{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "9236d0e1",
   "metadata": {},
   "source": [
    "## Automated Hyperparameter Selection \n",
    "\n",
    "This tutorial demonstrates how to automatically select optimal Gaussian Process hyperparameters for your surrogate model by testing multiple configurations and choosing the one with the best test set performance.\n",
    "\n",
    "### Why This Matters\n",
    "\n",
    "GP surrogate model performance is highly sensitive to:\n",
    "- **Kernel choice** (ExpSquared, Matern32, Matern52, etc.)\n",
    "- **Data scaling** (no scaling, MinMax, StandardScaler)\n",
    "- **Other hyperparameters** (white noise, amplitude, length scales)\n",
    "\n",
    "Rather than manually tuning these, we can systematically test combinations and select the best configuration based on test set mean squared error (MSE)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b7306461",
   "metadata": {},
   "source": [
    "### Step 1: Import Required Libraries\n",
    "\n",
    "Import `alabi` and its submodules along with standard scientific computing libraries."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "06ae9c1d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-02-21T07:39:36.024546Z",
     "iopub.status.busy": "2026-02-21T07:39:36.024375Z",
     "iopub.status.idle": "2026-02-21T07:39:37.178518Z",
     "shell.execute_reply": "2026-02-21T07:39:37.177817Z"
    }
   },
   "outputs": [],
   "source": [
    "import alabi\n",
    "import alabi.utility as ut\n",
    "import alabi.benchmarks as bm\n",
    "from alabi.core import SurrogateModel\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn import preprocessing\n",
    "from itertools import product\n",
    "\n",
    "from matplotlib import rcParams\n",
    "# rcParams['font.family'] = 'serif'\n",
    "# rcParams['text.usetex'] = True\n",
    "\n",
    "random_state = 101\n",
    "np.random.seed(random_state)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "478c87bf",
   "metadata": {},
   "source": [
    "### Step 2: Define Problem and Base Configuration\n",
    "\n",
    "Set up the benchmark problem (eggbox function) and define a base configuration for GP hyperparameters. These settings will be partially overridden when we test different combinations.\n",
    "\n",
    "**Key parameters:**\n",
    "- `ninit`: Number of initial training points\n",
    "- `niter`: Number of active learning iterations\n",
    "- `ncore`: Number of CPU cores for parallel evaluation (use 1 if experiencing multiprocessing issues)\n",
    "- `white_noise`: Log-scale white noise parameter (increase if you see positive definiteness errors)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "076901c5",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-02-21T07:39:37.180912Z",
     "iopub.status.busy": "2026-02-21T07:39:37.180512Z",
     "iopub.status.idle": "2026-02-21T07:39:41.657673Z",
     "shell.execute_reply": "2026-02-21T07:39:41.657118Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 50/50 [00:01<00:00, 41.58it/s]\n",
      "100%|██████████| 1000/1000 [00:01<00:00, 883.55it/s]\n"
     ]
    }
   ],
   "source": [
    "ninit = 50\n",
    "niter = 100\n",
    "basedir = \"demo\"\n",
    "kernel = \"ExpSquaredKernel\"\n",
    "benchmark = \"rosenbrock\"\n",
    "savedir = f\"{basedir}/{benchmark}/{kernel}/{ninit}_{niter}\"\n",
    "\n",
    "gp_kwargs = {\"kernel\": kernel, \n",
    "             \"fit_amp\": True, \n",
    "             \"fit_mean\": True, \n",
    "             \"fit_white_noise\": False, \n",
    "             \"white_noise\": -12,\n",
    "             \"gp_opt_method\": \"l-bfgs-b\",\n",
    "             \"hyperopt_method\": \"cv\",\n",
    "             \"cv_folds\": 8,\n",
    "             \"gp_amp_rng\": [-1,1],\n",
    "             \"gp_scale_rng\": [-2,2],\n",
    "             \"theta_scaler\": alabi.no_scaler,\n",
    "             \"y_scaler\": alabi.no_scaler}\n",
    "    \n",
    "sm = SurrogateModel(lnlike_fn=bm.eggbox[\"fn\"], \n",
    "                    bounds=bm.eggbox[\"bounds\"], \n",
    "                    savedir=savedir,\n",
    "                    ncore=8, \n",
    "                    pool_method=\"forkserver\",\n",
    "                    verbose=True,\n",
    "                    random_state=random_state)\n",
    "\n",
    "sm.init_samples(ntrain=ninit, ntest=1000, sampler=\"lhs\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9de6dade",
   "metadata": {},
   "source": [
    "### Step 3: Create Hyperparameter Grid\n",
    "\n",
    "Generate all combinations of the hyperparameters we want to test:\n",
    "- **Kernels**: Different covariance functions capture different smoothness assumptions\n",
    "- **Theta scaler**: How to scale input parameters\n",
    "- **Y scaler**: How to scale output values\n",
    "\n",
    "The `dict_to_combinations` function creates a Cartesian product of all options."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "0362ceb5",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-02-21T07:39:41.659382Z",
     "iopub.status.busy": "2026-02-21T07:39:41.659234Z",
     "iopub.status.idle": "2026-02-21T07:39:41.664387Z",
     "shell.execute_reply": "2026-02-21T07:39:41.663426Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "27 combinations to test\n"
     ]
    }
   ],
   "source": [
    "def dict_to_combinations(options_dict):\n",
    "    keys = options_dict.keys()\n",
    "    values = options_dict.values()\n",
    "    return [dict(zip(keys, combo)) for combo in product(*values)]\n",
    "\n",
    "def combinations_to_dict(combinations):\n",
    "    result = {key: [] for key in combinations[0].keys()}\n",
    "    for combo in combinations:\n",
    "        for key, value in combo.items():\n",
    "            result[key].append(value)\n",
    "    return result\n",
    "\n",
    "gp_kwarg_options = {\"kernel\": [\"ExpSquaredKernel\", \"Matern32Kernel\", \"Matern52Kernel\"],\n",
    "                    \"theta_scaler\": [ut.no_scaler, preprocessing.MinMaxScaler(), preprocessing.StandardScaler()],\n",
    "                    \"y_scaler\": [ut.no_scaler, preprocessing.MinMaxScaler(), preprocessing.StandardScaler()]}\n",
    "\n",
    "variable_settings = dict_to_combinations(gp_kwarg_options)\n",
    "print(len(variable_settings), \"combinations to test\")\n",
    "\n",
    "# get a list of dictionaries with all combinations of settings, where each dictionary is a copy of the original gp_kwargs with the variable settings updated\n",
    "setting_combos = []\n",
    "for settings in variable_settings:\n",
    "    new_settings = gp_kwargs.copy()\n",
    "    for key in settings.keys():\n",
    "        new_settings[key] = settings[key]\n",
    "    setting_combos.append(new_settings)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5de84eb1",
   "metadata": {},
   "source": [
    "### Step 4: Test All Hyperparameter Combinations\n",
    "\n",
    "Loop through all combinations and fit a GP for each. The `init_gp` method returns the test set MSE, which we use to evaluate performance.\n",
    "\n",
    "**Note:** This can take several minutes depending on the number of combinations and problem dimensionality. Use `try/except` to handle configurations that fail to converge."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8899aaf0",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-02-21T07:39:41.665929Z",
     "iopub.status.busy": "2026-02-21T07:39:41.665721Z",
     "iopub.status.idle": "2026-02-21T07:40:32.161385Z",
     "shell.execute_reply": "2026-02-21T07:40:32.160748Z"
    }
   },
   "outputs": [],
   "source": [
    "# docs: collapse output\n",
    "for ii in range(len(setting_combos)):\n",
    "    try:\n",
    "        test_mse = sm.init_gp(**setting_combos[ii], overwrite=True)\n",
    "    except:\n",
    "        test_mse = np.nan\n",
    "    setting_combos[ii][\"test_mse\"] = test_mse"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10e1a149",
   "metadata": {},
   "source": [
    "### Step 5: Analyze Results\n",
    "\n",
    "Convert results to a pandas DataFrame and inspect the top-performing configurations. Lower test MSE indicates better generalization performance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "6d8e114f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-02-21T07:40:32.162870Z",
     "iopub.status.busy": "2026-02-21T07:40:32.162743Z",
     "iopub.status.idle": "2026-02-21T07:40:32.175199Z",
     "shell.execute_reply": "2026-02-21T07:40:32.174726Z"
    }
   },
   "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>kernel</th>\n",
       "      <th>fit_amp</th>\n",
       "      <th>fit_mean</th>\n",
       "      <th>fit_white_noise</th>\n",
       "      <th>white_noise</th>\n",
       "      <th>gp_opt_method</th>\n",
       "      <th>hyperopt_method</th>\n",
       "      <th>cv_folds</th>\n",
       "      <th>gp_amp_rng</th>\n",
       "      <th>gp_scale_rng</th>\n",
       "      <th>theta_scaler</th>\n",
       "      <th>y_scaler</th>\n",
       "      <th>test_mse</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>ExpSquaredKernel</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>-12</td>\n",
       "      <td>l-bfgs-b</td>\n",
       "      <td>cv</td>\n",
       "      <td>8</td>\n",
       "      <td>[-1, 1]</td>\n",
       "      <td>[-2, 2]</td>\n",
       "      <td>StandardScaler()</td>\n",
       "      <td>MinMaxScaler()</td>\n",
       "      <td>1993.447717</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>ExpSquaredKernel</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>-12</td>\n",
       "      <td>l-bfgs-b</td>\n",
       "      <td>cv</td>\n",
       "      <td>8</td>\n",
       "      <td>[-1, 1]</td>\n",
       "      <td>[-2, 2]</td>\n",
       "      <td>StandardScaler()</td>\n",
       "      <td>StandardScaler()</td>\n",
       "      <td>1998.471961</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>Matern52Kernel</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>-12</td>\n",
       "      <td>l-bfgs-b</td>\n",
       "      <td>cv</td>\n",
       "      <td>8</td>\n",
       "      <td>[-1, 1]</td>\n",
       "      <td>[-2, 2]</td>\n",
       "      <td>StandardScaler()</td>\n",
       "      <td>no_scaler</td>\n",
       "      <td>2029.055811</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>ExpSquaredKernel</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>-12</td>\n",
       "      <td>l-bfgs-b</td>\n",
       "      <td>cv</td>\n",
       "      <td>8</td>\n",
       "      <td>[-1, 1]</td>\n",
       "      <td>[-2, 2]</td>\n",
       "      <td>StandardScaler()</td>\n",
       "      <td>no_scaler</td>\n",
       "      <td>2066.363090</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Matern32Kernel</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>-12</td>\n",
       "      <td>l-bfgs-b</td>\n",
       "      <td>cv</td>\n",
       "      <td>8</td>\n",
       "      <td>[-1, 1]</td>\n",
       "      <td>[-2, 2]</td>\n",
       "      <td>no_scaler</td>\n",
       "      <td>StandardScaler()</td>\n",
       "      <td>2067.461220</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              kernel  fit_amp  fit_mean  fit_white_noise  white_noise  \\\n",
       "7   ExpSquaredKernel     True      True            False          -12   \n",
       "8   ExpSquaredKernel     True      True            False          -12   \n",
       "24    Matern52Kernel     True      True            False          -12   \n",
       "6   ExpSquaredKernel     True      True            False          -12   \n",
       "11    Matern32Kernel     True      True            False          -12   \n",
       "\n",
       "   gp_opt_method hyperopt_method  cv_folds gp_amp_rng gp_scale_rng  \\\n",
       "7       l-bfgs-b              cv         8    [-1, 1]      [-2, 2]   \n",
       "8       l-bfgs-b              cv         8    [-1, 1]      [-2, 2]   \n",
       "24      l-bfgs-b              cv         8    [-1, 1]      [-2, 2]   \n",
       "6       l-bfgs-b              cv         8    [-1, 1]      [-2, 2]   \n",
       "11      l-bfgs-b              cv         8    [-1, 1]      [-2, 2]   \n",
       "\n",
       "        theta_scaler          y_scaler     test_mse  \n",
       "7   StandardScaler()    MinMaxScaler()  1993.447717  \n",
       "8   StandardScaler()  StandardScaler()  1998.471961  \n",
       "24  StandardScaler()         no_scaler  2029.055811  \n",
       "6   StandardScaler()         no_scaler  2066.363090  \n",
       "11         no_scaler  StandardScaler()  2067.461220  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results = pd.DataFrame(data=combinations_to_dict(setting_combos))\n",
    "top_fits = results.sort_values(\"test_mse\").head(5)\n",
    "top_fits"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "66d2bf18",
   "metadata": {},
   "source": [
    "### Step 6: Extract Best Configuration\n",
    "\n",
    "Identify the hyperparameter configuration with the lowest test MSE. This will be used for active learning."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "728d862d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-02-21T07:40:32.177024Z",
     "iopub.status.busy": "2026-02-21T07:40:32.176876Z",
     "iopub.status.idle": "2026-02-21T07:40:32.199112Z",
     "shell.execute_reply": "2026-02-21T07:40:32.198678Z"
    }
   },
   "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>kernel</th>\n",
       "      <th>fit_amp</th>\n",
       "      <th>fit_mean</th>\n",
       "      <th>fit_white_noise</th>\n",
       "      <th>white_noise</th>\n",
       "      <th>gp_opt_method</th>\n",
       "      <th>hyperopt_method</th>\n",
       "      <th>cv_folds</th>\n",
       "      <th>gp_amp_rng</th>\n",
       "      <th>gp_scale_rng</th>\n",
       "      <th>theta_scaler</th>\n",
       "      <th>y_scaler</th>\n",
       "      <th>test_mse</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>ExpSquaredKernel</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>-12</td>\n",
       "      <td>l-bfgs-b</td>\n",
       "      <td>cv</td>\n",
       "      <td>8</td>\n",
       "      <td>[-1, 1]</td>\n",
       "      <td>[-2, 2]</td>\n",
       "      <td>StandardScaler()</td>\n",
       "      <td>MinMaxScaler()</td>\n",
       "      <td>1993.447717</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             kernel  fit_amp  fit_mean  fit_white_noise  white_noise  \\\n",
       "7  ExpSquaredKernel     True      True            False          -12   \n",
       "\n",
       "  gp_opt_method hyperopt_method  cv_folds gp_amp_rng gp_scale_rng  \\\n",
       "7      l-bfgs-b              cv         8    [-1, 1]      [-2, 2]   \n",
       "\n",
       "       theta_scaler        y_scaler     test_mse  \n",
       "7  StandardScaler()  MinMaxScaler()  1993.447717  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_gp_results = results[results[\"test_mse\"] == results[\"test_mse\"].min()]\n",
    "best_gp_results"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4feaa4b6",
   "metadata": {},
   "source": [
    "### Step 7: Run Active Learning\n",
    "\n",
    "Use the optimal GP configuration for active learning. The BAPE (Bayesian Active Posterior Estimation) algorithm iteratively selects new training points to improve the surrogate model.\n",
    "\n",
    "**Key active learning parameters:**\n",
    "- `algorithm`: Acquisition function (\"bape\", \"agp\", or \"jones\")\n",
    "- `gp_opt_freq`: How often to reoptimize GP hyperparameters during training\n",
    "- `obj_opt_method`: Optimization method for acquisition function\n",
    "- `nopt`: Number of optimization restarts for acquisition function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "52294152",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-02-21T07:40:32.200651Z",
     "iopub.status.busy": "2026-02-21T07:40:32.200540Z",
     "iopub.status.idle": "2026-02-21T07:41:00.567133Z",
     "shell.execute_reply": "2026-02-21T07:41:00.566549Z"
    }
   },
   "outputs": [],
   "source": [
    "# docs: collapse output\n",
    "best_gp_kwargs = best_gp_results[gp_kwargs.keys()].to_dict(orient=\"records\")[0]\n",
    "\n",
    "al_kwargs = {\"algorithm\": \"bape\", \n",
    "             \"gp_opt_freq\": 20, \n",
    "             \"obj_opt_method\": \"nelder-mead\", \n",
    "             \"nopt\": 6}\n",
    "\n",
    "sm.init_gp(**best_gp_kwargs, overwrite=True)\n",
    "sm.active_train(niter=200, **al_kwargs)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6a223fdd",
   "metadata": {},
   "source": [
    "### Step 8: Visualize Training Progress\n",
    "\n",
    "Plot the test set MSE over active learning iterations. A decreasing trend indicates the surrogate model is improving. We can also highlight which iterations the GP hyperparameters are re-optimized (vertical gray lines)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "8d453f35",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-02-21T07:41:00.568671Z",
     "iopub.status.busy": "2026-02-21T07:41:00.568548Z",
     "iopub.status.idle": "2026-02-21T07:41:01.133854Z",
     "shell.execute_reply": "2026-02-21T07:41:01.133371Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(sm.training_results[\"iteration\"], sm.training_results[\"test_mse\"])\n",
    "for ii in range(0, sm.nactive, sm.gp_opt_freq+1):\n",
    "    plt.axvline(ii, color=\"gray\", linestyle=\"--\", alpha=0.5)\n",
    "plt.xlabel(\"Iteration\", fontsize=18)\n",
    "plt.ylabel(\"Test MSE\", fontsize=18)\n",
    "plt.xlim(0, sm.nactive)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c7ee12ee",
   "metadata": {},
   "source": [
    "How much does random variance affect the performance?\n",
    "\n",
    "Let's do a trial where we do 10 attempts with the same hyperparameter configuration and track the test mse:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f09fb10e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# docs: collapse output\n",
    "best_gp_kwargs = best_gp_results[gp_kwargs.keys()].to_dict(orient=\"records\")[0]\n",
    "\n",
    "test_mse_trials = []\n",
    "for ii in range(10):\n",
    "    al_kwargs = {\"algorithm\": \"bape\", \n",
    "                \"gp_opt_freq\": 20, \n",
    "                \"obj_opt_method\": \"nelder-mead\", \n",
    "                \"nopt\": 6}\n",
    "\n",
    "    sm.init_gp(**best_gp_kwargs, overwrite=True)\n",
    "    sm.active_train(niter=200, **al_kwargs)\n",
    "    test_mse_trials.append(sm.training_results[\"test_mse\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "4db3866c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for test_mse in test_mse_trials:\n",
    "    plt.plot(sm.training_results[\"iteration\"], test_mse, color=\"C0\", alpha=0.5)\n",
    "for ii in range(0, sm.nactive, sm.gp_opt_freq+1):\n",
    "    plt.axvline(ii, color=\"gray\", linestyle=\"--\", alpha=0.5)\n",
    "plt.xlabel(\"Iteration\", fontsize=18)\n",
    "plt.ylabel(\"Test MSE\", fontsize=18)\n",
    "plt.xlim(0, sm.nactive)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1cfb76e5",
   "metadata": {},
   "source": [
    "How do the other top initial fits perform during active learning?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "01e7f962",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-02-21T07:41:01.135513Z",
     "iopub.status.busy": "2026-02-21T07:41:01.135370Z",
     "iopub.status.idle": "2026-02-21T07:43:35.019319Z",
     "shell.execute_reply": "2026-02-21T07:43:35.018661Z"
    }
   },
   "outputs": [],
   "source": [
    "# docs: collapse output\n",
    "al_kwargs = {\"algorithm\": \"bape\", \n",
    "             \"gp_opt_freq\": 20, \n",
    "             \"obj_opt_method\": \"nelder-mead\", \n",
    "             \"nopt\": 6}\n",
    "\n",
    "mse_results = {}\n",
    "for idx in top_fits.index:\n",
    "    gp_kwargs_idx = top_fits[top_fits.index == idx][gp_kwargs.keys()].to_dict(orient=\"records\")[0]\n",
    "    sm.init_gp(**gp_kwargs_idx, overwrite=True)\n",
    "    sm.active_train(niter=200, **al_kwargs)\n",
    "    mse_results[idx] = sm.training_results[\"test_mse\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "5eec2bb2",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-02-21T07:43:35.020866Z",
     "iopub.status.busy": "2026-02-21T07:43:35.020644Z",
     "iopub.status.idle": "2026-02-21T07:43:35.200135Z",
     "shell.execute_reply": "2026-02-21T07:43:35.199636Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>kernel</th>\n",
       "      <th>fit_amp</th>\n",
       "      <th>fit_mean</th>\n",
       "      <th>fit_white_noise</th>\n",
       "      <th>white_noise</th>\n",
       "      <th>gp_opt_method</th>\n",
       "      <th>hyperopt_method</th>\n",
       "      <th>cv_folds</th>\n",
       "      <th>gp_amp_rng</th>\n",
       "      <th>gp_scale_rng</th>\n",
       "      <th>theta_scaler</th>\n",
       "      <th>y_scaler</th>\n",
       "      <th>test_mse</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>ExpSquaredKernel</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>-12</td>\n",
       "      <td>l-bfgs-b</td>\n",
       "      <td>cv</td>\n",
       "      <td>8</td>\n",
       "      <td>[-1, 1]</td>\n",
       "      <td>[-2, 2]</td>\n",
       "      <td>StandardScaler()</td>\n",
       "      <td>MinMaxScaler()</td>\n",
       "      <td>1993.447717</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>ExpSquaredKernel</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>-12</td>\n",
       "      <td>l-bfgs-b</td>\n",
       "      <td>cv</td>\n",
       "      <td>8</td>\n",
       "      <td>[-1, 1]</td>\n",
       "      <td>[-2, 2]</td>\n",
       "      <td>StandardScaler()</td>\n",
       "      <td>StandardScaler()</td>\n",
       "      <td>1998.471961</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>Matern52Kernel</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>-12</td>\n",
       "      <td>l-bfgs-b</td>\n",
       "      <td>cv</td>\n",
       "      <td>8</td>\n",
       "      <td>[-1, 1]</td>\n",
       "      <td>[-2, 2]</td>\n",
       "      <td>StandardScaler()</td>\n",
       "      <td>no_scaler</td>\n",
       "      <td>2029.055811</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>ExpSquaredKernel</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>-12</td>\n",
       "      <td>l-bfgs-b</td>\n",
       "      <td>cv</td>\n",
       "      <td>8</td>\n",
       "      <td>[-1, 1]</td>\n",
       "      <td>[-2, 2]</td>\n",
       "      <td>StandardScaler()</td>\n",
       "      <td>no_scaler</td>\n",
       "      <td>2066.363090</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Matern32Kernel</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>-12</td>\n",
       "      <td>l-bfgs-b</td>\n",
       "      <td>cv</td>\n",
       "      <td>8</td>\n",
       "      <td>[-1, 1]</td>\n",
       "      <td>[-2, 2]</td>\n",
       "      <td>no_scaler</td>\n",
       "      <td>StandardScaler()</td>\n",
       "      <td>2067.461220</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              kernel  fit_amp  fit_mean  fit_white_noise  white_noise  \\\n",
       "7   ExpSquaredKernel     True      True            False          -12   \n",
       "8   ExpSquaredKernel     True      True            False          -12   \n",
       "24    Matern52Kernel     True      True            False          -12   \n",
       "6   ExpSquaredKernel     True      True            False          -12   \n",
       "11    Matern32Kernel     True      True            False          -12   \n",
       "\n",
       "   gp_opt_method hyperopt_method  cv_folds gp_amp_rng gp_scale_rng  \\\n",
       "7       l-bfgs-b              cv         8    [-1, 1]      [-2, 2]   \n",
       "8       l-bfgs-b              cv         8    [-1, 1]      [-2, 2]   \n",
       "24      l-bfgs-b              cv         8    [-1, 1]      [-2, 2]   \n",
       "6       l-bfgs-b              cv         8    [-1, 1]      [-2, 2]   \n",
       "11      l-bfgs-b              cv         8    [-1, 1]      [-2, 2]   \n",
       "\n",
       "        theta_scaler          y_scaler     test_mse  \n",
       "7   StandardScaler()    MinMaxScaler()  1993.447717  \n",
       "8   StandardScaler()  StandardScaler()  1998.471961  \n",
       "24  StandardScaler()         no_scaler  2029.055811  \n",
       "6   StandardScaler()         no_scaler  2066.363090  \n",
       "11         no_scaler  StandardScaler()  2067.461220  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,6))\n",
    "for idx in top_fits.index:\n",
    "    plt.plot(sm.training_results[\"iteration\"], mse_results[idx], label=f\"config {idx}\")\n",
    "for ii in range(0, sm.nactive, sm.gp_opt_freq+1):\n",
    "    plt.axvline(ii, color=\"gray\", linestyle=\"--\", alpha=0.5)\n",
    "plt.legend(loc=\"upper right\", fontsize=16)\n",
    "plt.xlabel(\"Iteration\", fontsize=18)\n",
    "plt.ylabel(\"Test MSE\", fontsize=18)\n",
    "plt.xlim(0, sm.nactive)\n",
    "plt.show()\n",
    "\n",
    "top_fits"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e6056be3",
   "metadata": {},
   "source": [
    "Here we see that the configuration for best initial fit isn't always necessarily the best fit after running active learning, but it is not a bad place to start. In practice, if the test error is not converging, it may be an indication to try a different configuration of hyperparameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "08bf8304",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "34cf407c",
   "metadata": {},
   "source": [
    "### Next Steps\n",
    "\n",
    "Now that you have an optimized surrogate model, you can:\n",
    "\n",
    "1. **Run posterior sampling** with `sm.run_emcee()` or `sm.run_dynesty()`\n",
    "2. **Visualize the surrogate** using `alabi.visualization` tools\n",
    "3. **Test on new problems** by changing the benchmark function\n",
    "4. **Expand the hyperparameter grid** to include more options like:\n",
    "   - Different `white_noise` values\n",
    "   - Different `cv_folds` settings\n",
    "   - Different data scaling functions for `theta_scaler` or `y_scaler`"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ae5764e9",
   "metadata": {},
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "py314",
   "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": "3.14.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}