{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Generation of supplemental materials for the MRIQC paper\n",
    "\n",
    "This notebook is associated to the paper:\n",
    "\n",
    "Esteban O, Birman D, Schaer M, Koyejo OO, Poldrack RA, Gorgolewski KJ; MRIQC: Predicting Quality in Manual MRI Assessment Protocols Using No-Reference Image Quality Measures; bioRxiv 111294; doi:[10.1101/111294](https://doi.org/10.1101/111294)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "import os.path as op\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from mriqc.viz import misc as mviz\n",
    "from pkg_resources import resource_filename as pkgrf\n",
    "outputs_path = '../../mriqc-data/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x_path = pkgrf('mriqc', 'data/csv/x_abide.csv')\n",
    "y_path = pkgrf('mriqc', 'data/csv/y_abide.csv')\n",
    "ds030_x_path = pkgrf('mriqc', 'data/csv/x_ds030.csv')\n",
    "ds030_y_path = pkgrf('mriqc', 'data/csv/y_ds030.csv')\n",
    "\n",
    "rater_types = {'rater_1': float, 'rater_2': float, 'rater_3': float}\n",
    "mdata = pd.read_csv(y_path, index_col=False, dtype=rater_types)\n",
    "\n",
    "sites = list(sorted(list(set(mdata.site.values.ravel().tolist()))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fig = mviz.raters_variability_plot(\n",
    "    mdata, raters=['rater_1', 'rater_2', 'rater_3'], \n",
    "    rater_names=['Rater 1', 'Rater 2A', 'Rater 2B'],\n",
    "    out_file=op.join(outputs_path, 'figures', 'suppl-fig02.pdf'),\n",
    "    only_overlap=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import cohen_kappa_score\n",
    "overlap = mdata[np.all(~np.isnan(mdata[['rater_2', 'rater_3']]), axis=1)]\n",
    "y1 = overlap.rater_2.values.ravel().tolist()\n",
    "y2 = overlap.rater_3.values.ravel().tolist()\n",
    "\n",
    "fig = mviz.inter_rater_variability(y1, y2, raters=['Protocol A', 'Protocol B'],\n",
    "                                   out_file=op.join(outputs_path, 'figures', 'suppl-intrarv.pdf'))\n",
    "\n",
    "print(\"Cohen's Kappa %f\" % cohen_kappa_score(y1, y2))\n",
    "\n",
    "y1 = overlap.rater_2.values.ravel()\n",
    "y1[y1 == 0] = 1\n",
    "\n",
    "y2 = overlap.rater_3.values.ravel()\n",
    "y2[y2 == 0] = 1\n",
    "print(\"Cohen's Kappa (binarized): %f\" % cohen_kappa_score(y1, y2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fig = mviz.plot_corrmat(x_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fig = mviz.plot_histograms(x_path, y_path)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}