experiments/manifold/stats.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sb\n",
    "import pandas as pd\n",
    "import wandb\n",
    "\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fetch data from W&B"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 319/319 [02:29<00:00,  2.14it/s]\n"
     ]
    }
   ],
   "source": [
    "api = wandb.Api()\n",
    "\n",
    "runs = api.runs('francois-rozet/priors-manifold')\n",
    "\n",
    "data = []\n",
    "\n",
    "for run in tqdm(runs):\n",
    "    if 'finished' not in run.state:\n",
    "        continue\n",
    "\n",
    "    df = run.history()\n",
    "    df = df[['sigma', 'divergence']]\n",
    "    df = df[1:]\n",
    "    df['name'] = run.name\n",
    "\n",
    "    for key, value in run.config.items():\n",
    "        df[key] = value\n",
    "\n",
    "    data.append(df)\n",
    "\n",
    "df = pd.concat(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>sigma</th>\n",
       "      <th>divergence</th>\n",
       "      <th>name</th>\n",
       "      <th>m</th>\n",
       "      <th>n</th>\n",
       "      <th>p</th>\n",
       "      <th>seed</th>\n",
       "      <th>alpha</th>\n",
       "      <th>noise</th>\n",
       "      <th>method</th>\n",
       "      <th>samples</th>\n",
       "      <th>thickness</th>\n",
       "      <th>components</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.01</td>\n",
       "      <td>0.000001</td>\n",
       "      <td>playful-armadillo-969</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>63</td>\n",
       "      <td>3</td>\n",
       "      <td>0.01</td>\n",
       "      <td>hessian</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.02</td>\n",
       "      <td>0.000001</td>\n",
       "      <td>playful-armadillo-969</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>63</td>\n",
       "      <td>3</td>\n",
       "      <td>0.01</td>\n",
       "      <td>hessian</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.05</td>\n",
       "      <td>0.007084</td>\n",
       "      <td>playful-armadillo-969</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>63</td>\n",
       "      <td>3</td>\n",
       "      <td>0.01</td>\n",
       "      <td>hessian</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.10</td>\n",
       "      <td>0.058440</td>\n",
       "      <td>playful-armadillo-969</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>63</td>\n",
       "      <td>3</td>\n",
       "      <td>0.01</td>\n",
       "      <td>hessian</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.20</td>\n",
       "      <td>0.180728</td>\n",
       "      <td>playful-armadillo-969</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>63</td>\n",
       "      <td>3</td>\n",
       "      <td>0.01</td>\n",
       "      <td>hessian</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.50</td>\n",
       "      <td>0.237771</td>\n",
       "      <td>lyric-dream-708</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0.01</td>\n",
       "      <td>zero</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.00</td>\n",
       "      <td>2.344846</td>\n",
       "      <td>lyric-dream-708</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0.01</td>\n",
       "      <td>zero</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2.00</td>\n",
       "      <td>40.463051</td>\n",
       "      <td>lyric-dream-708</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0.01</td>\n",
       "      <td>zero</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>5.00</td>\n",
       "      <td>673.022711</td>\n",
       "      <td>lyric-dream-708</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0.01</td>\n",
       "      <td>zero</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10.00</td>\n",
       "      <td>682.878042</td>\n",
       "      <td>lyric-dream-708</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0.01</td>\n",
       "      <td>zero</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>256</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2550 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    sigma  divergence                   name  m  n    p  seed  alpha  noise  \\\n",
       "1    0.01    0.000001  playful-armadillo-969  1  3  1.0    63      3   0.01   \n",
       "2    0.02    0.000001  playful-armadillo-969  1  3  1.0    63      3   0.01   \n",
       "3    0.05    0.007084  playful-armadillo-969  1  3  1.0    63      3   0.01   \n",
       "4    0.10    0.058440  playful-armadillo-969  1  3  1.0    63      3   0.01   \n",
       "5    0.20    0.180728  playful-armadillo-969  1  3  1.0    63      3   0.01   \n",
       "..    ...         ...                    ... .. ..  ...   ...    ...    ...   \n",
       "6    0.50    0.237771        lyric-dream-708  1  3  1.0     0      3   0.01   \n",
       "7    1.00    2.344846        lyric-dream-708  1  3  1.0     0      3   0.01   \n",
       "8    2.00   40.463051        lyric-dream-708  1  3  1.0     0      3   0.01   \n",
       "9    5.00  673.022711        lyric-dream-708  1  3  1.0     0      3   0.01   \n",
       "10  10.00  682.878042        lyric-dream-708  1  3  1.0     0      3   0.01   \n",
       "\n",
       "     method    samples  thickness  components  \n",
       "1   hessian  1048576.0       0.01         256  \n",
       "2   hessian  1048576.0       0.01         256  \n",
       "3   hessian  1048576.0       0.01         256  \n",
       "4   hessian  1048576.0       0.01         256  \n",
       "5   hessian  1048576.0       0.01         256  \n",
       "..      ...        ...        ...         ...  \n",
       "6      zero  1048576.0       0.01         256  \n",
       "7      zero  1048576.0       0.01         256  \n",
       "8      zero  1048576.0       0.01         256  \n",
       "9      zero  1048576.0       0.01         256  \n",
       "10     zero  1048576.0       0.01         256  \n",
       "\n",
       "[2550 rows x 13 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = df[(df.m == 1) & (df.n == 3) & (df.components == 256)]\n",
    "data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "rc = {\n",
    "    'axes.axisbelow': True,\n",
    "    'axes.linewidth': 1.0,\n",
    "    'figure.dpi': 150,\n",
    "    'figure.figsize': (3.2, 3.2),\n",
    "    'font.size': 12.0,\n",
    "    'legend.fontsize': 'xx-small',\n",
    "    'lines.linewidth': 1.0,\n",
    "    'savefig.transparent': True,\n",
    "    'xtick.labelsize': 'x-small',\n",
    "    'xtick.major.width': 1.0,\n",
    "    'ytick.labelsize': 'x-small',\n",
    "    'ytick.major.width': 1.0,\n",
    "    'font.family': ['serif'],\n",
    "    'text.usetex': True,\n",
    "    'text.latex.preamble': r'\\usepackage{amsfonts}',\n",
    "}\n",
    "\n",
    "plt.rcParams.update(rc)\n",
    "sb.set_palette(sb.color_palette('deep'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 480x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = sb.lineplot(\n",
    "    data,\n",
    "    x='sigma',\n",
    "    y='divergence',\n",
    "    hue='heuristic',\n",
    "    estimator='median',\n",
    "    errorbar=('pi', 50),\n",
    "    err_kws={'linewidth': 0},\n",
    ")\n",
    "\n",
    "ax.grid()\n",
    "ax.set_xscale('log')\n",
    "ax.set_yscale('log')\n",
    "ax.set_xlabel('$\\sigma_t$')\n",
    "ax.set_ylabel('$S_\\lambda(p(x_t \\mid y) \\, \\| \\, q(x_t \\mid y))$')\n",
    "\n",
    "handles, _ = ax.get_legend_handles_labels()\n",
    "labels = ['$\\mathbb{V}[x | x_t]$', '$(\\Sigma_x^{-1} + \\Sigma_t^{-1})^{-1}$', '$\\Sigma_t$', '$0$']\n",
    "\n",
    "ax.legend(handles=handles, labels=labels)\n",
    "\n",
    "fig = plt.gcf()\n",
    "fig.tight_layout(pad=0.5)\n",
    "fig.savefig('divergence.pdf', dpi=300)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "inox",
   "language": "python",
   "name": "inox"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}