{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Banded ridge regression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This tutorial builds on the previous encoding model tutorial to map _multiple_ feature spaces onto brain activity during natural language comprehension. We will use two language models in one joint model: speech embeddings from Whisper's encoder and language embeddings from GPT2-XL. The [Himalaya](https://gallantlab.org/himalaya/index.html) package ([Dupré La Tour et al., 2022](https://doi.org/10.1016/j.neuroimage.2022.119728)) provides code to run and evaluate these joint encoding models (example [tutorial](https://gallantlab.org/voxelwise_tutorials/_auto_examples/shortclips/06_plot_banded_ridge_model.html)).\n",
    "\n",
    "[![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/hassonlab/podcast-ecog-tutorials/blob/main/notebooks/06-banded-ridge.ipynb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# only run this cell in colab\n",
    "!pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu124\n",
    "!pip install mne mne_bids himalaya scikit-learn pandas matplotlib nilearn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import mne\n",
    "import h5py\n",
    "import torch\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from nilearn.plotting import plot_markers\n",
    "from mne_bids import BIDSPath\n",
    "\n",
    "from himalaya.backend import set_backend, get_backend\n",
    "from himalaya.kernel_ridge import Kernelizer, ColumnKernelizer, MultipleKernelRidgeCV\n",
    "from himalaya.scoring import correlation_score_split\n",
    "\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "from sklearn import set_config\n",
    "set_config(display='diagram')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Use a GPU for fitting an encoding model, if available."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using cuda!\n"
     ]
    }
   ],
   "source": [
    "if torch.cuda.is_available():\n",
    "    set_backend(\"torch_cuda\")\n",
    "    print(\"Using cuda!\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load GPT-2 embeddings"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Similar to previous tutorials, we will load the contextual word embeddings from GPT-2."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using embedding file path: ../../monkey/stimuli/gpt2-xl/features.hdf5\n"
     ]
    }
   ],
   "source": [
    "bids_root = \"\"  # if using a local dataset, set this variable accordingly\n",
    "\n",
    "# Download the transcript, if required\n",
    "embedding_path = f\"{bids_root}stimuli/gpt2-xl/features.hdf5\"\n",
    "if not len(bids_root):\n",
    "    !wget -nc https://s3.amazonaws.com/openneuro.org/ds005574/$embedding_path\n",
    "    embedding_path = \"features.hdf5\"\n",
    "\n",
    "print(f\"Using embedding file path: {embedding_path}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LLM embedding matrix has shape: (5491, 1600)\n"
     ]
    }
   ],
   "source": [
    "modelname, layer = 'gpt2-xl', 24\n",
    "with h5py.File(embedding_path, \"r\") as f:\n",
    "    contextual_embeddings = f[f\"layer-{layer}\"][...]\n",
    "print(f\"LLM embedding matrix has shape: {contextual_embeddings.shape}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model accuracy: 30.942%\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>word_idx</th>\n",
       "      <th>word</th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "      <th>hftoken</th>\n",
       "      <th>token_id</th>\n",
       "      <th>rank</th>\n",
       "      <th>true_prob</th>\n",
       "      <th>top_pred</th>\n",
       "      <th>entropy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>Act</td>\n",
       "      <td>3.710</td>\n",
       "      <td>3.790</td>\n",
       "      <td>ĠAct</td>\n",
       "      <td>2191</td>\n",
       "      <td>1644</td>\n",
       "      <td>0.000012</td>\n",
       "      <td>0</td>\n",
       "      <td>2.402717</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>one,</td>\n",
       "      <td>3.990</td>\n",
       "      <td>4.190</td>\n",
       "      <td>Ġone</td>\n",
       "      <td>530</td>\n",
       "      <td>92</td>\n",
       "      <td>0.000342</td>\n",
       "      <td>352</td>\n",
       "      <td>3.732053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>one,</td>\n",
       "      <td>3.990</td>\n",
       "      <td>4.190</td>\n",
       "      <td>,</td>\n",
       "      <td>11</td>\n",
       "      <td>3</td>\n",
       "      <td>0.059520</td>\n",
       "      <td>25</td>\n",
       "      <td>4.259335</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>monkey</td>\n",
       "      <td>4.651</td>\n",
       "      <td>4.931</td>\n",
       "      <td>Ġmonkey</td>\n",
       "      <td>21657</td>\n",
       "      <td>4022</td>\n",
       "      <td>0.000018</td>\n",
       "      <td>3715</td>\n",
       "      <td>6.621269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>in</td>\n",
       "      <td>4.951</td>\n",
       "      <td>5.011</td>\n",
       "      <td>Ġin</td>\n",
       "      <td>287</td>\n",
       "      <td>15</td>\n",
       "      <td>0.004237</td>\n",
       "      <td>0</td>\n",
       "      <td>4.444838</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   word_idx    word  start    end  hftoken  token_id  rank  true_prob  \\\n",
       "0         0     Act  3.710  3.790     ĠAct      2191  1644   0.000012   \n",
       "1         1    one,  3.990  4.190     Ġone       530    92   0.000342   \n",
       "2         1    one,  3.990  4.190        ,        11     3   0.059520   \n",
       "3         2  monkey  4.651  4.931  Ġmonkey     21657  4022   0.000018   \n",
       "4         3      in  4.951  5.011      Ġin       287    15   0.004237   \n",
       "\n",
       "   top_pred   entropy  \n",
       "0         0  2.402717  \n",
       "1       352  3.732053  \n",
       "2        25  4.259335  \n",
       "3      3715  6.621269  \n",
       "4         0  4.444838  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Download the transcript, if required\n",
    "transcript_path = f\"{bids_root}stimuli/gpt2-xl/transcript.tsv\"\n",
    "if not len(bids_root):\n",
    "    !wget -nc https://s3.amazonaws.com/openneuro.org/ds005574/$transcript_path\n",
    "    transcript_path = \"transcript.tsv\"\n",
    "\n",
    "# Load transcript\n",
    "df_contextual = pd.read_csv(transcript_path, sep=\"\\t\", index_col=0)\n",
    "if \"rank\" in df_contextual.columns:\n",
    "    model_acc = (df_contextual[\"rank\"] == 0).mean()\n",
    "    print(f\"Model accuracy: {model_acc*100:.3f}%\")\n",
    "\n",
    "df_contextual.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LLM embeddings matrix has shape: (5136, 1600)\n"
     ]
    }
   ],
   "source": [
    "aligned_gpt_embeddings = []\n",
    "for _, group in df_contextual.groupby(\"word_idx\"): # group by word index\n",
    "    indices = group.index.to_numpy()\n",
    "    average_emb = contextual_embeddings[indices].mean(0) # average features\n",
    "    aligned_gpt_embeddings.append(average_emb)\n",
    "aligned_gpt_embeddings = np.stack(aligned_gpt_embeddings)\n",
    "print(f\"LLM embeddings matrix has shape: {aligned_gpt_embeddings.shape}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>word</th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>word_idx</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Act</td>\n",
       "      <td>3.710</td>\n",
       "      <td>3.790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>one,</td>\n",
       "      <td>3.990</td>\n",
       "      <td>4.190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>monkey</td>\n",
       "      <td>4.651</td>\n",
       "      <td>4.931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>in</td>\n",
       "      <td>4.951</td>\n",
       "      <td>5.011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>the</td>\n",
       "      <td>5.051</td>\n",
       "      <td>5.111</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            word  start    end\n",
       "word_idx                      \n",
       "0            Act  3.710  3.790\n",
       "1           one,  3.990  4.190\n",
       "2         monkey  4.651  4.931\n",
       "3             in  4.951  5.011\n",
       "4            the  5.051  5.111"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_word_gpt = df_contextual.groupby(\"word_idx\").agg(dict(word=\"first\", start=\"first\", end=\"last\"))\n",
    "df_word_gpt.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Whisper encoder embeddings"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will also load speech embeddings from Whisper (medium size) from the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using embedding file path: ../../monkey/stimuli/whisper-medium/features.hdf5\n"
     ]
    }
   ],
   "source": [
    "# Download the transcript, if required\n",
    "embedding_path = f\"{bids_root}stimuli/whisper-medium/features.hdf5\"\n",
    "if not len(bids_root):\n",
    "    !wget -nc https://s3.amazonaws.com/openneuro.org/ds005574/$embedding_path\n",
    "    embedding_path = \"features.hdf5\"\n",
    "\n",
    "print(f\"Using embedding file path: {embedding_path}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LLM embedding matrix has shape: (5134, 2048)\n"
     ]
    }
   ],
   "source": [
    "with h5py.File(embedding_path, \"r\") as f:\n",
    "    whisper_embeddings = f[\"vectors\"][...]\n",
    "print(f\"LLM embedding matrix has shape: {whisper_embeddings.shape}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>word_idx</th>\n",
       "      <th>word</th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>Act</td>\n",
       "      <td>3.710</td>\n",
       "      <td>3.790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>one,</td>\n",
       "      <td>3.990</td>\n",
       "      <td>4.190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>monkey</td>\n",
       "      <td>4.651</td>\n",
       "      <td>4.931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>in</td>\n",
       "      <td>4.951</td>\n",
       "      <td>5.011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>the</td>\n",
       "      <td>5.051</td>\n",
       "      <td>5.111</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   word_idx    word  start    end\n",
       "0         0     Act  3.710  3.790\n",
       "1         1    one,  3.990  4.190\n",
       "2         2  monkey  4.651  4.931\n",
       "3         3      in  4.951  5.011\n",
       "4         4     the  5.051  5.111"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Download the transcript, if required\n",
    "transcript_path = f\"{bids_root}stimuli/whisper-medium/transcript.tsv\"\n",
    "if not len(bids_root):\n",
    "    !wget -nc https://s3.amazonaws.com/openneuro.org/ds005574/$transcript_path\n",
    "    transcript_path = \"transcript.tsv\"\n",
    "\n",
    "# Load transcript\n",
    "df_whisper = pd.read_csv(transcript_path, sep=\"\\t\", index_col=0)\n",
    "df_whisper.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LLM embeddings matrix has shape: (5134, 2048)\n"
     ]
    }
   ],
   "source": [
    "aligned_whisper_embeddings = []\n",
    "for _, group in df_whisper.groupby(\"word_idx\"): # group by word index\n",
    "    indices = group.index.to_numpy()\n",
    "    average_emb = whisper_embeddings[indices].mean(0) # average features\n",
    "    aligned_whisper_embeddings.append(average_emb)\n",
    "aligned_whisper_embeddings = np.stack(aligned_whisper_embeddings)\n",
    "print(f\"LLM embeddings matrix has shape: {aligned_whisper_embeddings.shape}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>word</th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>word_idx</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Act</td>\n",
       "      <td>3.710</td>\n",
       "      <td>3.790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>one,</td>\n",
       "      <td>3.990</td>\n",
       "      <td>4.190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>monkey</td>\n",
       "      <td>4.651</td>\n",
       "      <td>4.931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>in</td>\n",
       "      <td>4.951</td>\n",
       "      <td>5.011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>the</td>\n",
       "      <td>5.051</td>\n",
       "      <td>5.111</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            word  start    end\n",
       "word_idx                      \n",
       "0            Act  3.710  3.790\n",
       "1           one,  3.990  4.190\n",
       "2         monkey  4.651  4.931\n",
       "3             in  4.951  5.011\n",
       "4            the  5.051  5.111"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_word_whisper = df_whisper.groupby(\"word_idx\").agg(dict(word=\"first\", start=\"first\", end=\"last\"))\n",
    "df_word_whisper.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Because we are using two different kinds of language models, they may have different tokenizers and thus their embeddings may not be aligned. We'll aligns their transcripts together here."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>word_x</th>\n",
       "      <th>start_x</th>\n",
       "      <th>end_x</th>\n",
       "      <th>word_y</th>\n",
       "      <th>start_y</th>\n",
       "      <th>end_y</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>word_idx</th>\n",
       "      <th></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>0</th>\n",
       "      <td>Act</td>\n",
       "      <td>3.710</td>\n",
       "      <td>3.790</td>\n",
       "      <td>Act</td>\n",
       "      <td>3.710</td>\n",
       "      <td>3.790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>one,</td>\n",
       "      <td>3.990</td>\n",
       "      <td>4.190</td>\n",
       "      <td>one,</td>\n",
       "      <td>3.990</td>\n",
       "      <td>4.190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>monkey</td>\n",
       "      <td>4.651</td>\n",
       "      <td>4.931</td>\n",
       "      <td>monkey</td>\n",
       "      <td>4.651</td>\n",
       "      <td>4.931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>in</td>\n",
       "      <td>4.951</td>\n",
       "      <td>5.011</td>\n",
       "      <td>in</td>\n",
       "      <td>4.951</td>\n",
       "      <td>5.011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>the</td>\n",
       "      <td>5.051</td>\n",
       "      <td>5.111</td>\n",
       "      <td>the</td>\n",
       "      <td>5.051</td>\n",
       "      <td>5.111</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5131</th>\n",
       "      <td>go</td>\n",
       "      <td>1798.546</td>\n",
       "      <td>1798.646</td>\n",
       "      <td>go</td>\n",
       "      <td>1798.546</td>\n",
       "      <td>1798.646</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5132</th>\n",
       "      <td>to</td>\n",
       "      <td>1798.666</td>\n",
       "      <td>1798.746</td>\n",
       "      <td>to</td>\n",
       "      <td>1798.666</td>\n",
       "      <td>1798.746</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5133</th>\n",
       "      <td>court</td>\n",
       "      <td>1798.786</td>\n",
       "      <td>1799.006</td>\n",
       "      <td>court</td>\n",
       "      <td>1798.786</td>\n",
       "      <td>1799.006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5134</th>\n",
       "      <td>over</td>\n",
       "      <td>1799.046</td>\n",
       "      <td>1799.226</td>\n",
       "      <td>over</td>\n",
       "      <td>1799.046</td>\n",
       "      <td>1799.226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5135</th>\n",
       "      <td>it.</td>\n",
       "      <td>1799.327</td>\n",
       "      <td>1799.367</td>\n",
       "      <td>it.</td>\n",
       "      <td>1799.327</td>\n",
       "      <td>1799.367</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5134 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          word_x   start_x     end_x  word_y   start_y     end_y\n",
       "word_idx                                                        \n",
       "0            Act     3.710     3.790     Act     3.710     3.790\n",
       "1           one,     3.990     4.190    one,     3.990     4.190\n",
       "2         monkey     4.651     4.931  monkey     4.651     4.931\n",
       "3             in     4.951     5.011      in     4.951     5.011\n",
       "4            the     5.051     5.111     the     5.051     5.111\n",
       "...          ...       ...       ...     ...       ...       ...\n",
       "5131          go  1798.546  1798.646      go  1798.546  1798.646\n",
       "5132          to  1798.666  1798.746      to  1798.666  1798.746\n",
       "5133       court  1798.786  1799.006   court  1798.786  1799.006\n",
       "5134        over  1799.046  1799.226    over  1799.046  1799.226\n",
       "5135         it.  1799.327  1799.367     it.  1799.327  1799.367\n",
       "\n",
       "[5134 rows x 6 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_merged = pd.merge(df_word_gpt, df_word_whisper, left_index=True, right_index=True)\n",
    "df_merged"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading brain data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we will load the preprocessed high-gamma ECoG data using MNE. Here, we will demonstrate loading data from our third subject."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "File path within the dataset: ../../monkey/derivatives/ecogprep/sub-03/ieeg/sub-03_task-podcast_desc-highgamma_ieeg.fif\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<script type=\"text/javascript\">\n",
       "    const toggleVisibility = (className) => {\n",
       "\n",
       "  const elements = document.querySelectorAll(`.${className}`)\n",
       "\n",
       "  elements.forEach(element => {\n",
       "    if (element.classList.contains('repr-section-header')) {\n",
       "      // Don't collapse the section header row.\n",
       "       return\n",
       "    }\n",
       "    if (element.classList.contains('repr-element-collapsed')) {\n",
       "      // Force a reflow to ensure the display change takes effect before removing the class\n",
       "      element.classList.remove('repr-element-collapsed')\n",
       "      element.offsetHeight // This forces the browser to recalculate layout\n",
       "      element.classList.remove('repr-element-faded')\n",
       "    } else {\n",
       "      // Start transition to hide the element\n",
       "      element.classList.add('repr-element-faded')\n",
       "      element.addEventListener('transitionend', handler = (e) => {\n",
       "        if (e.propertyName === 'opacity' && getComputedStyle(element).opacity === '0.2') {\n",
       "          element.classList.add('repr-element-collapsed')\n",
       "          element.removeEventListener('transitionend', handler)\n",
       "        }\n",
       "      });\n",
       "    }\n",
       "  });\n",
       "\n",
       "  // Take care of button (adjust caret)\n",
       "  const button = document.querySelectorAll(`.repr-section-header.${className} > th.repr-section-toggle-col > button`)[0]\n",
       "  button.classList.toggle('collapsed')\n",
       "\n",
       "  // Take care of the tooltip of the section header row\n",
       "  const sectionHeaderRow = document.querySelectorAll(`tr.repr-section-header.${className}`)[0]\n",
       "  sectionHeaderRow.classList.toggle('collapsed')\n",
       "  sectionHeaderRow.title = sectionHeaderRow.title === 'Hide section' ? 'Show section' : 'Hide section'\n",
       "}\n",
       "</script>\n",
       "\n",
       "<style type=\"text/css\">\n",
       "    table.repr.table.table-hover.table-striped.table-sm.table-responsive.small {\n",
       "  /* Don't make rows wider than they need to be. */\n",
       "  display: inline;\n",
       "}\n",
       "\n",
       "table > tbody > tr.repr-element > td {\n",
       "  /* Apply a tighter layout to the table cells. */\n",
       "  padding-top: 0.1rem;\n",
       "  padding-bottom: 0.1rem;\n",
       "  padding-right: 1rem;\n",
       "}\n",
       "\n",
       "table > tbody > tr > td.repr-section-toggle-col {\n",
       "  /* Remove background and border of the first cell in every row\n",
       "     (this row is only used for the collapse / uncollapse caret)\n",
       "\n",
       "     TODO: Need to find a good solution for VS Code that works in both\n",
       "           light and dark mode. */\n",
       "  border-color: transparent;\n",
       "  --bs-table-accent-bg: transparent;\n",
       "}\n",
       "\n",
       "tr.repr-section-header {\n",
       "  /* Remove stripes from section header rows */\n",
       "  background-color: transparent;\n",
       "  border-color: transparent;\n",
       "  --bs-table-striped-bg: transparent;\n",
       "  cursor: pointer;\n",
       "}\n",
       "\n",
       "tr.repr-section-header > th {\n",
       "  text-align: left !important;\n",
       "  vertical-align: middle;\n",
       "}\n",
       "\n",
       ".repr-element, tr.repr-element > td {\n",
       "  opacity: 1;\n",
       "  text-align: left !important;\n",
       "}\n",
       "\n",
       ".repr-element-faded {\n",
       "  transition: 0.3s ease;\n",
       "  opacity: 0.2;\n",
       "}\n",
       "\n",
       ".repr-element-collapsed {\n",
       "  display: none;\n",
       "}\n",
       "\n",
       "/* Collapse / uncollapse button and the caret it contains. */\n",
       ".repr-section-toggle-col button {\n",
       "  cursor: pointer;\n",
       "  width: 1rem;\n",
       "  background-color: transparent;\n",
       "  border-color: transparent;\n",
       "}\n",
       "\n",
       "span.collapse-uncollapse-caret {\n",
       "  width: 1rem;\n",
       "  height: 1rem;\n",
       "  display: block;\n",
       "  background-repeat: no-repeat;\n",
       "  background-position: left;\n",
       "  background-size: contain;\n",
       "}\n",
       "\n",
       "/* The collapse / uncollapse carets were copied from the free Font Awesome collection and adjusted. */\n",
       "\n",
       "/* Default to black carets for light mode */\n",
       ".repr-section-toggle-col > button.collapsed > span.collapse-uncollapse-caret {\n",
       "  background-image: url('data:image/svg+xml;charset=utf8,<svg xmlns=\"http://www.w3.org/2000/svg\" viewBox=\"0 0 256 512\"><!--!Font Awesome Free 6.5.2 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license/free Copyright 2024 Fonticons, Inc.--><path fill=\"black\" d=\"M246.6 278.6c12.5-12.5 12.5-32.8 0-45.3l-128-128c-9.2-9.2-22.9-11.9-34.9-6.9s-19.8 16.6-19.8 29.6l0 256c0 12.9 7.8 24.6 19.8 29.6s25.7 2.2 34.9-6.9l128-128z\"/></svg>');\n",
       "}\n",
       "\n",
       ".repr-section-toggle-col\n",
       "  > button:not(.collapsed)\n",
       "  > span.collapse-uncollapse-caret {\n",
       "  background-image: url('data:image/svg+xml;charset=utf8,<svg xmlns=\"http://www.w3.org/2000/svg\" viewBox=\"0 0 320 512\"><!--!Font Awesome Free 6.5.2 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license/free Copyright 2024 Fonticons, Inc.--><path fill=\"black\" d=\"M137.4 374.6c12.5 12.5 32.8 12.5 45.3 0l128-128c9.2-9.2 11.9-22.9 6.9-34.9s-16.6-19.8-29.6-19.8L32 192c-12.9 0-24.6 7.8-29.6 19.8s-2.2 25.7 6.9 34.9l128 128z\"/></svg>');\n",
       "}\n",
       "\n",
       "/* Use white carets for dark mode */\n",
       "@media (prefers-color-scheme: dark) {\n",
       "  .repr-section-toggle-col > button.collapsed > span.collapse-uncollapse-caret {\n",
       "    background-image: url('data:image/svg+xml;charset=utf8,<svg xmlns=\"http://www.w3.org/2000/svg\" viewBox=\"0 0 256 512\"><!--!Font Awesome Free 6.5.2 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license/free Copyright 2024 Fonticons, Inc.--><path fill=\"white\" d=\"M246.6 278.6c12.5-12.5 12.5-32.8 0-45.3l-128-128c-9.2-9.2-22.9-11.9-34.9-6.9s-19.8 16.6-19.8 29.6l0 256c0 12.9 7.8 24.6 19.8 29.6s25.7 2.2 34.9-6.9l128-128z\"/></svg>');\n",
       "  }\n",
       "\n",
       "  .repr-section-toggle-col\n",
       "    > button:not(.collapsed)\n",
       "    > span.collapse-uncollapse-caret {\n",
       "    background-image: url('data:image/svg+xml;charset=utf8,<svg xmlns=\"http://www.w3.org/2000/svg\" viewBox=\"0 0 320 512\"><!--!Font Awesome Free 6.5.2 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license/free Copyright 2024 Fonticons, Inc.--><path fill=\"white\" d=\"M137.4 374.6c12.5 12.5 32.8 12.5 45.3 0l128-128c9.2-9.2 11.9-22.9 6.9-34.9s-16.6-19.8-29.6-19.8L32 192c-12.9 0-24.6 7.8-29.6 19.8s-2.2 25.7 6.9 34.9l128 128z\"/></svg>');\n",
       "  }\n",
       "}\n",
       "\n",
       ".channel-names-btn {\n",
       "  padding: 0;\n",
       "  border: none;\n",
       "  background: none;\n",
       "  text-decoration: underline;\n",
       "  text-decoration-style: dashed;\n",
       "  cursor: pointer;\n",
       "  color: #0d6efd;\n",
       "}\n",
       "\n",
       ".channel-names-btn:hover {\n",
       "  color: #0a58ca;\n",
       "}\n",
       "</style>\n",
       "\n",
       "\n",
       "\n",
       "<table class=\"repr table table-hover table-striped table-sm table-responsive small\">\n",
       "    \n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "<tr class=\"repr-section-header general-4d66a7a3-a0e4-4708-9091-4e5e27f1b970\"  title=\"Hide section\" \n",
       "    onclick=\"toggleVisibility('general-4d66a7a3-a0e4-4708-9091-4e5e27f1b970')\">\n",
       "    <th class=\"repr-section-toggle-col\">\n",
       "        <button>\n",
       "            \n",
       "            <span class=\"collapse-uncollapse-caret\"></span>\n",
       "        </button>\n",
       "    </th>\n",
       "    <th colspan=\"2\">\n",
       "        <strong>General</strong>\n",
       "    </th>\n",
       "</tr>\n",
       "\n",
       "<tr class=\"repr-element general-4d66a7a3-a0e4-4708-9091-4e5e27f1b970 \">\n",
       "    <td class=\"repr-section-toggle-col\"></td>\n",
       "    <td>Filename(s)</td>\n",
       "    <td>\n",
       "        \n",
       "        sub-03_task-podcast_desc-highgamma_ieeg.fif\n",
       "        \n",
       "        \n",
       "    </td>\n",
       "</tr>\n",
       "\n",
       "<tr class=\"repr-element general-4d66a7a3-a0e4-4708-9091-4e5e27f1b970 \">\n",
       "    <td class=\"repr-section-toggle-col\"></td>\n",
       "    <td>MNE object type</td>\n",
       "    <td>Raw</td>\n",
       "</tr>\n",
       "<tr class=\"repr-element general-4d66a7a3-a0e4-4708-9091-4e5e27f1b970 \">\n",
       "    <td class=\"repr-section-toggle-col\"></td>\n",
       "    <td>Measurement date</td>\n",
       "    \n",
       "    <td>2019-03-11 at 10:54:21 UTC</td>\n",
       "    \n",
       "</tr>\n",
       "<tr class=\"repr-element general-4d66a7a3-a0e4-4708-9091-4e5e27f1b970 \">\n",
       "    <td class=\"repr-section-toggle-col\"></td>\n",
       "    <td>Participant</td>\n",
       "    \n",
       "    \n",
       "    <td>sub-03</td>\n",
       "    \n",
       "    \n",
       "</tr>\n",
       "<tr class=\"repr-element general-4d66a7a3-a0e4-4708-9091-4e5e27f1b970 \">\n",
       "    <td class=\"repr-section-toggle-col\"></td>\n",
       "    <td>Experimenter</td>\n",
       "    \n",
       "    <td>Unknown</td>\n",
       "    \n",
       "</tr>\n",
       "    \n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "<tr class=\"repr-section-header acquisition-c2fdbc1f-dd70-4983-a5c0-ef8e0a3ece81\" \n",
       "    title=\"Hide section\"  onclick=\"toggleVisibility('acquisition-c2fdbc1f-dd70-4983-a5c0-ef8e0a3ece81')\">\n",
       "    <th class=\"repr-section-toggle-col\">\n",
       "        <button>\n",
       "            \n",
       "            <span class=\"collapse-uncollapse-caret\"></span>\n",
       "        </button>\n",
       "    </th>\n",
       "    <th colspan=\"2\">\n",
       "        <strong>Acquisition</strong>\n",
       "    </th>\n",
       "</tr>\n",
       "\n",
       "<tr class=\"repr-element acquisition-c2fdbc1f-dd70-4983-a5c0-ef8e0a3ece81 \">\n",
       "    <td class=\"repr-section-toggle-col\"></td>\n",
       "    <td>Duration</td>\n",
       "    <td>00:29:60 (HH:MM:SS)</td>\n",
       "</tr>\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "<tr class=\"repr-element acquisition-c2fdbc1f-dd70-4983-a5c0-ef8e0a3ece81 \">\n",
       "    <td class=\"repr-section-toggle-col\"></td>\n",
       "    <td>Sampling frequency</td>\n",
       "    <td>512.00 Hz</td>\n",
       "</tr>\n",
       "\n",
       "\n",
       "<tr class=\"repr-element acquisition-c2fdbc1f-dd70-4983-a5c0-ef8e0a3ece81 \">\n",
       "    <td class=\"repr-section-toggle-col\"></td>\n",
       "    <td>Time points</td>\n",
       "    <td>921,600</td>\n",
       "</tr>\n",
       "\n",
       "\n",
       "    \n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "<tr class=\"repr-section-header channels-3ef46833-4f4e-4c5d-a467-05f1dc4566d4\"  title=\"Hide section\" \n",
       "    onclick=\"toggleVisibility('channels-3ef46833-4f4e-4c5d-a467-05f1dc4566d4')\">\n",
       "    <th class=\"repr-section-toggle-col\">\n",
       "        <button>\n",
       "            \n",
       "            <span class=\"collapse-uncollapse-caret\"></span>\n",
       "        </button>\n",
       "    </th>\n",
       "    <th colspan=\"2\">\n",
       "        <strong>Channels</strong>\n",
       "    </th>\n",
       "</tr>\n",
       "\n",
       "\n",
       "<tr class=\"repr-element channels-3ef46833-4f4e-4c5d-a467-05f1dc4566d4 \">\n",
       "    <td class=\"repr-section-toggle-col\"></td>\n",
       "    <td>ECoG</td>\n",
       "    <td>\n",
       "        <button class=\"channel-names-btn\" onclick=\"alert('Good ECoG:\\n\\nLGA2, LGA3, LGA4, LGA5, LGA6, LGA7, LGA8, LGA9, LGA10, LGA11, LGA12, LGA13, LGA14, LGA15, LGA16, LGA17, LGA18, LGA19, LGA20, LGA21, LGA22, LGA23, LGA24, LGA25, LGA26, LGA27, LGA28, LGA29, LGA30, LGA31, LGA32, LGA33, LGA34, LGA35, LGA36, LGA37, LGA38, LGA39, LGA40, LGA41, LGA42, LGA43, LGA44, LGA45, LGA46, LGA47, LGA48, LGA49, LGA50, LGA51, LGA52, LGA53, LGA54, LGA55, LGA56, LGA57, LGA58, LGA59, LGA60, LGA61, LGA62, LGA63, LGA64, LGB65, LGB66, LGB67, LGB68, LGB69, LGB70, LGB71, LGB72, LGB73, LGB74, LGB75, LGB76, LGB77, LGB78, LGB79, LGB80, LGB81, LGB82, LGB83, LGB84, LGB85, LGB86, LGB87, LGB88, LGB89, LGB90, LGB91, LGB92, LGB93, LGB94, LGB95, LGB96, LGB97, LGB98, LGB99, LGB100, LGB101, LGB102, LGB103, LGB104, LGB105, LGB106, LGB107, LGB108, LGB109, LGB110, LGB111, LGB112, LGB113, LGB114, LGB115, LGB116, LGB117, LGB118, LGB119, LGB120, LGB121, LGB122, LGB123, LGB124, LGB125, LGB126, LGB127, LGB128')\" title=\"(Click to open in popup)&#13;&#13;LGA2, LGA3, LGA4, LGA5, LGA6, LGA7, LGA8, LGA9, LGA10, LGA11, LGA12, LGA13, LGA14, LGA15, LGA16, LGA17, LGA18, LGA19, LGA20, LGA21, LGA22, LGA23, LGA24, LGA25, LGA26, LGA27, LGA28, LGA29, LGA30, LGA31, LGA32, LGA33, LGA34, LGA35, LGA36, LGA37, LGA38, LGA39, LGA40, LGA41, LGA42, LGA43, LGA44, LGA45, LGA46, LGA47, LGA48, LGA49, LGA50, LGA51, LGA52, LGA53, LGA54, LGA55, LGA56, LGA57, LGA58, LGA59, LGA60, LGA61, LGA62, LGA63, LGA64, LGB65, LGB66, LGB67, LGB68, LGB69, LGB70, LGB71, LGB72, LGB73, LGB74, LGB75, LGB76, LGB77, LGB78, LGB79, LGB80, LGB81, LGB82, LGB83, LGB84, LGB85, LGB86, LGB87, LGB88, LGB89, LGB90, LGB91, LGB92, LGB93, LGB94, LGB95, LGB96, LGB97, LGB98, LGB99, LGB100, LGB101, LGB102, LGB103, LGB104, LGB105, LGB106, LGB107, LGB108, LGB109, LGB110, LGB111, LGB112, LGB113, LGB114, LGB115, LGB116, LGB117, LGB118, LGB119, LGB120, LGB121, LGB122, LGB123, LGB124, LGB125, LGB126, LGB127, LGB128\">\n",
       "            127\n",
       "        </button>\n",
       "\n",
       "        \n",
       "    </td>\n",
       "</tr>\n",
       "\n",
       "\n",
       "<tr class=\"repr-element channels-3ef46833-4f4e-4c5d-a467-05f1dc4566d4 \">\n",
       "    <td class=\"repr-section-toggle-col\"></td>\n",
       "    <td>Head & sensor digitization</td>\n",
       "    \n",
       "    <td>238 points</td>\n",
       "    \n",
       "</tr>\n",
       "    \n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "<tr class=\"repr-section-header filters-15417875-b7f7-44ad-ad18-28278d2a6dcb\"  title=\"Hide section\" \n",
       "    onclick=\"toggleVisibility('filters-15417875-b7f7-44ad-ad18-28278d2a6dcb')\">\n",
       "    <th class=\"repr-section-toggle-col\">\n",
       "        <button>\n",
       "            \n",
       "            <span class=\"collapse-uncollapse-caret\"></span>\n",
       "        </button>\n",
       "    </th>\n",
       "    <th colspan=\"2\">\n",
       "        <strong>Filters</strong>\n",
       "    </th>\n",
       "</tr>\n",
       "\n",
       "<tr class=\"repr-element filters-15417875-b7f7-44ad-ad18-28278d2a6dcb \">\n",
       "    <td class=\"repr-section-toggle-col\"></td>\n",
       "    <td>Highpass</td>\n",
       "    <td>70.00 Hz</td>\n",
       "</tr>\n",
       "\n",
       "\n",
       "<tr class=\"repr-element filters-15417875-b7f7-44ad-ad18-28278d2a6dcb \">\n",
       "    <td class=\"repr-section-toggle-col\"></td>\n",
       "    <td>Lowpass</td>\n",
       "    <td>200.00 Hz</td>\n",
       "</tr>\n",
       "\n",
       "\n",
       "</table>"
      ],
      "text/plain": [
       "<Raw | sub-03_task-podcast_desc-highgamma_ieeg.fif, 127 x 921600 (1800.0 s), ~187 kB, data not loaded>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "file_path = BIDSPath(root=f\"{bids_root}derivatives/ecogprep\",\n",
    "                    subject=\"03\", task=\"podcast\", datatype=\"ieeg\", description=\"highgamma\",\n",
    "                    suffix=\"ieeg\", extension=\".fif\")\n",
    "print(f\"File path within the dataset: {file_path}\")\n",
    "\n",
    "# You only need to run this if using Colab (i.e. if you did not set bids_root to a local directory)\n",
    "if not len(bids_root):\n",
    "    !wget -nc https://s3.amazonaws.com/openneuro.org/ds005574/$file_path\n",
    "    file_path = file_path.basename\n",
    "\n",
    "raw = mne.io.read_raw_fif(file_path, verbose=False)\n",
    "picks = mne.pick_channels_regexp(raw.ch_names, \"LG[AB]*\")\n",
    "raw = raw.pick(picks)\n",
    "raw"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will map the start information (in seconds) of each word in the dataframe onto the brain signal data by multiplying by the sampling rate. Here the first column of `events` mark the start of each word on the brain signal data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5134, 3)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "events = np.zeros((len(df_merged), 3), dtype=int)\n",
    "events[:, 0] = (df_merged.start_x * raw.info['sfreq']).astype(int)\n",
    "events.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then we'll take advantage of MNE's tools for creating epochs around stimulus events, which here are the starts (onsets) of each word, to visualize brain signal that respond to word onsets. Here, we take a fixed-width window ranging from -2 seconds to +2 seconds relative to word onset. Since the sampling rate is 512 Hz (512 samples per second), we have 2049 lags total. The ECoG data is a numpy array with the shape of (number of words * number of ECoG electrodes * number of lags)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Not setting metadata\n",
      "5134 matching events found\n",
      "No baseline correction applied\n",
      "Loading data for 5134 events and 2049 original time points ...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_3099603/3482921024.py:1: RuntimeWarning: The events passed to the Epochs constructor are not chronologically ordered.\n",
      "  epochs = mne.Epochs(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6 bad epochs dropped\n",
      "Epochs object has a shape of: (5128, 127, 2049)\n"
     ]
    }
   ],
   "source": [
    "epochs = mne.Epochs(\n",
    "    raw,\n",
    "    events,\n",
    "    tmin=-2.0,\n",
    "    tmax=2.0,\n",
    "    baseline=None,\n",
    "    proj=False,\n",
    "    event_id=None,\n",
    "    preload=True,\n",
    "    event_repeated=\"merge\",\n",
    ")\n",
    "print(f\"Epochs object has a shape of: {epochs._data.shape}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we'll downsample the temporal resolution to 32 Hz, which reduces the number of lags to 32 * 4 = 128.\n",
    "\n",
    "<div class=\"alert alert-info\">\n",
    "\n",
    "**Note**\n",
    "\n",
    "This code block may take ~3 minutes to run.\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epochs object has a shape of: (5128, 127, 128)\n"
     ]
    }
   ],
   "source": [
    "epochs = epochs.resample(sfreq=32, npad='auto', method='fft', window='hamming')\n",
    "print(f\"Epochs object has a shape of: {epochs._data.shape}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setting up feature and brain data\n",
    "\n",
    "Now we have both the features and the ECoG data ready. We plan to fit encoding models at each electrode and for each lag, so we'll reshape our target matrix `Y` to horizontally stack both electrodes and lags along the second dimension."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ECoG data matrix shape: (5128, 16256)\n"
     ]
    }
   ],
   "source": [
    "epochs_data = epochs.get_data(copy=True)\n",
    "epochs_data = epochs_data.reshape(len(epochs), -1)\n",
    "print(f\"ECoG data matrix shape: {epochs_data.shape}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will also align our features with the ECoG data. At the same time, we need to construct one design matrix as predictor variables in our encoding model. We do this by horizontally stacking the embeddings together, to get one wide model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Combined model embeddings size: (5128, 3648)\n"
     ]
    }
   ],
   "source": [
    "gpt2_embeddings = aligned_gpt_embeddings[epochs.selection]\n",
    "whisper_embeddings = aligned_whisper_embeddings[epochs.selection]\n",
    "input_embeddings = np.hstack((gpt2_embeddings, whisper_embeddings))\n",
    "\n",
    "print(f\"Combined model embeddings size: {input_embeddings.shape}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will change the float precision to float32 for all data to take advantage of the GPU memory and computational speed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((5128, 3648), (5128, 16256))"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = input_embeddings\n",
    "Y = epochs_data\n",
    "\n",
    "if \"torch\" in get_backend().__name__:\n",
    "    X = X.astype(np.float32)\n",
    "    Y = Y.astype(np.float32)\n",
    "\n",
    "X.shape, Y.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Building the encoding model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This section closely follows the Himalaya [banded ridge tutorial](https://gallantlab.org/voxelwise_tutorials/_auto_examples/shortclips/06_plot_banded_ridge_model.html) to construct an encoding model pipeline prior to fitting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-1 {\n",
       "  /* Definition of color scheme common for light and dark mode */\n",
       "  --sklearn-color-text: black;\n",
       "  --sklearn-color-line: gray;\n",
       "  /* Definition of color scheme for unfitted estimators */\n",
       "  --sklearn-color-unfitted-level-0: #fff5e6;\n",
       "  --sklearn-color-unfitted-level-1: #f6e4d2;\n",
       "  --sklearn-color-unfitted-level-2: #ffe0b3;\n",
       "  --sklearn-color-unfitted-level-3: chocolate;\n",
       "  /* Definition of color scheme for fitted estimators */\n",
       "  --sklearn-color-fitted-level-0: #f0f8ff;\n",
       "  --sklearn-color-fitted-level-1: #d4ebff;\n",
       "  --sklearn-color-fitted-level-2: #b3dbfd;\n",
       "  --sklearn-color-fitted-level-3: cornflowerblue;\n",
       "\n",
       "  /* Specific color for light theme */\n",
       "  --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
       "  --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, white)));\n",
       "  --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
       "  --sklearn-color-icon: #696969;\n",
       "\n",
       "  @media (prefers-color-scheme: dark) {\n",
       "    /* Redefinition of color scheme for dark theme */\n",
       "    --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
       "    --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, #111)));\n",
       "    --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
       "    --sklearn-color-icon: #878787;\n",
       "  }\n",
       "}\n",
       "\n",
       "#sk-container-id-1 {\n",
       "  color: var(--sklearn-color-text);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 pre {\n",
       "  padding: 0;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 input.sk-hidden--visually {\n",
       "  border: 0;\n",
       "  clip: rect(1px 1px 1px 1px);\n",
       "  clip: rect(1px, 1px, 1px, 1px);\n",
       "  height: 1px;\n",
       "  margin: -1px;\n",
       "  overflow: hidden;\n",
       "  padding: 0;\n",
       "  position: absolute;\n",
       "  width: 1px;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-dashed-wrapped {\n",
       "  border: 1px dashed var(--sklearn-color-line);\n",
       "  margin: 0 0.4em 0.5em 0.4em;\n",
       "  box-sizing: border-box;\n",
       "  padding-bottom: 0.4em;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-container {\n",
       "  /* jupyter's `normalize.less` sets `[hidden] { display: none; }`\n",
       "     but bootstrap.min.css set `[hidden] { display: none !important; }`\n",
       "     so we also need the `!important` here to be able to override the\n",
       "     default hidden behavior on the sphinx rendered scikit-learn.org.\n",
       "     See: https://github.com/scikit-learn/scikit-learn/issues/21755 */\n",
       "  display: inline-block !important;\n",
       "  position: relative;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-text-repr-fallback {\n",
       "  display: none;\n",
       "}\n",
       "\n",
       "div.sk-parallel-item,\n",
       "div.sk-serial,\n",
       "div.sk-item {\n",
       "  /* draw centered vertical line to link estimators */\n",
       "  background-image: linear-gradient(var(--sklearn-color-text-on-default-background), var(--sklearn-color-text-on-default-background));\n",
       "  background-size: 2px 100%;\n",
       "  background-repeat: no-repeat;\n",
       "  background-position: center center;\n",
       "}\n",
       "\n",
       "/* Parallel-specific style estimator block */\n",
       "\n",
       "#sk-container-id-1 div.sk-parallel-item::after {\n",
       "  content: \"\";\n",
       "  width: 100%;\n",
       "  border-bottom: 2px solid var(--sklearn-color-text-on-default-background);\n",
       "  flex-grow: 1;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-parallel {\n",
       "  display: flex;\n",
       "  align-items: stretch;\n",
       "  justify-content: center;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  position: relative;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-parallel-item {\n",
       "  display: flex;\n",
       "  flex-direction: column;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-parallel-item:first-child::after {\n",
       "  align-self: flex-end;\n",
       "  width: 50%;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-parallel-item:last-child::after {\n",
       "  align-self: flex-start;\n",
       "  width: 50%;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-parallel-item:only-child::after {\n",
       "  width: 0;\n",
       "}\n",
       "\n",
       "/* Serial-specific style estimator block */\n",
       "\n",
       "#sk-container-id-1 div.sk-serial {\n",
       "  display: flex;\n",
       "  flex-direction: column;\n",
       "  align-items: center;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  padding-right: 1em;\n",
       "  padding-left: 1em;\n",
       "}\n",
       "\n",
       "\n",
       "/* Toggleable style: style used for estimator/Pipeline/ColumnTransformer box that is\n",
       "clickable and can be expanded/collapsed.\n",
       "- Pipeline and ColumnTransformer use this feature and define the default style\n",
       "- Estimators will overwrite some part of the style using the `sk-estimator` class\n",
       "*/\n",
       "\n",
       "/* Pipeline and ColumnTransformer style (default) */\n",
       "\n",
       "#sk-container-id-1 div.sk-toggleable {\n",
       "  /* Default theme specific background. It is overwritten whether we have a\n",
       "  specific estimator or a Pipeline/ColumnTransformer */\n",
       "  background-color: var(--sklearn-color-background);\n",
       "}\n",
       "\n",
       "/* Toggleable label */\n",
       "#sk-container-id-1 label.sk-toggleable__label {\n",
       "  cursor: pointer;\n",
       "  display: block;\n",
       "  width: 100%;\n",
       "  margin-bottom: 0;\n",
       "  padding: 0.5em;\n",
       "  box-sizing: border-box;\n",
       "  text-align: center;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 label.sk-toggleable__label-arrow:before {\n",
       "  /* Arrow on the left of the label */\n",
       "  content: \"▸\";\n",
       "  float: left;\n",
       "  margin-right: 0.25em;\n",
       "  color: var(--sklearn-color-icon);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {\n",
       "  color: var(--sklearn-color-text);\n",
       "}\n",
       "\n",
       "/* Toggleable content - dropdown */\n",
       "\n",
       "#sk-container-id-1 div.sk-toggleable__content {\n",
       "  max-height: 0;\n",
       "  max-width: 0;\n",
       "  overflow: hidden;\n",
       "  text-align: left;\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-toggleable__content.fitted {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-toggleable__content pre {\n",
       "  margin: 0.2em;\n",
       "  border-radius: 0.25em;\n",
       "  color: var(--sklearn-color-text);\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-toggleable__content.fitted pre {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {\n",
       "  /* Expand drop-down */\n",
       "  max-height: 200px;\n",
       "  max-width: 100%;\n",
       "  overflow: auto;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {\n",
       "  content: \"▾\";\n",
       "}\n",
       "\n",
       "/* Pipeline/ColumnTransformer-specific style */\n",
       "\n",
       "#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-label.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Estimator-specific style */\n",
       "\n",
       "/* Colorize estimator box */\n",
       "#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-estimator.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-label label.sk-toggleable__label,\n",
       "#sk-container-id-1 div.sk-label label {\n",
       "  /* The background is the default theme color */\n",
       "  color: var(--sklearn-color-text-on-default-background);\n",
       "}\n",
       "\n",
       "/* On hover, darken the color of the background */\n",
       "#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "/* Label box, darken color on hover, fitted */\n",
       "#sk-container-id-1 div.sk-label.fitted:hover label.sk-toggleable__label.fitted {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Estimator label */\n",
       "\n",
       "#sk-container-id-1 div.sk-label label {\n",
       "  font-family: monospace;\n",
       "  font-weight: bold;\n",
       "  display: inline-block;\n",
       "  line-height: 1.2em;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-label-container {\n",
       "  text-align: center;\n",
       "}\n",
       "\n",
       "/* Estimator-specific */\n",
       "#sk-container-id-1 div.sk-estimator {\n",
       "  font-family: monospace;\n",
       "  border: 1px dotted var(--sklearn-color-border-box);\n",
       "  border-radius: 0.25em;\n",
       "  box-sizing: border-box;\n",
       "  margin-bottom: 0.5em;\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-estimator.fitted {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "/* on hover */\n",
       "#sk-container-id-1 div.sk-estimator:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-1 div.sk-estimator.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Specification for estimator info (e.g. \"i\" and \"?\") */\n",
       "\n",
       "/* Common style for \"i\" and \"?\" */\n",
       "\n",
       ".sk-estimator-doc-link,\n",
       "a:link.sk-estimator-doc-link,\n",
       "a:visited.sk-estimator-doc-link {\n",
       "  float: right;\n",
       "  font-size: smaller;\n",
       "  line-height: 1em;\n",
       "  font-family: monospace;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  border-radius: 1em;\n",
       "  height: 1em;\n",
       "  width: 1em;\n",
       "  text-decoration: none !important;\n",
       "  margin-left: 1ex;\n",
       "  /* unfitted */\n",
       "  border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-unfitted-level-1);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link.fitted,\n",
       "a:link.sk-estimator-doc-link.fitted,\n",
       "a:visited.sk-estimator-doc-link.fitted {\n",
       "  /* fitted */\n",
       "  border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-fitted-level-1);\n",
       "}\n",
       "\n",
       "/* On hover */\n",
       "div.sk-estimator:hover .sk-estimator-doc-link:hover,\n",
       ".sk-estimator-doc-link:hover,\n",
       "div.sk-label-container:hover .sk-estimator-doc-link:hover,\n",
       ".sk-estimator-doc-link:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "div.sk-estimator.fitted:hover .sk-estimator-doc-link.fitted:hover,\n",
       ".sk-estimator-doc-link.fitted:hover,\n",
       "div.sk-label-container:hover .sk-estimator-doc-link.fitted:hover,\n",
       ".sk-estimator-doc-link.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "/* Span, style for the box shown on hovering the info icon */\n",
       ".sk-estimator-doc-link span {\n",
       "  display: none;\n",
       "  z-index: 9999;\n",
       "  position: relative;\n",
       "  font-weight: normal;\n",
       "  right: .2ex;\n",
       "  padding: .5ex;\n",
       "  margin: .5ex;\n",
       "  width: min-content;\n",
       "  min-width: 20ex;\n",
       "  max-width: 50ex;\n",
       "  color: var(--sklearn-color-text);\n",
       "  box-shadow: 2pt 2pt 4pt #999;\n",
       "  /* unfitted */\n",
       "  background: var(--sklearn-color-unfitted-level-0);\n",
       "  border: .5pt solid var(--sklearn-color-unfitted-level-3);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link.fitted span {\n",
       "  /* fitted */\n",
       "  background: var(--sklearn-color-fitted-level-0);\n",
       "  border: var(--sklearn-color-fitted-level-3);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link:hover span {\n",
       "  display: block;\n",
       "}\n",
       "\n",
       "/* \"?\"-specific style due to the `<a>` HTML tag */\n",
       "\n",
       "#sk-container-id-1 a.estimator_doc_link {\n",
       "  float: right;\n",
       "  font-size: 1rem;\n",
       "  line-height: 1em;\n",
       "  font-family: monospace;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  border-radius: 1rem;\n",
       "  height: 1rem;\n",
       "  width: 1rem;\n",
       "  text-decoration: none;\n",
       "  /* unfitted */\n",
       "  color: var(--sklearn-color-unfitted-level-1);\n",
       "  border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 a.estimator_doc_link.fitted {\n",
       "  /* fitted */\n",
       "  border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-fitted-level-1);\n",
       "}\n",
       "\n",
       "/* On hover */\n",
       "#sk-container-id-1 a.estimator_doc_link:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "#sk-container-id-1 a.estimator_doc_link.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-3);\n",
       "}\n",
       "</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;standardscaler&#x27;, StandardScaler(with_std=False)),\n",
       "                (&#x27;kernelizer&#x27;, Kernelizer())])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" ><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">&nbsp;&nbsp;Pipeline<a class=\"sk-estimator-doc-link \" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.5/modules/generated/sklearn.pipeline.Pipeline.html\">?<span>Documentation for Pipeline</span></a><span class=\"sk-estimator-doc-link \">i<span>Not fitted</span></span></label><div class=\"sk-toggleable__content \"><pre>Pipeline(steps=[(&#x27;standardscaler&#x27;, StandardScaler(with_std=False)),\n",
       "                (&#x27;kernelizer&#x27;, Kernelizer())])</pre></div> </div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" ><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">&nbsp;StandardScaler<a class=\"sk-estimator-doc-link \" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.5/modules/generated/sklearn.preprocessing.StandardScaler.html\">?<span>Documentation for StandardScaler</span></a></label><div class=\"sk-toggleable__content \"><pre>StandardScaler(with_std=False)</pre></div> </div></div><div class=\"sk-item\"><div class=\"sk-estimator  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" ><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">Kernelizer</label><div class=\"sk-toggleable__content \"><pre>Kernelizer()</pre></div> </div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('standardscaler', StandardScaler(with_std=False)),\n",
       "                ('kernelizer', Kernelizer())])"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preprocess_pipeline = make_pipeline(\n",
    "    StandardScaler(with_mean=True, with_std=False),\n",
    "    Kernelizer(kernel=\"linear\"),\n",
    ")\n",
    "preprocess_pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-2 {\n",
       "  /* Definition of color scheme common for light and dark mode */\n",
       "  --sklearn-color-text: black;\n",
       "  --sklearn-color-line: gray;\n",
       "  /* Definition of color scheme for unfitted estimators */\n",
       "  --sklearn-color-unfitted-level-0: #fff5e6;\n",
       "  --sklearn-color-unfitted-level-1: #f6e4d2;\n",
       "  --sklearn-color-unfitted-level-2: #ffe0b3;\n",
       "  --sklearn-color-unfitted-level-3: chocolate;\n",
       "  /* Definition of color scheme for fitted estimators */\n",
       "  --sklearn-color-fitted-level-0: #f0f8ff;\n",
       "  --sklearn-color-fitted-level-1: #d4ebff;\n",
       "  --sklearn-color-fitted-level-2: #b3dbfd;\n",
       "  --sklearn-color-fitted-level-3: cornflowerblue;\n",
       "\n",
       "  /* Specific color for light theme */\n",
       "  --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
       "  --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, white)));\n",
       "  --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
       "  --sklearn-color-icon: #696969;\n",
       "\n",
       "  @media (prefers-color-scheme: dark) {\n",
       "    /* Redefinition of color scheme for dark theme */\n",
       "    --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
       "    --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, #111)));\n",
       "    --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
       "    --sklearn-color-icon: #878787;\n",
       "  }\n",
       "}\n",
       "\n",
       "#sk-container-id-2 {\n",
       "  color: var(--sklearn-color-text);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 pre {\n",
       "  padding: 0;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 input.sk-hidden--visually {\n",
       "  border: 0;\n",
       "  clip: rect(1px 1px 1px 1px);\n",
       "  clip: rect(1px, 1px, 1px, 1px);\n",
       "  height: 1px;\n",
       "  margin: -1px;\n",
       "  overflow: hidden;\n",
       "  padding: 0;\n",
       "  position: absolute;\n",
       "  width: 1px;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-dashed-wrapped {\n",
       "  border: 1px dashed var(--sklearn-color-line);\n",
       "  margin: 0 0.4em 0.5em 0.4em;\n",
       "  box-sizing: border-box;\n",
       "  padding-bottom: 0.4em;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-container {\n",
       "  /* jupyter's `normalize.less` sets `[hidden] { display: none; }`\n",
       "     but bootstrap.min.css set `[hidden] { display: none !important; }`\n",
       "     so we also need the `!important` here to be able to override the\n",
       "     default hidden behavior on the sphinx rendered scikit-learn.org.\n",
       "     See: https://github.com/scikit-learn/scikit-learn/issues/21755 */\n",
       "  display: inline-block !important;\n",
       "  position: relative;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-text-repr-fallback {\n",
       "  display: none;\n",
       "}\n",
       "\n",
       "div.sk-parallel-item,\n",
       "div.sk-serial,\n",
       "div.sk-item {\n",
       "  /* draw centered vertical line to link estimators */\n",
       "  background-image: linear-gradient(var(--sklearn-color-text-on-default-background), var(--sklearn-color-text-on-default-background));\n",
       "  background-size: 2px 100%;\n",
       "  background-repeat: no-repeat;\n",
       "  background-position: center center;\n",
       "}\n",
       "\n",
       "/* Parallel-specific style estimator block */\n",
       "\n",
       "#sk-container-id-2 div.sk-parallel-item::after {\n",
       "  content: \"\";\n",
       "  width: 100%;\n",
       "  border-bottom: 2px solid var(--sklearn-color-text-on-default-background);\n",
       "  flex-grow: 1;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-parallel {\n",
       "  display: flex;\n",
       "  align-items: stretch;\n",
       "  justify-content: center;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  position: relative;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-parallel-item {\n",
       "  display: flex;\n",
       "  flex-direction: column;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-parallel-item:first-child::after {\n",
       "  align-self: flex-end;\n",
       "  width: 50%;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-parallel-item:last-child::after {\n",
       "  align-self: flex-start;\n",
       "  width: 50%;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-parallel-item:only-child::after {\n",
       "  width: 0;\n",
       "}\n",
       "\n",
       "/* Serial-specific style estimator block */\n",
       "\n",
       "#sk-container-id-2 div.sk-serial {\n",
       "  display: flex;\n",
       "  flex-direction: column;\n",
       "  align-items: center;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  padding-right: 1em;\n",
       "  padding-left: 1em;\n",
       "}\n",
       "\n",
       "\n",
       "/* Toggleable style: style used for estimator/Pipeline/ColumnTransformer box that is\n",
       "clickable and can be expanded/collapsed.\n",
       "- Pipeline and ColumnTransformer use this feature and define the default style\n",
       "- Estimators will overwrite some part of the style using the `sk-estimator` class\n",
       "*/\n",
       "\n",
       "/* Pipeline and ColumnTransformer style (default) */\n",
       "\n",
       "#sk-container-id-2 div.sk-toggleable {\n",
       "  /* Default theme specific background. It is overwritten whether we have a\n",
       "  specific estimator or a Pipeline/ColumnTransformer */\n",
       "  background-color: var(--sklearn-color-background);\n",
       "}\n",
       "\n",
       "/* Toggleable label */\n",
       "#sk-container-id-2 label.sk-toggleable__label {\n",
       "  cursor: pointer;\n",
       "  display: block;\n",
       "  width: 100%;\n",
       "  margin-bottom: 0;\n",
       "  padding: 0.5em;\n",
       "  box-sizing: border-box;\n",
       "  text-align: center;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 label.sk-toggleable__label-arrow:before {\n",
       "  /* Arrow on the left of the label */\n",
       "  content: \"▸\";\n",
       "  float: left;\n",
       "  margin-right: 0.25em;\n",
       "  color: var(--sklearn-color-icon);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {\n",
       "  color: var(--sklearn-color-text);\n",
       "}\n",
       "\n",
       "/* Toggleable content - dropdown */\n",
       "\n",
       "#sk-container-id-2 div.sk-toggleable__content {\n",
       "  max-height: 0;\n",
       "  max-width: 0;\n",
       "  overflow: hidden;\n",
       "  text-align: left;\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-toggleable__content.fitted {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-toggleable__content pre {\n",
       "  margin: 0.2em;\n",
       "  border-radius: 0.25em;\n",
       "  color: var(--sklearn-color-text);\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-toggleable__content.fitted pre {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {\n",
       "  /* Expand drop-down */\n",
       "  max-height: 200px;\n",
       "  max-width: 100%;\n",
       "  overflow: auto;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {\n",
       "  content: \"▾\";\n",
       "}\n",
       "\n",
       "/* Pipeline/ColumnTransformer-specific style */\n",
       "\n",
       "#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-label.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Estimator-specific style */\n",
       "\n",
       "/* Colorize estimator box */\n",
       "#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-estimator.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-label label.sk-toggleable__label,\n",
       "#sk-container-id-2 div.sk-label label {\n",
       "  /* The background is the default theme color */\n",
       "  color: var(--sklearn-color-text-on-default-background);\n",
       "}\n",
       "\n",
       "/* On hover, darken the color of the background */\n",
       "#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "/* Label box, darken color on hover, fitted */\n",
       "#sk-container-id-2 div.sk-label.fitted:hover label.sk-toggleable__label.fitted {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Estimator label */\n",
       "\n",
       "#sk-container-id-2 div.sk-label label {\n",
       "  font-family: monospace;\n",
       "  font-weight: bold;\n",
       "  display: inline-block;\n",
       "  line-height: 1.2em;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-label-container {\n",
       "  text-align: center;\n",
       "}\n",
       "\n",
       "/* Estimator-specific */\n",
       "#sk-container-id-2 div.sk-estimator {\n",
       "  font-family: monospace;\n",
       "  border: 1px dotted var(--sklearn-color-border-box);\n",
       "  border-radius: 0.25em;\n",
       "  box-sizing: border-box;\n",
       "  margin-bottom: 0.5em;\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-estimator.fitted {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "/* on hover */\n",
       "#sk-container-id-2 div.sk-estimator:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-estimator.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Specification for estimator info (e.g. \"i\" and \"?\") */\n",
       "\n",
       "/* Common style for \"i\" and \"?\" */\n",
       "\n",
       ".sk-estimator-doc-link,\n",
       "a:link.sk-estimator-doc-link,\n",
       "a:visited.sk-estimator-doc-link {\n",
       "  float: right;\n",
       "  font-size: smaller;\n",
       "  line-height: 1em;\n",
       "  font-family: monospace;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  border-radius: 1em;\n",
       "  height: 1em;\n",
       "  width: 1em;\n",
       "  text-decoration: none !important;\n",
       "  margin-left: 1ex;\n",
       "  /* unfitted */\n",
       "  border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-unfitted-level-1);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link.fitted,\n",
       "a:link.sk-estimator-doc-link.fitted,\n",
       "a:visited.sk-estimator-doc-link.fitted {\n",
       "  /* fitted */\n",
       "  border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-fitted-level-1);\n",
       "}\n",
       "\n",
       "/* On hover */\n",
       "div.sk-estimator:hover .sk-estimator-doc-link:hover,\n",
       ".sk-estimator-doc-link:hover,\n",
       "div.sk-label-container:hover .sk-estimator-doc-link:hover,\n",
       ".sk-estimator-doc-link:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "div.sk-estimator.fitted:hover .sk-estimator-doc-link.fitted:hover,\n",
       ".sk-estimator-doc-link.fitted:hover,\n",
       "div.sk-label-container:hover .sk-estimator-doc-link.fitted:hover,\n",
       ".sk-estimator-doc-link.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "/* Span, style for the box shown on hovering the info icon */\n",
       ".sk-estimator-doc-link span {\n",
       "  display: none;\n",
       "  z-index: 9999;\n",
       "  position: relative;\n",
       "  font-weight: normal;\n",
       "  right: .2ex;\n",
       "  padding: .5ex;\n",
       "  margin: .5ex;\n",
       "  width: min-content;\n",
       "  min-width: 20ex;\n",
       "  max-width: 50ex;\n",
       "  color: var(--sklearn-color-text);\n",
       "  box-shadow: 2pt 2pt 4pt #999;\n",
       "  /* unfitted */\n",
       "  background: var(--sklearn-color-unfitted-level-0);\n",
       "  border: .5pt solid var(--sklearn-color-unfitted-level-3);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link.fitted span {\n",
       "  /* fitted */\n",
       "  background: var(--sklearn-color-fitted-level-0);\n",
       "  border: var(--sklearn-color-fitted-level-3);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link:hover span {\n",
       "  display: block;\n",
       "}\n",
       "\n",
       "/* \"?\"-specific style due to the `<a>` HTML tag */\n",
       "\n",
       "#sk-container-id-2 a.estimator_doc_link {\n",
       "  float: right;\n",
       "  font-size: 1rem;\n",
       "  line-height: 1em;\n",
       "  font-family: monospace;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  border-radius: 1rem;\n",
       "  height: 1rem;\n",
       "  width: 1rem;\n",
       "  text-decoration: none;\n",
       "  /* unfitted */\n",
       "  color: var(--sklearn-color-unfitted-level-1);\n",
       "  border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 a.estimator_doc_link.fitted {\n",
       "  /* fitted */\n",
       "  border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-fitted-level-1);\n",
       "}\n",
       "\n",
       "/* On hover */\n",
       "#sk-container-id-2 a.estimator_doc_link:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 a.estimator_doc_link.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-3);\n",
       "}\n",
       "</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>ColumnKernelizer(transformers=[(&#x27;gpt2&#x27;,\n",
       "                                Pipeline(steps=[(&#x27;standardscaler&#x27;,\n",
       "                                                 StandardScaler(with_std=False)),\n",
       "                                                (&#x27;kernelizer&#x27;, Kernelizer())]),\n",
       "                                slice(0, 1600, None)),\n",
       "                               (&#x27;whisper&#x27;,\n",
       "                                Pipeline(steps=[(&#x27;standardscaler&#x27;,\n",
       "                                                 StandardScaler(with_std=False)),\n",
       "                                                (&#x27;kernelizer&#x27;, Kernelizer())]),\n",
       "                                slice(1600, None, None))])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-4\" type=\"checkbox\" ><label for=\"sk-estimator-id-4\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">&nbsp;ColumnKernelizer<span class=\"sk-estimator-doc-link \">i<span>Not fitted</span></span></label><div class=\"sk-toggleable__content \"><pre>ColumnKernelizer(transformers=[(&#x27;gpt2&#x27;,\n",
       "                                Pipeline(steps=[(&#x27;standardscaler&#x27;,\n",
       "                                                 StandardScaler(with_std=False)),\n",
       "                                                (&#x27;kernelizer&#x27;, Kernelizer())]),\n",
       "                                slice(0, 1600, None)),\n",
       "                               (&#x27;whisper&#x27;,\n",
       "                                Pipeline(steps=[(&#x27;standardscaler&#x27;,\n",
       "                                                 StandardScaler(with_std=False)),\n",
       "                                                (&#x27;kernelizer&#x27;, Kernelizer())]),\n",
       "                                slice(1600, None, None))])</pre></div> </div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-5\" type=\"checkbox\" ><label for=\"sk-estimator-id-5\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">gpt2</label><div class=\"sk-toggleable__content \"><pre>slice(0, 1600, None)</pre></div> </div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-6\" type=\"checkbox\" ><label for=\"sk-estimator-id-6\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">&nbsp;StandardScaler<a class=\"sk-estimator-doc-link \" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.5/modules/generated/sklearn.preprocessing.StandardScaler.html\">?<span>Documentation for StandardScaler</span></a></label><div class=\"sk-toggleable__content \"><pre>StandardScaler(with_std=False)</pre></div> </div></div><div class=\"sk-item\"><div class=\"sk-estimator  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-7\" type=\"checkbox\" ><label for=\"sk-estimator-id-7\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">Kernelizer</label><div class=\"sk-toggleable__content \"><pre>Kernelizer()</pre></div> </div></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-8\" type=\"checkbox\" ><label for=\"sk-estimator-id-8\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">whisper</label><div class=\"sk-toggleable__content \"><pre>slice(1600, None, None)</pre></div> </div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-9\" type=\"checkbox\" ><label for=\"sk-estimator-id-9\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">&nbsp;StandardScaler<a class=\"sk-estimator-doc-link \" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.5/modules/generated/sklearn.preprocessing.StandardScaler.html\">?<span>Documentation for StandardScaler</span></a></label><div class=\"sk-toggleable__content \"><pre>StandardScaler(with_std=False)</pre></div> </div></div><div class=\"sk-item\"><div class=\"sk-estimator  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-10\" type=\"checkbox\" ><label for=\"sk-estimator-id-10\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">Kernelizer</label><div class=\"sk-toggleable__content \"><pre>Kernelizer()</pre></div> </div></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "ColumnKernelizer(transformers=[('gpt2',\n",
       "                                Pipeline(steps=[('standardscaler',\n",
       "                                                 StandardScaler(with_std=False)),\n",
       "                                                ('kernelizer', Kernelizer())]),\n",
       "                                slice(0, 1600, None)),\n",
       "                               ('whisper',\n",
       "                                Pipeline(steps=[('standardscaler',\n",
       "                                                 StandardScaler(with_std=False)),\n",
       "                                                ('kernelizer', Kernelizer())]),\n",
       "                                slice(1600, None, None))])"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature_names = ['gpt2', 'whisper']\n",
    "slices = [slice(0, 1600), slice(1600, None)]\n",
    "\n",
    "kernelizers_tuples = [(name, preprocess_pipeline, slice_)\n",
    "                      for name, slice_ in zip(feature_names, slices)]\n",
    "column_kernelizer = ColumnKernelizer(kernelizers_tuples)\n",
    "column_kernelizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-3 {\n",
       "  /* Definition of color scheme common for light and dark mode */\n",
       "  --sklearn-color-text: black;\n",
       "  --sklearn-color-line: gray;\n",
       "  /* Definition of color scheme for unfitted estimators */\n",
       "  --sklearn-color-unfitted-level-0: #fff5e6;\n",
       "  --sklearn-color-unfitted-level-1: #f6e4d2;\n",
       "  --sklearn-color-unfitted-level-2: #ffe0b3;\n",
       "  --sklearn-color-unfitted-level-3: chocolate;\n",
       "  /* Definition of color scheme for fitted estimators */\n",
       "  --sklearn-color-fitted-level-0: #f0f8ff;\n",
       "  --sklearn-color-fitted-level-1: #d4ebff;\n",
       "  --sklearn-color-fitted-level-2: #b3dbfd;\n",
       "  --sklearn-color-fitted-level-3: cornflowerblue;\n",
       "\n",
       "  /* Specific color for light theme */\n",
       "  --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
       "  --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, white)));\n",
       "  --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
       "  --sklearn-color-icon: #696969;\n",
       "\n",
       "  @media (prefers-color-scheme: dark) {\n",
       "    /* Redefinition of color scheme for dark theme */\n",
       "    --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
       "    --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, #111)));\n",
       "    --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
       "    --sklearn-color-icon: #878787;\n",
       "  }\n",
       "}\n",
       "\n",
       "#sk-container-id-3 {\n",
       "  color: var(--sklearn-color-text);\n",
       "}\n",
       "\n",
       "#sk-container-id-3 pre {\n",
       "  padding: 0;\n",
       "}\n",
       "\n",
       "#sk-container-id-3 input.sk-hidden--visually {\n",
       "  border: 0;\n",
       "  clip: rect(1px 1px 1px 1px);\n",
       "  clip: rect(1px, 1px, 1px, 1px);\n",
       "  height: 1px;\n",
       "  margin: -1px;\n",
       "  overflow: hidden;\n",
       "  padding: 0;\n",
       "  position: absolute;\n",
       "  width: 1px;\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-dashed-wrapped {\n",
       "  border: 1px dashed var(--sklearn-color-line);\n",
       "  margin: 0 0.4em 0.5em 0.4em;\n",
       "  box-sizing: border-box;\n",
       "  padding-bottom: 0.4em;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-container {\n",
       "  /* jupyter's `normalize.less` sets `[hidden] { display: none; }`\n",
       "     but bootstrap.min.css set `[hidden] { display: none !important; }`\n",
       "     so we also need the `!important` here to be able to override the\n",
       "     default hidden behavior on the sphinx rendered scikit-learn.org.\n",
       "     See: https://github.com/scikit-learn/scikit-learn/issues/21755 */\n",
       "  display: inline-block !important;\n",
       "  position: relative;\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-text-repr-fallback {\n",
       "  display: none;\n",
       "}\n",
       "\n",
       "div.sk-parallel-item,\n",
       "div.sk-serial,\n",
       "div.sk-item {\n",
       "  /* draw centered vertical line to link estimators */\n",
       "  background-image: linear-gradient(var(--sklearn-color-text-on-default-background), var(--sklearn-color-text-on-default-background));\n",
       "  background-size: 2px 100%;\n",
       "  background-repeat: no-repeat;\n",
       "  background-position: center center;\n",
       "}\n",
       "\n",
       "/* Parallel-specific style estimator block */\n",
       "\n",
       "#sk-container-id-3 div.sk-parallel-item::after {\n",
       "  content: \"\";\n",
       "  width: 100%;\n",
       "  border-bottom: 2px solid var(--sklearn-color-text-on-default-background);\n",
       "  flex-grow: 1;\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-parallel {\n",
       "  display: flex;\n",
       "  align-items: stretch;\n",
       "  justify-content: center;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  position: relative;\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-parallel-item {\n",
       "  display: flex;\n",
       "  flex-direction: column;\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-parallel-item:first-child::after {\n",
       "  align-self: flex-end;\n",
       "  width: 50%;\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-parallel-item:last-child::after {\n",
       "  align-self: flex-start;\n",
       "  width: 50%;\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-parallel-item:only-child::after {\n",
       "  width: 0;\n",
       "}\n",
       "\n",
       "/* Serial-specific style estimator block */\n",
       "\n",
       "#sk-container-id-3 div.sk-serial {\n",
       "  display: flex;\n",
       "  flex-direction: column;\n",
       "  align-items: center;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  padding-right: 1em;\n",
       "  padding-left: 1em;\n",
       "}\n",
       "\n",
       "\n",
       "/* Toggleable style: style used for estimator/Pipeline/ColumnTransformer box that is\n",
       "clickable and can be expanded/collapsed.\n",
       "- Pipeline and ColumnTransformer use this feature and define the default style\n",
       "- Estimators will overwrite some part of the style using the `sk-estimator` class\n",
       "*/\n",
       "\n",
       "/* Pipeline and ColumnTransformer style (default) */\n",
       "\n",
       "#sk-container-id-3 div.sk-toggleable {\n",
       "  /* Default theme specific background. It is overwritten whether we have a\n",
       "  specific estimator or a Pipeline/ColumnTransformer */\n",
       "  background-color: var(--sklearn-color-background);\n",
       "}\n",
       "\n",
       "/* Toggleable label */\n",
       "#sk-container-id-3 label.sk-toggleable__label {\n",
       "  cursor: pointer;\n",
       "  display: block;\n",
       "  width: 100%;\n",
       "  margin-bottom: 0;\n",
       "  padding: 0.5em;\n",
       "  box-sizing: border-box;\n",
       "  text-align: center;\n",
       "}\n",
       "\n",
       "#sk-container-id-3 label.sk-toggleable__label-arrow:before {\n",
       "  /* Arrow on the left of the label */\n",
       "  content: \"▸\";\n",
       "  float: left;\n",
       "  margin-right: 0.25em;\n",
       "  color: var(--sklearn-color-icon);\n",
       "}\n",
       "\n",
       "#sk-container-id-3 label.sk-toggleable__label-arrow:hover:before {\n",
       "  color: var(--sklearn-color-text);\n",
       "}\n",
       "\n",
       "/* Toggleable content - dropdown */\n",
       "\n",
       "#sk-container-id-3 div.sk-toggleable__content {\n",
       "  max-height: 0;\n",
       "  max-width: 0;\n",
       "  overflow: hidden;\n",
       "  text-align: left;\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-toggleable__content.fitted {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-toggleable__content pre {\n",
       "  margin: 0.2em;\n",
       "  border-radius: 0.25em;\n",
       "  color: var(--sklearn-color-text);\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-toggleable__content.fitted pre {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-3 input.sk-toggleable__control:checked~div.sk-toggleable__content {\n",
       "  /* Expand drop-down */\n",
       "  max-height: 200px;\n",
       "  max-width: 100%;\n",
       "  overflow: auto;\n",
       "}\n",
       "\n",
       "#sk-container-id-3 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {\n",
       "  content: \"▾\";\n",
       "}\n",
       "\n",
       "/* Pipeline/ColumnTransformer-specific style */\n",
       "\n",
       "#sk-container-id-3 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-label.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Estimator-specific style */\n",
       "\n",
       "/* Colorize estimator box */\n",
       "#sk-container-id-3 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-estimator.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-label label.sk-toggleable__label,\n",
       "#sk-container-id-3 div.sk-label label {\n",
       "  /* The background is the default theme color */\n",
       "  color: var(--sklearn-color-text-on-default-background);\n",
       "}\n",
       "\n",
       "/* On hover, darken the color of the background */\n",
       "#sk-container-id-3 div.sk-label:hover label.sk-toggleable__label {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "/* Label box, darken color on hover, fitted */\n",
       "#sk-container-id-3 div.sk-label.fitted:hover label.sk-toggleable__label.fitted {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Estimator label */\n",
       "\n",
       "#sk-container-id-3 div.sk-label label {\n",
       "  font-family: monospace;\n",
       "  font-weight: bold;\n",
       "  display: inline-block;\n",
       "  line-height: 1.2em;\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-label-container {\n",
       "  text-align: center;\n",
       "}\n",
       "\n",
       "/* Estimator-specific */\n",
       "#sk-container-id-3 div.sk-estimator {\n",
       "  font-family: monospace;\n",
       "  border: 1px dotted var(--sklearn-color-border-box);\n",
       "  border-radius: 0.25em;\n",
       "  box-sizing: border-box;\n",
       "  margin-bottom: 0.5em;\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-estimator.fitted {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "/* on hover */\n",
       "#sk-container-id-3 div.sk-estimator:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-3 div.sk-estimator.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Specification for estimator info (e.g. \"i\" and \"?\") */\n",
       "\n",
       "/* Common style for \"i\" and \"?\" */\n",
       "\n",
       ".sk-estimator-doc-link,\n",
       "a:link.sk-estimator-doc-link,\n",
       "a:visited.sk-estimator-doc-link {\n",
       "  float: right;\n",
       "  font-size: smaller;\n",
       "  line-height: 1em;\n",
       "  font-family: monospace;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  border-radius: 1em;\n",
       "  height: 1em;\n",
       "  width: 1em;\n",
       "  text-decoration: none !important;\n",
       "  margin-left: 1ex;\n",
       "  /* unfitted */\n",
       "  border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-unfitted-level-1);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link.fitted,\n",
       "a:link.sk-estimator-doc-link.fitted,\n",
       "a:visited.sk-estimator-doc-link.fitted {\n",
       "  /* fitted */\n",
       "  border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-fitted-level-1);\n",
       "}\n",
       "\n",
       "/* On hover */\n",
       "div.sk-estimator:hover .sk-estimator-doc-link:hover,\n",
       ".sk-estimator-doc-link:hover,\n",
       "div.sk-label-container:hover .sk-estimator-doc-link:hover,\n",
       ".sk-estimator-doc-link:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "div.sk-estimator.fitted:hover .sk-estimator-doc-link.fitted:hover,\n",
       ".sk-estimator-doc-link.fitted:hover,\n",
       "div.sk-label-container:hover .sk-estimator-doc-link.fitted:hover,\n",
       ".sk-estimator-doc-link.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "/* Span, style for the box shown on hovering the info icon */\n",
       ".sk-estimator-doc-link span {\n",
       "  display: none;\n",
       "  z-index: 9999;\n",
       "  position: relative;\n",
       "  font-weight: normal;\n",
       "  right: .2ex;\n",
       "  padding: .5ex;\n",
       "  margin: .5ex;\n",
       "  width: min-content;\n",
       "  min-width: 20ex;\n",
       "  max-width: 50ex;\n",
       "  color: var(--sklearn-color-text);\n",
       "  box-shadow: 2pt 2pt 4pt #999;\n",
       "  /* unfitted */\n",
       "  background: var(--sklearn-color-unfitted-level-0);\n",
       "  border: .5pt solid var(--sklearn-color-unfitted-level-3);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link.fitted span {\n",
       "  /* fitted */\n",
       "  background: var(--sklearn-color-fitted-level-0);\n",
       "  border: var(--sklearn-color-fitted-level-3);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link:hover span {\n",
       "  display: block;\n",
       "}\n",
       "\n",
       "/* \"?\"-specific style due to the `<a>` HTML tag */\n",
       "\n",
       "#sk-container-id-3 a.estimator_doc_link {\n",
       "  float: right;\n",
       "  font-size: 1rem;\n",
       "  line-height: 1em;\n",
       "  font-family: monospace;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  border-radius: 1rem;\n",
       "  height: 1rem;\n",
       "  width: 1rem;\n",
       "  text-decoration: none;\n",
       "  /* unfitted */\n",
       "  color: var(--sklearn-color-unfitted-level-1);\n",
       "  border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
       "}\n",
       "\n",
       "#sk-container-id-3 a.estimator_doc_link.fitted {\n",
       "  /* fitted */\n",
       "  border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-fitted-level-1);\n",
       "}\n",
       "\n",
       "/* On hover */\n",
       "#sk-container-id-3 a.estimator_doc_link:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "#sk-container-id-3 a.estimator_doc_link.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-3);\n",
       "}\n",
       "</style><div id=\"sk-container-id-3\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>MultipleKernelRidgeCV(cv=KFold(n_splits=5, random_state=None, shuffle=False),\n",
       "                      kernels=&#x27;precomputed&#x27;,\n",
       "                      solver_params={&#x27;alphas&#x27;: array([1.e+01, 1.e+02, 1.e+03, 1.e+04, 1.e+05, 1.e+06, 1.e+07, 1.e+08,\n",
       "       1.e+09, 1.e+10]),\n",
       "                                     &#x27;n_iter&#x27;: 20})</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-11\" type=\"checkbox\" checked><label for=\"sk-estimator-id-11\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">&nbsp;MultipleKernelRidgeCV<span class=\"sk-estimator-doc-link \">i<span>Not fitted</span></span></label><div class=\"sk-toggleable__content \"><pre>MultipleKernelRidgeCV(cv=KFold(n_splits=5, random_state=None, shuffle=False),\n",
       "                      kernels=&#x27;precomputed&#x27;,\n",
       "                      solver_params={&#x27;alphas&#x27;: array([1.e+01, 1.e+02, 1.e+03, 1.e+04, 1.e+05, 1.e+06, 1.e+07, 1.e+08,\n",
       "       1.e+09, 1.e+10]),\n",
       "                                     &#x27;n_iter&#x27;: 20})</pre></div> </div></div></div></div>"
      ],
      "text/plain": [
       "MultipleKernelRidgeCV(cv=KFold(n_splits=5, random_state=None, shuffle=False),\n",
       "                      kernels='precomputed',\n",
       "                      solver_params={'alphas': array([1.e+01, 1.e+02, 1.e+03, 1.e+04, 1.e+05, 1.e+06, 1.e+07, 1.e+08,\n",
       "       1.e+09, 1.e+10]),\n",
       "                                     'n_iter': 20})"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_iter = 20\n",
    "alphas = np.logspace(1, 10, 10) # specify alpha values\n",
    "inner_cv = KFold(n_splits=5, shuffle=False) # inner 5-fold cross-validation setup\n",
    "\n",
    "solver_params = dict(n_iter=n_iter, alphas=alphas)\n",
    "mkr_model = MultipleKernelRidgeCV(kernels=\"precomputed\", solver='random_search',\n",
    "                              solver_params=solver_params, cv=inner_cv)\n",
    "\n",
    "mkr_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-4 {\n",
       "  /* Definition of color scheme common for light and dark mode */\n",
       "  --sklearn-color-text: black;\n",
       "  --sklearn-color-line: gray;\n",
       "  /* Definition of color scheme for unfitted estimators */\n",
       "  --sklearn-color-unfitted-level-0: #fff5e6;\n",
       "  --sklearn-color-unfitted-level-1: #f6e4d2;\n",
       "  --sklearn-color-unfitted-level-2: #ffe0b3;\n",
       "  --sklearn-color-unfitted-level-3: chocolate;\n",
       "  /* Definition of color scheme for fitted estimators */\n",
       "  --sklearn-color-fitted-level-0: #f0f8ff;\n",
       "  --sklearn-color-fitted-level-1: #d4ebff;\n",
       "  --sklearn-color-fitted-level-2: #b3dbfd;\n",
       "  --sklearn-color-fitted-level-3: cornflowerblue;\n",
       "\n",
       "  /* Specific color for light theme */\n",
       "  --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
       "  --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, white)));\n",
       "  --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
       "  --sklearn-color-icon: #696969;\n",
       "\n",
       "  @media (prefers-color-scheme: dark) {\n",
       "    /* Redefinition of color scheme for dark theme */\n",
       "    --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
       "    --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, #111)));\n",
       "    --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
       "    --sklearn-color-icon: #878787;\n",
       "  }\n",
       "}\n",
       "\n",
       "#sk-container-id-4 {\n",
       "  color: var(--sklearn-color-text);\n",
       "}\n",
       "\n",
       "#sk-container-id-4 pre {\n",
       "  padding: 0;\n",
       "}\n",
       "\n",
       "#sk-container-id-4 input.sk-hidden--visually {\n",
       "  border: 0;\n",
       "  clip: rect(1px 1px 1px 1px);\n",
       "  clip: rect(1px, 1px, 1px, 1px);\n",
       "  height: 1px;\n",
       "  margin: -1px;\n",
       "  overflow: hidden;\n",
       "  padding: 0;\n",
       "  position: absolute;\n",
       "  width: 1px;\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-dashed-wrapped {\n",
       "  border: 1px dashed var(--sklearn-color-line);\n",
       "  margin: 0 0.4em 0.5em 0.4em;\n",
       "  box-sizing: border-box;\n",
       "  padding-bottom: 0.4em;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-container {\n",
       "  /* jupyter's `normalize.less` sets `[hidden] { display: none; }`\n",
       "     but bootstrap.min.css set `[hidden] { display: none !important; }`\n",
       "     so we also need the `!important` here to be able to override the\n",
       "     default hidden behavior on the sphinx rendered scikit-learn.org.\n",
       "     See: https://github.com/scikit-learn/scikit-learn/issues/21755 */\n",
       "  display: inline-block !important;\n",
       "  position: relative;\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-text-repr-fallback {\n",
       "  display: none;\n",
       "}\n",
       "\n",
       "div.sk-parallel-item,\n",
       "div.sk-serial,\n",
       "div.sk-item {\n",
       "  /* draw centered vertical line to link estimators */\n",
       "  background-image: linear-gradient(var(--sklearn-color-text-on-default-background), var(--sklearn-color-text-on-default-background));\n",
       "  background-size: 2px 100%;\n",
       "  background-repeat: no-repeat;\n",
       "  background-position: center center;\n",
       "}\n",
       "\n",
       "/* Parallel-specific style estimator block */\n",
       "\n",
       "#sk-container-id-4 div.sk-parallel-item::after {\n",
       "  content: \"\";\n",
       "  width: 100%;\n",
       "  border-bottom: 2px solid var(--sklearn-color-text-on-default-background);\n",
       "  flex-grow: 1;\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-parallel {\n",
       "  display: flex;\n",
       "  align-items: stretch;\n",
       "  justify-content: center;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  position: relative;\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-parallel-item {\n",
       "  display: flex;\n",
       "  flex-direction: column;\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-parallel-item:first-child::after {\n",
       "  align-self: flex-end;\n",
       "  width: 50%;\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-parallel-item:last-child::after {\n",
       "  align-self: flex-start;\n",
       "  width: 50%;\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-parallel-item:only-child::after {\n",
       "  width: 0;\n",
       "}\n",
       "\n",
       "/* Serial-specific style estimator block */\n",
       "\n",
       "#sk-container-id-4 div.sk-serial {\n",
       "  display: flex;\n",
       "  flex-direction: column;\n",
       "  align-items: center;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  padding-right: 1em;\n",
       "  padding-left: 1em;\n",
       "}\n",
       "\n",
       "\n",
       "/* Toggleable style: style used for estimator/Pipeline/ColumnTransformer box that is\n",
       "clickable and can be expanded/collapsed.\n",
       "- Pipeline and ColumnTransformer use this feature and define the default style\n",
       "- Estimators will overwrite some part of the style using the `sk-estimator` class\n",
       "*/\n",
       "\n",
       "/* Pipeline and ColumnTransformer style (default) */\n",
       "\n",
       "#sk-container-id-4 div.sk-toggleable {\n",
       "  /* Default theme specific background. It is overwritten whether we have a\n",
       "  specific estimator or a Pipeline/ColumnTransformer */\n",
       "  background-color: var(--sklearn-color-background);\n",
       "}\n",
       "\n",
       "/* Toggleable label */\n",
       "#sk-container-id-4 label.sk-toggleable__label {\n",
       "  cursor: pointer;\n",
       "  display: block;\n",
       "  width: 100%;\n",
       "  margin-bottom: 0;\n",
       "  padding: 0.5em;\n",
       "  box-sizing: border-box;\n",
       "  text-align: center;\n",
       "}\n",
       "\n",
       "#sk-container-id-4 label.sk-toggleable__label-arrow:before {\n",
       "  /* Arrow on the left of the label */\n",
       "  content: \"▸\";\n",
       "  float: left;\n",
       "  margin-right: 0.25em;\n",
       "  color: var(--sklearn-color-icon);\n",
       "}\n",
       "\n",
       "#sk-container-id-4 label.sk-toggleable__label-arrow:hover:before {\n",
       "  color: var(--sklearn-color-text);\n",
       "}\n",
       "\n",
       "/* Toggleable content - dropdown */\n",
       "\n",
       "#sk-container-id-4 div.sk-toggleable__content {\n",
       "  max-height: 0;\n",
       "  max-width: 0;\n",
       "  overflow: hidden;\n",
       "  text-align: left;\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-toggleable__content.fitted {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-toggleable__content pre {\n",
       "  margin: 0.2em;\n",
       "  border-radius: 0.25em;\n",
       "  color: var(--sklearn-color-text);\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-toggleable__content.fitted pre {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-4 input.sk-toggleable__control:checked~div.sk-toggleable__content {\n",
       "  /* Expand drop-down */\n",
       "  max-height: 200px;\n",
       "  max-width: 100%;\n",
       "  overflow: auto;\n",
       "}\n",
       "\n",
       "#sk-container-id-4 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {\n",
       "  content: \"▾\";\n",
       "}\n",
       "\n",
       "/* Pipeline/ColumnTransformer-specific style */\n",
       "\n",
       "#sk-container-id-4 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-label.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Estimator-specific style */\n",
       "\n",
       "/* Colorize estimator box */\n",
       "#sk-container-id-4 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-estimator.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-label label.sk-toggleable__label,\n",
       "#sk-container-id-4 div.sk-label label {\n",
       "  /* The background is the default theme color */\n",
       "  color: var(--sklearn-color-text-on-default-background);\n",
       "}\n",
       "\n",
       "/* On hover, darken the color of the background */\n",
       "#sk-container-id-4 div.sk-label:hover label.sk-toggleable__label {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "/* Label box, darken color on hover, fitted */\n",
       "#sk-container-id-4 div.sk-label.fitted:hover label.sk-toggleable__label.fitted {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Estimator label */\n",
       "\n",
       "#sk-container-id-4 div.sk-label label {\n",
       "  font-family: monospace;\n",
       "  font-weight: bold;\n",
       "  display: inline-block;\n",
       "  line-height: 1.2em;\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-label-container {\n",
       "  text-align: center;\n",
       "}\n",
       "\n",
       "/* Estimator-specific */\n",
       "#sk-container-id-4 div.sk-estimator {\n",
       "  font-family: monospace;\n",
       "  border: 1px dotted var(--sklearn-color-border-box);\n",
       "  border-radius: 0.25em;\n",
       "  box-sizing: border-box;\n",
       "  margin-bottom: 0.5em;\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-estimator.fitted {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "/* on hover */\n",
       "#sk-container-id-4 div.sk-estimator:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-4 div.sk-estimator.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Specification for estimator info (e.g. \"i\" and \"?\") */\n",
       "\n",
       "/* Common style for \"i\" and \"?\" */\n",
       "\n",
       ".sk-estimator-doc-link,\n",
       "a:link.sk-estimator-doc-link,\n",
       "a:visited.sk-estimator-doc-link {\n",
       "  float: right;\n",
       "  font-size: smaller;\n",
       "  line-height: 1em;\n",
       "  font-family: monospace;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  border-radius: 1em;\n",
       "  height: 1em;\n",
       "  width: 1em;\n",
       "  text-decoration: none !important;\n",
       "  margin-left: 1ex;\n",
       "  /* unfitted */\n",
       "  border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-unfitted-level-1);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link.fitted,\n",
       "a:link.sk-estimator-doc-link.fitted,\n",
       "a:visited.sk-estimator-doc-link.fitted {\n",
       "  /* fitted */\n",
       "  border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-fitted-level-1);\n",
       "}\n",
       "\n",
       "/* On hover */\n",
       "div.sk-estimator:hover .sk-estimator-doc-link:hover,\n",
       ".sk-estimator-doc-link:hover,\n",
       "div.sk-label-container:hover .sk-estimator-doc-link:hover,\n",
       ".sk-estimator-doc-link:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "div.sk-estimator.fitted:hover .sk-estimator-doc-link.fitted:hover,\n",
       ".sk-estimator-doc-link.fitted:hover,\n",
       "div.sk-label-container:hover .sk-estimator-doc-link.fitted:hover,\n",
       ".sk-estimator-doc-link.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "/* Span, style for the box shown on hovering the info icon */\n",
       ".sk-estimator-doc-link span {\n",
       "  display: none;\n",
       "  z-index: 9999;\n",
       "  position: relative;\n",
       "  font-weight: normal;\n",
       "  right: .2ex;\n",
       "  padding: .5ex;\n",
       "  margin: .5ex;\n",
       "  width: min-content;\n",
       "  min-width: 20ex;\n",
       "  max-width: 50ex;\n",
       "  color: var(--sklearn-color-text);\n",
       "  box-shadow: 2pt 2pt 4pt #999;\n",
       "  /* unfitted */\n",
       "  background: var(--sklearn-color-unfitted-level-0);\n",
       "  border: .5pt solid var(--sklearn-color-unfitted-level-3);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link.fitted span {\n",
       "  /* fitted */\n",
       "  background: var(--sklearn-color-fitted-level-0);\n",
       "  border: var(--sklearn-color-fitted-level-3);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link:hover span {\n",
       "  display: block;\n",
       "}\n",
       "\n",
       "/* \"?\"-specific style due to the `<a>` HTML tag */\n",
       "\n",
       "#sk-container-id-4 a.estimator_doc_link {\n",
       "  float: right;\n",
       "  font-size: 1rem;\n",
       "  line-height: 1em;\n",
       "  font-family: monospace;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  border-radius: 1rem;\n",
       "  height: 1rem;\n",
       "  width: 1rem;\n",
       "  text-decoration: none;\n",
       "  /* unfitted */\n",
       "  color: var(--sklearn-color-unfitted-level-1);\n",
       "  border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
       "}\n",
       "\n",
       "#sk-container-id-4 a.estimator_doc_link.fitted {\n",
       "  /* fitted */\n",
       "  border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-fitted-level-1);\n",
       "}\n",
       "\n",
       "/* On hover */\n",
       "#sk-container-id-4 a.estimator_doc_link:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "#sk-container-id-4 a.estimator_doc_link.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-3);\n",
       "}\n",
       "</style><div id=\"sk-container-id-4\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;columnkernelizer&#x27;,\n",
       "                 ColumnKernelizer(transformers=[(&#x27;gpt2&#x27;,\n",
       "                                                 Pipeline(steps=[(&#x27;standardscaler&#x27;,\n",
       "                                                                  StandardScaler(with_std=False)),\n",
       "                                                                 (&#x27;kernelizer&#x27;,\n",
       "                                                                  Kernelizer())]),\n",
       "                                                 slice(0, 1600, None)),\n",
       "                                                (&#x27;whisper&#x27;,\n",
       "                                                 Pipeline(steps=[(&#x27;standardscaler&#x27;,\n",
       "                                                                  StandardScaler(with_std=False)),\n",
       "                                                                 (&#x27;kernelizer&#x27;,\n",
       "                                                                  Kernelizer())]),\n",
       "                                                 slice(1600, None, None))])),\n",
       "                (&#x27;multiplekernelridgecv&#x27;,\n",
       "                 MultipleKernelRidgeCV(cv=KFold(n_splits=5, random_state=None, shuffle=False),\n",
       "                                       kernels=&#x27;precomputed&#x27;,\n",
       "                                       solver_params={&#x27;alphas&#x27;: array([1.e+01, 1.e+02, 1.e+03, 1.e+04, 1.e+05, 1.e+06, 1.e+07, 1.e+08,\n",
       "       1.e+09, 1.e+10]),\n",
       "                                                      &#x27;n_iter&#x27;: 20}))])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-12\" type=\"checkbox\" ><label for=\"sk-estimator-id-12\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">&nbsp;&nbsp;Pipeline<a class=\"sk-estimator-doc-link \" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.5/modules/generated/sklearn.pipeline.Pipeline.html\">?<span>Documentation for Pipeline</span></a><span class=\"sk-estimator-doc-link \">i<span>Not fitted</span></span></label><div class=\"sk-toggleable__content \"><pre>Pipeline(steps=[(&#x27;columnkernelizer&#x27;,\n",
       "                 ColumnKernelizer(transformers=[(&#x27;gpt2&#x27;,\n",
       "                                                 Pipeline(steps=[(&#x27;standardscaler&#x27;,\n",
       "                                                                  StandardScaler(with_std=False)),\n",
       "                                                                 (&#x27;kernelizer&#x27;,\n",
       "                                                                  Kernelizer())]),\n",
       "                                                 slice(0, 1600, None)),\n",
       "                                                (&#x27;whisper&#x27;,\n",
       "                                                 Pipeline(steps=[(&#x27;standardscaler&#x27;,\n",
       "                                                                  StandardScaler(with_std=False)),\n",
       "                                                                 (&#x27;kernelizer&#x27;,\n",
       "                                                                  Kernelizer())]),\n",
       "                                                 slice(1600, None, None))])),\n",
       "                (&#x27;multiplekernelridgecv&#x27;,\n",
       "                 MultipleKernelRidgeCV(cv=KFold(n_splits=5, random_state=None, shuffle=False),\n",
       "                                       kernels=&#x27;precomputed&#x27;,\n",
       "                                       solver_params={&#x27;alphas&#x27;: array([1.e+01, 1.e+02, 1.e+03, 1.e+04, 1.e+05, 1.e+06, 1.e+07, 1.e+08,\n",
       "       1.e+09, 1.e+10]),\n",
       "                                                      &#x27;n_iter&#x27;: 20}))])</pre></div> </div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-13\" type=\"checkbox\" ><label for=\"sk-estimator-id-13\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">columnkernelizer: ColumnKernelizer</label><div class=\"sk-toggleable__content \"><pre>ColumnKernelizer(transformers=[(&#x27;gpt2&#x27;,\n",
       "                                Pipeline(steps=[(&#x27;standardscaler&#x27;,\n",
       "                                                 StandardScaler(with_std=False)),\n",
       "                                                (&#x27;kernelizer&#x27;, Kernelizer())]),\n",
       "                                slice(0, 1600, None)),\n",
       "                               (&#x27;whisper&#x27;,\n",
       "                                Pipeline(steps=[(&#x27;standardscaler&#x27;,\n",
       "                                                 StandardScaler(with_std=False)),\n",
       "                                                (&#x27;kernelizer&#x27;, Kernelizer())]),\n",
       "                                slice(1600, None, None))])</pre></div> </div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-14\" type=\"checkbox\" ><label for=\"sk-estimator-id-14\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">gpt2</label><div class=\"sk-toggleable__content \"><pre>slice(0, 1600, None)</pre></div> </div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-15\" type=\"checkbox\" ><label for=\"sk-estimator-id-15\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">&nbsp;StandardScaler<a class=\"sk-estimator-doc-link \" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.5/modules/generated/sklearn.preprocessing.StandardScaler.html\">?<span>Documentation for StandardScaler</span></a></label><div class=\"sk-toggleable__content \"><pre>StandardScaler(with_std=False)</pre></div> </div></div><div class=\"sk-item\"><div class=\"sk-estimator  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-16\" type=\"checkbox\" ><label for=\"sk-estimator-id-16\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">Kernelizer</label><div class=\"sk-toggleable__content \"><pre>Kernelizer()</pre></div> </div></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-17\" type=\"checkbox\" ><label for=\"sk-estimator-id-17\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">whisper</label><div class=\"sk-toggleable__content \"><pre>slice(1600, None, None)</pre></div> </div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-18\" type=\"checkbox\" ><label for=\"sk-estimator-id-18\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">&nbsp;StandardScaler<a class=\"sk-estimator-doc-link \" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.5/modules/generated/sklearn.preprocessing.StandardScaler.html\">?<span>Documentation for StandardScaler</span></a></label><div class=\"sk-toggleable__content \"><pre>StandardScaler(with_std=False)</pre></div> </div></div><div class=\"sk-item\"><div class=\"sk-estimator  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-19\" type=\"checkbox\" ><label for=\"sk-estimator-id-19\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">Kernelizer</label><div class=\"sk-toggleable__content \"><pre>Kernelizer()</pre></div> </div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator  sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-20\" type=\"checkbox\" ><label for=\"sk-estimator-id-20\" class=\"sk-toggleable__label  sk-toggleable__label-arrow \">MultipleKernelRidgeCV</label><div class=\"sk-toggleable__content \"><pre>MultipleKernelRidgeCV(cv=KFold(n_splits=5, random_state=None, shuffle=False),\n",
       "                      kernels=&#x27;precomputed&#x27;,\n",
       "                      solver_params={&#x27;alphas&#x27;: array([1.e+01, 1.e+02, 1.e+03, 1.e+04, 1.e+05, 1.e+06, 1.e+07, 1.e+08,\n",
       "       1.e+09, 1.e+10]),\n",
       "                                     &#x27;n_iter&#x27;: 20})</pre></div> </div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('columnkernelizer',\n",
       "                 ColumnKernelizer(transformers=[('gpt2',\n",
       "                                                 Pipeline(steps=[('standardscaler',\n",
       "                                                                  StandardScaler(with_std=False)),\n",
       "                                                                 ('kernelizer',\n",
       "                                                                  Kernelizer())]),\n",
       "                                                 slice(0, 1600, None)),\n",
       "                                                ('whisper',\n",
       "                                                 Pipeline(steps=[('standardscaler',\n",
       "                                                                  StandardScaler(with_std=False)),\n",
       "                                                                 ('kernelizer',\n",
       "                                                                  Kernelizer())]),\n",
       "                                                 slice(1600, None, None))])),\n",
       "                ('multiplekernelridgecv',\n",
       "                 MultipleKernelRidgeCV(cv=KFold(n_splits=5, random_state=None, shuffle=False),\n",
       "                                       kernels='precomputed',\n",
       "                                       solver_params={'alphas': array([1.e+01, 1.e+02, 1.e+03, 1.e+04, 1.e+05, 1.e+06, 1.e+07, 1.e+08,\n",
       "       1.e+09, 1.e+10]),\n",
       "                                                      'n_iter': 20}))])"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = make_pipeline(\n",
    "    column_kernelizer,\n",
    "    mkr_model,\n",
    ")\n",
    "model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training encoding models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-info\">\n",
    "\n",
    "This code block may take a while to run. Make sure you are using a GPU if you have one (verify by running `nvidia-smi`). You may also consider resampling the epochs even further to use fewer lags, and/or choose specific electrodes to run to use fewer electrodes.\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[........................................] 100% | 22.45 sec | 20 random sampling with cv | \n",
      "[........................................] 100% | 19.45 sec | 20 random sampling with cv | \n",
      "Encoding performance correlating matrix shape: (2, 2, 127, 128)\n"
     ]
    }
   ],
   "source": [
    "epochs_shape = epochs._data.shape[1:] # number of electrodes * number of lags\n",
    "\n",
    "def train_encoding(X, Y):\n",
    "\n",
    "    corrs = [] # empty array to store correlation results\n",
    "    kfold = KFold(2, shuffle=False) # outer 2-fold cross-validation setup\n",
    "    for train_index, test_index in kfold.split(X): # loop through folds\n",
    "\n",
    "        # Split train and test datasets\n",
    "        X1_train, X1_test = X[train_index], X[test_index]\n",
    "        Y_train, Y_test = Y[train_index], Y[test_index]\n",
    "\n",
    "        # Standardize Y\n",
    "        scaler = StandardScaler()\n",
    "        Y_train = scaler.fit_transform(Y_train)\n",
    "        Y_test = scaler.transform(Y_test)\n",
    "\n",
    "        model.fit(X1_train, Y_train) # Fit pipeline with transforms and ridge estimator\n",
    "        Y_preds = model.predict(X1_test, split=True) # Use trained model to predict on test set\n",
    "        corr = correlation_score_split(Y_test, Y_preds) # Compute correlation score\n",
    "\n",
    "        if \"torch\" in get_backend().__name__: # if using gpu, transform tensor back to numpy\n",
    "            corr = corr.numpy(force=True)\n",
    "\n",
    "        corrs.append(corr) # append fold correlation results to final results\n",
    "    return np.stack(corrs)\n",
    "\n",
    "# set_backend(\"torch\") # resort to torch or numpy if cuda out of memory\n",
    "corrs_embedding = train_encoding(X, Y)\n",
    "corrs_embedding = corrs_embedding.reshape(2, 2, *epochs_shape)\n",
    "print(f\"Encoding performance correlating matrix shape: {corrs_embedding.shape}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plotting encoding performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Coordinate matrix shape:  (127, 3)\n"
     ]
    }
   ],
   "source": [
    "ch2loc = {ch['ch_name']: ch['loc'][:3] for ch in raw.info['chs']}\n",
    "coords = np.vstack([ch2loc[ch] for ch in raw.info['ch_names']])\n",
    "coords *= 1000  # nilearn likes to plot in meters, not mm\n",
    "print(\"Coordinate matrix shape: \", coords.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x400 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1, 2, figsize=(12, 4))\n",
    "\n",
    "values = corrs_embedding[:, 0].mean(0).max(-1)\n",
    "order = values.argsort()\n",
    "plot_markers(values[order], coords[order],\n",
    "             node_size=30, display_mode='l',\n",
    "             figure=fig, axes=axes[0],\n",
    "             node_vmin=0,  node_vmax=0.2,\n",
    "             title=\"GPT-2\",\n",
    "             node_cmap='inferno_r', colorbar=True)\n",
    "\n",
    "\n",
    "values = corrs_embedding[:, 1].mean(0).max(-1)\n",
    "order = values.argsort()\n",
    "plot_markers(values[order], coords[order],\n",
    "             node_size=30, display_mode='l',\n",
    "             figure=fig, axes=axes[1],\n",
    "             node_vmin=0,  node_vmax=0.2,\n",
    "             title=\"Whisper\",\n",
    "             node_cmap='inferno_r', colorbar=True)\n",
    "\n",
    "fig.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "mne",
   "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": "3.12.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}