vis_algorithms.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "from os.path import join as joinpath\n",
    "from glob import glob\n",
    "\n",
    "import torch\n",
    "import pandas as pd\n",
    "\n",
    "from IPython import display\n",
    "import matplotlib\n",
    "import matplotlib.cm as cm\n",
    "import matplotlib.pyplot as plt\n",
    "matplotlib.rcParams['font.family'] = 'Times New Roman'\n",
    "matplotlib.rcParams['font.size'] = 13"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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></th>\n",
       "      <th>Algorithm</th>\n",
       "      <th>LR</th>\n",
       "      <th>FR mean</th>\n",
       "      <th>FR std</th>\n",
       "      <th># exp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>spsa_sgd</td>\n",
       "      <td>1.0E-1</td>\n",
       "      <td>10.470000</td>\n",
       "      <td>0.697264</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>spsa_sgd</td>\n",
       "      <td>1.0E-2</td>\n",
       "      <td>17.688999</td>\n",
       "      <td>3.977468</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>spsa_sgd</td>\n",
       "      <td>1.0E-3</td>\n",
       "      <td>33.221001</td>\n",
       "      <td>6.536767</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>spsa_sgd</td>\n",
       "      <td>1.0E-4</td>\n",
       "      <td>1.674000</td>\n",
       "      <td>0.119648</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>spsa_nag</td>\n",
       "      <td>1.0E-1</td>\n",
       "      <td>25.290001</td>\n",
       "      <td>4.608335</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>spsa_nag</td>\n",
       "      <td>1.0E-2</td>\n",
       "      <td>27.375999</td>\n",
       "      <td>3.643899</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>spsa_nag</td>\n",
       "      <td>1.0E-3</td>\n",
       "      <td>29.534000</td>\n",
       "      <td>3.752914</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>spsa_nag</td>\n",
       "      <td>1.0E-4</td>\n",
       "      <td>37.139000</td>\n",
       "      <td>2.973369</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>spsa_nag</td>\n",
       "      <td>1.0E-5</td>\n",
       "      <td>44.630001</td>\n",
       "      <td>4.716783</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>spsa_nag</td>\n",
       "      <td>1.0E-6</td>\n",
       "      <td>25.365999</td>\n",
       "      <td>4.318454</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>spsa_adam</td>\n",
       "      <td>1.0E-2</td>\n",
       "      <td>11.151000</td>\n",
       "      <td>2.737149</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>spsa_adam</td>\n",
       "      <td>1.0E-3</td>\n",
       "      <td>39.195000</td>\n",
       "      <td>3.247208</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>spsa_adam</td>\n",
       "      <td>1.0E-4</td>\n",
       "      <td>4.087000</td>\n",
       "      <td>0.751222</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>spsa_adam</td>\n",
       "      <td>1.0E-5</td>\n",
       "      <td>0.260000</td>\n",
       "      <td>0.036818</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Algorithm      LR    FR mean    FR std  # exp\n",
       "0    spsa_sgd  1.0E-1  10.470000  0.697264     10\n",
       "1    spsa_sgd  1.0E-2  17.688999  3.977468     10\n",
       "2    spsa_sgd  1.0E-3  33.221001  6.536767     10\n",
       "3    spsa_sgd  1.0E-4   1.674000  0.119648     10\n",
       "4    spsa_nag  1.0E-1  25.290001  4.608335     10\n",
       "5    spsa_nag  1.0E-2  27.375999  3.643899     10\n",
       "6    spsa_nag  1.0E-3  29.534000  3.752914     10\n",
       "7    spsa_nag  1.0E-4  37.139000  2.973369     10\n",
       "8    spsa_nag  1.0E-5  44.630001  4.716783     10\n",
       "9    spsa_nag  1.0E-6  25.365999  4.318454     10\n",
       "10  spsa_adam  1.0E-2  11.151000  2.737149     10\n",
       "11  spsa_adam  1.0E-3  39.195000  3.247208     10\n",
       "12  spsa_adam  1.0E-4   4.087000  0.751222     10\n",
       "13  spsa_adam  1.0E-5   0.260000  0.036818     10"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pattern_tags = [\n",
    "    ('./log/cifar10_vit_1.0/alg/find_lr/spsa_sgd_*_*/ckpt.pth', 'spsa_sgd'),\n",
    "    ('./log/cifar10_vit_1.0/alg/find_lr/spsa_nag_*_*/ckpt.pth', 'spsa_nag'),\n",
    "    ('./log/cifar10_vit_1.0/alg/find_lr/spsa_adam_*_*/ckpt.pth', 'spsa_adam'),\n",
    "]\n",
    "\n",
    "tag_power_fr = {}\n",
    "for pattern, tag in pattern_tags:\n",
    "    paths = sorted(glob(pattern))\n",
    "    for path in paths:\n",
    "        power_idx = path.find(tag) + len(tag) + 1\n",
    "        power = path[power_idx:power_idx+2]\n",
    "        with open('temp.txt', 'ta') as file:\n",
    "            print(tag, power, path, file=file)\n",
    "        ckpts = torch.load(path)\n",
    "        fr = ckpts['fr']\n",
    "        \n",
    "        key = (tag, power)\n",
    "        if key in tag_power_fr:\n",
    "            tag_power_fr[key].append(fr)\n",
    "        else:\n",
    "            tag_power_fr[key] = [fr]\n",
    "\n",
    "table = []\n",
    "for key in tag_power_fr.keys():\n",
    "    fr_list = torch.tensor(tag_power_fr[key])\n",
    "    std, mean = torch.std_mean(fr_list)\n",
    "    count = len(fr_list)\n",
    "    \n",
    "    tag = key[0]\n",
    "    lr = f'1.0E{key[1]}'\n",
    "    \n",
    "    table.append((tag, lr, mean.item(), std.item(), count))\n",
    "    \n",
    "df = pd.DataFrame(table, columns=('Algorithm', 'LR', 'FR mean', 'FR std', '# exp'))\n",
    "display.display(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def bar_from_table(ax, mean, std, columns=None, legend=None, \n",
    "                   column_width=0.9, bar_padding=0.1, cmap='tab10'):\n",
    "    num_bars, num_columns = mean.shape\n",
    "    bar_width = column_width / num_bars\n",
    "    \n",
    "    right = 0.5 * (column_width - bar_width)\n",
    "    left = -right\n",
    "    \n",
    "    bar_width *= (1 - bar_padding)\n",
    "    centers = torch.arange(num_columns).float()\n",
    "    offsets = torch.linspace(left, right, num_bars)\n",
    "    \n",
    "    cmap = cm.get_cmap(cmap)\n",
    "    for i, (row_mean, row_std) in enumerate(zip(mean, std)):\n",
    "        ax.bar(centers + offsets[i], row_mean, yerr=row_std, \n",
    "               width=bar_width, capsize=20 * bar_width, color=cmap(i))\n",
    "    ax.set_ylabel('ASR (%)')\n",
    "        \n",
    "    if columns:\n",
    "        ax.set_xticks(centers, columns)\n",
    "    else:\n",
    "        ax.set_xticks(centers, (centers.int() + 1).tolist())\n",
    "        \n",
    "    if legend:\n",
    "        ax.legend(legend)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ResNet18]\n"
     ]
    },
    {
     "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></th>\n",
       "      <th>Algorithm</th>\n",
       "      <th>ASR (%) mean</th>\n",
       "      <th>ASR (%) std</th>\n",
       "      <th>l2-norm</th>\n",
       "      <th># exp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RGF</td>\n",
       "      <td>81.539001</td>\n",
       "      <td>0.684389</td>\n",
       "      <td>382.666443</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>RGF+Adam</td>\n",
       "      <td>80.855499</td>\n",
       "      <td>0.664004</td>\n",
       "      <td>374.750549</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>SPSA</td>\n",
       "      <td>87.401497</td>\n",
       "      <td>0.222764</td>\n",
       "      <td>433.427155</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SPSA-GC</td>\n",
       "      <td>83.434502</td>\n",
       "      <td>0.375030</td>\n",
       "      <td>437.224121</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>SPSA-AM (ours)</td>\n",
       "      <td>85.961998</td>\n",
       "      <td>0.180192</td>\n",
       "      <td>408.272522</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Algorithm  ASR (%) mean  ASR (%) std     l2-norm  # exp\n",
       "0             RGF     81.539001     0.684389  382.666443     20\n",
       "1        RGF+Adam     80.855499     0.664004  374.750549     20\n",
       "2            SPSA     87.401497     0.222764  433.427155     20\n",
       "3         SPSA-GC     83.434502     0.375030  437.224121     20\n",
       "4  SPSA-AM (ours)     85.961998     0.180192  408.272522     20"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ViT]\n"
     ]
    },
    {
     "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></th>\n",
       "      <th>Algorithm</th>\n",
       "      <th>ASR (%) mean</th>\n",
       "      <th>ASR (%) std</th>\n",
       "      <th>l2-norm</th>\n",
       "      <th># exp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RGF</td>\n",
       "      <td>44.865002</td>\n",
       "      <td>2.178176</td>\n",
       "      <td>369.928528</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>RGF+Adam</td>\n",
       "      <td>40.737000</td>\n",
       "      <td>0.986476</td>\n",
       "      <td>355.989410</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>SPSA</td>\n",
       "      <td>58.985001</td>\n",
       "      <td>5.414708</td>\n",
       "      <td>420.890564</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SPSA-GC</td>\n",
       "      <td>50.075001</td>\n",
       "      <td>0.483145</td>\n",
       "      <td>434.950531</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>SPSA-AM (ours)</td>\n",
       "      <td>57.678001</td>\n",
       "      <td>0.487535</td>\n",
       "      <td>405.038086</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Algorithm  ASR (%) mean  ASR (%) std     l2-norm  # exp\n",
       "0             RGF     44.865002     2.178176  369.928528     10\n",
       "1        RGF+Adam     40.737000     0.986476  355.989410     10\n",
       "2            SPSA     58.985001     5.414708  420.890564     10\n",
       "3         SPSA-GC     50.075001     0.483145  434.950531     10\n",
       "4  SPSA-AM (ours)     57.678001     0.487535  405.038086     20"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x350 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "tags = ['RGF_SGD',  'RGF_Adam', 'SPSA_SGD', 'SPSA_NAG', 'SPSA_Adam']\n",
    "algs = ['RGF',      'RGF+Adam', 'SPSA',     'SPSA-GC',  'SPSA-AM (ours)']\n",
    "\n",
    "def make_dataframe(arch: str):\n",
    "    root = f'./log/cifar10_{arch.lower()}_1.0/alg'\n",
    "    table = []\n",
    "    std_list, mean_list = [], []\n",
    "    for tag, alg in zip(tags, algs):\n",
    "        ckpt_filenames = sorted(glob(joinpath(root, tag.lower(), '*', 'ckpt.pth')))\n",
    "        ckpts = [torch.load(fn) for fn in ckpt_filenames]\n",
    "        \n",
    "        fr_list = []\n",
    "        for ckpt in ckpts:\n",
    "            short = 800 - len(ckpt['fr_list'])\n",
    "            fr_list.append(ckpt['fr_list'] + ([ckpt['fr_list'][-1]] * short))\n",
    "        fr_list = torch.tensor(fr_list)\n",
    "        std, mean = torch.std_mean(fr_list, dim=0)\n",
    "        \n",
    "        uaps = torch.cat([ckpt['uap'] for ckpt in ckpts], dim=0)\n",
    "        norm = torch.norm(uaps * 255.0, p=2, dim=(1, 2, 3))\n",
    "        norm = torch.mean(norm)\n",
    "\n",
    "        std_list.append(std)\n",
    "        mean_list.append(mean)\n",
    "        table.append((alg, mean[-1].item(), std[-1].item(), norm.item(), len(ckpts)))\n",
    "        df = pd.DataFrame(table, columns=('Algorithm', 'ASR (%) mean', 'ASR (%) std', 'l2-norm', '# exp'))\n",
    "    return table, df, std_list, mean_list\n",
    "\n",
    "\n",
    "arch_list = ['ResNet18', 'ViT']\n",
    "arch_plot_idx = 1\n",
    "std = 1.0\n",
    "title_pos = -0.35\n",
    "\n",
    "# table, df, std_list, mean_list\n",
    "info = [make_dataframe(arch) for arch in arch_list]\n",
    "\n",
    "#region display\n",
    "for arch, (_, df, _, _) in zip(arch_list, info):\n",
    "    print(f'[{arch}]')\n",
    "    display.display(df)\n",
    "\n",
    "algs_x = range(len(table))\n",
    "epochs_x = range(len(info[0][3][0]))\n",
    "plt.figure(figsize=(5, 3.5))\n",
    "\n",
    "# Draw bar chart\n",
    "# ax = plt.subplot(121)\n",
    "# if title_pos != None:\n",
    "#     plt.title('(a) ASR over Optimization Algorithms', y=title_pos)\n",
    "# mean_table, std_table = [], []\n",
    "# for i in info:\n",
    "#     for row in i[0]:\n",
    "#         mean_table.append(row[1])\n",
    "#         std_table.append(row[2])\n",
    "# mean_table = torch.tensor(mean_table).view(len(arch_list), -1)\n",
    "# std_table = torch.tensor(std_table).view(len(arch_list), -1)\n",
    "\n",
    "# plot_kwargs = dict(\n",
    "#     column_width=0.9, \n",
    "#     bar_padding=0.1,\n",
    "#     cmap='Pastel1',\n",
    "# )\n",
    "# bar_from_table(ax=ax, mean=mean_table, std=std_table, columns=algs, legend=arch_list, **plot_kwargs)\n",
    "# ax.set_xlabel('Algorithms')\n",
    "\n",
    "# Draw epoch plot\n",
    "_, _, std_list, mean_list = info[arch_plot_idx]\n",
    "ax = plt.subplot(111)\n",
    "# if title_pos != None:\n",
    "#     plt.title(f'(b) ASR against {arch_list[arch_plot_idx]} over Optimization Algorithms', y=title_pos)\n",
    "plots = []\n",
    "for best_means, best_stds in zip(mean_list, std_list):\n",
    "    plots.append(ax.plot(epochs_x, best_means))\n",
    "for best_means, best_stds in zip(mean_list, std_list):\n",
    "    best_stds *= std\n",
    "    ax.fill_between(epochs_x, best_means - best_stds, best_means + best_stds, \n",
    "                    alpha=0.3, edgecolor='face', linewidth=1, linestyle='-', antialiased=True)\n",
    "ax.legend(algs)\n",
    "ax.set_xlabel('Epochs')\n",
    "ax.set_ylabel('ASR (%)')\n",
    "\n",
    "matplotlib.rcParams['font.size'] = 12\n",
    "plt.tight_layout()\n",
    "plt.savefig('temp/algs.pdf')\n",
    "plt.show()\n",
    "#endregion"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "torch-env",
   "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": "3.9.15"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "f1ae0c47510530b88e9c8438f722b7e2a23916e3745af6fc20fb70ed8c8a0aa2"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}