{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# The distinct iterative prior updating process in ASD and TD individuals \n",
    "\n",
    "Research has shown individuals with autism spectrum disorder (ASD) display unique patterns in predictive processing, yet it remains controversial regarding what causes these atypical behaviors.  Both ASD individuals and typically developing (TD) counterparts participated in a task where they reproduced time durations over two sessions, one characterized by high volatility and the other by predictable sequence. Both sessions involved the same time durations, but the sequences differed in volatility. A visual stimulus (a disk) appeared for a given duration, and participants were asked to reproduce the duration by pressing a key.\n",
    "\n",
    "This repository contains the data and analysis scripts for this study.   The codes and data are organized as follows:\n",
    "\n",
    "## 1. Folder Structure\n",
    "\n",
    "1. `/experiments`: Experimental codes and instructions\n",
    "\n",
    "This sub-folder contains Matlab codes and instructions for the duration reproduction task. The sequences of the duration reproductions are stored in the sub-folder `/experiments/seqs`. Those sequences were used for matched participants. \n",
    "\n",
    "2. `/data`: raw data files\n",
    "\n",
    "- `rawdata.csv`: Raw reproduction trials of all participants\n",
    "- `parinfo.csv`: Participant information, including measured scores AQ, EQ, SQ, IQ etc. \n",
    "\n",
    "3. `/figures`: output figures. \n",
    "4. analysis scripts\n",
    "- `analysis-notebook.ipynb`: Jupyter notebook for data analysis\n",
    "- `kmodelY.py`: Python script for the Kalman filter two-state model\n",
    "- \n",
    "\n",
    "## 2. Data Analysis\n",
    "\n",
    "### 2.1 Import raw data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Nothing done.\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>Unnamed: 0</th>\n",
       "      <th>Duration</th>\n",
       "      <th>Volatility</th>\n",
       "      <th>trlNo</th>\n",
       "      <th>blkNo</th>\n",
       "      <th>dur1</th>\n",
       "      <th>pdur</th>\n",
       "      <th>production</th>\n",
       "      <th>vrep</th>\n",
       "      <th>Reproduction</th>\n",
       "      <th>sub</th>\n",
       "      <th>group</th>\n",
       "      <th>itd</th>\n",
       "      <th>rep_err</th>\n",
       "      <th>sequence</th>\n",
       "      <th>preDuration</th>\n",
       "      <th>blk1st</th>\n",
       "      <th>Order</th>\n",
       "      <th>preErr</th>\n",
       "      <th>outlier</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0.400</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.400</td>\n",
       "      <td>0.400</td>\n",
       "      <td>0.409</td>\n",
       "      <td>0.377</td>\n",
       "      <td>0.374</td>\n",
       "      <td>A31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.026</td>\n",
       "      <td>31</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>LV First</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0.500</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0.500</td>\n",
       "      <td>0.506</td>\n",
       "      <td>0.507</td>\n",
       "      <td>0.353</td>\n",
       "      <td>0.342</td>\n",
       "      <td>A31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>0.100</td>\n",
       "      <td>-0.158</td>\n",
       "      <td>31</td>\n",
       "      <td>0.400</td>\n",
       "      <td>0</td>\n",
       "      <td>LV First</td>\n",
       "      <td>-0.026</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>0.400</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0.400</td>\n",
       "      <td>0.400</td>\n",
       "      <td>0.406</td>\n",
       "      <td>0.342</td>\n",
       "      <td>0.326</td>\n",
       "      <td>A31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>-0.100</td>\n",
       "      <td>-0.074</td>\n",
       "      <td>31</td>\n",
       "      <td>0.506</td>\n",
       "      <td>0</td>\n",
       "      <td>LV First</td>\n",
       "      <td>-0.158</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0.400</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0.400</td>\n",
       "      <td>0.400</td>\n",
       "      <td>0.397</td>\n",
       "      <td>0.412</td>\n",
       "      <td>0.406</td>\n",
       "      <td>A31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.006</td>\n",
       "      <td>31</td>\n",
       "      <td>0.400</td>\n",
       "      <td>0</td>\n",
       "      <td>LV First</td>\n",
       "      <td>-0.074</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0.500</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>0.500</td>\n",
       "      <td>0.506</td>\n",
       "      <td>0.501</td>\n",
       "      <td>0.412</td>\n",
       "      <td>0.398</td>\n",
       "      <td>A31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>0.100</td>\n",
       "      <td>-0.102</td>\n",
       "      <td>31</td>\n",
       "      <td>0.400</td>\n",
       "      <td>0</td>\n",
       "      <td>LV First</td>\n",
       "      <td>0.006</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0  Duration Volatility  trlNo  blkNo  dur1  pdur  production  \\\n",
       "0           1     0.400  Low Vola.      1      1 0.400 0.400       0.409   \n",
       "1           2     0.500  Low Vola.      2      1 0.500 0.506       0.507   \n",
       "2           3     0.400  Low Vola.      3      1 0.400 0.400       0.406   \n",
       "3           4     0.400  Low Vola.      4      1 0.400 0.400       0.397   \n",
       "4           5     0.500  Low Vola.      5      1 0.500 0.506       0.501   \n",
       "\n",
       "   vrep  Reproduction  sub group    itd  rep_err  sequence  preDuration  \\\n",
       "0 0.377         0.374  A31   ASD    NaN   -0.026        31          NaN   \n",
       "1 0.353         0.342  A31   ASD  0.100   -0.158        31        0.400   \n",
       "2 0.342         0.326  A31   ASD -0.100   -0.074        31        0.506   \n",
       "3 0.412         0.406  A31   ASD  0.000    0.006        31        0.400   \n",
       "4 0.412         0.398  A31   ASD  0.100   -0.102        31        0.400   \n",
       "\n",
       "   blk1st     Order  preErr  outlier  \n",
       "0       1  LV First     NaN     True  \n",
       "1       0  LV First  -0.026    False  \n",
       "2       0  LV First  -0.158    False  \n",
       "3       0  LV First  -0.074    False  \n",
       "4       0  LV First   0.006    False  "
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# load data analysis packages\n",
    "%reset\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import os\n",
    "import statsmodels.api as sm\n",
    "import scipy.stats as stats\n",
    "from scipy.optimize import least_squares\n",
    "# pingouin ANOVA\n",
    "import pingouin as pg\n",
    "# keep output precision to 3 decimal places\n",
    "pd.options.display.float_format = '{:,.3f}'.format\n",
    "\n",
    "# read data from ./data/rawdata.csv\n",
    "rawdata = pd.read_csv('./data/rawdata.csv')\n",
    "# change rawdata.group to upper case\n",
    "rawdata['group'] = rawdata['group'].str.upper()\n",
    "# add a new column preErr, indicating the previous trial's error\n",
    "rawdata['preErr'] = rawdata.groupby(['sub', 'Volatility'])['rep_err'].shift(1)\n",
    "# mark the outliers that exceed [Duration/3, Duration * 3] or preErr is nan\n",
    "rawdata['outlier'] = (rawdata['Reproduction'] < rawdata['Duration']/3) | (rawdata['Reproduction'] > rawdata['Duration']*3) | (rawdata['preErr'].isna())\n",
    "\n",
    "# show the first 5 rows of rawdata\n",
    "rawdata.head()\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The key columns in `rawdata.csv` are:\n",
    "1. Duration: the duration of the visual stimulus\n",
    "2. Reproduction: the reproduced duration\n",
    "3. rep_err: the reproduction error (Reproduction - Duration)\n",
    "4. preDuration: the duration of the previous visual stimulus\n",
    "5. sub: subject ID\n",
    "6. group: ASD or TD\n",
    "7. Volatility: Low Vola. or High Vola."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The outlier trials were generally low for both groups: 2.3% for the ASD group and 1% for the TD group. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "group\n",
       "ASD   0.023\n",
       "TD    0.010\n",
       "Name: outlier, dtype: float64"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show the percentange of outliers in each group\n",
    "rawdata.groupby(['sub', 'group'])['outlier'].mean().reset_index().groupby('group')['outlier'].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.2 Duration sequences\n",
    "\n",
    "The experiment was structured into two distinct sessions, characterized by high and low volatility respectively. Each session included 500 trials, all following the same distribution, yet they varied in terms of sequential volatility. The provided figure illustrates a typical sequence of durations for a single participant.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# sns \n",
    "sns.set(style='ticks', context='paper', rc={'figure.figsize': (10, 5)})\n",
    "#sns.color_palette(palette='colorblind')\n",
    "sns.set_palette(\"Dark2\")\n",
    "\n",
    "# select a subset of rawdata (sub == 'ara27') for illustration\n",
    "subdata = rawdata.query('sub == \"ara27\"')\n",
    "# plot Duration as a function of trlNo, using different color for Volatility\n",
    "# use color palette 'colorblind' from seaborn\n",
    "#sns.set_palette('colorblind')\n",
    "sns.lineplot(x='trlNo', y='Duration', hue='Volatility', data=subdata)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.3 Central tendency and autocorrelation\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x350 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#  calculate the mean and standard deviation of Duration for each sub, group, Volatility, and Duration\n",
    "mdata = rawdata.query('outlier == False').groupby(\n",
    "    ['sub', 'group', 'Volatility', 'Duration','sequence','Order']).agg(\n",
    "        {'rep_err': ['mean', 'std']}).reset_index()\n",
    "mdata.columns = ['sub', 'group', 'Volatility', 'Duration', 'sequence','Order','rep_err', 'rep_err_std']\n",
    "\n",
    "# visualize the mean reproduction error as a function of Duration\n",
    "ax = sns.lmplot(x='Duration', y='rep_err', hue='Volatility', col='group', \n",
    "                scatter_kws = {'alpha':0.3, 's':5}, legend = False, height = 3.5, \n",
    "                data=mdata, hue_order = ['Low Vola.', 'High Vola.'])\n",
    "\n",
    "# add dashed line 0 to each subplot\n",
    "for ax1 in ax.axes.flat:\n",
    "    ax1.axhline(0, ls='--')\n",
    "plt.ylim(-0.5, 1.)\n",
    "# save the figure to ./figures/rep_err_vs_Duration.png\n",
    "plt.savefig('./figures/rep_err_vs_Duration_b.png', dpi=300)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In both group, the high volatility session showed a larger central tendency than the low volatility session. Let's do linear regression and durbin-watson test for the autocorrelation in two environments."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>sub</th>\n",
       "      <th>Volatility</th>\n",
       "      <th>sequence</th>\n",
       "      <th>group</th>\n",
       "      <th>level_4</th>\n",
       "      <th>r2</th>\n",
       "      <th>intercept</th>\n",
       "      <th>slope</th>\n",
       "      <th>cti</th>\n",
       "      <th>ar_dw</th>\n",
       "      <th>aic</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A31</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>0</td>\n",
       "      <td>0.389</td>\n",
       "      <td>0.284</td>\n",
       "      <td>-0.451</td>\n",
       "      <td>0.451</td>\n",
       "      <td>1.710</td>\n",
       "      <td>9.836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A31</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>0</td>\n",
       "      <td>0.459</td>\n",
       "      <td>0.279</td>\n",
       "      <td>-0.506</td>\n",
       "      <td>0.506</td>\n",
       "      <td>1.017</td>\n",
       "      <td>-4.688</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A32</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>32</td>\n",
       "      <td>ASD</td>\n",
       "      <td>0</td>\n",
       "      <td>0.606</td>\n",
       "      <td>0.550</td>\n",
       "      <td>-0.904</td>\n",
       "      <td>0.904</td>\n",
       "      <td>1.405</td>\n",
       "      <td>-86.745</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A32</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>32</td>\n",
       "      <td>ASD</td>\n",
       "      <td>0</td>\n",
       "      <td>0.020</td>\n",
       "      <td>0.222</td>\n",
       "      <td>-0.227</td>\n",
       "      <td>0.227</td>\n",
       "      <td>1.701</td>\n",
       "      <td>166.866</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A33</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>33</td>\n",
       "      <td>ASD</td>\n",
       "      <td>0</td>\n",
       "      <td>0.182</td>\n",
       "      <td>0.180</td>\n",
       "      <td>-0.239</td>\n",
       "      <td>0.239</td>\n",
       "      <td>1.783</td>\n",
       "      <td>-113.065</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sub  Volatility  sequence group  level_4    r2  intercept  slope   cti  \\\n",
       "0  A31  High Vola.        31   ASD        0 0.389      0.284 -0.451 0.451   \n",
       "1  A31   Low Vola.        31   ASD        0 0.459      0.279 -0.506 0.506   \n",
       "2  A32  High Vola.        32   ASD        0 0.606      0.550 -0.904 0.904   \n",
       "3  A32   Low Vola.        32   ASD        0 0.020      0.222 -0.227 0.227   \n",
       "4  A33  High Vola.        33   ASD        0 0.182      0.180 -0.239 0.239   \n",
       "\n",
       "   ar_dw      aic  \n",
       "0  1.710    9.836  \n",
       "1  1.017   -4.688  \n",
       "2  1.405  -86.745  \n",
       "3  1.701  166.866  \n",
       "4  1.783 -113.065  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# define a regression function with input of a dataframe\n",
    "from statsmodels.stats.stattools import durbin_watson\n",
    "# define a regression function, input is a dataframe and y and x \n",
    "def reg_func(data, x='Duration'):\n",
    "   # Check and remove NaNs or infinite values\n",
    "    data = data.replace([np.inf, -np.inf], np.nan).dropna(subset=[x, 'rep_err'])\n",
    "    \n",
    "    if data.empty or data[x].isnull().any() or data['rep_err'].isnull().any():\n",
    "        # Return None or some indication that the data was not valid for regression\n",
    "        return pd.DataFrame({'r2': [None], 'intercept': [None], 'slope': [None],\n",
    "            'cti': [None], 'ar_dw': [None], 'aic': [None]})\n",
    "\n",
    "    reg = sm.OLS(data['rep_err'], sm.add_constant(data[x])).fit()\n",
    "    dw = durbin_watson(reg.resid)\n",
    "    aic = reg.aic\n",
    "    # return goodness of fit, intercept, slope, and dw\n",
    "    return pd.DataFrame({'r2': reg.rsquared, 'intercept': reg.params.iloc[0], 'slope':reg.params.iloc[1],\n",
    "        'cti': -reg.params.iloc[1], 'ar_dw': dw, 'aic':reg.aic}, index=[0])\n",
    "# apply the regression function to each sub, Volatility, and group\n",
    "df_coef = rawdata.query('outlier == False').groupby(['sub', 'Volatility', 'sequence', 'group']).apply(reg_func).reset_index()\n",
    "# show the first 5 rows of df_coef\n",
    "df_coef.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.003</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.003</td>\n",
       "      <td>0.071</td>\n",
       "      <td>0.790</td>\n",
       "      <td>0.001</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>1.611</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>1.611</td>\n",
       "      <td>80.105</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.564</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.005</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.228</td>\n",
       "      <td>0.634</td>\n",
       "      <td>0.004</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS      F  p-unc   np2   eps\n",
       "0        group 0.003    1   62 0.003  0.071  0.790 0.001   NaN\n",
       "1   Volatility 1.611    1   62 1.611 80.105  0.000 0.564 1.000\n",
       "2  Interaction 0.005    1   62 0.005  0.228  0.634 0.004   NaN"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Mixed ANOVA on the central tendency effect cti\n",
    "# use pingouin to perform mixed ANOVA\n",
    "aov = pg.mixed_anova(data=df_coef, dv='cti', within='Volatility', between='group', subject='sub')\n",
    "# show the ANOVA table\n",
    "aov"
   ]
  },
  {
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.009</td>\n",
       "      <td>0.926</td>\n",
       "      <td>0.000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>1.370</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>1.370</td>\n",
       "      <td>84.976</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.578</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.010</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.010</td>\n",
       "      <td>0.650</td>\n",
       "      <td>0.423</td>\n",
       "      <td>0.010</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS      F  p-unc   np2   eps\n",
       "0        group 0.000    1   62 0.000  0.009  0.926 0.000   NaN\n",
       "1   Volatility 1.370    1   62 1.370 84.976  0.000 0.578 1.000\n",
       "2  Interaction 0.010    1   62 0.010  0.650  0.423 0.010   NaN"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Mixed ANOVA on the intercept\n",
    "aov = pg.mixed_anova(data=df_coef, \n",
    "                     dv='intercept', within='Volatility', between='group', subject='sub')\n",
    "# show the ANOVA table\n",
    "aov"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "However, autoregressive durbins-watson test values were significant different between the two groups, even with the outliers. This suggests that we need to consider the inter-trial updating of the prior information."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.314</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.314</td>\n",
       "      <td>4.498</td>\n",
       "      <td>0.038</td>\n",
       "      <td>0.068</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.059</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.059</td>\n",
       "      <td>2.142</td>\n",
       "      <td>0.148</td>\n",
       "      <td>0.033</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.060</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.060</td>\n",
       "      <td>2.183</td>\n",
       "      <td>0.145</td>\n",
       "      <td>0.034</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS     F  p-unc   np2   eps\n",
       "0        group 0.314    1   62 0.314 4.498  0.038 0.068   NaN\n",
       "1   Volatility 0.059    1   62 0.059 2.142  0.148 0.033 1.000\n",
       "2  Interaction 0.060    1   62 0.060 2.183  0.145 0.034   NaN"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pg.mixed_anova(data=df_coef, \n",
    "               dv='ar_dw', within='Volatility', between = 'group', subject='sub')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot mean and error bar of aw_dw as a function of group, Volatility\n",
    "#sns.barplot(data=df_coef.query(\"sequence not in @outliers_regress\"), \n",
    "#            x='group', y='ar_dw', hue='Volatility',capsize = .1,\n",
    "#            zorder = 5, errorbar=('ci', 68))\n",
    "# change y axis as 'DW index'\n",
    "# y axis from 1 to 2\n",
    "#plt.ylim(1, 2.4)\n",
    "# add stripplot\n",
    "sns.boxplot(data=df_coef, y='group', x='ar_dw', hue='Volatility', orient = 'h', hue_order=['Low Vola.', 'High Vola.'])\n",
    "# only show the last two legends\n",
    "handles, labels = plt.gca().get_legend_handles_labels()\n",
    "#plt.legend(handles[2:], labels[2:])\n",
    "# add dashed line 2 to indicate the 0 autocorrelation\n",
    "plt.axvline(2, ls='--', c='k')\n",
    "# remove box around the plot\n",
    "sns.despine()\n",
    "plt.xlabel('DW index')\n",
    "\n",
    "# save the figure to vector file ./figures/ar_dw.pdf\n",
    "plt.savefig('./figures/ar_dw.pdf', dpi=300)\n",
    "# save the figure to ./figures/ar_dw.png\n",
    "plt.savefig('./figures/ar_dw.png', dpi=300)\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "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>T</th>\n",
       "      <th>dof</th>\n",
       "      <th>alternative</th>\n",
       "      <th>p-val</th>\n",
       "      <th>CI95%</th>\n",
       "      <th>cohen-d</th>\n",
       "      <th>BF10</th>\n",
       "      <th>power</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>T-test</th>\n",
       "      <td>-7.913</td>\n",
       "      <td>63</td>\n",
       "      <td>two-sided</td>\n",
       "      <td>0.000</td>\n",
       "      <td>[1.72, 1.83]</td>\n",
       "      <td>0.989</td>\n",
       "      <td>1.893e+08</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            T  dof alternative  p-val         CI95%  cohen-d       BF10  power\n",
       "T-test -7.913   63   two-sided  0.000  [1.72, 1.83]    0.989  1.893e+08  1.000"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# let's compare if ar_dw was significantly lower than 2 in the two groups. \n",
    "# We will use a one-sample t-test to compare the mean of ar_dw to 2.\n",
    "# t-test for ar_dw\n",
    "pg.ttest(df_coef.query(\"group == 'ASD'\")['ar_dw'], 2)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>T</th>\n",
       "      <th>dof</th>\n",
       "      <th>alternative</th>\n",
       "      <th>p-val</th>\n",
       "      <th>CI95%</th>\n",
       "      <th>cohen-d</th>\n",
       "      <th>BF10</th>\n",
       "      <th>power</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>T-test</th>\n",
       "      <td>-11.877</td>\n",
       "      <td>63</td>\n",
       "      <td>two-sided</td>\n",
       "      <td>0.000</td>\n",
       "      <td>[1.62, 1.73]</td>\n",
       "      <td>1.485</td>\n",
       "      <td>5.748e+14</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             T  dof alternative  p-val         CI95%  cohen-d       BF10  \\\n",
       "T-test -11.877   63   two-sided  0.000  [1.62, 1.73]    1.485  5.748e+14   \n",
       "\n",
       "        power  \n",
       "T-test  1.000  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# for TD group\n",
    "pg.ttest(df_coef.query(\"group == 'TD'\")['ar_dw'], 2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.4. Outliers \n",
    "There was no difference in cti between groups, partly because there were three outliers in the ASD groups, as the CTI > 0.9. "
   ]
  },
  {
   "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>sub</th>\n",
       "      <th>Volatility</th>\n",
       "      <th>sequence</th>\n",
       "      <th>group</th>\n",
       "      <th>level_4</th>\n",
       "      <th>r2</th>\n",
       "      <th>intercept</th>\n",
       "      <th>slope</th>\n",
       "      <th>cti</th>\n",
       "      <th>ar_dw</th>\n",
       "      <th>aic</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A32</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>32</td>\n",
       "      <td>ASD</td>\n",
       "      <td>0</td>\n",
       "      <td>0.606</td>\n",
       "      <td>0.550</td>\n",
       "      <td>-0.904</td>\n",
       "      <td>0.904</td>\n",
       "      <td>1.405</td>\n",
       "      <td>-86.745</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>aril02</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>2</td>\n",
       "      <td>ASD</td>\n",
       "      <td>0</td>\n",
       "      <td>0.647</td>\n",
       "      <td>1.051</td>\n",
       "      <td>-0.997</td>\n",
       "      <td>0.997</td>\n",
       "      <td>1.313</td>\n",
       "      <td>-131.330</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>arm13</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>13</td>\n",
       "      <td>ASD</td>\n",
       "      <td>0</td>\n",
       "      <td>0.617</td>\n",
       "      <td>0.669</td>\n",
       "      <td>-0.907</td>\n",
       "      <td>0.907</td>\n",
       "      <td>1.650</td>\n",
       "      <td>-133.904</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       sub  Volatility  sequence group  level_4    r2  intercept  slope   cti  \\\n",
       "2      A32  High Vola.        32   ASD        0 0.606      0.550 -0.904 0.904   \n",
       "36  aril02  High Vola.         2   ASD        0 0.647      1.051 -0.997 0.997   \n",
       "50   arm13  High Vola.        13   ASD        0 0.617      0.669 -0.907 0.907   \n",
       "\n",
       "    ar_dw      aic  \n",
       "2   1.405  -86.745  \n",
       "36  1.313 -131.330  \n",
       "50  1.650 -133.904  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_coef.query(\"cti > 0.9\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's visualize these three outliers. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1050x350 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "outliers_regress = df_coef.query(\"cti > 0.9\").sequence\n",
    "# from mdata, select rows with sequence in outliers_regress, and plot rep_err vs. Duration\n",
    "# separate for individual sub, Volatility\n",
    "moutliers = rawdata.query('sequence in @outliers_regress and group == \"ASD\" and outlier == False')\n",
    "sns.set_palette('Dark2')\n",
    "ax = sns.lmplot(x='Duration', y='Reproduction', hue='Volatility', col='sub',\n",
    "                scatter_kws = {'alpha':0.3, 's':5}, legend = False, height = 3.5,\n",
    "                data=moutliers)\n",
    "ax.set(ylabel = 'Reproduction (s)')\n",
    "# remove subplots titles\n",
    "ax.set_titles('')\n",
    "# add diagonal dashed line to each subplot\n",
    "for ax in ax.axes.flat:\n",
    "    ax.plot(ax.get_xlim(), ax.get_xlim(), ls='--', c='k')\n",
    "\n",
    "# put legend in the upper right corner\n",
    "plt.legend(loc='upper right')\n",
    "# save the figure to ./figures/rep_err_vs_Duration_outliers.png\n",
    "plt.savefig('./figures/outliers.png', dpi=300)\n",
    "plt.savefig('./figures/outliers.pdf', dpi=300)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.047</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.047</td>\n",
       "      <td>1.734</td>\n",
       "      <td>0.193</td>\n",
       "      <td>0.030</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>1.204</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>1.204</td>\n",
       "      <td>81.366</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.592</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.049</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.049</td>\n",
       "      <td>3.324</td>\n",
       "      <td>0.074</td>\n",
       "      <td>0.056</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS      F  p-unc   np2   eps\n",
       "0        group 0.047    1   56 0.047  1.734  0.193 0.030   NaN\n",
       "1   Volatility 1.204    1   56 1.204 81.366  0.000 0.592 1.000\n",
       "2  Interaction 0.049    1   56 0.049  3.324  0.074 0.056   NaN"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Let's see what happens if we remove the outliers from the regression analysis.\n",
    "# ANOVA for slopes with group as between-subject factor and Volatility as within-subject factor\n",
    "pg.mixed_anova(data=df_coef.query(\"sequence not in @outliers_regress\"), \n",
    "               dv='cti', within='Volatility', between = 'group', subject='sub')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It remains not significant between groups, suggesting both groups had similar acquisition of the prior (Volatility) information. But let's check the residual autocorrelation.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.559</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.559</td>\n",
       "      <td>8.357</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.130</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.087</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.087</td>\n",
       "      <td>3.338</td>\n",
       "      <td>0.073</td>\n",
       "      <td>0.056</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.081</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.081</td>\n",
       "      <td>3.103</td>\n",
       "      <td>0.084</td>\n",
       "      <td>0.052</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS     F  p-unc   np2   eps\n",
       "0        group 0.559    1   56 0.559 8.357  0.005 0.130   NaN\n",
       "1   Volatility 0.087    1   56 0.087 3.338  0.073 0.056 1.000\n",
       "2  Interaction 0.081    1   56 0.081 3.103  0.084 0.052   NaN"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pg.mixed_anova(data=df_coef.query(\"sequence not in @outliers_regress\"), \n",
    "               dv='ar_dw', within='Volatility', between = 'group', subject='sub')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "By excluding the outliers, the difference between groups became even more significant, p = .005. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.5 Visualize the behavioral results \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x600 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# now we combine plots together for the paper\n",
    "sns.set(style='ticks', context='paper')\n",
    "fig, axes = plt.subplots(nrows = 2, ncols = 2, figsize = (7, 6))\n",
    "sns.set_palette('Dark2')\n",
    "# for each group, on seperate subplot, plot rep_err as a function of Duration, using different color for Volatility with regplot \n",
    "for i, group in enumerate(df_coef.group.unique()):\n",
    "    for j, vol in enumerate(df_coef.Volatility.unique()):\n",
    "        sns.regplot(data=mdata.query(\"group == @group and Volatility == @vol\"), \n",
    "                    x='Duration', y='rep_err', ax = axes[0,i], color = sns.color_palette('Dark2')[1-j], label = vol, \n",
    "                    ci = None, scatter_kws = {'alpha':0.3, 's':5})\n",
    "    if i == 1:\n",
    "        axes[0,i].legend(loc='lower left')\n",
    "    else:\n",
    "        # remove legend\n",
    "        axes[0,i].legend().remove()\n",
    "    # add dashed line 0 to each subplot\n",
    "    axes[0,i].axhline(0, ls='--', c='k')\n",
    "    # set y axis limit\n",
    "    axes[0,i].set_ylim(-0.6, 0.6)\n",
    "    # set x label to 'Duration (s)', y label to 'Reproduction error (s)'\n",
    "    axes[0,i].set(xlabel = 'Duration (s)', ylabel = 'Reproduction error (s)')\n",
    "    # set title to group\n",
    "    axes[0,i].set_title(group)\n",
    "# add horizontal boxplot for cti from df_coef on the second row first column\n",
    "sns.boxplot(data=df_coef, \n",
    "            y='group', x='cti', hue='Volatility', ax = axes[1,0],  width = 0.5, \n",
    "            orient = 'h', hue_order=['Low Vola.', 'High Vola.'])\n",
    "axes[1,0].set(xlabel = 'Central Tendency Index')\n",
    "axes[1,0].legend().remove()\n",
    "# add a vertical line at 0.0\n",
    "axes[1,0].axvline(0, ls='--', c='k')\n",
    "# remove y axis label\n",
    "axes[1,0].set(ylabel = '')\n",
    "# add horizontal boxplot for ar_dw from df_coef on the second row second column\n",
    "sns.boxplot(data=df_coef, y='group', x='ar_dw', hue='Volatility', width = 0.5, \n",
    "            ax = axes[1,1], orient = 'h', hue_order=['Low Vola.', 'High Vola.'])\n",
    "# legend off\n",
    "#axes[1,1].legend().remove()\n",
    "axes[1,1].legend(loc='lower right')\n",
    "axes[1,1].set(xlabel = 'Autocorrelation (DW) Index')\n",
    "# add a vertical line at 2.0\n",
    "axes[1,1].axvline(2, ls='--', c='k')\n",
    "# x axis from 0.9 to 3\n",
    "axes[1,1].set_xlim(0.9, 3)\n",
    "axes[1,1].set(ylabel = '')\n",
    "# remove box around the plot\n",
    "sns.despine()\n",
    "# add labels to subplots a, b, c, d\n",
    "for i, label in enumerate(['a', 'b', 'c', 'd']):\n",
    "    axes[int(i/2),i%2].text(-0.1, 1.1, label, transform=axes[int(i/2),i%2].transAxes, \n",
    "                            fontsize=16, fontweight='bold', va='top', ha='right')\n",
    "    \n",
    "\n",
    "# Adjust layout and show plot\n",
    "plt.tight_layout()\n",
    "\n",
    "# save fig to vector file ./figures/rep_err_vs_Duration.png\n",
    "plt.savefig('./figures/rep_err_vs_Duration.png', dpi=300, bbox_inches='tight', facecolor='white')\n",
    "plt.savefig('./figures/rep_err_vs_Duration.pdf', dpi=300, bbox_inches='tight', facecolor='white')\n",
    "\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.6 General bias (over- or under-reproduction) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.005</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.385</td>\n",
       "      <td>0.537</td>\n",
       "      <td>0.006</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.010</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.010</td>\n",
       "      <td>5.573</td>\n",
       "      <td>0.021</td>\n",
       "      <td>0.082</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.001</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.698</td>\n",
       "      <td>0.407</td>\n",
       "      <td>0.011</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS     F  p-unc   np2   eps\n",
       "0        group 0.005    1   62 0.005 0.385  0.537 0.006   NaN\n",
       "1   Volatility 0.010    1   62 0.010 5.573  0.021 0.082 1.000\n",
       "2  Interaction 0.001    1   62 0.001 0.698  0.407 0.011   NaN"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# average the reproduction errors for general biases\n",
    "mrep_err = rawdata.query(\"outlier == False\").groupby(['sub','group', 'Volatility'])['rep_err'].mean().reset_index()\n",
    "# pingouin mixed ANOVA on mrep_err\n",
    "aov = pg.mixed_anova(data=mrep_err, dv='rep_err', within='Volatility', between='group', subject='sub')\n",
    "# show the ANOVA table\n",
    "aov"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>group</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ASD</td>\n",
       "      <td>0.024</td>\n",
       "      <td>0.099</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>TD</td>\n",
       "      <td>0.036</td>\n",
       "      <td>0.062</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  group  mean   std\n",
       "0   ASD 0.024 0.099\n",
       "1    TD 0.036 0.062"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# calculate the mean reproduction error and standard deviation for each group and Volatility\n",
    "mrep_err.groupby(['group'])['rep_err'].agg(['mean', 'std']).reset_index()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.7 Two-state Iterative model\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# include in one notebook (the following code from KmodelY.py)\n",
    "from scipy.optimize import least_squares\n",
    "from statsmodels.stats.stattools import durbin_watson\n",
    "import numpy as np\n",
    "\n",
    "def fitKmodel(subdata, nolog=None, pfit=None, p0=None):\n",
    "    \"\"\"\n",
    "   Parameters:\n",
    "    - subdata : subject data\n",
    "    - nolog : If zero, logarithm is not used (default is 0)\n",
    "    - pfit : A list of logical values to indicate which parameters to fit (default is [True, True, True])\n",
    "    - p0 : Initial parameters (must always have length 3)\n",
    "    \n",
    "    Returns:\n",
    "    - px : Parameters of the model\n",
    "\n",
    "    By S.Glasauer 2019 (matlab), translated to Python by Strongway\n",
    "    # add AIC and DW\n",
    "\n",
    "    \"\"\"\n",
    "    \n",
    "    # Handle default arguments\n",
    "    if p0 is None:\n",
    "        p0 = [1., 1, 0]\n",
    "    if pfit is None:\n",
    "        pfit = [True, True, True]\n",
    "    if nolog is None:\n",
    "        nolog = 0\n",
    "\n",
    "    # Convert pfit to logical and filter p0\n",
    "    pfit = np.array(pfit, dtype=bool)\n",
    "    p0 = np.array(p0)[pfit]\n",
    "\n",
    "    # Lower bounds (lb) for the optimization\n",
    "    lb = np.array([0, 0, -np.inf])[pfit]\n",
    "    \n",
    "\n",
    "    # extract Duration and Reproduction from subdata as 2d array\n",
    "    x = subdata['Duration'].values\n",
    "    y = subdata['Reproduction'].values\n",
    "    # replace extreme y with nan with y > 3 * x or y < x/3\n",
    "    y[(y > 3 * x) | (y < x/3)] = np.nan\n",
    "    # combine x,y as 2d array\n",
    "    stimrep = np.vstack([x,y]).T\n",
    "    # Perform the optimization using least_squares (equivalent to lsqnonlin in MATLAB)\n",
    "    result = least_squares(kmodelY,  p0, args = (stimrep, 1),\n",
    "                           bounds=(lb, np.inf), method='trf')\n",
    "    \n",
    "    # calculate kalmann filter parameters\n",
    "    q11 = result.x[0]\n",
    "    q22 = result.x[1]\n",
    "    r = 1\n",
    "    # calculate residual sum of squares\n",
    "    rss = np.sum(result.fun**2)\n",
    "    dw = durbin_watson(result.fun)\n",
    "    # number of parameters\n",
    "    k = len(result.x)\n",
    "    # number of observations\n",
    "    n = len(stimrep)\n",
    "    # calculate the log-likelihood\n",
    "    ll = -n/2*(np.log(2*np.pi) + np.log(rss/n) + 1)\n",
    "    # calculate the Akaike information criterion (AIC)\n",
    "    aic = 2*k - 2*ll\n",
    "    # steady state solution\n",
    "    p22 = (q22+np.sqrt(q22*q22+4*(q11+r)*q22))/2\n",
    "    K = np.array([p22 + q11, p22])/(p22+q11+r)\n",
    "    # return the optimized parameters, steady state solution, and AIC\n",
    "    return np.append(np.append(result.x, K), [aic, dw])  # Optimized parameters\n",
    "\n",
    "\n",
    "def kmodelY(par, stimrep, nolog=1, pfit=[1, 1, 1]):\n",
    "    \"\"\"\n",
    "    Function to perform Kalman filter-based estimation.\n",
    "    \n",
    "    Parameters:\n",
    "    - par: Model parameters (if pfit = [1,1,1], then par = [q1/r, q2/r, cost-related parameter (0 for median)])\n",
    "    - stimrep: Stimulus representation\n",
    "    - nolog: Flag to decide if logarithm transformation is needed\n",
    "    - pfit: Parameter fitting list (note: len(par) = sum(pfit))\n",
    "    \n",
    "    Returns:\n",
    "    - sres: Stimulus residuals\n",
    "    - xest: Estimated state\n",
    "    - pest: Estimate error covariance\n",
    "    - resp: Response\n",
    "    - perr: Prediction error\n",
    "\n",
    "    S.Glasauer 2019/2023, translated to Python by Strongway\n",
    "    \"\"\"\n",
    "\n",
    "    # Convert pfit to a boolean array\n",
    "    pfit = np.array(pfit, dtype=bool)\n",
    "\n",
    "    # Adjust pfit based on the size of par\n",
    "    if len(par) < 3:\n",
    "        pfit[len(par):] = False\n",
    "    \n",
    "    # Adjust stimrep's shape for further processing\n",
    "    if stimrep.shape[1] == 1:\n",
    "        stimrep = np.tile(stimrep, (1, 2))\n",
    "    # the first column is the stimulus, the second column is the response, \n",
    "    # and add the third column to indicate the start of a new sequence\n",
    "    #if stimrep.shape[1] == 2:\n",
    "    #    stimrep = np.hstack((stimrep, np.zeros((stimrep.shape[0], 1))))\n",
    "    #    stimrep[0, 2] = 1\n",
    "    \n",
    "    # Initialize pars and overwrite with provided parameters based on pfit\n",
    "    pars = np.array([0.0, 0.0, 0.0])\n",
    "    pars[pfit] = par\n",
    "    par = pars\n",
    "\n",
    "    # Constants for the model\n",
    "    a = 10.0\n",
    "    off = 1.\n",
    "    r = 1.\n",
    "    q1 = par[0] * r\n",
    "    q2 = par[1] * r\n",
    "\n",
    "    # Define matrices Q, P, H, and F for the Kalman filter of two-state model\n",
    "    # details see Glasauer & Shi, 2022, Sci. Rep., https://doi.org/10.1038/s41598-022-14939-8\n",
    "    Q = np.array([[q1, 0], [0, q2]])\n",
    "    P = np.array([[r, 0], [0, r]])\n",
    "    H = np.array([[1., 0]])\n",
    "    F = np.array([[0, 1.], [0, 1.]])\n",
    "\n",
    "    # Apply logarithm transformation if nolog is false\n",
    "    if nolog:\n",
    "        z = stimrep[:, 0]\n",
    "    else: # log transformation\n",
    "        z = np.log(a * stimrep[:, 0] + off)\n",
    "\n",
    "    # Initialize state vector x\n",
    "    x = np.array([[z[0]], [z[0]]])\n",
    "\n",
    "    # Initialize matrices for storing results\n",
    "    xest = np.zeros((len(z), 2))\n",
    "    pest = np.zeros((len(z), 2))\n",
    "    perr = np.zeros(len(z))\n",
    "\n",
    "    # Kalman filter estimation loop\n",
    "    for i in range(len(z)):\n",
    "        \n",
    "        x = F@x\n",
    "        P = F@P@F.T + Q\n",
    "        K = P@H.T/(H@P@H.T + r)\n",
    "        perr[i] = z[i] - H@x\n",
    "        x = x + K*perr[i]\n",
    "        P = (np.eye(2) - K@H)@P\n",
    "\n",
    "        pest[i, :] = np.diag(P)\n",
    "        xest[i, :] = x.reshape(-1)\n",
    "\n",
    "    # Adjust for third parameter, if present\n",
    "    if len(par) == 3:\n",
    "        sh = par[2]\n",
    "    else:\n",
    "        sh = 0\n",
    "\n",
    "    # Compute response, adjusting for logarithm if needed\n",
    "    if nolog:\n",
    "        resp = xest[:, 0] + sh\n",
    "    else: # log transformation\n",
    "        resp = (np.exp(xest[:, 0] + sh) - off)/a \n",
    "        \n",
    "\n",
    "    # Calculate stimulus residuals\n",
    "    sres = stimrep[:, 1] - resp\n",
    "\n",
    "    # Remove NaNs from sres\n",
    "    sres = sres[np.isfinite(sres)]\n",
    "\n",
    "    return sres\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/dc/hksrz0yj5bb8n7f4_yptkcmw0000gn/T/ipykernel_88937/1675682148.py:146: DeprecationWarning: Conversion of an array with ndim > 0 to a scalar is deprecated, and will error in future. Ensure you extract a single element from your array before performing this operation. (Deprecated NumPy 1.25.)\n",
      "  perr[i] = z[i] - H@x\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>sub</th>\n",
       "      <th>Volatility</th>\n",
       "      <th>sequence</th>\n",
       "      <th>group</th>\n",
       "      <th>par</th>\n",
       "      <th>p1</th>\n",
       "      <th>p2</th>\n",
       "      <th>tau</th>\n",
       "      <th>K1</th>\n",
       "      <th>K2</th>\n",
       "      <th>AIC_2S</th>\n",
       "      <th>DW_2S</th>\n",
       "      <th>Kd</th>\n",
       "      <th>kseq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A31</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[0.9185124640252414, 0.05243938464301662, -0.1...</td>\n",
       "      <td>0.919</td>\n",
       "      <td>0.052</td>\n",
       "      <td>-0.170</td>\n",
       "      <td>0.558</td>\n",
       "      <td>0.152</td>\n",
       "      <td>-1.177</td>\n",
       "      <td>1.728</td>\n",
       "      <td>-0.406</td>\n",
       "      <td>0.067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A31</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[0.7563162466196017, 2.049556565459247e-05, -0...</td>\n",
       "      <td>0.756</td>\n",
       "      <td>0.000</td>\n",
       "      <td>-0.290</td>\n",
       "      <td>0.433</td>\n",
       "      <td>0.003</td>\n",
       "      <td>-42.879</td>\n",
       "      <td>1.181</td>\n",
       "      <td>-0.429</td>\n",
       "      <td>0.002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A32</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>32</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[0.09834201160694395, 1.1492562435182896e-18, ...</td>\n",
       "      <td>0.098</td>\n",
       "      <td>0.000</td>\n",
       "      <td>-0.330</td>\n",
       "      <td>0.090</td>\n",
       "      <td>0.000</td>\n",
       "      <td>-132.762</td>\n",
       "      <td>1.382</td>\n",
       "      <td>-0.090</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A32</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>32</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[3.342491279429646e-10, 0.4012756813450186, -0...</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.401</td>\n",
       "      <td>-0.001</td>\n",
       "      <td>0.464</td>\n",
       "      <td>0.464</td>\n",
       "      <td>158.794</td>\n",
       "      <td>1.669</td>\n",
       "      <td>-0.000</td>\n",
       "      <td>0.249</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A33</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>33</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[1.0307068777913497, 1.3110276998645982, -0.06...</td>\n",
       "      <td>1.031</td>\n",
       "      <td>1.311</td>\n",
       "      <td>-0.061</td>\n",
       "      <td>0.775</td>\n",
       "      <td>0.543</td>\n",
       "      <td>-134.099</td>\n",
       "      <td>1.838</td>\n",
       "      <td>-0.232</td>\n",
       "      <td>0.122</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sub  Volatility  sequence group  \\\n",
       "0  A31  High Vola.        31   ASD   \n",
       "1  A31   Low Vola.        31   ASD   \n",
       "2  A32  High Vola.        32   ASD   \n",
       "3  A32   Low Vola.        32   ASD   \n",
       "4  A33  High Vola.        33   ASD   \n",
       "\n",
       "                                                 par    p1    p2    tau    K1  \\\n",
       "0  [0.9185124640252414, 0.05243938464301662, -0.1... 0.919 0.052 -0.170 0.558   \n",
       "1  [0.7563162466196017, 2.049556565459247e-05, -0... 0.756 0.000 -0.290 0.433   \n",
       "2  [0.09834201160694395, 1.1492562435182896e-18, ... 0.098 0.000 -0.330 0.090   \n",
       "3  [3.342491279429646e-10, 0.4012756813450186, -0... 0.000 0.401 -0.001 0.464   \n",
       "4  [1.0307068777913497, 1.3110276998645982, -0.06... 1.031 1.311 -0.061 0.775   \n",
       "\n",
       "     K2   AIC_2S  DW_2S     Kd  kseq  \n",
       "0 0.152   -1.177  1.728 -0.406 0.067  \n",
       "1 0.003  -42.879  1.181 -0.429 0.002  \n",
       "2 0.000 -132.762  1.382 -0.090 0.000  \n",
       "3 0.464  158.794  1.669 -0.000 0.249  \n",
       "4 0.543 -134.099  1.838 -0.232 0.122  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fit two-state model\n",
    "# from kmodelY import * #(when use kmodelY.py)\n",
    "\n",
    "# fit two-state model\n",
    "# subsecting each subject, volatility, and estimate two-state model parameters\n",
    "df_kmodel = rawdata.groupby(\n",
    "    ['sub', 'Volatility', 'sequence', 'group']).apply(\n",
    "        fitKmodel).reset_index()\n",
    "df_kmodel.columns = ['sub', 'Volatility', 'sequence', 'group', 'par']\n",
    "# split the parameters to columns\n",
    "df_kmodel[['p1','p2','tau','K1', 'K2','AIC_2S','DW_2S']] = pd.DataFrame(df_kmodel['par'].tolist(), index=df_kmodel.index)\n",
    "# add a column for the difference between K1 and K2\n",
    "df_kmodel['Kd'] = df_kmodel['K2'] - df_kmodel['K1']\n",
    "# sequential dependence analytical results for randomized sequences\n",
    "df_kmodel['kseq'] = df_kmodel['K2']*(1-df_kmodel['K1'])\n",
    "df_kmodel.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>sub</th>\n",
       "      <th>Volatility</th>\n",
       "      <th>sequence</th>\n",
       "      <th>group</th>\n",
       "      <th>par</th>\n",
       "      <th>p1</th>\n",
       "      <th>p2</th>\n",
       "      <th>tau</th>\n",
       "      <th>K1</th>\n",
       "      <th>K2</th>\n",
       "      <th>...</th>\n",
       "      <th>Kd</th>\n",
       "      <th>kseq</th>\n",
       "      <th>level_4</th>\n",
       "      <th>r2</th>\n",
       "      <th>intercept</th>\n",
       "      <th>slope</th>\n",
       "      <th>cti</th>\n",
       "      <th>ar_dw</th>\n",
       "      <th>aic</th>\n",
       "      <th>dAIC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A31</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[0.9185124640252414, 0.05243938464301662, -0.1...</td>\n",
       "      <td>0.919</td>\n",
       "      <td>0.052</td>\n",
       "      <td>-0.170</td>\n",
       "      <td>0.558</td>\n",
       "      <td>0.152</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.406</td>\n",
       "      <td>0.067</td>\n",
       "      <td>0</td>\n",
       "      <td>0.389</td>\n",
       "      <td>0.284</td>\n",
       "      <td>-0.451</td>\n",
       "      <td>0.451</td>\n",
       "      <td>1.710</td>\n",
       "      <td>9.836</td>\n",
       "      <td>-11.013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A31</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[0.7563162466196017, 2.049556565459247e-05, -0...</td>\n",
       "      <td>0.756</td>\n",
       "      <td>0.000</td>\n",
       "      <td>-0.290</td>\n",
       "      <td>0.433</td>\n",
       "      <td>0.003</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.429</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0</td>\n",
       "      <td>0.459</td>\n",
       "      <td>0.279</td>\n",
       "      <td>-0.506</td>\n",
       "      <td>0.506</td>\n",
       "      <td>1.017</td>\n",
       "      <td>-4.688</td>\n",
       "      <td>-38.192</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A32</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>32</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[0.09834201160694395, 1.1492562435182896e-18, ...</td>\n",
       "      <td>0.098</td>\n",
       "      <td>0.000</td>\n",
       "      <td>-0.330</td>\n",
       "      <td>0.090</td>\n",
       "      <td>0.000</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.090</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0</td>\n",
       "      <td>0.606</td>\n",
       "      <td>0.550</td>\n",
       "      <td>-0.904</td>\n",
       "      <td>0.904</td>\n",
       "      <td>1.405</td>\n",
       "      <td>-86.745</td>\n",
       "      <td>-46.017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A32</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>32</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[3.342491279429646e-10, 0.4012756813450186, -0...</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.401</td>\n",
       "      <td>-0.001</td>\n",
       "      <td>0.464</td>\n",
       "      <td>0.464</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.000</td>\n",
       "      <td>0.249</td>\n",
       "      <td>0</td>\n",
       "      <td>0.020</td>\n",
       "      <td>0.222</td>\n",
       "      <td>-0.227</td>\n",
       "      <td>0.227</td>\n",
       "      <td>1.701</td>\n",
       "      <td>166.866</td>\n",
       "      <td>-8.072</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A33</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>33</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[1.0307068777913497, 1.3110276998645982, -0.06...</td>\n",
       "      <td>1.031</td>\n",
       "      <td>1.311</td>\n",
       "      <td>-0.061</td>\n",
       "      <td>0.775</td>\n",
       "      <td>0.543</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.232</td>\n",
       "      <td>0.122</td>\n",
       "      <td>0</td>\n",
       "      <td>0.182</td>\n",
       "      <td>0.180</td>\n",
       "      <td>-0.239</td>\n",
       "      <td>0.239</td>\n",
       "      <td>1.783</td>\n",
       "      <td>-113.065</td>\n",
       "      <td>-21.033</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   sub  Volatility  sequence group  \\\n",
       "0  A31  High Vola.        31   ASD   \n",
       "1  A31   Low Vola.        31   ASD   \n",
       "2  A32  High Vola.        32   ASD   \n",
       "3  A32   Low Vola.        32   ASD   \n",
       "4  A33  High Vola.        33   ASD   \n",
       "\n",
       "                                                 par    p1    p2    tau    K1  \\\n",
       "0  [0.9185124640252414, 0.05243938464301662, -0.1... 0.919 0.052 -0.170 0.558   \n",
       "1  [0.7563162466196017, 2.049556565459247e-05, -0... 0.756 0.000 -0.290 0.433   \n",
       "2  [0.09834201160694395, 1.1492562435182896e-18, ... 0.098 0.000 -0.330 0.090   \n",
       "3  [3.342491279429646e-10, 0.4012756813450186, -0... 0.000 0.401 -0.001 0.464   \n",
       "4  [1.0307068777913497, 1.3110276998645982, -0.06... 1.031 1.311 -0.061 0.775   \n",
       "\n",
       "     K2  ...     Kd  kseq  level_4    r2  intercept  slope   cti  ar_dw  \\\n",
       "0 0.152  ... -0.406 0.067        0 0.389      0.284 -0.451 0.451  1.710   \n",
       "1 0.003  ... -0.429 0.002        0 0.459      0.279 -0.506 0.506  1.017   \n",
       "2 0.000  ... -0.090 0.000        0 0.606      0.550 -0.904 0.904  1.405   \n",
       "3 0.464  ... -0.000 0.249        0 0.020      0.222 -0.227 0.227  1.701   \n",
       "4 0.543  ... -0.232 0.122        0 0.182      0.180 -0.239 0.239  1.783   \n",
       "\n",
       "       aic    dAIC  \n",
       "0    9.836 -11.013  \n",
       "1   -4.688 -38.192  \n",
       "2  -86.745 -46.017  \n",
       "3  166.866  -8.072  \n",
       "4 -113.065 -21.033  \n",
       "\n",
       "[5 rows x 22 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# combine df_kmodel and df_seq by sub, Volatility, and group\n",
    "kpars = df_kmodel.merge(df_coef, on=['sub', 'Volatility', 'sequence', 'group'])\n",
    "# show the first 5 rows of df_kmodel\n",
    "kpars['dAIC'] = kpars['AIC_2S'] - kpars['aic']\n",
    "kpars.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-10.75263200035032"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# mean benefits of two-state model in terms of AIC\n",
    "kpars['dAIC'].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "the AIC difference between the two models was significant, suggesting the two-state model was a better fit to the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.lines.Line2D at 0x28de8a210>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x350 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(ncols = 2, figsize = (7, 3.5))\n",
    "# barplot for DW_2S as a function of group, Volatility\n",
    "sns.barplot(data=kpars, x='group', y='DW_2S', hue='Volatility', capsize = .1,\n",
    "            zorder = 5, errorbar=('ci', 68), ax = axes[0])\n",
    "# change y axis to 1.5 to 2\n",
    "axes[0].set_ylim(1.5, 2)\n",
    "# add dashed line 2 to indicate the 0 autocorrelation\n",
    "axes[0].axhline(2, ls='--', c='k')\n",
    "axes[0].axhline(1.7, ls='--', c='k')\n",
    "# second subplot for the ar_dw as a function of group, Volatility\n",
    "sns.barplot(data=kpars, x='group', y='ar_dw', hue='Volatility', capsize = .1,\n",
    "            zorder = 5, errorbar=('ci', 68), ax = axes[1])\n",
    "# change y axis to 1.5 to 2\n",
    "axes[1].set_ylim(1.5, 2)\n",
    "# add dashed line 2 to indicate the 0 autocorrelation\n",
    "axes[1].axhline(2, ls='--', c='k')\n",
    "axes[1].axhline(1.7, ls='--', c='k')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">K1</th>\n",
       "      <th colspan=\"2\" halign=\"left\">K2</th>\n",
       "      <th colspan=\"2\" halign=\"left\">kseq</th>\n",
       "      <th colspan=\"2\" halign=\"left\">tau</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>group</th>\n",
       "      <th>Volatility</th>\n",
       "      <th></th>\n",
       "      <th></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 rowspan=\"2\" valign=\"top\">ASD</th>\n",
       "      <th>High Vola.</th>\n",
       "      <td>0.681</td>\n",
       "      <td>0.043</td>\n",
       "      <td>0.446</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.100</td>\n",
       "      <td>0.013</td>\n",
       "      <td>0.013</td>\n",
       "      <td>0.019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Low Vola.</th>\n",
       "      <td>0.642</td>\n",
       "      <td>0.036</td>\n",
       "      <td>0.344</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.089</td>\n",
       "      <td>0.016</td>\n",
       "      <td>0.041</td>\n",
       "      <td>0.017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">TD</th>\n",
       "      <th>High Vola.</th>\n",
       "      <td>0.682</td>\n",
       "      <td>0.026</td>\n",
       "      <td>0.396</td>\n",
       "      <td>0.039</td>\n",
       "      <td>0.113</td>\n",
       "      <td>0.012</td>\n",
       "      <td>0.031</td>\n",
       "      <td>0.010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Low Vola.</th>\n",
       "      <td>0.568</td>\n",
       "      <td>0.035</td>\n",
       "      <td>0.198</td>\n",
       "      <td>0.038</td>\n",
       "      <td>0.088</td>\n",
       "      <td>0.016</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.010</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    K1          K2        kseq         tau      \n",
       "                  mean   sem  mean   sem  mean   sem  mean   sem\n",
       "group Volatility                                                \n",
       "ASD   High Vola. 0.681 0.043 0.446 0.047 0.100 0.013 0.013 0.019\n",
       "      Low Vola.  0.642 0.036 0.344 0.057 0.089 0.016 0.041 0.017\n",
       "TD    High Vola. 0.682 0.026 0.396 0.039 0.113 0.012 0.031 0.010\n",
       "      Low Vola.  0.568 0.035 0.198 0.038 0.088 0.016 0.057 0.010"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "# average K1, K2, and tau for each Volatility, group, with standard error\n",
    "df_kmodel.groupby(['group', 'Volatility']).agg({'K1': ['mean', 'sem'], 'K2': ['mean', 'sem'], 'kseq': ['mean', 'sem'], 'tau': ['mean', 'sem']})\n",
    "\n",
    "# save kmodel_v to csv file ./data/kmodel_v.csv\n",
    "#kmodel_v.to_csv('./data/kmodel_v.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Statistics for those parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.043</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.043</td>\n",
       "      <td>0.685</td>\n",
       "      <td>0.411</td>\n",
       "      <td>0.011</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.188</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.188</td>\n",
       "      <td>10.613</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.146</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.046</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.046</td>\n",
       "      <td>2.620</td>\n",
       "      <td>0.111</td>\n",
       "      <td>0.041</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS      F  p-unc   np2   eps\n",
       "0        group 0.043    1   62 0.043  0.685  0.411 0.011   NaN\n",
       "1   Volatility 0.188    1   62 0.188 10.613  0.002 0.146 1.000\n",
       "2  Interaction 0.046    1   62 0.046  2.620  0.111 0.041   NaN"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# statistics for K1\n",
    "pg.mixed_anova(data=df_kmodel, dv='K1', within='Volatility', between = 'group', subject='sub')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.307</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.307</td>\n",
       "      <td>4.232</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.064</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.725</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.725</td>\n",
       "      <td>11.754</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.159</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.075</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.075</td>\n",
       "      <td>1.209</td>\n",
       "      <td>0.276</td>\n",
       "      <td>0.019</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS      F  p-unc   np2   eps\n",
       "0        group 0.307    1   62 0.307  4.232  0.044 0.064   NaN\n",
       "1   Volatility 0.725    1   62 0.725 11.754  0.001 0.159 1.000\n",
       "2  Interaction 0.075    1   62 0.075  1.209  0.276 0.019   NaN"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# statistics for K2\n",
    "pg.mixed_anova(data=df_kmodel, dv='K2', within='Volatility', between = 'group', subject='sub')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "K2 shows a significant difference between groups!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.009</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.009</td>\n",
       "      <td>0.794</td>\n",
       "      <td>0.376</td>\n",
       "      <td>0.013</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.024</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.024</td>\n",
       "      <td>11.835</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.160</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.013</td>\n",
       "      <td>0.911</td>\n",
       "      <td>0.000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS      F  p-unc   np2   eps\n",
       "0        group 0.009    1   62 0.009  0.794  0.376 0.013   NaN\n",
       "1   Volatility 0.024    1   62 0.024 11.835  0.001 0.160 1.000\n",
       "2  Interaction 0.000    1   62 0.000  0.013  0.911 0.000   NaN"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# statistics for tau\n",
    "pg.mixed_anova(data=df_kmodel, dv='tau', within='Volatility', between = 'group', subject='sub')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# exclude outliers\n",
    "kmodel_v = df_kmodel.query(\"sequence not in @outliers_regress\")\n",
    "#kmodel_v = df_kmodel"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualize the parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x350 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualize the mean and standard error of K1, K2, and tau as a function of group, Volatility\n",
    "sns.set(style='ticks', context='paper')\n",
    "sns.set_palette('Dark2')\n",
    "\n",
    "# three subplots side by side\n",
    "fig, axes = plt.subplots(ncols = 3, figsize = (9, 3.5))\n",
    "# barplot for K1 as a function of group, Volatility\n",
    "sns.barplot(data=df_kmodel, x='group', y='K1', hue='Volatility', capsize = .1,\n",
    "            zorder = 5, errorbar=('se'), ax = axes[0], hue_order=['Low Vola.', 'High Vola.'])\n",
    "# y axis from 0.5 to 1\n",
    "axes[0].set_ylim(0.1, .8)\n",
    "# legend top right\n",
    "axes[0].legend(loc='upper right')\n",
    "axes[0].legend().remove()\n",
    "# remove x axis label\n",
    "axes[0].set(xlabel = '')\n",
    "# barplot for K2 as a function of group, Volatility\n",
    "sns.barplot(data=df_kmodel, x='group', y='K2', hue='Volatility', capsize = .1,\n",
    "            zorder = 5, errorbar=('se'), ax = axes[1], hue_order=['Low Vola.', 'High Vola.'])\n",
    "# legend off\n",
    "#axes[1].legend().remove()\n",
    "axes[1].set_ylim(0.1, .8)\n",
    "axes[1].set(xlabel = '')\n",
    "\n",
    "# barplot for tau as a function of group, Volatility\n",
    "# change the tau from seconds to milliseconds\n",
    "df_kmodel['Tau'] = df_kmodel['tau']*1000\n",
    "sns.barplot(data=df_kmodel, x='group', y='Tau', hue='Volatility', capsize = .1,\n",
    "            zorder = 5, errorbar=('se'), ax = axes[2], hue_order=['Low Vola.', 'High Vola.'])\n",
    "# y label to 'tau (ms)'\n",
    "axes[2].set_ylabel('General Bias (ms)')\n",
    "# legend off\n",
    "axes[2].legend().remove()\n",
    "axes[2].set(xlabel = '')\n",
    "axes[2].set_ylim(0, 70)\n",
    "\n",
    "# remove box around the plot\n",
    "sns.despine()\n",
    "# add labels to subplots a, b, c, d\n",
    "for i, label in enumerate(['a', 'b', 'c']):\n",
    "    axes[i].text(-0.1, 1.1, label, transform=axes[i].transAxes, \n",
    "                            fontsize=16, fontweight='bold', va='top', ha='right')\n",
    "# tight layout\n",
    "plt.tight_layout()\n",
    "\n",
    "# save the figure to vector file ./figures/kmodel.png\n",
    "plt.savefig('./figures/kmodel.png', dpi=300)\n",
    "plt.savefig('./figures/kmodel.pdf', dpi=300)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">K1</th>\n",
       "      <th colspan=\"2\" halign=\"left\">K2</th>\n",
       "      <th colspan=\"2\" halign=\"left\">kseq</th>\n",
       "      <th colspan=\"2\" halign=\"left\">tau</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>group</th>\n",
       "      <th>Volatility</th>\n",
       "      <th></th>\n",
       "      <th></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 rowspan=\"2\" valign=\"top\">ASD</th>\n",
       "      <th>High Vola.</th>\n",
       "      <td>0.681</td>\n",
       "      <td>0.043</td>\n",
       "      <td>0.446</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.100</td>\n",
       "      <td>0.013</td>\n",
       "      <td>0.013</td>\n",
       "      <td>0.019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Low Vola.</th>\n",
       "      <td>0.642</td>\n",
       "      <td>0.036</td>\n",
       "      <td>0.344</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.089</td>\n",
       "      <td>0.016</td>\n",
       "      <td>0.041</td>\n",
       "      <td>0.017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">TD</th>\n",
       "      <th>High Vola.</th>\n",
       "      <td>0.682</td>\n",
       "      <td>0.026</td>\n",
       "      <td>0.396</td>\n",
       "      <td>0.039</td>\n",
       "      <td>0.113</td>\n",
       "      <td>0.012</td>\n",
       "      <td>0.031</td>\n",
       "      <td>0.010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Low Vola.</th>\n",
       "      <td>0.568</td>\n",
       "      <td>0.035</td>\n",
       "      <td>0.198</td>\n",
       "      <td>0.038</td>\n",
       "      <td>0.088</td>\n",
       "      <td>0.016</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.010</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    K1          K2        kseq         tau      \n",
       "                  mean   sem  mean   sem  mean   sem  mean   sem\n",
       "group Volatility                                                \n",
       "ASD   High Vola. 0.681 0.043 0.446 0.047 0.100 0.013 0.013 0.019\n",
       "      Low Vola.  0.642 0.036 0.344 0.057 0.089 0.016 0.041 0.017\n",
       "TD    High Vola. 0.682 0.026 0.396 0.039 0.113 0.012 0.031 0.010\n",
       "      Low Vola.  0.568 0.035 0.198 0.038 0.088 0.016 0.057 0.010"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# display those outliers parameters\n",
    "df_kmodel.groupby(['group', 'Volatility']).agg({'K1': ['mean', 'sem'], 'K2': ['mean', 'sem'], 'kseq': ['mean', 'sem'], 'tau': ['mean', 'sem']})\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.8 Would large K2 lead to a slow updating?\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Given that K2 was significantly larger in the ASD group, we would expect a slower updating of the prior information. So we split the trials into two: the first half and the second half. Due to reduction of the sample trials, we excluded the outliers in this analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# let's check the first half of the trials from each session from the rawdata\n",
    "# select the first half of the trials (trlNo 1-125 for the first session, 251-375 for the second) \n",
    "firsthalf_raw = rawdata.query('trlNo <= 125 or (trlNo >= 251 and trlNo <= 375)')\n",
    "# and second half of the trials (trlNo 126-250 for the first session, 376-500 for the second)\n",
    "secondhalf_raw = rawdata.query('trlNo >= 126 or trlNo >= 376')\n",
    "# calculate coefficient for the first half of the trials\n",
    "df_coef1 = firsthalf_raw.query('outlier == False and sequence not in @outliers_regress').groupby(['sub', 'Volatility', 'sequence', 'group']).apply(reg_func).reset_index()\n",
    "# calculate coefficient for the second half of the trials\n",
    "df_coef2 = secondhalf_raw.query('outlier == False and sequence not in @outliers_regress').groupby(['sub', 'Volatility', 'sequence', 'group']).apply(reg_func).reset_index()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x350 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "sns.set(style='ticks', context='paper')\n",
    "sns.set_palette('Dark2')\n",
    "# two subplots side by side\n",
    "fig, axes = plt.subplots(ncols = 2, figsize = (7, 3.5))\n",
    "# barplot for cti as a function of group, Volatility, and half\n",
    "sns.barplot(data=df_coef1, x='group', y='cti', hue='Volatility', capsize = .1,\n",
    "            zorder = 5, errorbar=('ci', 68), ax = axes[0], hue_order=['Low Vola.', 'High Vola.'])\n",
    "axes[0].set_ylim(0, 0.45)\n",
    "# legend off\n",
    "axes[0].legend().remove()\n",
    "# remove x axis label\n",
    "axes[0].set(xlabel = '')\n",
    "axes[0].set(ylabel = 'Central Tendency Index (CTI)')\n",
    "# add title to the subplot\n",
    "axes[0].set_title('The first half of trials')\n",
    "# second subplot for the ar_dw as a function of group, Volatility, and half\n",
    "sns.barplot(data=df_coef1, x='group', y='ar_dw', hue='Volatility', capsize = .1,\n",
    "            zorder = 5, errorbar=('ci', 68), ax = axes[1], hue_order=['Low Vola.', 'High Vola.'])\n",
    "axes[1].set(xlabel = '')\n",
    "axes[1].set(ylabel = 'DW index')\n",
    "# y axis from 1.5 to 2.5\n",
    "axes[1].set_ylim(1.5, 2.2)\n",
    "# add dashed line 2 to indicate the 0 autocorrelation\n",
    "#axes[1].axhline(2, ls='--', c='k')\n",
    "axes[1].set_title('The first half of trials')\n",
    "# add title to the subplot\n",
    "#axes[1].set_title('Changes in CTI')\n",
    "# remove box around the plot\n",
    "sns.despine()\n",
    "# add labels to subplots a, b, c, d\n",
    "for i, label in enumerate(['a', 'b']):\n",
    "    axes[i].text(-0.1, 1.1, label, transform=axes[i].transAxes, \n",
    "                            fontsize=16, fontweight='bold', va='top', ha='right')\n",
    "# tight layout\n",
    "plt.tight_layout()\n",
    "# save the figure to vector file ./figures/cti_half.png\n",
    "plt.savefig('./figures/cti_half.png', dpi=300)\n",
    "plt.savefig('./figures/cti_half.pdf', dpi=300)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.222</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.222</td>\n",
       "      <td>6.236</td>\n",
       "      <td>0.015</td>\n",
       "      <td>0.100</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.480</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.480</td>\n",
       "      <td>24.738</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.306</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.013</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.013</td>\n",
       "      <td>0.682</td>\n",
       "      <td>0.412</td>\n",
       "      <td>0.012</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS      F  p-unc   np2   eps\n",
       "0        group 0.222    1   56 0.222  6.236  0.015 0.100   NaN\n",
       "1   Volatility 0.480    1   56 0.480 24.738  0.000 0.306 1.000\n",
       "2  Interaction 0.013    1   56 0.013  0.682  0.412 0.012   NaN"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pg.mixed_anova(data=df_coef1, \n",
    "               dv='cti', within='Volatility', between = 'group', subject='sub')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The above statistics showed that the central tendency was significantly difference between groups in the first half! This suggests that the ASD group had a slower updating of the prior information, which might not have fully updated in the first half."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.544</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.544</td>\n",
       "      <td>7.202</td>\n",
       "      <td>0.010</td>\n",
       "      <td>0.114</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.036</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.036</td>\n",
       "      <td>0.694</td>\n",
       "      <td>0.408</td>\n",
       "      <td>0.012</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.027</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.027</td>\n",
       "      <td>0.522</td>\n",
       "      <td>0.473</td>\n",
       "      <td>0.009</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS     F  p-unc   np2   eps\n",
       "0        group 0.544    1   56 0.544 7.202  0.010 0.114   NaN\n",
       "1   Volatility 0.036    1   56 0.036 0.694  0.408 0.012 1.000\n",
       "2  Interaction 0.027    1   56 0.027 0.522  0.473 0.009   NaN"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pg.mixed_anova(data=df_coef1, \n",
    "               dv='ar_dw', within='Volatility', between = 'group', subject='sub')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "the autocorrelation index (DW) for the first half of the trials was also significant between two groups. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Compared to the entire sessions: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.047</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.047</td>\n",
       "      <td>1.734</td>\n",
       "      <td>0.193</td>\n",
       "      <td>0.030</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>1.204</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>1.204</td>\n",
       "      <td>81.366</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.592</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.049</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.049</td>\n",
       "      <td>3.324</td>\n",
       "      <td>0.074</td>\n",
       "      <td>0.056</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS      F  p-unc   np2   eps\n",
       "0        group 0.047    1   56 0.047  1.734  0.193 0.030   NaN\n",
       "1   Volatility 1.204    1   56 1.204 81.366  0.000 0.592 1.000\n",
       "2  Interaction 0.049    1   56 0.049  3.324  0.074 0.056   NaN"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pg.mixed_anova(data=df_coef.query(\"sequence not in @outliers_regress\"), \n",
    "               dv='cti', within='Volatility', between = 'group', subject='sub')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.559</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.559</td>\n",
       "      <td>8.357</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.130</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.087</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.087</td>\n",
       "      <td>3.338</td>\n",
       "      <td>0.073</td>\n",
       "      <td>0.056</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.081</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.081</td>\n",
       "      <td>3.103</td>\n",
       "      <td>0.084</td>\n",
       "      <td>0.052</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS     F  p-unc   np2   eps\n",
       "0        group 0.559    1   56 0.559 8.357  0.005 0.130   NaN\n",
       "1   Volatility 0.087    1   56 0.087 3.338  0.073 0.056 1.000\n",
       "2  Interaction 0.081    1   56 0.081 3.103  0.084 0.052   NaN"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pg.mixed_anova(data=df_coef.query(\"sequence not in @outliers_regress\"), \n",
    "               dv='ar_dw', within='Volatility', between = 'group', subject='sub')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fit the two-state model to the first half of the trials"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/dc/hksrz0yj5bb8n7f4_yptkcmw0000gn/T/ipykernel_88937/1675682148.py:146: DeprecationWarning: Conversion of an array with ndim > 0 to a scalar is deprecated, and will error in future. Ensure you extract a single element from your array before performing this operation. (Deprecated NumPy 1.25.)\n",
      "  perr[i] = z[i] - H@x\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>sub</th>\n",
       "      <th>Volatility</th>\n",
       "      <th>sequence</th>\n",
       "      <th>group</th>\n",
       "      <th>par</th>\n",
       "      <th>p1</th>\n",
       "      <th>p2</th>\n",
       "      <th>tau</th>\n",
       "      <th>K1</th>\n",
       "      <th>K2</th>\n",
       "      <th>AIC_2S</th>\n",
       "      <th>DW_2S</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A31</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[0.9714549830363363, 0.01693385718297349, -0.1...</td>\n",
       "      <td>0.971</td>\n",
       "      <td>0.017</td>\n",
       "      <td>-0.199</td>\n",
       "      <td>0.538</td>\n",
       "      <td>0.088</td>\n",
       "      <td>-7.870</td>\n",
       "      <td>1.933</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A31</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[2.0532000286870844e-22, 12.169718779060531, -...</td>\n",
       "      <td>0.000</td>\n",
       "      <td>12.170</td>\n",
       "      <td>-0.438</td>\n",
       "      <td>0.929</td>\n",
       "      <td>0.929</td>\n",
       "      <td>66.842</td>\n",
       "      <td>0.850</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A33</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>33</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[1.204485814584274, 0.7122651732928083, -0.058...</td>\n",
       "      <td>1.204</td>\n",
       "      <td>0.712</td>\n",
       "      <td>-0.058</td>\n",
       "      <td>0.741</td>\n",
       "      <td>0.429</td>\n",
       "      <td>-60.205</td>\n",
       "      <td>1.666</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A33</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>33</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[5.735340853415004e-12, 0.4293106313521714, -0...</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.429</td>\n",
       "      <td>-0.028</td>\n",
       "      <td>0.475</td>\n",
       "      <td>0.475</td>\n",
       "      <td>-117.713</td>\n",
       "      <td>2.117</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A34</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>34</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[0.06397846678007811, 0.4849299971316463, 0.00...</td>\n",
       "      <td>0.064</td>\n",
       "      <td>0.485</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.516</td>\n",
       "      <td>0.485</td>\n",
       "      <td>-133.310</td>\n",
       "      <td>1.814</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sub  Volatility  sequence group  \\\n",
       "0  A31  High Vola.        31   ASD   \n",
       "1  A31   Low Vola.        31   ASD   \n",
       "2  A33  High Vola.        33   ASD   \n",
       "3  A33   Low Vola.        33   ASD   \n",
       "4  A34  High Vola.        34   ASD   \n",
       "\n",
       "                                                 par    p1     p2    tau  \\\n",
       "0  [0.9714549830363363, 0.01693385718297349, -0.1... 0.971  0.017 -0.199   \n",
       "1  [2.0532000286870844e-22, 12.169718779060531, -... 0.000 12.170 -0.438   \n",
       "2  [1.204485814584274, 0.7122651732928083, -0.058... 1.204  0.712 -0.058   \n",
       "3  [5.735340853415004e-12, 0.4293106313521714, -0... 0.000  0.429 -0.028   \n",
       "4  [0.06397846678007811, 0.4849299971316463, 0.00... 0.064  0.485  0.005   \n",
       "\n",
       "     K1    K2   AIC_2S  DW_2S  \n",
       "0 0.538 0.088   -7.870  1.933  \n",
       "1 0.929 0.929   66.842  0.850  \n",
       "2 0.741 0.429  -60.205  1.666  \n",
       "3 0.475 0.475 -117.713  2.117  \n",
       "4 0.516 0.485 -133.310  1.814  "
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# now fit two-state model to the first half of the trials\n",
    "# subsecting each subject, volatility, and estimate two-state model parameters\n",
    "df_kmodel1 = firsthalf_raw.query(\"sequence not in @outliers_regress\").groupby(\n",
    "    ['sub', 'Volatility', 'sequence', 'group']).apply(\n",
    "        fitKmodel).reset_index()\n",
    "df_kmodel1.columns = ['sub', 'Volatility', 'sequence', 'group', 'par']\n",
    "# split the parameters to columns\n",
    "df_kmodel1[['p1','p2','tau','K1', 'K2','AIC_2S','DW_2S']] = pd.DataFrame(df_kmodel1['par'].tolist(), \n",
    "                                                                         index=df_kmodel1.index)\n",
    "df_kmodel1.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">K1</th>\n",
       "      <th colspan=\"2\" halign=\"left\">K2</th>\n",
       "      <th colspan=\"2\" halign=\"left\">tau</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>group</th>\n",
       "      <th>Volatility</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 rowspan=\"2\" valign=\"top\">ASD</th>\n",
       "      <th>High Vola.</th>\n",
       "      <td>0.750</td>\n",
       "      <td>0.029</td>\n",
       "      <td>0.466</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.028</td>\n",
       "      <td>0.016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Low Vola.</th>\n",
       "      <td>0.610</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.333</td>\n",
       "      <td>0.058</td>\n",
       "      <td>0.034</td>\n",
       "      <td>0.022</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">TD</th>\n",
       "      <th>High Vola.</th>\n",
       "      <td>0.685</td>\n",
       "      <td>0.026</td>\n",
       "      <td>0.396</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.012</td>\n",
       "      <td>0.011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Low Vola.</th>\n",
       "      <td>0.554</td>\n",
       "      <td>0.042</td>\n",
       "      <td>0.178</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.052</td>\n",
       "      <td>0.013</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    K1          K2         tau      \n",
       "                  mean   sem  mean   sem  mean   sem\n",
       "group Volatility                                    \n",
       "ASD   High Vola. 0.750 0.029 0.466 0.057 0.028 0.016\n",
       "      Low Vola.  0.610 0.046 0.333 0.058 0.034 0.022\n",
       "TD    High Vola. 0.685 0.026 0.396 0.048 0.012 0.011\n",
       "      Low Vola.  0.554 0.042 0.178 0.047 0.052 0.013"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show the average K1, K2 and tau for each group and Volatility\n",
    "df_kmodel1.groupby(['group', 'Volatility']).agg({'K1': ['mean', 'sem'], 'K2': ['mean', 'sem'], 'tau': ['mean', 'sem']})\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/dc/hksrz0yj5bb8n7f4_yptkcmw0000gn/T/ipykernel_88937/1675682148.py:146: DeprecationWarning: Conversion of an array with ndim > 0 to a scalar is deprecated, and will error in future. Ensure you extract a single element from your array before performing this operation. (Deprecated NumPy 1.25.)\n",
      "  perr[i] = z[i] - H@x\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>sub</th>\n",
       "      <th>Volatility</th>\n",
       "      <th>sequence</th>\n",
       "      <th>group</th>\n",
       "      <th>par</th>\n",
       "      <th>p1</th>\n",
       "      <th>p2</th>\n",
       "      <th>tau</th>\n",
       "      <th>K1</th>\n",
       "      <th>K2</th>\n",
       "      <th>AIC_2S</th>\n",
       "      <th>DW_2S</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A31</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[0.9185124640252414, 0.05243938464301662, -0.1...</td>\n",
       "      <td>0.919</td>\n",
       "      <td>0.052</td>\n",
       "      <td>-0.170</td>\n",
       "      <td>0.558</td>\n",
       "      <td>0.152</td>\n",
       "      <td>-1.177</td>\n",
       "      <td>1.728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A31</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>31</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[1.511976006043292, 0.5687978606678477, -0.024...</td>\n",
       "      <td>1.512</td>\n",
       "      <td>0.569</td>\n",
       "      <td>-0.025</td>\n",
       "      <td>0.752</td>\n",
       "      <td>0.376</td>\n",
       "      <td>-112.774</td>\n",
       "      <td>1.116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A33</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>33</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[1.0307068777913497, 1.3110276998645982, -0.06...</td>\n",
       "      <td>1.031</td>\n",
       "      <td>1.311</td>\n",
       "      <td>-0.061</td>\n",
       "      <td>0.775</td>\n",
       "      <td>0.543</td>\n",
       "      <td>-134.099</td>\n",
       "      <td>1.838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A33</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>33</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[1.9645484931303274, 0.18705402981725533, -0.0...</td>\n",
       "      <td>1.965</td>\n",
       "      <td>0.187</td>\n",
       "      <td>-0.045</td>\n",
       "      <td>0.737</td>\n",
       "      <td>0.222</td>\n",
       "      <td>-86.835</td>\n",
       "      <td>1.842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A34</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>34</td>\n",
       "      <td>ASD</td>\n",
       "      <td>[0.25349416480762293, 0.40788636070138073, -0....</td>\n",
       "      <td>0.253</td>\n",
       "      <td>0.408</td>\n",
       "      <td>-0.020</td>\n",
       "      <td>0.546</td>\n",
       "      <td>0.430</td>\n",
       "      <td>-272.994</td>\n",
       "      <td>1.652</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sub  Volatility  sequence group  \\\n",
       "0  A31  High Vola.        31   ASD   \n",
       "1  A31   Low Vola.        31   ASD   \n",
       "2  A33  High Vola.        33   ASD   \n",
       "3  A33   Low Vola.        33   ASD   \n",
       "4  A34  High Vola.        34   ASD   \n",
       "\n",
       "                                                 par    p1    p2    tau    K1  \\\n",
       "0  [0.9185124640252414, 0.05243938464301662, -0.1... 0.919 0.052 -0.170 0.558   \n",
       "1  [1.511976006043292, 0.5687978606678477, -0.024... 1.512 0.569 -0.025 0.752   \n",
       "2  [1.0307068777913497, 1.3110276998645982, -0.06... 1.031 1.311 -0.061 0.775   \n",
       "3  [1.9645484931303274, 0.18705402981725533, -0.0... 1.965 0.187 -0.045 0.737   \n",
       "4  [0.25349416480762293, 0.40788636070138073, -0.... 0.253 0.408 -0.020 0.546   \n",
       "\n",
       "     K2   AIC_2S  DW_2S  \n",
       "0 0.152   -1.177  1.728  \n",
       "1 0.376 -112.774  1.116  \n",
       "2 0.543 -134.099  1.838  \n",
       "3 0.222  -86.835  1.842  \n",
       "4 0.430 -272.994  1.652  "
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Fit the second half of the trials\n",
    "df_kmodel2 = secondhalf_raw.query(\"sequence not in @outliers_regress\").groupby(\n",
    "    ['sub', 'Volatility', 'sequence', 'group']).apply(\n",
    "        fitKmodel).reset_index()\n",
    "df_kmodel2.columns = ['sub', 'Volatility', 'sequence', 'group', 'par']\n",
    "# split the parameters to columns\n",
    "df_kmodel2[['p1','p2','tau','K1', 'K2','AIC_2S','DW_2S']] = pd.DataFrame(df_kmodel2['par'].tolist(), \n",
    "                                                                         index=df_kmodel2.index)\n",
    "df_kmodel2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">K1</th>\n",
       "      <th colspan=\"2\" halign=\"left\">K2</th>\n",
       "      <th colspan=\"2\" halign=\"left\">tau</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>group</th>\n",
       "      <th>Volatility</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 rowspan=\"2\" valign=\"top\">ASD</th>\n",
       "      <th>High Vola.</th>\n",
       "      <td>0.728</td>\n",
       "      <td>0.028</td>\n",
       "      <td>0.482</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.038</td>\n",
       "      <td>0.015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Low Vola.</th>\n",
       "      <td>0.673</td>\n",
       "      <td>0.034</td>\n",
       "      <td>0.360</td>\n",
       "      <td>0.059</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">TD</th>\n",
       "      <th>High Vola.</th>\n",
       "      <td>0.655</td>\n",
       "      <td>0.027</td>\n",
       "      <td>0.385</td>\n",
       "      <td>0.036</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Low Vola.</th>\n",
       "      <td>0.576</td>\n",
       "      <td>0.036</td>\n",
       "      <td>0.267</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.012</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    K1          K2         tau      \n",
       "                  mean   sem  mean   sem  mean   sem\n",
       "group Volatility                                    \n",
       "ASD   High Vola. 0.728 0.028 0.482 0.044 0.038 0.015\n",
       "      Low Vola.  0.673 0.034 0.360 0.059 0.057 0.014\n",
       "TD    High Vola. 0.655 0.027 0.385 0.036 0.044 0.013\n",
       "      Low Vola.  0.576 0.036 0.267 0.044 0.060 0.012"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_kmodel2.groupby(['group', 'Volatility']).agg({'K1': ['mean', 'sem'], 'K2': ['mean', 'sem'], 'tau': ['mean', 'sem']})\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">K1</th>\n",
       "      <th colspan=\"2\" halign=\"left\">K2</th>\n",
       "      <th colspan=\"2\" halign=\"left\">tau</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>group</th>\n",
       "      <th>Volatility</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 rowspan=\"2\" valign=\"top\">ASD</th>\n",
       "      <th>High Vola.</th>\n",
       "      <td>0.681</td>\n",
       "      <td>0.043</td>\n",
       "      <td>0.446</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.013</td>\n",
       "      <td>0.019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Low Vola.</th>\n",
       "      <td>0.642</td>\n",
       "      <td>0.036</td>\n",
       "      <td>0.344</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.041</td>\n",
       "      <td>0.017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">TD</th>\n",
       "      <th>High Vola.</th>\n",
       "      <td>0.682</td>\n",
       "      <td>0.026</td>\n",
       "      <td>0.396</td>\n",
       "      <td>0.039</td>\n",
       "      <td>0.031</td>\n",
       "      <td>0.010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Low Vola.</th>\n",
       "      <td>0.568</td>\n",
       "      <td>0.035</td>\n",
       "      <td>0.198</td>\n",
       "      <td>0.038</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.010</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    K1          K2         tau      \n",
       "                  mean   sem  mean   sem  mean   sem\n",
       "group Volatility                                    \n",
       "ASD   High Vola. 0.681 0.043 0.446 0.047 0.013 0.019\n",
       "      Low Vola.  0.642 0.036 0.344 0.057 0.041 0.017\n",
       "TD    High Vola. 0.682 0.026 0.396 0.039 0.031 0.010\n",
       "      Low Vola.  0.568 0.035 0.198 0.038 0.057 0.010"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# compared to df_kmodel\n",
    "df_kmodel.groupby(['group', 'Volatility']).agg({'K1': ['mean', 'sem'], 'K2': ['mean', 'sem'], 'tau': ['mean', 'sem']})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">dK1</th>\n",
       "      <th colspan=\"2\" halign=\"left\">dK2</th>\n",
       "      <th colspan=\"2\" halign=\"left\">dtau</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>group</th>\n",
       "      <th>Volatility</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 rowspan=\"2\" valign=\"top\">ASD</th>\n",
       "      <th>High Vola.</th>\n",
       "      <td>0.022</td>\n",
       "      <td>0.015</td>\n",
       "      <td>-0.017</td>\n",
       "      <td>0.057</td>\n",
       "      <td>-0.010</td>\n",
       "      <td>0.008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Low Vola.</th>\n",
       "      <td>-0.063</td>\n",
       "      <td>0.038</td>\n",
       "      <td>-0.027</td>\n",
       "      <td>0.065</td>\n",
       "      <td>-0.023</td>\n",
       "      <td>0.017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">TD</th>\n",
       "      <th>High Vola.</th>\n",
       "      <td>0.031</td>\n",
       "      <td>0.015</td>\n",
       "      <td>0.011</td>\n",
       "      <td>0.051</td>\n",
       "      <td>-0.032</td>\n",
       "      <td>0.011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Low Vola.</th>\n",
       "      <td>-0.022</td>\n",
       "      <td>0.035</td>\n",
       "      <td>-0.089</td>\n",
       "      <td>0.060</td>\n",
       "      <td>-0.008</td>\n",
       "      <td>0.010</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    dK1          dK2         dtau      \n",
       "                   mean   sem   mean   sem   mean   sem\n",
       "group Volatility                                       \n",
       "ASD   High Vola.  0.022 0.015 -0.017 0.057 -0.010 0.008\n",
       "      Low Vola.  -0.063 0.038 -0.027 0.065 -0.023 0.017\n",
       "TD    High Vola.  0.031 0.015  0.011 0.051 -0.032 0.011\n",
       "      Low Vola.  -0.022 0.035 -0.089 0.060 -0.008 0.010"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# combine df_kmodel2 and df_kmodel1 by sub, Volatility, and group, for K1, K2, and tau, and add differences between two \n",
    "# K1, K2, and tau\n",
    "kmodel_v1 = df_kmodel1.merge(df_kmodel2, on=['sub', 'Volatility', 'sequence', 'group'])\n",
    "kmodel_v1['dK1'] = kmodel_v1['K1_x'] - kmodel_v1['K1_y']\n",
    "kmodel_v1['dK2'] = kmodel_v1['K2_x'] - kmodel_v1['K2_y']\n",
    "kmodel_v1['dtau'] = kmodel_v1['tau_x'] - kmodel_v1['tau_y']\n",
    "# show the average dk1, dk2, and dtau for each group and Volatility\n",
    "kmodel_v1.groupby(['group', 'Volatility']).agg({'dK1': ['mean', 'sem'], 'dK2': ['mean', 'sem'], 'dtau': ['mean', 'sem']})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.018</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.018</td>\n",
       "      <td>0.751</td>\n",
       "      <td>0.390</td>\n",
       "      <td>0.013</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.136</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.136</td>\n",
       "      <td>6.182</td>\n",
       "      <td>0.016</td>\n",
       "      <td>0.099</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.007</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.007</td>\n",
       "      <td>0.331</td>\n",
       "      <td>0.567</td>\n",
       "      <td>0.006</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS     F  p-unc   np2   eps\n",
       "0        group 0.018    1   56 0.018 0.751  0.390 0.013   NaN\n",
       "1   Volatility 0.136    1   56 0.136 6.182  0.016 0.099 1.000\n",
       "2  Interaction 0.007    1   56 0.007 0.331  0.567 0.006   NaN"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# anova test for dK1\n",
    "pg.mixed_anova(data=kmodel_v1, \n",
    "               dv='dK1', within='Volatility', between = 'group', subject='sub')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">dK1</th>\n",
       "      <th colspan=\"2\" halign=\"left\">dK2</th>\n",
       "      <th colspan=\"2\" halign=\"left\">dtau</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Volatility</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>High Vola.</th>\n",
       "      <td>0.026</td>\n",
       "      <td>0.011</td>\n",
       "      <td>-0.003</td>\n",
       "      <td>0.038</td>\n",
       "      <td>-0.021</td>\n",
       "      <td>0.007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Low Vola.</th>\n",
       "      <td>-0.042</td>\n",
       "      <td>0.026</td>\n",
       "      <td>-0.058</td>\n",
       "      <td>0.044</td>\n",
       "      <td>-0.016</td>\n",
       "      <td>0.010</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              dK1          dK2         dtau      \n",
       "             mean   sem   mean   sem   mean   sem\n",
       "Volatility                                       \n",
       "High Vola.  0.026 0.011 -0.003 0.038 -0.021 0.007\n",
       "Low Vola.  -0.042 0.026 -0.058 0.044 -0.016 0.010"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show the average dk1, dk2, and dtau for each group and Volatility\n",
    "kmodel_v1.groupby([ 'Volatility']).agg({'dK1': ['mean', 'sem'], 'dK2': ['mean', 'sem'], 'dtau': ['mean', 'sem']})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.105</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.105</td>\n",
       "      <td>1.914</td>\n",
       "      <td>0.172</td>\n",
       "      <td>0.033</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.532</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.532</td>\n",
       "      <td>21.887</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.281</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.001</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.021</td>\n",
       "      <td>0.885</td>\n",
       "      <td>0.000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS      F  p-unc   np2   eps\n",
       "0        group 0.105    1   56 0.105  1.914  0.172 0.033   NaN\n",
       "1   Volatility 0.532    1   56 0.532 21.887  0.000 0.281 1.000\n",
       "2  Interaction 0.001    1   56 0.001  0.021  0.885 0.000   NaN"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pg.mixed_anova(data=df_kmodel1, \n",
    "               dv='K1', within='Volatility', between = 'group', subject='sub')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.009</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.009</td>\n",
       "      <td>0.077</td>\n",
       "      <td>0.782</td>\n",
       "      <td>0.001</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.089</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.089</td>\n",
       "      <td>1.107</td>\n",
       "      <td>0.297</td>\n",
       "      <td>0.019</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.059</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.059</td>\n",
       "      <td>0.736</td>\n",
       "      <td>0.395</td>\n",
       "      <td>0.013</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS     F  p-unc   np2   eps\n",
       "0        group 0.009    1   56 0.009 0.077  0.782 0.001   NaN\n",
       "1   Volatility 0.089    1   56 0.089 1.107  0.297 0.019 1.000\n",
       "2  Interaction 0.059    1   56 0.059 0.736  0.395 0.013   NaN"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# anova test for dK2\n",
    "pg.mixed_anova(data=kmodel_v1, \n",
    "               dv='dK2', within='Volatility', between = 'group', subject='sub')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.369</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.369</td>\n",
       "      <td>4.514</td>\n",
       "      <td>0.038</td>\n",
       "      <td>0.075</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.893</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.893</td>\n",
       "      <td>11.179</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.166</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.053</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.053</td>\n",
       "      <td>0.661</td>\n",
       "      <td>0.420</td>\n",
       "      <td>0.012</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS      F  p-unc   np2   eps\n",
       "0        group 0.369    1   56 0.369  4.514  0.038 0.075   NaN\n",
       "1   Volatility 0.893    1   56 0.893 11.179  0.001 0.166 1.000\n",
       "2  Interaction 0.053    1   56 0.053  0.661  0.420 0.012   NaN"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# anova test for dK1\n",
    "pg.mixed_anova(data=df_kmodel1, \n",
    "               dv='K2', within='Volatility', between = 'group', subject='sub')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.217</td>\n",
       "      <td>0.643</td>\n",
       "      <td>0.004</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.078</td>\n",
       "      <td>0.781</td>\n",
       "      <td>0.001</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.003</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.003</td>\n",
       "      <td>1.626</td>\n",
       "      <td>0.208</td>\n",
       "      <td>0.028</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS     F  p-unc   np2   eps\n",
       "0        group 0.000    1   56 0.000 0.217  0.643 0.004   NaN\n",
       "1   Volatility 0.000    1   56 0.000 0.078  0.781 0.001 1.000\n",
       "2  Interaction 0.003    1   56 0.003 1.626  0.208 0.028   NaN"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# anova test for dtau\n",
    "pg.mixed_anova(data=kmodel_v1, \n",
    "               dv='dtau', within='Volatility', between = 'group', subject='sub')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.9 General Linear Model with the current and previous trial durations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Given that the low-volatility session used the random walk sequence, the current and previous durations were highly correlated. So we use the difference between the current and previous durations as a regressor. This won't affect the fully random sequence in the high-volatility session.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/dc/hksrz0yj5bb8n7f4_yptkcmw0000gn/T/ipykernel_88937/4187532016.py:34: FutureWarning: The behavior of DataFrame concatenation with empty or all-NA entries is deprecated. In a future version, this will no longer exclude empty or all-NA columns when determining the result dtypes. To retain the old behavior, exclude the relevant entries before the concat operation.\n",
      "  glm_results = pd.concat([glm_results, res], ignore_index=True)\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>sub</th>\n",
       "      <th>Volatility</th>\n",
       "      <th>group</th>\n",
       "      <th>Intercept</th>\n",
       "      <th>curDur</th>\n",
       "      <th>dDur</th>\n",
       "      <th>p_Intercept</th>\n",
       "      <th>p_CTI</th>\n",
       "      <th>p_dDur</th>\n",
       "      <th>r2</th>\n",
       "      <th>sequence</th>\n",
       "      <th>aic</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A31</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>ASD</td>\n",
       "      <td>-0.167</td>\n",
       "      <td>-0.393</td>\n",
       "      <td>-0.057</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.122</td>\n",
       "      <td>0.390</td>\n",
       "      <td>31.000</td>\n",
       "      <td>15.538</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A31</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>ASD</td>\n",
       "      <td>-0.224</td>\n",
       "      <td>-0.487</td>\n",
       "      <td>-0.130</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.450</td>\n",
       "      <td>0.447</td>\n",
       "      <td>31.000</td>\n",
       "      <td>-4.166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A33</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>ASD</td>\n",
       "      <td>-0.056</td>\n",
       "      <td>-0.131</td>\n",
       "      <td>-0.104</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.007</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.222</td>\n",
       "      <td>33.000</td>\n",
       "      <td>-117.297</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>A33</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>ASD</td>\n",
       "      <td>-0.032</td>\n",
       "      <td>-0.061</td>\n",
       "      <td>-0.382</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.023</td>\n",
       "      <td>0.004</td>\n",
       "      <td>0.060</td>\n",
       "      <td>33.000</td>\n",
       "      <td>-216.504</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>A34</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>ASD</td>\n",
       "      <td>-0.023</td>\n",
       "      <td>-0.379</td>\n",
       "      <td>-0.138</td>\n",
       "      <td>0.007</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.664</td>\n",
       "      <td>34.000</td>\n",
       "      <td>-284.612</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sub  Volatility group  Intercept  curDur   dDur  p_Intercept  p_CTI  \\\n",
       "0  A31  High Vola.   ASD     -0.167  -0.393 -0.057        0.000  0.000   \n",
       "1  A31   Low Vola.   ASD     -0.224  -0.487 -0.130        0.000  0.000   \n",
       "4  A33  High Vola.   ASD     -0.056  -0.131 -0.104        0.000  0.007   \n",
       "5  A33   Low Vola.   ASD     -0.032  -0.061 -0.382        0.001  0.023   \n",
       "6  A34  High Vola.   ASD     -0.023  -0.379 -0.138        0.007  0.000   \n",
       "\n",
       "   p_dDur    r2  sequence      aic  \n",
       "0   0.122 0.390    31.000   15.538  \n",
       "1   0.450 0.447    31.000   -4.166  \n",
       "4   0.002 0.222    33.000 -117.297  \n",
       "5   0.004 0.060    33.000 -216.504  \n",
       "6   0.000 0.664    34.000 -284.612  "
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# import OLS from statsmodels\n",
    "from statsmodels.regression.linear_model import OLS\n",
    "# creating an autorergressive model with lag 1, store the results in a new dataframe\n",
    "glm_results = pd.DataFrame(columns = ['sub', 'Volatility', 'group', \n",
    "                                      'Intercept', 'curDur', 'dDur',\n",
    "                                      'p_Intercept', 'p_CTI', 'p_dDur','r2'])\n",
    "\n",
    "# drop nan values in preDuration, preErr from vdata\n",
    "vdata = rawdata.query(\"outlier == False\").dropna(subset=['preDuration', 'preErr'])\n",
    "# center Duration to 1\n",
    "vdata['cDuration'] = vdata['Duration'] - 1\n",
    "vdata['cpreDuration'] = vdata['preDuration'] - 1\n",
    "vdata['dDuration'] = vdata['Duration'] - vdata['preDuration']\n",
    "groups = vdata.groupby(['sub', 'Volatility', 'group', 'sequence'])\n",
    "# drop nan values in preDuration, preErr\n",
    "\n",
    "for name, group in groups:\n",
    "    glm_mod = OLS(group['rep_err'], sm.add_constant(group[['cDuration', 'dDuration']])).fit()\n",
    "    summary = glm_mod.summary2()\n",
    "    r2 = glm_mod.rsquared\n",
    "    aic = glm_mod.aic\n",
    "    # estimate the prediction reproduced error\n",
    "    res = pd.DataFrame({\n",
    "        'sub': name[0], 'Volatility': name[1], 'group': name[2], 'sequence': name[3],\n",
    "        'Intercept': glm_mod.params['const'], \n",
    "        'curDur': glm_mod.params['cDuration'], #as CTI = -beta\n",
    "        'dDur': glm_mod.params['dDuration'],\n",
    "        'p_Intercept': summary.tables[1]['P>|t|']['const'],\n",
    "        'p_CTI': summary.tables[1]['P>|t|']['cDuration'],\n",
    "        'p_dDur': summary.tables[1]['P>|t|']['dDuration'],\n",
    "        'r2': r2,\n",
    "        'aic': aic\n",
    "        }, index=[0])\n",
    "    glm_results = pd.concat([glm_results, res], ignore_index=True)\n",
    "    \n",
    "# remove outliers\n",
    "glm_v = glm_results.query(\"sequence not in @outliers_regress\" )\n",
    "glm_v.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>group</th>\n",
       "      <th>Volatility</th>\n",
       "      <th colspan=\"2\" halign=\"left\">Intercept</th>\n",
       "      <th colspan=\"2\" halign=\"left\">curDur</th>\n",
       "      <th colspan=\"2\" halign=\"left\">dDur</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "      <th>mean</th>\n",
       "      <th>sem</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ASD</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>0.032</td>\n",
       "      <td>0.015</td>\n",
       "      <td>-0.166</td>\n",
       "      <td>0.026</td>\n",
       "      <td>-0.104</td>\n",
       "      <td>0.016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ASD</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>0.042</td>\n",
       "      <td>0.015</td>\n",
       "      <td>-0.092</td>\n",
       "      <td>0.030</td>\n",
       "      <td>-0.281</td>\n",
       "      <td>0.041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TD</td>\n",
       "      <td>High Vola.</td>\n",
       "      <td>0.029</td>\n",
       "      <td>0.011</td>\n",
       "      <td>-0.224</td>\n",
       "      <td>0.023</td>\n",
       "      <td>-0.123</td>\n",
       "      <td>0.012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TD</td>\n",
       "      <td>Low Vola.</td>\n",
       "      <td>0.038</td>\n",
       "      <td>0.013</td>\n",
       "      <td>-0.086</td>\n",
       "      <td>0.021</td>\n",
       "      <td>-0.394</td>\n",
       "      <td>0.049</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  group  Volatility Intercept       curDur         dDur      \n",
       "                         mean   sem   mean   sem   mean   sem\n",
       "0   ASD  High Vola.     0.032 0.015 -0.166 0.026 -0.104 0.016\n",
       "1   ASD   Low Vola.     0.042 0.015 -0.092 0.030 -0.281 0.041\n",
       "2    TD  High Vola.     0.029 0.011 -0.224 0.023 -0.123 0.012\n",
       "3    TD   Low Vola.     0.038 0.013 -0.086 0.021 -0.394 0.049"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# some unknown error with sub column in groupby and agg, so drop it first\n",
    "# show the mean and standard errors of Intercept, Duration, preErr for each group, Volatility\n",
    "glm_v.drop('sub', axis =1).groupby(['group', 'Volatility']).agg(['mean', 'sem'])[['Intercept', 'curDur', 'dDur']].reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.039</td>\n",
       "      <td>0.845</td>\n",
       "      <td>0.001</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.003</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.003</td>\n",
       "      <td>3.388</td>\n",
       "      <td>0.071</td>\n",
       "      <td>0.057</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.010</td>\n",
       "      <td>0.922</td>\n",
       "      <td>0.000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS     F  p-unc   np2   eps\n",
       "0        group 0.000    1   56 0.000 0.039  0.845 0.001   NaN\n",
       "1   Volatility 0.003    1   56 0.003 3.388  0.071 0.057 1.000\n",
       "2  Interaction 0.000    1   56 0.000 0.010  0.922 0.000   NaN"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pg.mixed_anova(data=glm_v, dv='Intercept', within='Volatility', between='group', subject='sub')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.020</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.020</td>\n",
       "      <td>0.906</td>\n",
       "      <td>0.345</td>\n",
       "      <td>0.016</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>0.326</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.326</td>\n",
       "      <td>24.570</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.305</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.029</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.029</td>\n",
       "      <td>2.218</td>\n",
       "      <td>0.142</td>\n",
       "      <td>0.038</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS      F  p-unc   np2   eps\n",
       "0        group 0.020    1   56 0.020  0.906  0.345 0.016   NaN\n",
       "1   Volatility 0.326    1   56 0.326 24.570  0.000 0.305 1.000\n",
       "2  Interaction 0.029    1   56 0.029  2.218  0.142 0.038   NaN"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# pingouin ANOVA test for CTI\n",
    "pg.mixed_anova(data=glm_v, dv='curDur', within='Volatility', between='group', subject='sub')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF1</th>\n",
       "      <th>DF2</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "      <th>eps</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>0.126</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.126</td>\n",
       "      <td>3.335</td>\n",
       "      <td>0.073</td>\n",
       "      <td>0.056</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Volatility</td>\n",
       "      <td>1.460</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>1.460</td>\n",
       "      <td>53.488</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.489</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Interaction</td>\n",
       "      <td>0.063</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>0.063</td>\n",
       "      <td>2.323</td>\n",
       "      <td>0.133</td>\n",
       "      <td>0.040</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Source    SS  DF1  DF2    MS      F  p-unc   np2   eps\n",
       "0        group 0.126    1   56 0.126  3.335  0.073 0.056   NaN\n",
       "1   Volatility 1.460    1   56 1.460 53.488  0.000 0.489 1.000\n",
       "2  Interaction 0.063    1   56 0.063  2.323  0.133 0.040   NaN"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "# pingouin ANOVA test for preDur\n",
    "pg.mixed_anova(data=glm_v, dv='dDur', within='Volatility', between='group', subject='sub')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.10 Correlation with EQ, AQ\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "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>T</th>\n",
       "      <th>dof</th>\n",
       "      <th>alternative</th>\n",
       "      <th>p-val</th>\n",
       "      <th>CI95%</th>\n",
       "      <th>cohen-d</th>\n",
       "      <th>BF10</th>\n",
       "      <th>power</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>T-test</th>\n",
       "      <td>-8.878</td>\n",
       "      <td>120.558</td>\n",
       "      <td>two-sided</td>\n",
       "      <td>0.000</td>\n",
       "      <td>[-39.01, -24.78]</td>\n",
       "      <td>1.577</td>\n",
       "      <td>7.799e+11</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            T     dof alternative  p-val             CI95%  cohen-d  \\\n",
       "T-test -8.878 120.558   two-sided  0.000  [-39.01, -24.78]    1.577   \n",
       "\n",
       "             BF10  power  \n",
       "T-test  7.799e+11  1.000  "
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import statsmodels.api as sm\n",
    "import statsmodels.formula.api as smf\n",
    "# load data/parinfo.csv as a dataframe\n",
    "pinfo = pd.read_csv('data/parinfo.csv')\n",
    "# add difference between EQ and SQ as a new column ES\n",
    "pinfo['ES'] = pinfo['EQ'] - pinfo['SQ'] # positive 'female brain', negative 'male brain'\n",
    "# merge glm_v and aq on sub\n",
    "res = pd.merge(kpars, pinfo, on=['group','sequence'])\n",
    "# t-test for DQ between ASD and TD\n",
    "pg.ttest(res.query(\"group == 'ASD'\")['ES'], res.query(\"group == 'TD'\")['ES'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>   <td>K1</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>        <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>        <td>0.0184</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>23.2472</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>        <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>              <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.801</td>   <td>0.129</td>   <td>6.218</td> <td>0.000</td>  <td>0.548</td>  <td>1.053</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>-0.097</td>   <td>0.083</td>  <td>-1.164</td> <td>0.245</td> <td>-0.261</td>  <td>0.066</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.078</td>   <td>0.024</td>  <td>-3.224</td> <td>0.001</td> <td>-0.125</td> <td>-0.031</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>AQ</th>                      <td>-0.003</td>   <td>0.003</td>  <td>-0.804</td> <td>0.421</td> <td>-0.009</td>  <td>0.004</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.023</td>   <td>0.060</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/strongway/miniforge3/envs/py312/lib/python3.12/site-packages/statsmodels/regression/mixed_linear_model.py:2238: ConvergenceWarning: The MLE may be on the boundary of the parameter space.\n",
      "  warnings.warn(msg, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>    <td>K2</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>         <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>         <td>0.0611</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>-17.5335</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>         <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>               <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.545</td>   <td>0.138</td>   <td>3.952</td> <td>0.000</td>  <td>0.275</td>  <td>0.815</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>-0.148</td>   <td>0.089</td>  <td>-1.666</td> <td>0.096</td> <td>-0.321</td>  <td>0.026</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.158</td>   <td>0.044</td>  <td>-3.587</td> <td>0.000</td> <td>-0.244</td> <td>-0.072</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>AQ</th>                      <td>-0.002</td>   <td>0.004</td>  <td>-0.538</td> <td>0.590</td> <td>-0.009</td>  <td>0.005</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.005</td>   <td>0.037</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/strongway/miniforge3/envs/py312/lib/python3.12/site-packages/statsmodels/regression/mixed_linear_model.py:2238: ConvergenceWarning: The MLE may be on the boundary of the parameter space.\n",
      "  warnings.warn(msg, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>    <td>tau</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>         <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>         <td>0.0020</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>143.1968</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>         <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>               <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.111</td>   <td>0.054</td>   <td>2.038</td> <td>0.042</td>  <td>0.004</td>  <td>0.218</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>-0.036</td>   <td>0.035</td>  <td>-1.024</td> <td>0.306</td> <td>-0.106</td>  <td>0.033</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th>  <td>0.027</td>   <td>0.008</td>   <td>3.366</td> <td>0.001</td>  <td>0.011</td>  <td>0.043</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>AQ</th>                      <td>-0.003</td>   <td>0.001</td>  <td>-1.849</td> <td>0.064</td> <td>-0.005</td>  <td>0.000</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.005</td>   <td>0.033</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>   <td>cti</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>        <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>        <td>0.0202</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>32.4618</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>        <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>              <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.286</td>   <td>0.106</td>   <td>2.706</td> <td>0.007</td>  <td>0.079</td>  <td>0.493</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>              <td>0.018</td>   <td>0.068</td>   <td>0.262</td> <td>0.793</td> <td>-0.116</td>  <td>0.152</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.225</td>   <td>0.025</td>  <td>-8.897</td> <td>0.000</td> <td>-0.275</td> <td>-0.176</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>AQ</th>                       <td>0.001</td>   <td>0.003</td>   <td>0.467</td> <td>0.641</td> <td>-0.004</td>  <td>0.007</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.011</td>   <td>0.038</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from IPython.display import display, HTML\n",
    "# fit the models with AQ\n",
    "display(HTML(smf.mixedlm(\"K1 ~ AQ + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"K2 ~ AQ + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"tau ~ AQ + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"cti ~ AQ + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "AQ was only marginal correlated with tau. Due to the multiple comparison, we did not consider this correlation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>   <td>K1</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>        <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>        <td>0.0184</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>23.2224</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>        <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>              <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.636</td>   <td>0.056</td>  <td>11.287</td> <td>0.000</td>  <td>0.526</td>  <td>0.747</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>-0.095</td>   <td>0.059</td>  <td>-1.617</td> <td>0.106</td> <td>-0.211</td>  <td>0.020</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.078</td>   <td>0.024</td>  <td>-3.224</td> <td>0.001</td> <td>-0.125</td> <td>-0.031</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>EQ</th>                       <td>0.002</td>   <td>0.002</td>   <td>1.415</td> <td>0.157</td> <td>-0.001</td>  <td>0.006</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.022</td>   <td>0.059</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/strongway/miniforge3/envs/py312/lib/python3.12/site-packages/statsmodels/regression/mixed_linear_model.py:2238: ConvergenceWarning: The MLE may be on the boundary of the parameter space.\n",
      "  warnings.warn(msg, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>    <td>K2</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>         <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>         <td>0.0611</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>-18.0032</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>         <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>               <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.432</td>   <td>0.063</td>   <td>6.874</td> <td>0.000</td>  <td>0.309</td>  <td>0.556</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>-0.143</td>   <td>0.063</td>  <td>-2.266</td> <td>0.023</td> <td>-0.266</td> <td>-0.019</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.158</td>   <td>0.044</td>  <td>-3.587</td> <td>0.000</td> <td>-0.244</td> <td>-0.072</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>EQ</th>                       <td>0.002</td>   <td>0.002</td>   <td>0.854</td> <td>0.393</td> <td>-0.002</td>  <td>0.005</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.005</td>   <td>0.037</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/strongway/miniforge3/envs/py312/lib/python3.12/site-packages/statsmodels/regression/mixed_linear_model.py:2238: ConvergenceWarning: The MLE may be on the boundary of the parameter space.\n",
      "  warnings.warn(msg, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>    <td>tau</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>         <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>         <td>0.0020</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>141.0893</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>         <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>               <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>               <td>-0.000</td>   <td>0.024</td>  <td>-0.010</td> <td>0.992</td> <td>-0.048</td>  <td>0.048</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>              <td>0.007</td>   <td>0.026</td>   <td>0.272</td> <td>0.785</td> <td>-0.044</td>  <td>0.058</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th>  <td>0.027</td>   <td>0.008</td>   <td>3.366</td> <td>0.001</td>  <td>0.011</td>  <td>0.043</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>EQ</th>                       <td>0.001</td>   <td>0.001</td>   <td>0.698</td> <td>0.485</td> <td>-0.001</td>  <td>0.002</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.005</td>   <td>0.034</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>   <td>cti</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>        <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>        <td>0.0202</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>31.7572</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>        <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>              <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.349</td>   <td>0.047</td>   <td>7.387</td> <td>0.000</td>  <td>0.257</td>  <td>0.442</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>              <td>0.005</td>   <td>0.049</td>   <td>0.098</td> <td>0.922</td> <td>-0.091</td>  <td>0.100</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.225</td>   <td>0.025</td>  <td>-8.897</td> <td>0.000</td> <td>-0.275</td> <td>-0.176</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>EQ</th>                      <td>-0.001</td>   <td>0.001</td>  <td>-0.429</td> <td>0.668</td> <td>-0.003</td>  <td>0.002</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.011</td>   <td>0.038</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(HTML(smf.mixedlm(\"K1 ~ EQ + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"K2 ~ EQ + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"tau ~ EQ + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"cti ~ EQ + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>   <td>K1</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>        <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>        <td>0.0184</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>22.2862</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>        <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>              <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.701</td>   <td>0.072</td>   <td>9.754</td> <td>0.000</td>  <td>0.560</td>  <td>0.841</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>-0.041</td>   <td>0.048</td>  <td>-0.846</td> <td>0.398</td> <td>-0.135</td>  <td>0.053</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.078</td>   <td>0.024</td>  <td>-3.224</td> <td>0.001</td> <td>-0.125</td> <td>-0.031</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>SQ</th>                       <td>0.000</td>   <td>0.002</td>   <td>0.001</td> <td>1.000</td> <td>-0.004</td>  <td>0.004</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.023</td>   <td>0.061</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/strongway/miniforge3/envs/py312/lib/python3.12/site-packages/statsmodels/regression/mixed_linear_model.py:2238: ConvergenceWarning: The MLE may be on the boundary of the parameter space.\n",
      "  warnings.warn(msg, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>    <td>K2</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>         <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>         <td>0.0611</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>-14.2098</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>         <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>               <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.660</td>   <td>0.073</td>   <td>8.988</td> <td>0.000</td>  <td>0.516</td>  <td>0.803</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>-0.155</td>   <td>0.047</td>  <td>-3.262</td> <td>0.001</td> <td>-0.248</td> <td>-0.062</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.158</td>   <td>0.044</td>  <td>-3.587</td> <td>0.000</td> <td>-0.244</td> <td>-0.072</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>SQ</th>                      <td>-0.005</td>   <td>0.002</td>  <td>-2.967</td> <td>0.003</td> <td>-0.009</td> <td>-0.002</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.001</td>   <td>0.033</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/strongway/miniforge3/envs/py312/lib/python3.12/site-packages/statsmodels/regression/mixed_linear_model.py:2238: ConvergenceWarning: The MLE may be on the boundary of the parameter space.\n",
      "  warnings.warn(msg, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>    <td>tau</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>         <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>         <td>0.0020</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>141.3026</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>         <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>               <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.038</td>   <td>0.031</td>   <td>1.248</td> <td>0.212</td> <td>-0.022</td>  <td>0.099</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>              <td>0.013</td>   <td>0.021</td>   <td>0.610</td> <td>0.542</td> <td>-0.028</td>  <td>0.053</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th>  <td>0.027</td>   <td>0.008</td>   <td>3.366</td> <td>0.001</td>  <td>0.011</td>  <td>0.043</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>SQ</th>                      <td>-0.001</td>   <td>0.001</td>  <td>-0.906</td> <td>0.365</td> <td>-0.002</td>  <td>0.001</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.005</td>   <td>0.034</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>   <td>cti</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>        <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>        <td>0.0202</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>31.7151</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>        <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>              <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.336</td>   <td>0.059</td>   <td>5.686</td> <td>0.000</td>  <td>0.220</td>  <td>0.452</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>-0.010</td>   <td>0.039</td>  <td>-0.246</td> <td>0.805</td> <td>-0.086</td>  <td>0.067</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.225</td>   <td>0.025</td>  <td>-8.897</td> <td>0.000</td> <td>-0.275</td> <td>-0.176</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>SQ</th>                      <td>-0.000</td>   <td>0.001</td>  <td>-0.053</td> <td>0.958</td> <td>-0.003</td>  <td>0.003</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.011</td>   <td>0.038</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(HTML(smf.mixedlm(\"K1 ~ SQ + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"K2 ~ SQ + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"tau ~ SQ + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"cti ~ SQ + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "SQ and K2 correlation was significant!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# scatterplot for K2 as a function of SQ, separated by group (different colors), Volatility (different symbols)\n",
    "#sns.scatterplot(data=res, x='SQ', y='K2', hue='group', style='Volatility', alpha=0.5)\n",
    "palette = sns.color_palette('Dark2')\n",
    "plt.figure(figsize=(6,4))\n",
    "# subplot 1\n",
    "#plt.subplot(1,2,1)\n",
    "plt.xlabel('SQ')\n",
    "plt.ylabel('K2')\n",
    "# now add regression lines for each group, Volatility\n",
    "sns.regplot(data=res.query(\"group == 'ASD' and Volatility == 'Low Vola.'\"), \n",
    "            x='SQ', y='K2', color = palette[0],  line_kws={'ls':'--'}, \n",
    "            scatter_kws ={'facecolors':'none','alpha':0.3}, marker='o', label = \"ASD/Low Vola.\") \n",
    "sns.regplot(data=res.query(\"group == 'ASD' and Volatility == 'High Vola.'\"),\n",
    "             x='SQ', y='K2', color = palette[1],  line_kws={'ls':'--'}, \n",
    "             scatter_kws ={'alpha':0.3}, marker='o', label = \"ASD/High Vola.\")\n",
    "sns.regplot(data=res.query(\"group == 'TD' and Volatility == 'Low Vola.'\"), \n",
    "            x='SQ', y='K2', color = palette[2], line_kws={'ls':'-'}, \n",
    "            scatter_kws ={'facecolors':'none','alpha':0.3}, marker = 's', label = \"TD/Low Vola.\")\n",
    "sns.regplot(data=res.query(\"group == 'TD' and Volatility == 'High Vola.'\"), \n",
    "            x='SQ', y='K2', color = palette[3],  line_kws={'ls':'-'}, \n",
    "            scatter_kws ={'alpha':0.3}, marker='s', label=\"TD/High Vola.\")\n",
    "# add legend\n",
    "plt.legend()\n",
    "# plot a label 'a' on the top left corner\n",
    "#plt.text(-0.1, 1.1, 'a', transform=plt.gca().transAxes, \n",
    "#                            fontsize=16, fontweight='bold', va='top', ha='right')\n",
    "# remove box around the plot\n",
    "sns.despine()\n",
    "# subplot 1\n",
    "#plt.subplot(1,2,2)\n",
    "#plt.xlabel('E-S')\n",
    "#plt.ylabel('K2')\n",
    "# now add regression lines for each group, Volatility\n",
    "#sns.regplot(data=res.query(\"group == 'ASD' and Volatility == 'Low Vola.'\"), \n",
    "#            x='ES', y='K2', color = palette[0], line_kws={'ls':'--'}, \n",
    "#            scatter_kws ={'facecolors':'none','alpha':0.3}, marker='o', label = \"ASD/Low Vola.\") \n",
    "#sns.regplot(data=res.query(\"group == 'ASD' and Volatility == 'High Vola.'\"),\n",
    "#             x='ES', y='K2', color = palette[1],  line_kws={'ls':'--'}, \n",
    "#             scatter_kws ={'alpha':0.3}, marker='o', label = \"ASD/High Vola.\")\n",
    "#sns.regplot(data=res.query(\"group == 'TD' and Volatility == 'Low Vola.'\"), \n",
    "#            x='ES', y='K2', color = palette[2], line_kws={'ls':'-'}, \n",
    "#            scatter_kws ={'facecolors':'none','alpha':0.3}, marker = 's', label = \"TD/Low Vola.\")\n",
    "#sns.regplot(data=res.query(\"group == 'TD' and Volatility == 'High Vola.'\"), \n",
    "#            x='ES', y='K2', color = palette[3],  line_kws={'ls':'-'}, \n",
    "#            scatter_kws ={'alpha':0.3}, marker='s', label=\"TD/High Vola.\")\n",
    "# add legend\n",
    "#plt.legend()\n",
    "# plot a label 'b' on the top left corner\n",
    "#plt.text(-0.1, 1.1, 'b', transform=plt.gca().transAxes, \n",
    "#                            fontsize=16, fontweight='bold', va='top', ha='right')\n",
    "# remove box around the plot\n",
    "sns.despine()\n",
    "# tight layout\n",
    "plt.tight_layout()\n",
    "# save the figure to vector file ./figures/K2_vs_SQ.pdf\n",
    "plt.savefig('./figures/K2_vs_SQ.png', dpi=300)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>   <td>K1</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>        <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>        <td>0.0184</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>23.3700</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>        <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>              <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.735</td>   <td>0.045</td>  <td>16.190</td> <td>0.000</td>  <td>0.646</td>  <td>0.823</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>-0.057</td>   <td>0.047</td>  <td>-1.214</td> <td>0.225</td> <td>-0.150</td>  <td>0.035</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.078</td>   <td>0.024</td>  <td>-3.224</td> <td>0.001</td> <td>-0.125</td> <td>-0.031</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>BDI</th>                     <td>-0.003</td>   <td>0.003</td>  <td>-1.112</td> <td>0.266</td> <td>-0.009</td>  <td>0.002</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.022</td>   <td>0.059</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/strongway/miniforge3/envs/py312/lib/python3.12/site-packages/statsmodels/regression/mixed_linear_model.py:2238: ConvergenceWarning: The MLE may be on the boundary of the parameter space.\n",
      "  warnings.warn(msg, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>    <td>K2</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>         <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>         <td>0.0611</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>-16.1921</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>         <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>               <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.532</td>   <td>0.051</td>  <td>10.531</td> <td>0.000</td>  <td>0.433</td>  <td>0.631</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>-0.136</td>   <td>0.049</td>  <td>-2.776</td> <td>0.006</td> <td>-0.232</td> <td>-0.040</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.158</td>   <td>0.044</td>  <td>-3.587</td> <td>0.000</td> <td>-0.244</td> <td>-0.072</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>BDI</th>                     <td>-0.005</td>   <td>0.003</td>  <td>-1.844</td> <td>0.065</td> <td>-0.011</td>  <td>0.000</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.004</td>   <td>0.035</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/strongway/miniforge3/envs/py312/lib/python3.12/site-packages/statsmodels/regression/mixed_linear_model.py:2238: ConvergenceWarning: The MLE may be on the boundary of the parameter space.\n",
      "  warnings.warn(msg, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>    <td>tau</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>         <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>         <td>0.0020</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>141.6521</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>         <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>               <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>             <th>Coef.</th> <th>Std.Err.</th>   <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>               <td>0.004</td>   <td>0.019</td>  <td>0.186</td> <td>0.852</td> <td>-0.035</td>  <td>0.042</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>0.024</td>   <td>0.021</td>  <td>1.158</td> <td>0.247</td> <td>-0.016</td>  <td>0.064</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>0.027</td>   <td>0.008</td>  <td>3.366</td> <td>0.001</td>  <td>0.011</td>  <td>0.043</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>BDI</th>                     <td>0.001</td>   <td>0.001</td>  <td>0.756</td> <td>0.450</td> <td>-0.001</td>  <td>0.003</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>               <td>0.005</td>   <td>0.034</td>    <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>   <td>cti</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>        <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>        <td>0.0202</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>32.6026</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>        <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>              <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.310</td>   <td>0.038</td>   <td>8.183</td> <td>0.000</td>  <td>0.236</td>  <td>0.385</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>              <td>0.002</td>   <td>0.039</td>   <td>0.060</td> <td>0.952</td> <td>-0.073</td>  <td>0.078</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.225</td>   <td>0.025</td>  <td>-8.897</td> <td>0.000</td> <td>-0.275</td> <td>-0.176</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>BDI</th>                      <td>0.002</td>   <td>0.002</td>   <td>0.916</td> <td>0.360</td> <td>-0.002</td>  <td>0.007</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.011</td>   <td>0.038</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(HTML(smf.mixedlm(\"K1 ~ BDI + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"K2 ~ BDI + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"tau ~ BDI + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"cti ~ BDI + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>   <td>K1</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>        <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>        <td>0.0184</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>22.2399</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>        <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>              <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.709</td>   <td>0.035</td>  <td>20.292</td> <td>0.000</td>  <td>0.641</td>  <td>0.778</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>-0.073</td>   <td>0.057</td>  <td>-1.281</td> <td>0.200</td> <td>-0.186</td>  <td>0.039</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.078</td>   <td>0.024</td>  <td>-3.224</td> <td>0.001</td> <td>-0.125</td> <td>-0.031</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ES</th>                       <td>0.001</td>   <td>0.001</td>   <td>0.921</td> <td>0.357</td> <td>-0.001</td>  <td>0.003</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.022</td>   <td>0.060</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/strongway/miniforge3/envs/py312/lib/python3.12/site-packages/statsmodels/regression/mixed_linear_model.py:2238: ConvergenceWarning: The MLE may be on the boundary of the parameter space.\n",
      "  warnings.warn(msg, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>    <td>K2</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>         <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>         <td>0.0611</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>-16.0307</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>         <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>               <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.496</td>   <td>0.040</td>  <td>12.412</td> <td>0.000</td>  <td>0.418</td>  <td>0.575</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>-0.195</td>   <td>0.058</td>  <td>-3.336</td> <td>0.001</td> <td>-0.309</td> <td>-0.080</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.158</td>   <td>0.044</td>  <td>-3.587</td> <td>0.000</td> <td>-0.244</td> <td>-0.072</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ES</th>                       <td>0.003</td>   <td>0.001</td>   <td>2.402</td> <td>0.016</td>  <td>0.001</td>  <td>0.005</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.002</td>   <td>0.034</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/strongway/miniforge3/envs/py312/lib/python3.12/site-packages/statsmodels/regression/mixed_linear_model.py:2238: ConvergenceWarning: The MLE may be on the boundary of the parameter space.\n",
      "  warnings.warn(msg, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>    <td>tau</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>         <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>         <td>0.0020</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>140.9529</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>         <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>               <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>             <th>Coef.</th> <th>Std.Err.</th>   <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>               <td>0.018</td>   <td>0.015</td>  <td>1.207</td> <td>0.228</td> <td>-0.011</td>  <td>0.047</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>0.003</td>   <td>0.025</td>  <td>0.120</td> <td>0.904</td> <td>-0.046</td>  <td>0.052</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>0.027</td>   <td>0.008</td>  <td>3.366</td> <td>0.001</td>  <td>0.011</td>  <td>0.043</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ES</th>                      <td>0.000</td>   <td>0.000</td>  <td>1.029</td> <td>0.303</td> <td>-0.000</td>  <td>0.001</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>               <td>0.005</td>   <td>0.034</td>    <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td>Model:</td>       <td>MixedLM</td> <td>Dependent Variable:</td>   <td>cti</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>No. Observations:</td>   <td>126</td>         <td>Method:</td>        <td>REML</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>No. Groups:</td>      <td>63</td>          <td>Scale:</td>        <td>0.0202</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Min. group size:</td>     <td>2</td>      <td>Log-Likelihood:</td>   <td>31.2769</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Max. group size:</td>     <td>2</td>        <td>Converged:</td>        <td>Yes</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Mean group size:</td>    <td>2.0</td>            <td></td>              <td></td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>              <th>Coef.</th> <th>Std.Err.</th>    <th>z</th>   <th>P>|z|</th> <th>[0.025</th> <th>0.975]</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th>                <td>0.331</td>   <td>0.030</td>  <td>11.144</td> <td>0.000</td>  <td>0.273</td>  <td>0.390</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>group[T.TD]</th>             <td>-0.002</td>   <td>0.047</td>  <td>-0.035</td> <td>0.972</td> <td>-0.094</td>  <td>0.090</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Volatility[T.Low Vola.]</th> <td>-0.225</td>   <td>0.025</td>  <td>-8.897</td> <td>0.000</td> <td>-0.275</td> <td>-0.176</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ES</th>                      <td>-0.000</td>   <td>0.001</td>  <td>-0.249</td> <td>0.804</td> <td>-0.002</td>  <td>0.002</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Group Var</th>                <td>0.011</td>   <td>0.038</td>     <td></td>      <td></td>       <td></td>       <td></td>   \n",
       "</tr>\n",
       "</table><br/>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(HTML(smf.mixedlm(\"K1 ~ ES + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"K2 ~ ES + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"tau ~ ES + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n",
    "display(HTML(smf.mixedlm(\"cti ~ ES + group + Volatility\", res, groups=res[\"sub\"]).fit().summary().as_html()))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "ES had a significant correlation with K2. Let's visualize it. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "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>sub</th>\n",
       "      <th>sequence</th>\n",
       "      <th>SQ</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A32</td>\n",
       "      <td>32</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A32</td>\n",
       "      <td>32</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>C32</td>\n",
       "      <td>32</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>C32</td>\n",
       "      <td>32</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>aril02</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>aril02</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>arm13</td>\n",
       "      <td>13</td>\n",
       "      <td>32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>arm13</td>\n",
       "      <td>13</td>\n",
       "      <td>32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>114</th>\n",
       "      <td>crs02</td>\n",
       "      <td>2</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>crs02</td>\n",
       "      <td>2</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>124</th>\n",
       "      <td>crw13</td>\n",
       "      <td>13</td>\n",
       "      <td>51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>crw13</td>\n",
       "      <td>13</td>\n",
       "      <td>51</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        sub  sequence  SQ\n",
       "2       A32        32  44\n",
       "3       A32        32  44\n",
       "12      C32        32  10\n",
       "13      C32        32  10\n",
       "36   aril02         2   8\n",
       "37   aril02         2   8\n",
       "50    arm13        13  32\n",
       "51    arm13        13  32\n",
       "114   crs02         2  21\n",
       "115   crs02         2  21\n",
       "124   crw13        13  51\n",
       "125   crw13        13  51"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show those outliers SQ scores\n",
    "res.query(\"sequence in @outliers_regress\")[['sub', 'sequence', 'SQ']]"
   ]
  }
 ],
 "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": "3.12.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}