{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Fannie Mae mortgage data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import cuxfilter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from cuxfilter import charts\n",
    "import cuxfilter\n",
    "from bokeh import palettes\n",
    "from cuxfilter.layouts import *\n",
    "import panel as pn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#update data_dir if you have downloaded datasets elsewhere\n",
    "DATA_DIR = './data'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Download required datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from cuxfilter.sampledata import datasets_check\n",
    "datasets_check('mortgage', base_dir=DATA_DIR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# add static legend to the dashboard widget section\n",
    "def generate_legend_markdown():\n",
    "    range_ = cux_df.data.delinquency_12_prediction.describe().round(2)\n",
    "    legend = pn.GridBox(ncols=2)\n",
    "    colors = palettes.Purples4\n",
    "\n",
    "    legend.append(f'{range_[\"min\"]} - {range_[\"25%\"]}')\n",
    "    legend.append(pn.pane.HTML(background=colors[0], width=30, height=10))\n",
    "    legend.append(f'{range_[\"25%\"]} - {range_[\"50%\"]}')\n",
    "    legend.append(pn.pane.HTML(background=colors[1], width=30, height=10))\n",
    "    legend.append(f'{range_[\"50%\"]} - {range_[\"75%\"]}')\n",
    "    legend.append(pn.pane.HTML(background=colors[2], width=30, height=10))\n",
    "    legend.append(f'{range_[\"75%\"]} <')\n",
    "    legend.append(pn.pane.HTML(background=colors[3], width=30, height=10))\n",
    "    return legend"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define charts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "cux_df = cuxfilter.DataFrame.from_arrow(DATA_DIR+'/146M_predictions_v2.arrow')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set map_api_key parameter on choropleth to get bg map tiles\n",
    "chart0 = charts.choropleth(x='zip', y='delinquency_12_prediction', color_column='delinquency_12_prediction', color_aggregate_fn='mean', geo_color_palette=palettes.Purples4,\n",
    "                                  geoJSONSource = 'https://raw.githubusercontent.com/rapidsai/cuxfilter/GTC-2018-mortgage-visualization/javascript/demos/GTC%20demo/src/data/zip3-ms-rhs-lessprops.json',\n",
    "                                  data_points=1000) #, mapbox_api_key=\"\")\n",
    "\n",
    "chart2 = charts.bar('delinquency_12_prediction',data_points=50)\n",
    "chart3 = charts.range_slider('borrower_credit_score',data_points=50)\n",
    "chart4 = charts.card(generate_legend_markdown(), widget=True, title=\"zip legend\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Add mappings for chart1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "mapper1 = {}\n",
    "\n",
    "for val in cux_df.data.dti.unique().to_pandas().tolist():\n",
    "    mapper1[int(val)] = 'l_'+str(val)\n",
    "\n",
    "chart1 = charts.panel_widgets.drop_down('dti', label_map=mapper1)\n",
    "#also available: panel_widgets.multi_select"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create a dashboard object"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d = cux_df.dashboard([chart0], sidebar=[chart1, chart4], layout=feature_and_base, theme=cuxfilter.themes.rapids_dark, title=\"Ajay's Dashboard\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#add more charts\n",
    "d.add_charts(charts=[chart2], sidebar=[chart3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#run dashboard inline within the notebook cell\n",
    "d.app()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Starting the dashboard"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. d.show('current_notebook_url:current_notebook_port') remote dashboard\n",
    "\n",
    "2. d.app() inline within the notebook cell\n",
    "\n",
    "Incase you need to stop the server:\n",
    "\n",
    "- d.stop()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![mortgage](../../../_images/mortgage.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Export the queried data into a dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "queried_df = d.export()"
   ]
  }
 ],
 "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": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}