nbwb_nbcst.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "2be84102-cfb3-4d74-a51f-bae772d2d173",
   "metadata": {},
   "source": [
    "Code to find the average FA/MD/AD/RD per track, and its significance\n",
    "\n",
    "Normalize by whole brain\n",
    "Normalize by corticospinal tract"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f61e9dba-7734-41cd-837a-84970d777478",
   "metadata": {},
   "outputs": [],
   "source": [
    "import csv\n",
    "import numpy as np\n",
    "import os\n",
    "import random as rd\n",
    "import re\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import scipy.stats as stats\n",
    "from scipy.stats import mannwhitneyu\n",
    "\n",
    "from matplotlib import pyplot as plt\n",
    "import seaborn as sns\n",
    "from matplotlib.ticker import AutoMinorLocator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4995d568-ebcc-4190-af82-1e54646af30a",
   "metadata": {},
   "outputs": [],
   "source": [
    "os.chdir('/data/Reback-DTI/')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a26cf73e-ea47-43f6-8e06-28bb2826ad1b",
   "metadata": {},
   "outputs": [],
   "source": [
    "outcomes = ['pcpcgp_6mo', 'mortality_6mos']\n",
    "measures = ['fa', 'rd', 'md', 'ad']\n",
    "agerange = ['all_ages','ages0_1','ages1_5', 'ages0_5', 'ages0_2']\n",
    "\n",
    "batch1_tracts = ['Cingulum_Parahippocampal_L','Cingulum_Parahippocampal_R','Inferior_Fronto_Occipital_Fasciculus_L','Inferior_Fronto_Occipital_Fasciculus_R','Inferior_Longitudinal_Fasciculus_L','Inferior_Longitudinal_Fasciculus_R','Corticospinal_Tract_L','Corticospinal_Tract_R','Corpus_Callosum_Body']\n",
    "batch2_tracts = ['Cingulum_Parahippocampal_L','Cingulum_Parahippocampal_R','Inferior_Fronto_Occipital_Fasciculus_L','Inferior_Fronto_Occipital_Fasciculus_R','Inferior_Longitudinal_Fasciculus_L','Inferior_Longitudinal_Fasciculus_R','Corticospinal_Tract_L','Corticospinal_Tract_R','Corpus_Callosum_Body','Cingulum_Frontal_Parahippocampal_L','Cingulum_Frontal_Parahippocampal_R','Cingulum_Frontal_Parietal_L','Cingulum_Frontal_Parietal_R','Cingulum_Parahippocampal_Parietal_L','Cingulum_Parahippocampal_Parietal_R','Cingulum_Rarolfactory_L','Cingulum_Rarolfactory_R','Corpus_Callosum_Forceps_Minor','Corpus_Callosum_Tapetum','Corpus_Callosum_Forceps_Major']\n",
    "batch3_tracts = ['Cingulum_Parahippocampal_L','Cingulum_Parahippocampal_R','Inferior_Fronto_Occipital_Fasciculus_L','Inferior_Fronto_Occipital_Fasciculus_R','Inferior_Longitudinal_Fasciculus_L','Inferior_Longitudinal_Fasciculus_R','Corticospinal_Tract_L','Corticospinal_Tract_R','Corpus_Callosum_Body','Cingulum_Frontal_Parahippocampal_L','Cingulum_Frontal_Parahippocampal_R','Cingulum_Frontal_Parietal_L','Cingulum_Frontal_Parietal_R','Cingulum_Parahippocampal_Parietal_L','Cingulum_Parahippocampal_Parietal_R','Cingulum_Rarolfactory_L','Cingulum_Rarolfactory_R','Corpus_Callosum_Forceps_Minor','Corpus_Callosum_Tapetum','Corpus_Callosum_Forceps_Major','whole_brain']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "21aab842-ed6a-4e88-aefc-4fb00aa7ae86",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "84eb55b7-1138-4dd3-bf14-0c2071516b48",
   "metadata": {},
   "outputs": [],
   "source": [
    "Eddy_corrected = '/data/Reback-DTI/resources/stats/Eddy_corrected/'\n",
    "Noncorrected = '/data/Reback-DTI/resources/stats/Noncorrected'\n",
    "Batch1 = os.path.join(Eddy_corrected, 'Batch1')\n",
    "Batch2 = os.path.join(Eddy_corrected, 'Batch2')\n",
    "Batch3 = os.path.join(Eddy_corrected, 'Batch3')\n",
    "\n",
    "#***********SELECT YOUR CHOICE****************#\n",
    "selected_agerange = agerange[0]\n",
    "Batch_dir  = Batch3\n",
    "#***********SELECT YOUR CHOICE****************#\n",
    "\n",
    "if Batch_dir == Batch1:\n",
    "    tracts = batch1_tracts\n",
    "elif Batch_dir == Batch2:\n",
    "    tracts = batch2_tracts\n",
    "elif Batch_dir == Batch3:\n",
    "    tracts = batch3_tracts\n",
    "\n",
    "datasrc_dir = os.path.join(Batch_dir, 'output_dir', 'strats', selected_agerange)\n",
    "datasrc_file = os.path.join(datasrc_dir, '{}.csv'.format(selected_agerange))\n",
    "\n",
    "outputBaseDir = os.path.join(datasrc_dir, 'normalBy')\n",
    "if not os.path.exists(outputBaseDir):\n",
    "    os.makedirs(outputBaseDir)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "96609f73-cabc-4f83-bb79-d01f7905f5ef",
   "metadata": {},
   "outputs": [],
   "source": [
    "def getData(src):\n",
    "    pdata = pd.read_csv(src)\n",
    "    pdata['id'] = pdata['StudyID'].map(str) + '_' + pdata['MR_num'].map(str) + '_' + pdata['Scan_num'].map(str)\n",
    "    #place_holder for future, won't currently delete anything probably\n",
    "    pdata = pdata.drop_duplicates(subset='id')  \n",
    "    pdata.head()\n",
    "    return pdata\n",
    "\n",
    "def makeEmptyResultsDf():\n",
    "    column_names = ['tract', 'outcome', 'measure', 'pvalue']\n",
    "    results_df = pd.DataFrame(columns = column_names)\n",
    "    return results_df\n",
    "\n",
    "#note, this is different from the one in statistics.ipynb\n",
    "def getTractMeasure(pdata, outcome=outcomes[0],tract='', measure='fa'): #default is all tracts FA\n",
    "    tract_measure= re.compile(r''+tract+'.{}$'.format(measure))\n",
    "    filtered = list(filter(tract_measure.search, pdata.columns))\n",
    "    #print(filtered)\n",
    "\n",
    "    ### USE MELT TO CONVERT FROM WIDE TO LONG FORM\n",
    "    df_meas = pd.melt(pdata, id_vars=['id', outcome, 'Sex'], value_vars=filtered, var_name='Tract', value_name='{}'.format(measure)) \n",
    "    df_meas['Tract'] = df_meas['Tract'].str.strip('.{}'.format(measure))\n",
    "    df_meas = df_meas.dropna()\n",
    "    df_meas.head()\n",
    "    return df_meas \n",
    "\n",
    "def getNormByMean(pdata, outcome=outcomes[0],tract='', measure='fa'): #default is all tracts FA\n",
    "    tract_measure= re.compile(r''+tract+'.{}$'.format(measure))\n",
    "    filtered = list(filter(tract_measure.search, pdata.columns))\n",
    "\n",
    "    reqFields = ['id', outcome, 'Sex', 'Race']\n",
    "    reqFields.extend(filtered)\n",
    "    df_nbm = pdata[reqFields]\n",
    "    \n",
    "    numtracts = len(tracts)\n",
    "    lbound = 0-numtracts\n",
    "    ubound = 0\n",
    "    df_nbm['globalAvg.{}'.format(measure)] = df_nbm.iloc[:, lbound:].sum(axis=1).divide(numtracts)\n",
    "    df_nbm_vals = df_nbm.iloc[:, lbound-1:ubound-1].divide(df_nbm['globalAvg.{}'.format(measure)], axis=0)\n",
    "    df_nbm[df_nbm_vals.columns] = df_nbm_vals\n",
    "    \n",
    "    return df_nbm \n",
    "\n",
    "\n",
    "def getNormByWholeBrain(pdata, outcome=outcomes[0],tract='', measure='fa'): #default is all tracts FA\n",
    "    tract_measure= re.compile(r''+tract+'.{}$'.format(measure))\n",
    "    filtered = list(filter(tract_measure.search, pdata.columns))\n",
    "\n",
    "    reqFields = ['id', outcome, 'Sex', 'Race']\n",
    "    reqFields.extend(filtered)\n",
    "    df_nbm = pdata[reqFields]\n",
    "    \n",
    "    numtracts = len(tracts)\n",
    "    lbound = 0-numtracts\n",
    "    ubound = 0\n",
    "    df_nbm_vals = df_nbm.iloc[:, lbound:].divide(df_nbm['whole_brain.{}'.format(measure)], axis=0)\n",
    "    df_nbm[df_nbm_vals.columns] = df_nbm_vals\n",
    "    \n",
    "    return df_nbm \n",
    "\n",
    "def getNormByCST(pdata, outcome=outcomes[0],tract='', measure='fa'): #default is all tracts FA\n",
    "    tract_measure= re.compile(r''+tract+'.{}$'.format(measure))\n",
    "    filtered = list(filter(tract_measure.search, pdata.columns))\n",
    "\n",
    "    reqFields = ['id', outcome, 'Sex', 'Race']\n",
    "    reqFields.extend(filtered)\n",
    "    df_nbm = pdata[reqFields]\n",
    "    \n",
    "    numtracts = len(tracts)\n",
    "    lbound = 0-numtracts\n",
    "    ubound = 0\n",
    "    df_nbm['CSTAvg.{}'.format(measure)] = datapd.loc[:,['Corticospinal_Tract_L.{}'.format(measure),'Corticospinal_Tract_R.{}'.format(measure)]].sum(axis=1).divide(2)\n",
    "    df_nbm_vals = df_nbm.iloc[:, lbound-1:ubound-1].divide(df_nbm['CSTAvg.{}'.format(measure)], axis=0)\n",
    "    df_nbm[df_nbm_vals.columns] = df_nbm_vals\n",
    "    \n",
    "    return df_nbm "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ef919716-826e-42a2-ba35-cf36ed5c56a0",
   "metadata": {},
   "outputs": [],
   "source": [
    "datapd     = getData(datasrc_file)\n",
    "#test\n",
    "datapd_nbm = getNormByCST(datapd, outcomes[0], measure=measures[0])\n",
    "datapd_nbm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8bd9aae8-464d-42f6-8c58-f2ea65376a3a",
   "metadata": {},
   "outputs": [],
   "source": [
    "def goGetTheNorms(normFunc, normFuncName):\n",
    "    datapd     = getData(datasrc_file)\n",
    "    reqFields  = ['id', outcomes[0], 'Sex', 'Race']\n",
    "    results_df = datapd[reqFields]\n",
    "    outputPath = os.path.join(outputBaseDir, normFuncName)\n",
    "    \n",
    "    for meas in measures:\n",
    "        datapd_nbm = normFunc(datapd, outcomes[0], measure=meas)\n",
    "        outfile = os.path.join(outputPath, '{}_{}.csv'.format(normFuncName, meas))\n",
    "        datapd_nbm.to_csv(outfile, sep=',', index=False)\n",
    "        results_df = results_df.merge(datapd_nbm, how='inner', on=reqFields)\n",
    "    print(normFuncName)\n",
    "    \n",
    "    outputPath = os.path.join(outputPath, '{}_{}.csv'.format(normFuncName, 'all'))\n",
    "    results_df.to_csv(outputPath, sep=',', index=False)\n",
    "    \n",
    "goGetTheNorms(getNormByWholeBrain, 'NBWB')\n",
    "goGetTheNorms(getNormByCST, 'NBCST')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "30cd9b4f-763f-4aa0-8ad8-defbfe8884ac",
   "metadata": {},
   "outputs": [],
   "source": [
    "#%%capture --no-display\n",
    "\n",
    "# normFunc = getNormByWholeBrain\n",
    "# normFuncName = 'NBWB'\n",
    "\n",
    "normFunc = getNormByWholeBrain\n",
    "normFuncName = 'NBWB'\n",
    "\n",
    "results_df = makeEmptyResultsDf()\n",
    "df_fa  = None #for future use\n",
    "df_rd  = None #for future use\n",
    "df_md  = None #for future use\n",
    "df_ad  = None #for future use\n",
    "df_age = None #for future use\n",
    "df_sex = None #for future use\n",
    "#get norm for every measure\n",
    "for meas in measures:\n",
    "    measure    = meas\n",
    "    datapd     = getData(datasrc_file)\n",
    "    datapd_nbm = normFunc(datapd, outcomes[0], measure=measure)\n",
    "    \n",
    "    ####for future use###\n",
    "    if measure=='fa':\n",
    "        df_fa = datapd_nbm\n",
    "    if measure=='rd':\n",
    "        df_rd = datapd_nbm\n",
    "    if measure=='md':\n",
    "        df_md = datapd_nbm\n",
    "    if measure=='ad':\n",
    "        df_ad = datapd_nbm\n",
    "    df_age = datapd[['id', outcomes[0], 'Age_y']]\n",
    "    df_age['Tract'] = 'All'\n",
    "    df_sex = datapd[['id', outcomes[0], 'Sex']]\n",
    "    df_sex['Tract'] = 'All'\n",
    "    #####################\n",
    "    \n",
    "    outputFilename = os.path.join(outputBaseDir, 'nbm_{}.csv'.format(measure))\n",
    "    #datapd_nbm.to_csv(outputFilename, sep=',', index=False)\n",
    "\n",
    "    outcome0 = datapd_nbm.loc[(datapd_nbm[outcomes[0]] == 0)]# & (datapd_nbm['Sex'] == 0) ]\n",
    "    outcome1 = datapd_nbm.loc[(datapd_nbm[outcomes[0]] == 1)]# & (datapd_nbm['Sex'] == 0) ]\n",
    "\n",
    "    tracts2 = tracts[:-1]\n",
    "    for tract in tracts2:\n",
    "        U1, pval = stats.mannwhitneyu(outcome0['{}.{}'.format(tract,measure)], outcome1['{}.{}'.format(tract,measure)])\n",
    "        \n",
    "        row = {'tract': tract, 'outcome': outcomes[0], 'measure': measure, 'pvalue': pval}\n",
    "        results_df = results_df.append(row, ignore_index=True)\n",
    "    \n",
    "        \n",
    "results_wide = results_df.pivot(index=['tract'], columns='measure', values='pvalue')\n",
    "results_wide.reset_index()\n",
    "\n",
    "pvalue_output = os.path.join(outputBaseDir, normFuncName, 'pvals_{}.csv'.format(selected_agerange)) \n",
    "results_wide.to_csv(pvalue_output)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d418a0b7-e160-49f0-a37e-e15e424f467a",
   "metadata": {},
   "outputs": [],
   "source": [
    "results_wide"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "66992e0e-b263-4382-9705-68eca1f3430a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# #below uses t-tests\n",
    "# results_df = makeEmptyResultsDf()\n",
    "# df_fa  = None #for future use\n",
    "# df_rd  = None #for future use\n",
    "# df_md  = None #for future use\n",
    "# df_ad  = None #for future use\n",
    "# df_age = None #for future use\n",
    "# df_sex = None #for future use\n",
    "# for meas in measures:\n",
    "#     measure    = meas\n",
    "#     datapd     = getData(datasrc_file)\n",
    "#     datapd_nbm = getNormByCST(datapd, outcomes[0], measure=measure)\n",
    "    \n",
    "#     ####for future use###\n",
    "#     if measure=='fa':\n",
    "#         df_fa = datapd_nbm\n",
    "#     if measure=='rd':\n",
    "#         df_rd = datapd_nbm\n",
    "#     if measure=='md':\n",
    "#         df_md = datapd_nbm\n",
    "#     if measure=='ad':\n",
    "#         df_ad = datapd_nbm\n",
    "#     df_age = datapd[['id', outcomes[0], 'Age_y']]\n",
    "#     df_age['Tract'] = 'All'\n",
    "#     df_sex = datapd[['id', outcomes[0], 'Sex']]\n",
    "#     df_sex['Tract'] = 'All'\n",
    "#     #####################\n",
    "    \n",
    "#     outputFilename = os.path.join(outputBaseDir, 'nbm_{}.csv'.format(measure))\n",
    "#     datapd_nbm.to_csv(outputFilename, sep=',', index=False)\n",
    "\n",
    "#     outcome0 = datapd_nbm.loc[(datapd_nbm[outcomes[0]] == 0)]# & (datapd_nbm['Sex'] == 0) ]\n",
    "#     outcome1 = datapd_nbm.loc[(datapd_nbm[outcomes[0]] == 1)]# & (datapd_nbm['Sex'] == 0) ]\n",
    "\n",
    "#     for tract in tracts:\n",
    "#         result = stats.ttest_ind(outcome0['{}.{}'.format(tract,measure)], outcome1['{}.{}'.format(tract,measure)])\n",
    "        \n",
    "#         row = {'tract': tract, 'outcome': outcomes[0], 'measure': measure, 'pvalue': result.pvalue}\n",
    "#         results_df = results_df.append(row, ignore_index=True)\n",
    "        \n",
    "# results_wide = results_df.pivot(index=['tract'], columns='measure', values='pvalue')\n",
    "# results_wide.reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "528e6b60-ed7d-4fc9-8b61-fd30f38261e5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# pvalue_output = os.path.join(outputBaseDir, , 'pvals_{}.txt'.format(selected_agerange)) \n",
    "# results_wide.to_csv(pvalue_output)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fde152d4-6b30-4b0c-ac13-41d2c42cbe80",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "6ea8ac76-6d90-452e-8b89-95a07adb2a39",
   "metadata": {},
   "source": [
    "Violin Plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6b37001b-9f16-4008-a863-19c7006f03f4",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plotViolins(df_fa, df_rd, df_md, df_ad, saveFile=None):\n",
    "    plt.style.use('ggplot')\n",
    "    fig = plt.figure(figsize=(16, 16))\n",
    "\n",
    "    plt.subplot(411) \n",
    "    ax1 = sns.violinplot(x='Tract', y='fa', data=df_fa, hue=outcomes[0], split=True, scale=\"count\", inner=\"stick\", scale_hue=False, bw=.2)\n",
    "    ax1.minorticks_on()\n",
    "    ax1.xaxis.set_minor_locator(AutoMinorLocator(2))\n",
    "    ax1.grid(which='minor', axis='x', linewidth=1)\n",
    "    ax1.axhline(1.00, alpha=0.5, c='#000000', ls='-')\n",
    "    ax1.set(xticklabels=[])\n",
    "    ax1.set(xlabel=None)\n",
    "\n",
    "    plt.subplot(412) \n",
    "    ax1 = sns.violinplot(x='Tract', y='rd', data=df_rd, hue=outcomes[0], split=True)\n",
    "    ax1.minorticks_on()\n",
    "    ax1.xaxis.set_minor_locator(AutoMinorLocator(2))\n",
    "    ax1.grid(which='minor', axis='x', linewidth=1)\n",
    "    ax1.axhline(1.00, alpha=0.5, c='#000000', ls='-')\n",
    "    ax1.set(xticklabels=[])\n",
    "    ax1.set(xlabel=None)\n",
    "\n",
    "    plt.subplot(413) \n",
    "    ax1 = sns.violinplot(x='Tract', y='md', data=df_md, hue=outcomes[0], split=True)\n",
    "    ax1.minorticks_on()\n",
    "    ax1.xaxis.set_minor_locator(AutoMinorLocator(2))\n",
    "    ax1.grid(which='minor', axis='x', linewidth=1)\n",
    "    ax1.axhline(1.00, alpha=0.5, c='#000000', ls='-')\n",
    "    ax1.set(xticklabels=[])\n",
    "    ax1.set(xlabel=None)\n",
    "\n",
    "    plt.subplot(414) \n",
    "    ax1 = sns.violinplot(x='Tract', y='ad', data=df_ad, hue=outcomes[0], split=True)\n",
    "    ax1.minorticks_on()\n",
    "    ax1.xaxis.set_minor_locator(AutoMinorLocator(2))\n",
    "    ax1.grid(which='minor', axis='x', linewidth=1)\n",
    "    ax1.axhline(1.00, alpha=0.5, c='#000000', ls='-')\n",
    "    plt.setp(ax1.get_xticklabels(), rotation=30, horizontalalignment='right')\n",
    "    \n",
    "    if saveFile != None:\n",
    "        plt.savefig(saveFile)\n",
    "\n",
    "def plotCovariates(df_age, df_sex, saveFile=None):\n",
    "    plt.style.use('ggplot')\n",
    "    fig = plt.figure(figsize=(9, 16))\n",
    "\n",
    "    plt.subplot(311) \n",
    "    ax1 = sns.boxplot(x=outcomes[0], y=\"Age_y\", data=df_age)\n",
    "    ax1.minorticks_on()\n",
    "    ax1.xaxis.set_minor_locator(AutoMinorLocator(2))\n",
    "    ax1.grid(which='minor', axis='x', linewidth=1)\n",
    "    sns.stripplot(x = outcomes[0], y = \"Age_y\", color = 'black', data = df_age)\n",
    "\n",
    "    plt.subplot(312) \n",
    "    ax1 = sns.barplot(x='Tract', y='Sex', data=df_sex, hue=outcomes[0])\n",
    "    ax1.minorticks_on()\n",
    "    ax1.xaxis.set_minor_locator(AutoMinorLocator(2))\n",
    "    ax1.grid(which='minor', axis='x', linewidth=1)\n",
    "\n",
    "    plt.subplot(313) \n",
    "    ax1 = sns.countplot(x=outcomes[0], data=df_age)\n",
    "    \n",
    "    if saveFile != None:\n",
    "        plt.savefig(saveFile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f3924759-1fdb-4b50-b9d6-bdc3cda16e2d",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plotGroupBoxplots(df_fa, df_rd, df_md, df_ad, saveFile=None):\n",
    "    \n",
    "    plt.style.use('ggplot')\n",
    "    fig = plt.figure(figsize=(16, 16))\n",
    "    \n",
    "    plt.subplot(411)\n",
    "    ax = sns.boxplot(x=\"Tract\", y=\"fa\",\n",
    "            hue=outcomes[0], palette='Pastel1',\n",
    "            data=df_fa)\n",
    "    ax = sns.despine(offset=10, trim=True)\n",
    "    ax = sns.swarmplot(x=\"Tract\", y=\"fa\", data=df_fa, hue=outcomes[0], dodge=True)\n",
    "    ax.axhline(1.00, alpha=0.5, c='#000000', ls='-')\n",
    "    ax.set(xticklabels=[])\n",
    "    ax.set(xlabel=None)\n",
    "\n",
    "    \n",
    "    plt.subplot(412)\n",
    "    ax = sns.boxplot(x=\"Tract\", y=\"rd\",\n",
    "            hue=outcomes[0], palette='Pastel1',\n",
    "            data=df_rd)\n",
    "    ax = sns.despine(offset=10, trim=True)\n",
    "    ax = sns.swarmplot(x=\"Tract\", y=\"rd\", data=df_rd, hue=outcomes[0], dodge=True)\n",
    "    ax.axhline(1.00, alpha=0.5, c='#000000', ls='-')\n",
    "    ax.set(xticklabels=[])\n",
    "    ax.set(xlabel=None)\n",
    "\n",
    "    \n",
    "    plt.subplot(413)\n",
    "    ax = sns.boxplot(x=\"Tract\", y=\"md\",\n",
    "            hue=outcomes[0], palette='Pastel1',\n",
    "            data=df_md)\n",
    "    ax = sns.despine(offset=10, trim=True)\n",
    "    ax = sns.swarmplot(x=\"Tract\", y=\"md\", data=df_md, hue=outcomes[0], dodge=True)\n",
    "    ax.axhline(1.00, alpha=0.5, c='#000000', ls='-')\n",
    "    ax.set(xticklabels=[])\n",
    "    ax.set(xlabel=None)\n",
    "    \n",
    "    plt.subplot(414)\n",
    "    ax = sns.boxplot(x=\"Tract\", y=\"ad\",\n",
    "            hue=outcomes[0], palette='Pastel1',\n",
    "            data=df_ad)\n",
    "    ax = sns.despine(offset=10, trim=True)\n",
    "    ax = sns.swarmplot(x=\"Tract\", y=\"ad\", data=df_ad, hue=outcomes[0], dodge=True)\n",
    "    ax.axhline(1.00, alpha=0.5, c='#000000', ls='-')\n",
    "    #ax.set(xticklabels=[])\n",
    "    ax.set(xlabel=None)\n",
    "    \n",
    "    \n",
    "    plt.setp(ax.get_xticklabels(), rotation=30, horizontalalignment='right') \n",
    "\n",
    "    \n",
    "    if saveFile != None:\n",
    "        plt.savefig(saveFile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9a75275f-7814-4c6c-b772-03bf3d8b26cb",
   "metadata": {},
   "outputs": [],
   "source": [
    "# def plotGroupBoxplots_PerMeas(df_meas, meas, saveFile=None):\n",
    "    \n",
    "#     plt.style.use('ggplot')\n",
    "#     fig = plt.figure(figsize=(4, 16))\n",
    "    \n",
    "#     ax = sns.boxplot(x=\"Tract\", y=meas,\n",
    "#             hue=outcomes[0], palette='Pastel1',\n",
    "#             data=df_meas)\n",
    "#     ax = sns.despine(offset=10, trim=True)\n",
    "#     ax = sns.swarmplot(x=\"Tract\", y=meas, data=df_meas, hue=outcomes[0], dodge=True)\n",
    "#     ax.axhline(1.00, alpha=0.5, c='#000000', ls='-')\n",
    "\n",
    "#     plt.setp(ax.get_xticklabels(), rotation=30, horizontalalignment='right') \n",
    "\n",
    "    \n",
    "#     if saveFile != None:\n",
    "#         plt.savefig(saveFile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7adbd5d5-2128-445f-b6b3-7d1790b0becb",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_fa_l = getTractMeasure(df_fa, measure=measures[0])\n",
    "df_rd_l = getTractMeasure(df_rd, measure=measures[1])\n",
    "df_md_l = getTractMeasure(df_md, measure=measures[2])\n",
    "df_ad_l = getTractMeasure(df_ad, measure=measures[3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "173bb20a-8761-4ff7-81e1-139319f389d8",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "29f81b26-1fd1-4b9a-9cb0-0c4f1b88f14c",
   "metadata": {},
   "outputs": [],
   "source": [
    "violin_output    = None\n",
    "covariate_output = None\n",
    "#violin_output    = os.path.join(outputBaseDir, 'normByAvgBoxplotsFA_{}.png'.format(selected_agerange))\n",
    "boxplot_output    = os.path.join(outputBaseDir, normFuncName,'normByBoxplots_{}.png'.format(selected_agerange))\n",
    "covariate_output = os.path.join(outputBaseDir, normFuncName, 'normByViolin_covariates_{}.png'.format(selected_agerange))\n",
    "\n",
    "#plotViolins(df_fa_l, df_rd_l, df_md_l, df_ad_l, violin_output)\n",
    "plotCovariates(df_age, df_sex, covariate_output)\n",
    "plotGroupBoxplots(df_fa_l, df_rd_l, df_md_l, df_ad_l, boxplot_output)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "93f97b44-8033-4163-92e5-ba43d5bc6316",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "00981246-fefd-4a62-a352-ea786aca43d0",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2308ced5-8c88-42c9-9939-1cc2fd041918",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d628a04d-969b-4c4b-ad82-b4299d9fd7bd",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.8.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}