2D2 paper','2-way 1-shot',83.4],\n",
    "                   ['R2D2 paper','2-way 5-shot',91.1]],columns=['code','wayshot','val'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>code</th>\n",
       "      <th>MAML code</th>\n",
       "      <th>MAML paper</th>\n",
       "      <th>our R2D2*</th>\n",
       "      <th>our R2D2</th>\n",
       "      <th>R2D2 paper</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>wayshot</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5-way 1-shot</th>\n",
       "      <td>56.8</td>\n",
       "      <td>58.9</td>\n",
       "      <td>54.3</td>\n",
       "      <td>60.2</td>\n",
       "      <td>65.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5-way 5-shot</th>\n",
       "      <td>70.8</td>\n",
       "      <td>71.5</td>\n",
       "      <td>69.7</td>\n",
       "      <td>70.9</td>\n",
       "      <td>79.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2-way 1-shot</th>\n",
       "      <td>83.1</td>\n",
       "      <td>82.8</td>\n",
       "      <td>78.3</td>\n",
       "      <td>83.6</td>\n",
       "      <td>83.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2-way 5-shot</th>\n",
       "      <td>88.5</td>\n",
       "      <td>88.3</td>\n",
       "      <td>87.7</td>\n",
       "      <td>89.0</td>\n",
       "      <td>91.1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "code          MAML code  MAML paper  our R2D2*  our R2D2  R2D2 paper\n",
       "wayshot                                                             \n",
       "5-way 1-shot       56.8        58.9       54.3      60.2        65.3\n",
       "5-way 5-shot       70.8        71.5       69.7      70.9        79.4\n",
       "2-way 1-shot       83.1        82.8       78.3      83.6        83.4\n",
       "2-way 5-shot       88.5        88.3       87.7      89.0        91.1"
      ]
     },
     "execution_count": 200,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfp = dfcifar.pivot(\"wayshot\", \"code\", \"val\")\n",
    "dfplot = dfp[['MAML code', 'MAML paper', 'our R2D2*','our R2D2','R2D2 paper']]\n",
    "dfplotcifar = dfplot.reindex(['5-way 1-shot', '5-way 5-shot', '2-way 1-shot','2-way 5-shot'])\n",
    "dfplotcifar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plotz = dfplotcifar.plot(kind='bar', rot=0,ylim=(40,100), grid=True, width=0.7, figsize=(15,4), color=tuple([\"palegreen\", \"green\",\"mistyrose\",\"coral\",\"red\"]))\n",
    "plotz.set_xlabel(\"CIFAR-FS few-shot results\")\n",
    "plotz.set_ylabel(\"accuracy (%)\")\n",
    "\n",
    "for p in plotz.patches:\n",
    "    plotz.annotate(str(p.get_height()), (p.get_x() * 1.001, p.get_height() * 1.01))\n",
    "\n",
    "plt.savefig('plots/CIFAR-FS.pdf', bbox_inches = 'tight', pad_inches = 0)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}