templates_for_plots/time_series_multi_plots.ipynb

{
 "metadata": {
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.0"
  },
  "orig_nbformat": 2,
  "kernelspec": {
   "name": "python390jvsc74a57bd0e2bfb1b1dd0bcdebdb315279aa118b1f834444d4ba3ba6d660e9f6ce7703f6a2",
   "display_name": "Python 3.9.0 64-bit"
  },
  "metadata": {
   "interpreter": {
    "hash": "e2bfb1b1dd0bcdebdb315279aa118b1f834444d4ba3ba6d660e9f6ce7703f6a2"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2,
 "cells": [
  {
   "source": [
    "# Time Series Multi-Plots\n",
    "\n",
    "For selected ICU stays, create a time series for each parameter (heart rate, blood pressure, and O2 saturation). The time series are arranged one below the other and share the same x-axis, so it is easy to compare all parameters at a certain time point."
   ],
   "cell_type": "markdown",
   "metadata": {}
  },
  {
   "source": [
    "## Load & Prepare Data"
   ],
   "cell_type": "markdown",
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import pyarrow as pa\n",
    "\n",
    "# Read chartevents_subset from parquet file to pandas data frame\n",
    "chartevents_subset = pd.read_parquet('../data/chartevents_subset.parquet', engine='pyarrow')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create set of relevant ICUSTAY_IDs to filter by\n",
    "icustayid_filter = [260223, 266144, 216834] # e.g. [260223,266144,216834]\n",
    "\n",
    "# Filter by ITEMIDs\n",
    "plotdata = chartevents_subset[chartevents_subset.ICUSTAY_ID.isin(icustayid_filter)].copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Add new column with ITEMID_LABEL, which can be used for the legend of the plot.\n",
    "import numpy as np\n",
    "plotdata.insert(loc=len(plotdata.columns), column='ITEMID_LABEL', value=np.nan)\n",
    "# Heart rate ITEMID_LABELs\n",
    "plotdata.loc[plotdata['ITEMID'] == 220045, 'ITEMID_LABEL'] = \"Heart rate (bpm)\"\n",
    "plotdata.loc[plotdata['ITEMID'] == 220046, 'ITEMID_LABEL'] = \"Alarm threshold: High heart rate (bpm)\"\n",
    "plotdata.loc[plotdata['ITEMID'] == 220047, 'ITEMID_LABEL'] = \"Alarm threshold: Low heart rate (bpm)\"\n",
    "# Blood pressure ITEMID_LABELs\n",
    "plotdata.loc[plotdata['ITEMID'] == 220179, 'ITEMID_LABEL'] = \"Non-invasive systolic blood pressure (mmHg)\"\n",
    "plotdata.loc[plotdata['ITEMID'] == 223751, 'ITEMID_LABEL'] = \"Alarm threshold: High systolic blood pressure (mmHg)\"\n",
    "plotdata.loc[plotdata['ITEMID'] == 223752, 'ITEMID_LABEL'] = \"Alarm threshold: Low systolic blood pressure (mmHg)\"\n",
    "# O2 saturation ITEMID_LABELs\n",
    "plotdata.loc[plotdata['ITEMID'] == 220277, 'ITEMID_LABEL'] = \"O2 saturation pulseoxymetry (%)\"\n",
    "plotdata.loc[plotdata['ITEMID'] == 223769, 'ITEMID_LABEL'] = \"Alarm threshold: High O2 saturation (%)\"\n",
    "plotdata.loc[plotdata['ITEMID'] == 223770, 'ITEMID_LABEL'] = \"Alarm threshold: Low O2 saturation (%)\""
   ]
  },
  {
   "source": [
    "## Create Plots"
   ],
   "cell_type": "markdown",
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import datetime as dt\n",
    "\n",
    "# Set shared variables\n",
    "fig_title = \"Vital sign histories for ICU stay \"\n",
    "xlabel = \"Time\"\n",
    "huevalue = \"ITEMID_LABEL\"\n",
    "\n",
    "# Set heart rate variables\n",
    "hr_title = \"Heart rate\"\n",
    "hr_ylabel = \"Beats per minute\"\n",
    "\n",
    "# Set blood pressure variables\n",
    "bp_title = \"Non-invasive systolic blood pressure\"\n",
    "bp_ylabel = \"Millimeter of mercury\"\n",
    "\n",
    "# Set O2 saturation variables\n",
    "o2_title = \"O2 saturation pulseoxymetry\"\n",
    "o2_ylabel = \"Percent\"\n",
    "\n",
    "for icustay in icustayid_filter:\n",
    "\n",
    "    # Set variables\n",
    "    xlimits = [plotdata[plotdata['ICUSTAY_ID'] == icustay].CHARTTIME.min()-pd.Timedelta('1 day'), plotdata[plotdata['ICUSTAY_ID'] == icustay].CHARTTIME.max()+pd.Timedelta('1 day')]\n",
    "\n",
    "    # Config figure\n",
    "    sns.set_style(\"whitegrid\")\n",
    "    fig, axs = plt.subplots(\n",
    "        3,\n",
    "        1,\n",
    "        figsize = (12, 15), \n",
    "        # sharex = True, \n",
    "        dpi = 72 # e.g. 72 for screen, 300 for print\n",
    "        )\n",
    "    fig.suptitle(str(fig_title+str(icustay)), fontweight='bold', color= 'black', fontsize=14, y=0.93)\n",
    "    fig.subplots_adjust(hspace = 0.6)\n",
    "\n",
    "    # Heart rate plot\n",
    "    ax = sns.lineplot(\n",
    "        ax = axs[0],\n",
    "        data = plotdata[(plotdata['ITEMID'].isin([220045,220046,220047])) & (plotdata['ICUSTAY_ID'] == icustay)], \n",
    "        x = \"CHARTTIME\",\n",
    "        y = \"VALUENUM\", \n",
    "        hue = huevalue,\n",
    "        hue_order = [\n",
    "            str(plotdata[plotdata['ITEMID'] == 220045].ITEMID_LABEL.iloc[0]),\n",
    "            str(plotdata[plotdata['ITEMID'] == 220046].ITEMID_LABEL.iloc[0]),\n",
    "            str(plotdata[plotdata['ITEMID'] == 220047].ITEMID_LABEL.iloc[0])],\n",
    "        style= huevalue,\n",
    "        drawstyle = 'steps-post', # Interpolate missing values by using the last available value\n",
    "        markers = ['p','^','v'],\n",
    "        markersize = 5,\n",
    "        dashes = False,\n",
    "        palette = [sns.color_palette(\"colorblind\")[0],sns.color_palette(\"colorblind\")[1],sns.color_palette(\"colorblind\")[2]]\n",
    "        )\n",
    "    axs[0].legend(loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "    axs[0].set_xlim(xlimits)\n",
    "    axs[0].set_title(hr_title, fontweight='bold', color= 'black', fontsize=13, y=1.025)\n",
    "    axs[0].set_xlabel(xlabel, fontsize=12, labelpad=15)\n",
    "    axs[0].set_ylabel(hr_ylabel, fontsize=12, labelpad=15)\n",
    "    axs[0].tick_params(axis=\"x\", rotation=90)\n",
    "\n",
    "    # Blood pressure plot\n",
    "    ax = sns.lineplot(\n",
    "        ax = axs[1],\n",
    "        data = plotdata[(plotdata['ITEMID'].isin([220179,223751,223752])) & (plotdata['ICUSTAY_ID'] == icustay)], \n",
    "        x = \"CHARTTIME\",\n",
    "        y = \"VALUENUM\", \n",
    "        hue = huevalue,\n",
    "        hue_order = [\n",
    "            str(plotdata[plotdata['ITEMID'] == 220179].ITEMID_LABEL.iloc[0]),\n",
    "            str(plotdata[plotdata['ITEMID'] == 223751].ITEMID_LABEL.iloc[0]),\n",
    "            str(plotdata[plotdata['ITEMID'] == 223752].ITEMID_LABEL.iloc[0])],\n",
    "        style= huevalue,\n",
    "        drawstyle = 'steps-post', # Interpolate missing values by using the last available value\n",
    "        markers = ['p','^','v'],\n",
    "        markersize = 5,\n",
    "        dashes = False,\n",
    "        palette = [sns.color_palette(\"colorblind\")[0],sns.color_palette(\"colorblind\")[1],sns.color_palette(\"colorblind\")[2]]\n",
    "        )\n",
    "    # plt.legend(title = None, bbox_to_anchor=(1.02, 0.3), loc='upper left', borderaxespad=0)\n",
    "    axs[1].legend(loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "    axs[1].set_xlim(xlimits) \n",
    "    axs[1].set_title(bp_title, fontweight='bold', color= 'black', fontsize=13, y=1.025)\n",
    "    axs[1].set_xlabel(xlabel, fontsize=12, labelpad=15)\n",
    "    axs[1].set_ylabel(bp_ylabel, fontsize=12, labelpad=15)\n",
    "    axs[1].tick_params(axis=\"x\", rotation=90)\n",
    "\n",
    "    # O2 saturation plot\n",
    "    ax = sns.lineplot(\n",
    "        ax = axs[2],\n",
    "        data = plotdata[(plotdata['ITEMID'].isin([220277,223769,223770])) & (plotdata['ICUSTAY_ID'] == icustay)], \n",
    "        x = \"CHARTTIME\",\n",
    "        y = \"VALUENUM\", \n",
    "        hue = huevalue,\n",
    "        hue_order = [\n",
    "            str(plotdata[plotdata['ITEMID'] == 220277].ITEMID_LABEL.iloc[0]),\n",
    "            str(plotdata[plotdata['ITEMID'] == 223769].ITEMID_LABEL.iloc[0]),\n",
    "            str(plotdata[plotdata['ITEMID'] == 223770].ITEMID_LABEL.iloc[0])],\n",
    "        style= huevalue,\n",
    "        drawstyle = 'steps-post', # Interpolate missing values by using the last available value\n",
    "        markers = ['p','^','v'],\n",
    "        markersize = 5,\n",
    "        dashes = False,\n",
    "        palette = [sns.color_palette(\"colorblind\")[0],sns.color_palette(\"colorblind\")[1],sns.color_palette(\"colorblind\")[2]]\n",
    "        )\n",
    "    axs[2].legend(loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "    axs[2].set_xlim(xlimits) \n",
    "    axs[2].set_title(o2_title, fontweight='bold', color= 'black', fontsize=13, y=1.025)\n",
    "    axs[2].set_xlabel(xlabel, fontsize=12, labelpad=15)\n",
    "    axs[2].set_ylabel(o2_ylabel, fontsize=12, labelpad=15)\n",
    "    axs[2].tick_params(axis=\"x\", rotation=90)\n",
    "\n",
    "    # Save plot as PNG file\n",
    "    path = '../plots/'\n",
    "    timestamp = dt.datetime.today().strftime('%Y-%m-%dT%H-%M-%S')\n",
    "    plt.savefig(str(path + 'icustay-' + str(icustay) + '_time-series-multi-plot_72-DPI_' + timestamp + '.png'), dpi=72, bbox_inches='tight')\n",
    "    plt.savefig(str(path + 'icustay-' + str(icustay) + '_time-series-multi-plot_300-DPI_' + timestamp + '.png'), dpi=300, bbox_inches='tight')\n",
    "\n",
    "    # Plot figure\n",
    "    plt.show(fig)\n"
   ]
  }
 ]
}