notebooks/14_automatic_sleep_staging.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Automatic sleep staging\n",
    "\n",
    "This notebook demonstrates how to perform automatic sleep staging of polysomnography data in YASA. For more details, make sure to read the [eLife publication](https://elifesciences.org/articles/70092).\n",
    "\n",
    "Please install the latest version of YASA first with: `pip install --upgrade yasa`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import mne\n",
    "import yasa\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data loading\n",
    "\n",
    "The automatic sleep staging function requires that the data are loaded using the [MNE Python package](https://mne.tools/stable/index.html). For instance, if your polysomnography data are stored in the standard European Data Fomat (.edf), you can use [this function](https://mne.tools/stable/generated/mne.io.read_raw_edf.html?highlight=read_raw_edf#mne.io.read_raw_edf) to load the data into Python."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The channels are: ['F3', 'F4', 'C3', 'C4', 'O1', 'O2', 'EOG1', 'EOG2', 'EMG1']\n",
      "The sampling frequency is: 100.0\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<script type=\"text/javascript\">\n",
       "    // must be `var` (not `const`) because this can get embedded multiple times on a page\n",
       "var toggleVisibility = (className) => {\n",
       "\n",
       "    const elements = document.querySelectorAll(`.${className}`);\n",
       "\n",
       "    elements.forEach(element => {\n",
       "        if (element.classList.contains(\"mne-repr-section-header\")) {\n",
       "            return  // Don't collapse the section header row\n",
       "        }\n",
       "        element.classList.toggle(\"mne-repr-collapsed\");\n",
       "    });\n",
       "\n",
       "    // trigger caret to rotate\n",
       "    var sel = `.mne-repr-section-header.${className} > th.mne-repr-section-toggle > button`;\n",
       "    const button = document.querySelector(sel);\n",
       "    button.classList.toggle(\"collapsed\");\n",
       "\n",
       "    // adjust tooltip\n",
       "    sel = `tr.mne-repr-section-header.${className}`;\n",
       "    const secHeadRow = document.querySelector(sel);\n",
       "    secHeadRow.classList.toggle(\"collapsed\");\n",
       "    secHeadRow.title = secHeadRow.title === \"Hide section\" ? \"Show section\" : \"Hide section\";\n",
       "}\n",
       "</script>\n",
       "\n",
       "<style type=\"text/css\">\n",
       "    /*\n",
       "Styles in this section apply both to the sphinx-built website docs and to notebooks\n",
       "rendered in an IDE or in Jupyter. In our web docs, styles here are complemented by\n",
       "doc/_static/styles.css and other CSS files (e.g. from the sphinx theme, sphinx-gallery,\n",
       "or bootstrap). In IDEs/Jupyter, those style files are unavailable, so only the rules in\n",
       "this file apply (plus whatever default styling the IDE applies).\n",
       "*/\n",
       ".mne-repr-table {\n",
       "    display: inline;  /* prevent using full container width */\n",
       "}\n",
       ".mne-repr-table tr.mne-repr-section-header > th {\n",
       "    padding-top: 1rem;\n",
       "    text-align: left;\n",
       "    vertical-align: middle;\n",
       "}\n",
       ".mne-repr-section-toggle > button {\n",
       "    all: unset;\n",
       "    display: block;\n",
       "    height: 1rem;\n",
       "    width: 1rem;\n",
       "}\n",
       ".mne-repr-section-toggle > button > svg {\n",
       "    height: 60%;\n",
       "}\n",
       "\n",
       "/* transition (rotation) effects on the collapser button */\n",
       ".mne-repr-section-toggle > button.collapsed > svg {\n",
       "    transition: 0.1s ease-out;\n",
       "    transform: rotate(-90deg);\n",
       "}\n",
       ".mne-repr-section-toggle > button:not(.collapsed) > svg {\n",
       "    transition: 0.1s ease-out;\n",
       "    transform: rotate(0deg);\n",
       "}\n",
       "\n",
       "/* hide collapsed table rows */\n",
       ".mne-repr-collapsed {\n",
       "    display: none;\n",
       "}\n",
       "\n",
       "\n",
       "@layer {\n",
       "    /*\n",
       "    Selectors in a `@layer` will always be lower-precedence than selectors outside the\n",
       "    layer. So even though e.g. `div.output_html` is present in the sphinx-rendered\n",
       "    website docs, the styles here won't take effect there as long as some other rule\n",
       "    somewhere in the page's CSS targets the same element.\n",
       "\n",
       "    In IDEs or Jupyter notebooks, though, the CSS files from the sphinx theme,\n",
       "    sphinx-gallery, and bootstrap are unavailable, so these styles will apply.\n",
       "\n",
       "    Notes:\n",
       "\n",
       "    - the selector `.accordion-body` is for MNE Reports\n",
       "    - the selector `.output_html` is for VSCode's notebook interface\n",
       "    - the selector `.jp-RenderedHTML` is for Jupyter notebook\n",
       "    - variables starting with `--theme-` are VSCode-specific.\n",
       "    - variables starting with `--jp-` are Jupyter styles, *some of which* are also\n",
       "      available in VSCode. Here we try the `--theme-` variable first, then fall back to\n",
       "      the `--jp-` ones.\n",
       "    */\n",
       "    .mne-repr-table {\n",
       "        --mne-toggle-color: var(--theme-foreground, var(--jp-ui-font-color1));\n",
       "        --mne-button-bg-color: var(--theme-button-background, var(--jp-info-color0, var(--jp-content-link-color)));\n",
       "        --mne-button-fg-color: var(--theme-button-foreground, var(--jp-ui-inverse-font-color0, var(--jp-editor-background)));\n",
       "        --mne-button-hover-bg-color: var(--theme-button-hover-background, var(--jp-info-color1));\n",
       "        --mne-button-radius: var(--jp-border-radius, 0.25rem);\n",
       "    }\n",
       "    /* chevron position/alignment; in VSCode it looks ok without adjusting */\n",
       "    .accordion-body .mne-repr-section-toggle > button,\n",
       "    .jp-RenderedHTML .mne-repr-section-toggle > button {\n",
       "        padding: 0 0 45% 25% !important;\n",
       "    }\n",
       "    /* chevron color; MNE Report doesn't have light/dark mode */\n",
       "    div.output_html .mne-repr-section-toggle > button > svg > path,\n",
       "    .jp-RenderedHTML .mne-repr-section-toggle > button > svg > path {\n",
       "        fill: var(--mne-toggle-color);\n",
       "    }\n",
       "    .accordion-body .mne-ch-names-btn,\n",
       "    div.output_html .mne-ch-names-btn,\n",
       "    .jp-RenderedHTML .mne-ch-names-btn {\n",
       "        -webkit-border-radius: var(--mne-button-radius);\n",
       "        -moz-border-radius: var(--mne-button-radius);\n",
       "        border-radius: var(--mne-button-radius);\n",
       "        border: none;\n",
       "        background-image: none;\n",
       "        background-color: var(--mne-button-bg-color);\n",
       "        color: var(--mne-button-fg-color);\n",
       "        font-size: inherit;\n",
       "        min-width: 1.5rem;\n",
       "        padding: 0.25rem;\n",
       "        text-align: center;\n",
       "        text-decoration: none;\n",
       "    }\n",
       "    .accordion-body .mne-ch-names-btn:hover,\n",
       "    div.output_html .mne.ch-names-btn:hover,\n",
       "    .jp-RenderedHTML .mne-ch-names-btn:hover {\n",
       "        background-color: var(--mne-button-hover-bg-color);\n",
       "        text-decoration: underline;\n",
       "    }\n",
       "    .accordion-body .mne-ch-names-btn:focus-visible,\n",
       "    div.output_html .mne-ch-names-btn:focus-visible,\n",
       "    .jp-RenderedHTML .mne-ch-names-btn:focus-visible {\n",
       "        outline: 0.1875rem solid var(--mne-button-bg-color) !important;\n",
       "        outline-offset: 0.1875rem !important;\n",
       "    }\n",
       "}\n",
       "</style>\n",
       "\n",
       "\n",
       "\n",
       "<table class=\"table mne-repr-table\">\n",
       "    \n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "<tr class=\"mne-repr-section-header general-b118500d-c347-47ae-b9d5-db41c1f53ebc\"\n",
       "     title=\"Hide section\" \n",
       "    onclick=\"toggleVisibility('general-b118500d-c347-47ae-b9d5-db41c1f53ebc')\">\n",
       "    <th class=\"mne-repr-section-toggle\">\n",
       "        <button >\n",
       "            <svg xmlns=\"http://www.w3.org/2000/svg\" viewBox=\"0 0 512 512\"><!--!Font Awesome Free 6.6.0 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license/free Copyright 2024 Fonticons, Inc.--><path d=\"M233.4 406.6c12.5 12.5 32.8 12.5 45.3 0l192-192c12.5-12.5 12.5-32.8 0-45.3s-32.8-12.5-45.3 0L256 338.7 86.6 169.4c-12.5-12.5-32.8-12.5-45.3 0s-12.5 32.8 0 45.3l192 192z\"/></svg>\n",
       "        </button>\n",
       "    </th>\n",
       "    <th colspan=\"2\">\n",
       "        <strong>General</strong>\n",
       "    </th>\n",
       "</tr>\n",
       "\n",
       "\n",
       "<tr class=\"repr-element general-b118500d-c347-47ae-b9d5-db41c1f53ebc \">\n",
       "    <td class=\"mne-repr-section-toggle\"></td>\n",
       "    <td>Filename(s)</td>\n",
       "    <td>\n",
       "        \n",
       "        sub-02_mne_raw.fif\n",
       "        \n",
       "        \n",
       "    </td>\n",
       "</tr>\n",
       "\n",
       "<tr class=\"repr-element general-b118500d-c347-47ae-b9d5-db41c1f53ebc \">\n",
       "    <td class=\"mne-repr-section-toggle\"></td>\n",
       "    <td>MNE object type</td>\n",
       "    <td>Raw</td>\n",
       "</tr>\n",
       "<tr class=\"repr-element general-b118500d-c347-47ae-b9d5-db41c1f53ebc \">\n",
       "    <td class=\"mne-repr-section-toggle\"></td>\n",
       "    <td>Measurement date</td>\n",
       "    \n",
       "    <td>2016-01-15 at 14:01:00 UTC</td>\n",
       "    \n",
       "</tr>\n",
       "<tr class=\"repr-element general-b118500d-c347-47ae-b9d5-db41c1f53ebc \">\n",
       "    <td class=\"mne-repr-section-toggle\"></td>\n",
       "    <td>Participant</td>\n",
       "    \n",
       "    <td>Unknown</td>\n",
       "    \n",
       "</tr>\n",
       "<tr class=\"repr-element general-b118500d-c347-47ae-b9d5-db41c1f53ebc \">\n",
       "    <td class=\"mne-repr-section-toggle\"></td>\n",
       "    <td>Experimenter</td>\n",
       "    \n",
       "    <td>Unknown</td>\n",
       "    \n",
       "</tr>\n",
       "    \n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "<tr class=\"mne-repr-section-header acquisition-da62099a-7378-4dcb-bbbc-b8fe3b134f0e\"\n",
       "     title=\"Hide section\" \n",
       "    onclick=\"toggleVisibility('acquisition-da62099a-7378-4dcb-bbbc-b8fe3b134f0e')\">\n",
       "    <th class=\"mne-repr-section-toggle\">\n",
       "        <button >\n",
       "            <svg xmlns=\"http://www.w3.org/2000/svg\" viewBox=\"0 0 512 512\"><!--!Font Awesome Free 6.6.0 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license/free Copyright 2024 Fonticons, Inc.--><path d=\"M233.4 406.6c12.5 12.5 32.8 12.5 45.3 0l192-192c12.5-12.5 12.5-32.8 0-45.3s-32.8-12.5-45.3 0L256 338.7 86.6 169.4c-12.5-12.5-32.8-12.5-45.3 0s-12.5 32.8 0 45.3l192 192z\"/></svg>\n",
       "        </button>\n",
       "    </th>\n",
       "    <th colspan=\"2\">\n",
       "        <strong>Acquisition</strong>\n",
       "    </th>\n",
       "</tr>\n",
       "\n",
       "\n",
       "<tr class=\"repr-element acquisition-da62099a-7378-4dcb-bbbc-b8fe3b134f0e \">\n",
       "    <td class=\"mne-repr-section-toggle\"></td>\n",
       "    <td>Duration</td>\n",
       "    <td>00:49:00 (HH:MM:SS)</td>\n",
       "</tr>\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "<tr class=\"repr-element acquisition-da62099a-7378-4dcb-bbbc-b8fe3b134f0e \">\n",
       "    <td class=\"mne-repr-section-toggle\"></td>\n",
       "    <td>Sampling frequency</td>\n",
       "    <td>100.00 Hz</td>\n",
       "</tr>\n",
       "\n",
       "\n",
       "<tr class=\"repr-element acquisition-da62099a-7378-4dcb-bbbc-b8fe3b134f0e \">\n",
       "    <td class=\"mne-repr-section-toggle\"></td>\n",
       "    <td>Time points</td>\n",
       "    <td>294,000</td>\n",
       "</tr>\n",
       "\n",
       "\n",
       "    \n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "<tr class=\"mne-repr-section-header channels-94fe5c29-bd99-46bf-9481-918ab3bbaa4c\"\n",
       "     title=\"Hide section\" \n",
       "    onclick=\"toggleVisibility('channels-94fe5c29-bd99-46bf-9481-918ab3bbaa4c')\">\n",
       "    <th class=\"mne-repr-section-toggle\">\n",
       "        <button >\n",
       "            <svg xmlns=\"http://www.w3.org/2000/svg\" viewBox=\"0 0 512 512\"><!--!Font Awesome Free 6.6.0 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license/free Copyright 2024 Fonticons, Inc.--><path d=\"M233.4 406.6c12.5 12.5 32.8 12.5 45.3 0l192-192c12.5-12.5 12.5-32.8 0-45.3s-32.8-12.5-45.3 0L256 338.7 86.6 169.4c-12.5-12.5-32.8-12.5-45.3 0s-12.5 32.8 0 45.3l192 192z\"/></svg>\n",
       "        </button>\n",
       "    </th>\n",
       "    <th colspan=\"2\">\n",
       "        <strong>Channels</strong>\n",
       "    </th>\n",
       "</tr>\n",
       "\n",
       "\n",
       "    \n",
       "<tr class=\"repr-element channels-94fe5c29-bd99-46bf-9481-918ab3bbaa4c \">\n",
       "    <td class=\"mne-repr-section-toggle\"></td>\n",
       "    <td>EEG</td>\n",
       "    <td>\n",
       "        <button class=\"mne-ch-names-btn sd-sphinx-override sd-btn sd-btn-info sd-text-wrap sd-shadow-sm\" onclick=\"alert('Good EEG:\\n\\nF3, F4, C3, C4, O1, O2')\" title=\"(Click to open in popup)&#13;&#13;F3, F4, C3, C4, O1, O2\">\n",
       "            6\n",
       "        </button>\n",
       "\n",
       "        \n",
       "    </td>\n",
       "</tr>\n",
       "\n",
       "    \n",
       "<tr class=\"repr-element channels-94fe5c29-bd99-46bf-9481-918ab3bbaa4c \">\n",
       "    <td class=\"mne-repr-section-toggle\"></td>\n",
       "    <td>EOG</td>\n",
       "    <td>\n",
       "        <button class=\"mne-ch-names-btn sd-sphinx-override sd-btn sd-btn-info sd-text-wrap sd-shadow-sm\" onclick=\"alert('Good EOG:\\n\\nEOG1, EOG2')\" title=\"(Click to open in popup)&#13;&#13;EOG1, EOG2\">\n",
       "            2\n",
       "        </button>\n",
       "\n",
       "        \n",
       "    </td>\n",
       "</tr>\n",
       "\n",
       "    \n",
       "<tr class=\"repr-element channels-94fe5c29-bd99-46bf-9481-918ab3bbaa4c \">\n",
       "    <td class=\"mne-repr-section-toggle\"></td>\n",
       "    <td>EMG</td>\n",
       "    <td>\n",
       "        <button class=\"mne-ch-names-btn sd-sphinx-override sd-btn sd-btn-info sd-text-wrap sd-shadow-sm\" onclick=\"alert('Good EMG:\\n\\nEMG1')\" title=\"(Click to open in popup)&#13;&#13;EMG1\">\n",
       "            1\n",
       "        </button>\n",
       "\n",
       "        \n",
       "    </td>\n",
       "</tr>\n",
       "\n",
       "\n",
       "<tr class=\"repr-element channels-94fe5c29-bd99-46bf-9481-918ab3bbaa4c \">\n",
       "    <td class=\"mne-repr-section-toggle\"></td>\n",
       "    <td>Head & sensor digitization</td>\n",
       "    \n",
       "    <td>15 points</td>\n",
       "    \n",
       "</tr>\n",
       "    \n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "\n",
       "<tr class=\"mne-repr-section-header filters-7ae80fee-f7a2-4f39-849a-960bd04bf311\"\n",
       "     title=\"Hide section\" \n",
       "    onclick=\"toggleVisibility('filters-7ae80fee-f7a2-4f39-849a-960bd04bf311')\">\n",
       "    <th class=\"mne-repr-section-toggle\">\n",
       "        <button >\n",
       "            <svg xmlns=\"http://www.w3.org/2000/svg\" viewBox=\"0 0 512 512\"><!--!Font Awesome Free 6.6.0 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license/free Copyright 2024 Fonticons, Inc.--><path d=\"M233.4 406.6c12.5 12.5 32.8 12.5 45.3 0l192-192c12.5-12.5 12.5-32.8 0-45.3s-32.8-12.5-45.3 0L256 338.7 86.6 169.4c-12.5-12.5-32.8-12.5-45.3 0s-12.5 32.8 0 45.3l192 192z\"/></svg>\n",
       "        </button>\n",
       "    </th>\n",
       "    <th colspan=\"2\">\n",
       "        <strong>Filters</strong>\n",
       "    </th>\n",
       "</tr>\n",
       "\n",
       "\n",
       "<tr class=\"repr-element filters-7ae80fee-f7a2-4f39-849a-960bd04bf311 \">\n",
       "    <td class=\"mne-repr-section-toggle\"></td>\n",
       "    <td>Highpass</td>\n",
       "    <td>0.00 Hz</td>\n",
       "</tr>\n",
       "\n",
       "\n",
       "<tr class=\"repr-element filters-7ae80fee-f7a2-4f39-849a-960bd04bf311 \">\n",
       "    <td class=\"mne-repr-section-toggle\"></td>\n",
       "    <td>Lowpass</td>\n",
       "    <td>50.00 Hz</td>\n",
       "</tr>\n",
       "\n",
       "\n",
       "</table>"
      ],
      "text/plain": [
       "<Raw | sub-02_mne_raw.fif, 9 x 294000 (2940.0 s), ~20.2 MiB, data loaded>"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let's load a nap recording, directly as an MNE Raw object\n",
    "raw = mne.io.read_raw_fif('sub-02_mne_raw.fif', preload=True, verbose=False)\n",
    "print('The channels are:', raw.ch_names)\n",
    "print('The sampling frequency is:', raw.info['sfreq'])\n",
    "raw"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['W', 'W', 'W', 'W', 'W', 'W', 'W', 'W', 'W', 'W', 'W', 'W', 'W',\n",
       "       'W', 'W', 'W', 'W', 'W', 'W', 'W', 'W', 'W', 'N1', 'W', 'W', 'W',\n",
       "       'W', 'N1', 'N1', 'N1', 'W', 'W', 'N1', 'N1', 'N1', 'N1', 'N2',\n",
       "       'N2', 'N2', 'W', 'N1', 'N2', 'N2', 'N2', 'N2', 'N2', 'N2', 'N2',\n",
       "       'N2', 'N2', 'N2', 'N2', 'N2', 'N2', 'N2', 'N2', 'N2', 'N2', 'N2',\n",
       "       'N2', 'N2', 'N2', 'N2', 'N2', 'N2', 'N2', 'N2', 'N2', 'N2', 'N3',\n",
       "       'N3', 'N3', 'N3', 'N3', 'N3', 'N3', 'N3', 'N3', 'N3', 'N3', 'N3',\n",
       "       'N3', 'N3', 'N3', 'N3', 'N3', 'N3', 'N3', 'N3', 'N3', 'N3', 'W',\n",
       "       'W', 'W', 'W', 'W', 'W', 'W'], dtype='<U2')"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let's now load the human-scored hypnogram, where each value represents a 30-sec epoch.\n",
    "hyp = np.loadtxt('sub-02_hypno_30s.txt', dtype=str)\n",
    "hyp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Epoch\n",
       "0     WAKE\n",
       "1     WAKE\n",
       "2     WAKE\n",
       "3     WAKE\n",
       "4     WAKE\n",
       "      ... \n",
       "93    WAKE\n",
       "94    WAKE\n",
       "95    WAKE\n",
       "96    WAKE\n",
       "97    WAKE\n",
       "Name: Stage, Length: 98, dtype: category\n",
       "Categories (7, object): ['WAKE', 'N1', 'N2', 'N3', 'REM', 'ART', 'UNS']"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Convert it to a Hypnogram instance, which is the preferred way to manipulate hypnograms since v0.7\n",
    "hyp = yasa.Hypnogram(hyp, freq=\"30s\")\n",
    "# The hypnogram values can be obtained with\n",
    "hyp.hypno"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 560x240 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Let's plot it\n",
    "fig, ax = plt.subplots(1, 1, figsize=(7, 3), constrained_layout=True, dpi=80)\n",
    "ax = hyp.plot_hypnogram(fill_color=\"gainsboro\", ax=ax)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Sleep staging\n",
    "\n",
    "Automatic sleep stages classification can be done since YASA 0.4.0 using the [SleepStaging](https://raphaelvallat.com/yasa/build/html/generated/yasa.SleepStaging.html#yasa.SleepStaging) class. Make sure to read the [documentation](https://raphaelvallat.com/yasa/build/html/generated/yasa.SleepStaging.html#yasa.SleepStaging), which explains how the algorithm works."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<SleepStaging | 3 x 294000 samples (49.0 minutes), 100.0 Hz>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We first need to specify the channel names and, optionally, the age and sex of the participant\n",
    "# - \"raw\" is the name of the variable containing the polysomnography data loaded with MNE.\n",
    "# - \"eeg_name\" is the name of the EEG channel, preferentially a central derivation (e.g. C4-M1). This is always required to run the sleep staging algorithm.\n",
    "# - \"eog_name\" is the name of the EOG channel (e.g. LOC-M1). This is optional.\n",
    "# - \"eog_name\" is the name of the EOG channel (e.g. EMG1-EMG3). This is optional.\n",
    "# - \"metadata\" is a dictionary containing the age and sex of the participant. This is optional.\n",
    "sls = yasa.SleepStaging(raw, eeg_name=\"C4\", eog_name=\"EOG1\", emg_name=\"EMG1\", metadata=dict(age=21, male=False))\n",
    "sls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/lib/python3.12/site-packages/sklearn/base.py:376: InconsistentVersionWarning: Trying to unpickle estimator LabelEncoder from version 0.24.2 when using version 1.5.1. This might lead to breaking code or invalid results. Use at your own risk. For more info please refer to:\n",
      "https://scikit-learn.org/stable/model_persistence.html#security-maintainability-limitations\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Hypnogram | 98 epochs x 30s (49.00 minutes), 5 unique stages>\n",
       " - Use `.hypno` to get the string values as a pandas.Series\n",
       " - Use `.as_int()` to get the integer values as a pandas.Series\n",
       " - Use `.plot_hypnogram()` to plot the hypnogram\n",
       "See the online documentation for more details."
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Getting the predicted sleep stages is now as easy as:\n",
    "y_pred = sls.predict()\n",
    "y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Epoch\n",
       "0     WAKE\n",
       "1     WAKE\n",
       "2     WAKE\n",
       "3     WAKE\n",
       "4     WAKE\n",
       "      ... \n",
       "93    WAKE\n",
       "94    WAKE\n",
       "95    WAKE\n",
       "96    WAKE\n",
       "97    WAKE\n",
       "Name: Stage, Length: 98, dtype: category\n",
       "Categories (7, object): ['WAKE', 'N1', 'N2', 'N3', 'REM', 'ART', 'UNS']"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred.hypno"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The overall agreement is 83.67%\n"
     ]
    }
   ],
   "source": [
    "# What is the accuracy of the prediction, compared to the human scoring\n",
    "accuracy = 100 * (hyp.hypno == y_pred.hypno).mean()\n",
    "print(f\"The overall agreement is {accuracy:.2f}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Plot and sleep statistics**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 560x240 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the predicted hypnogram\n",
    "fig, ax = plt.subplots(1, 1, figsize=(7, 3), constrained_layout=True, dpi=80)\n",
    "ax = y_pred.plot_hypnogram(fill_color=\"gainsboro\", ax=ax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'TIB': 49.0,\n",
       " 'SPT': 28.0,\n",
       " 'WASO': 0.0,\n",
       " 'TST': 28.0,\n",
       " 'SE': 57.1429,\n",
       " 'SME': 100.0,\n",
       " 'SFI': 1.0714,\n",
       " 'SOL': 17.0,\n",
       " 'SOL_5min': 17.0,\n",
       " 'Lat_REM': nan,\n",
       " 'WAKE': 21.0,\n",
       " 'N1': 0.0,\n",
       " 'N2': 15.0,\n",
       " 'N3': 13.0,\n",
       " 'REM': 0.0,\n",
       " '%N1': 0.0,\n",
       " '%N2': 53.5714,\n",
       " '%N3': 46.4286,\n",
       " '%REM': 0.0}"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Calculate the summary sleep statistics of the predicted hypnogram\n",
    "y_pred.sleep_statistics()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Stage probabilities and confidence of the algorithm at each epoch**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/raphael/GitHub/yasa/yasa/staging.py:484: FutureWarning: The `predict_proba` function is deprecated and will be removed in v0.8. The predicted probabilities can now be accessed with `yasa.Hypnogram.proba` instead, e.g `SleepStaging.predict().proba`\n",
      "  warnings.warn(\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>N1</th>\n",
       "      <th>N2</th>\n",
       "      <th>N3</th>\n",
       "      <th>REM</th>\n",
       "      <th>WAKE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Epoch</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>0.002170</td>\n",
       "      <td>0.005012</td>\n",
       "      <td>0.000683</td>\n",
       "      <td>1.772861e-18</td>\n",
       "      <td>0.992135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.002470</td>\n",
       "      <td>0.003121</td>\n",
       "      <td>0.002585</td>\n",
       "      <td>8.013632e-05</td>\n",
       "      <td>0.991744</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.003882</td>\n",
       "      <td>0.003285</td>\n",
       "      <td>0.000097</td>\n",
       "      <td>6.435026e-04</td>\n",
       "      <td>0.992092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.001994</td>\n",
       "      <td>0.001806</td>\n",
       "      <td>0.000051</td>\n",
       "      <td>6.712369e-04</td>\n",
       "      <td>0.995478</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.002609</td>\n",
       "      <td>0.008254</td>\n",
       "      <td>0.000255</td>\n",
       "      <td>5.924781e-04</td>\n",
       "      <td>0.988289</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>0.003944</td>\n",
       "      <td>0.009049</td>\n",
       "      <td>0.004683</td>\n",
       "      <td>9.824195e-05</td>\n",
       "      <td>0.982225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>0.002002</td>\n",
       "      <td>0.029846</td>\n",
       "      <td>0.135356</td>\n",
       "      <td>2.641568e-04</td>\n",
       "      <td>0.832531</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>0.001389</td>\n",
       "      <td>0.001854</td>\n",
       "      <td>0.000503</td>\n",
       "      <td>4.100423e-05</td>\n",
       "      <td>0.996213</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>0.001921</td>\n",
       "      <td>0.000878</td>\n",
       "      <td>0.000088</td>\n",
       "      <td>5.482605e-05</td>\n",
       "      <td>0.997057</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>0.000855</td>\n",
       "      <td>0.000934</td>\n",
       "      <td>0.000024</td>\n",
       "      <td>2.945145e-05</td>\n",
       "      <td>0.998157</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>98 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             N1        N2        N3           REM      WAKE\n",
       "Epoch                                                      \n",
       "0      0.002170  0.005012  0.000683  1.772861e-18  0.992135\n",
       "1      0.002470  0.003121  0.002585  8.013632e-05  0.991744\n",
       "2      0.003882  0.003285  0.000097  6.435026e-04  0.992092\n",
       "3      0.001994  0.001806  0.000051  6.712369e-04  0.995478\n",
       "4      0.002609  0.008254  0.000255  5.924781e-04  0.988289\n",
       "...         ...       ...       ...           ...       ...\n",
       "93     0.003944  0.009049  0.004683  9.824195e-05  0.982225\n",
       "94     0.002002  0.029846  0.135356  2.641568e-04  0.832531\n",
       "95     0.001389  0.001854  0.000503  4.100423e-05  0.996213\n",
       "96     0.001921  0.000878  0.000088  5.482605e-05  0.997057\n",
       "97     0.000855  0.000934  0.000024  2.945145e-05  0.998157\n",
       "\n",
       "[98 rows x 5 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# What are the predicted probabilities of each sleep stage at each epoch?\n",
    "proba = sls.predict_proba()\n",
    "proba"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the predicted probabilities\n",
    "sls.plot_predict_proba();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Epoch\n",
       "0     0.992135\n",
       "1     0.991744\n",
       "2     0.992092\n",
       "3     0.995478\n",
       "4     0.988289\n",
       "        ...   \n",
       "93    0.982225\n",
       "94    0.832531\n",
       "95    0.996213\n",
       "96    0.997057\n",
       "97    0.998157\n",
       "Length: 98, dtype: float64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# From the probabilities, we can extract a confidence level (ranging from 0 to 1) for each epoch.\n",
    "confidence = proba.max(1)\n",
    "confidence"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Exporting to a CSV file**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The predicted sleep stages can be exported to a CSV file with:\n",
    "hyp.hypno.to_csv(\"my_hypno.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>Stage</th>\n",
       "      <th>Confidence</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Epoch</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>WAKE</td>\n",
       "      <td>0.992135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>WAKE</td>\n",
       "      <td>0.991744</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>WAKE</td>\n",
       "      <td>0.992092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>WAKE</td>\n",
       "      <td>0.995478</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>WAKE</td>\n",
       "      <td>0.988289</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>WAKE</td>\n",
       "      <td>0.987672</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Stage  Confidence\n",
       "Epoch                  \n",
       "0      WAKE    0.992135\n",
       "1      WAKE    0.991744\n",
       "2      WAKE    0.992092\n",
       "3      WAKE    0.995478\n",
       "4      WAKE    0.988289\n",
       "5      WAKE    0.987672"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We can also add the confidence level:\n",
    "df_pred = hyp.hypno.to_frame()\n",
    "df_pred[\"Confidence\"] = confidence\n",
    "df_pred.head(6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Now export to a CSV file\n",
    "df_pred.to_csv(\"my_hypno.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Applying the detection using only a single EEG derivation**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/lib/python3.12/site-packages/sklearn/base.py:376: InconsistentVersionWarning: Trying to unpickle estimator LabelEncoder from version 0.24.2 when using version 1.5.1. This might lead to breaking code or invalid results. Use at your own risk. For more info please refer to:\n",
      "https://scikit-learn.org/stable/model_persistence.html#security-maintainability-limitations\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Epoch\n",
       "0     WAKE\n",
       "1     WAKE\n",
       "2     WAKE\n",
       "3     WAKE\n",
       "4     WAKE\n",
       "      ... \n",
       "93    WAKE\n",
       "94    WAKE\n",
       "95    WAKE\n",
       "96    WAKE\n",
       "97    WAKE\n",
       "Name: Stage, Length: 98, dtype: category\n",
       "Categories (7, object): ['WAKE', 'N1', 'N2', 'N3', 'REM', 'ART', 'UNS']"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Using just an EEG channel (= no EOG or EMG)\n",
    "y_pred = yasa.SleepStaging(raw, eeg_name=\"C4\").predict()\n",
    "y_pred.hypno"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": 4
}