7_extra plots.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "4382ad81-b2cb-41c4-9fc2-8c0de49ec4aa",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from scipy import stats\n",
    "from matplotlib import pyplot as plt\n",
    "import seaborn as sns\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ffb3a741-c1f9-452f-8de3-b66f483dafc5",
   "metadata": {},
   "source": [
    "# Dylan's plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "ef674fae-e2fa-47ca-be8c-99569e11a2fe",
   "metadata": {},
   "outputs": [],
   "source": [
    "d3 = {'dataset': ['ABCD:parent', 'ABCD:parent', 'ABCD:parent', 'ABCD:child', 'ABCD:child', 'ABCD:child', 'HBN', 'HBN', 'HBN', 'BHRC', 'BHRC', 'BHRC'],\n",
    "     'hypothesis': ['Sensitivity', 'Specificity', 'Strict Specificity', 'Sensitivity', 'Specificity', 'Strict Specificity', 'Sensitivity', 'Specificity', 'Strict Specificity', 'Sensitivity', 'Specificity', 'Strict Specificity'],\n",
    "     'AUCROC': [0.953, 0.874, 0,  0.619, 0.478, 0.459, 0.861, 0.712, 0,  0.896, 0.802, 0.781],\n",
    "     'lci':    [0.927, 0.828, 0,  0.565, 0.398, 0.372, 0.804, 0.648, 0,  0.855, 0.736, 0.697],\n",
    "     'uci':    [0.974, 0.917, 0,  0.672, 0.558, 0.550, 0.915, 0.774, 0,  0.930, 0.860, 0.855],\n",
    "     'thresh': [0.8, 0.8,  np.nan,  0.8,   0.8,   0.8, 0.8, 0.8, np.nan,  0.8,   0.8,  0.8]  \n",
    "     }\n",
    "df3 = pd.DataFrame(data=d3)\n",
    "df3['los'] = df3.AUCROC - df3.lci\n",
    "df3['uos'] = df3.uci - df3.AUCROC\n",
    "\n",
    "title_lut = {\n",
    "    'ABCD:parent': 'ABCD: K-SADS Parent',\n",
    "    'ABCD:child': 'ABCD: K-SADS Child',\n",
    "    'HBN': 'HBN: K-SADS Clinician',\n",
    "    'BHRC': 'BHRC: DAWBA Clinician'\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "b2497ae4-6ee6-421b-bc10-7d12574d4110",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1206.36x400 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plotting:\n",
    "colors=['darkgray','gray','dimgray','lightgray']\n",
    "g = sns.catplot(x=\"hypothesis\", y=\"AUCROC\",\n",
    "                 hue = \"hypothesis\", col=\"dataset\",\n",
    "                 data=df3, kind=\"bar\",\n",
    "                 height=4, aspect=.7,\n",
    "                 palette='Greys');\n",
    "g.set_xticklabels(rotation=30)\n",
    "\n",
    "\n",
    "# get some color bar colors so they stand out\n",
    "foo = sns.palettes.color_palette('Greys', 2)\n",
    "ecolors =[foo[1], foo[0], foo[0]] \n",
    "\n",
    "for ax, (dsname, title) in zip(g.axes[0], title_lut.items()):\n",
    "    ax.set_title(title)\n",
    "    ax.set_xlabel('Hypothesis')\n",
    "    ax.errorbar(\n",
    "        x=[0,1,2],\n",
    "        y=df3.loc[df3.dataset == dsname, 'AUCROC'].values,\n",
    "        yerr=df3.loc[df3.dataset == dsname, ['los', 'uos']].values.T,\n",
    "        ecolor=ecolors,\n",
    "        elinewidth=2,\n",
    "        linestyle='none',\n",
    "    )\n",
    "    thrs = df3.loc[df3.dataset == dsname, 'thresh'].values\n",
    "    for thr, center, color in zip(thrs, [0,1,2], ecolors):\n",
    "        ax.hlines(thr, center-0.5, center+0.5, colors=color, ls='--')\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1d0f1d1f-f4d8-4dd7-9b8f-33746ba0d1bd",
   "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": "3.9.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}