{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Cascading search\n",
    "\n",
    "In the [Quickstart](./quickstart.ipynb) notebook we covered searching datasets one-by-one and gradually reducing the spatial domain of our search based on overlapping footprints. \n",
    "\n",
    "`coincident` also provides a  `cascading_search()`[#coincident.search.cascading_search] method as a convenience to perform this same type of search in one go.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import coincident\n",
    "import geopandas as gpd\n",
    "import xyzservices.providers as xyz\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Identify a primary dataset\n",
    "\n",
    "Start by loading a full resolution polygon of a 3DEP LiDAR workunit which has a known start_datetime and end_datatime:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "workunit = \"CO_WestCentral_2019\"\n",
    "df_wesm = coincident.search.wesm.read_wesm_csv()\n",
    "gf_lidar = coincident.search.wesm.load_by_fid(\n",
    "    df_wesm[df_wesm.workunit == workunit].index\n",
    ")\n",
    "gf_lidar"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Search secondary datasets\n",
    "\n",
    "Provide a list that will be searched in order. The list contains tuples of dataset aliases and the temporal pad in days to search before the primary dataset start and end dates"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "secondary_datasets = [\n",
    "    (\"maxar\", 14),  # +/- 14 days from lidar\n",
    "    (\"gedi\", 40),  # +/- 40 days from lidar\n",
    "    (\"icesat-2\", 60),\n",
    "]\n",
    "\n",
    "dataframes = coincident.search.cascading_search(\n",
    "    gf_lidar,\n",
    "    secondary_datasets,\n",
    "    min_overlap_area=30,  # km^2\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualize results\n",
    "\n",
    "Below we visualize cropped footprints from each secondary dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "gf_maxar = dataframes[0]\n",
    "print(len(gf_maxar))\n",
    "m = gf_lidar.explore(\n",
    "    style_kwds=dict(fill=None, color=\"black\"), tiles=xyz.CartoDB.Positron\n",
    ")  # basemap\n",
    "gf_maxar.explore(m=m, column=\"datetime\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "gf_gedi = dataframes[1]\n",
    "print(len(gf_gedi))\n",
    "m = gf_lidar.explore(\n",
    "    style_kwds=dict(fill=None, color=\"black\"), tiles=xyz.CartoDB.Positron\n",
    ")  # basemap\n",
    "gf_gedi.explore(m=m, column=\"datetime\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "gf_is2 = dataframes[2]\n",
    "print(len(gf_maxar))\n",
    "m = gf_lidar.explore(\n",
    "    style_kwds=dict(fill=None, color=\"black\"), tiles=xyz.CartoDB.Positron\n",
    ")  # basemap\n",
    "gf_is2.explore(m=m, column=\"datetime\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(figsize=(10, 5))\n",
    "for df in dataframes:\n",
    "    if \"constellation\" in df.columns:\n",
    "        label = df.constellation.iloc[0]\n",
    "    else:\n",
    "        label = df.collection.iloc[0]\n",
    "    plt.scatter(x=df[\"datetime\"], y=df[\"collection\"], s=50, marker=\"d\", label=label)\n",
    "\n",
    "# NOTE: probably a more robust way to set aspect depending on date range\n",
    "ax.set_aspect(6)\n",
    "plt.axvline(\n",
    "    gf_lidar.start_datetime.iloc[0],\n",
    "    color=\"black\",\n",
    "    linestyle=\"--\",\n",
    "    linewidth=0.5,\n",
    "    label=\"LiDAR\",\n",
    ")\n",
    "plt.axvline(gf_lidar.end_datetime.iloc[0], color=\"black\", linestyle=\"--\", linewidth=0.5)\n",
    "plt.title(\"CO_WestCentral_2019 Overlaps\")\n",
    "plt.legend(loc=\"lower right\")\n",
    "fig.autofmt_xdate()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Save results\n",
    "\n",
    "The footprints of the last secondary dataset show where we have *spatial* intersections across all datasets. We save this a single MultiPolygon to use in QGIS or [geojson.io](https://geojson.io)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "gf_is2.dissolve()[[\"geometry\"]].to_file(\"/tmp/CO_WestCentral_2019_overlaps.geojson\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Summary \n",
    "\n",
    "- The CO_WestCentral_2019 USGS 3DEP LiDAR was acquired between 2019-08-21 and 2019-09-19\n",
    "- We found 7 Maxar Stereo acquisitions with 14 days of the LiDAR\n",
    "- We found 14 GEDI acquisitions that overlap the lidar+stereo footprints within 40 days of LiDAR \n",
    "- We found 7 ICESat-2 acquisitions that overlap combined lidar+stereo+GEDI footprints within 60 days of LiDAR\n",
    "\n",
    "The final 'overlap' polygons have at least a 30k^2 area in which at least two of the datasets intersect. Acquisition dates for any given footprint vary from 60 days before the LiDAR was acquired through 60 days afterwards. "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "dev",
   "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": "3.12.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}