{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "39bf1de3-cba6-475a-a988-ad48e5af4a04",
   "metadata": {},
   "source": [
    "# Get zonal stats "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ba047a55-642d-4c27-a367-5f35f4406218",
   "metadata": {},
   "outputs": [],
   "source": [
    "import ibis\n",
    "import ibis.selectors as s\n",
    "from ibis import _\n",
    "import fiona\n",
    "import geopandas as gpd\n",
    "import rioxarray\n",
    "from shapely.geometry import box\n",
    "\n",
    "import rasterio\n",
    "from rasterio.mask import mask\n",
    "from rasterstats import zonal_stats\n",
    "import pandas as pd\n",
    "from joblib import Parallel, delayed\n",
    "\n",
    "con = ibis.duckdb.connect()\n",
    "con.load_extension(\"spatial\")\n",
    "threads = -1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8b5656db-2d1d-4ca8-826d-7588126e52e8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# cropping US data to only CA \n",
    "def crop_raster_to_bounds(tif_file, vector_gdf):\n",
    "    with rasterio.open(tif_file) as src:\n",
    "        # Get California's bounding box in the same CRS as the raster\n",
    "        california_bounds = vector_gdf.total_bounds\n",
    "        california_bounds = rasterio.coords.BoundingBox(\n",
    "            *california_bounds\n",
    "        )\n",
    "        # Crop the raster to the California bounding box\n",
    "        out_image, out_transform = mask(src, [california_bounds], crop=True)\n",
    "        out_meta = src.meta.copy()\n",
    "        out_meta.update({\n",
    "            \"driver\": \"GTiff\",\n",
    "            \"height\": out_image.shape[1],\n",
    "            \"width\": out_image.shape[2],\n",
    "            \"transform\": out_transform\n",
    "        })\n",
    "        print(\"Unique values in cropped raster:\", np.unique(out_image))\n",
    "\n",
    "    return out_image, out_meta\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9a0e3446-16ac-40b0-9e34-db0157038c5a",
   "metadata": {},
   "outputs": [],
   "source": [
    "def big_zonal_stats(vec_file, tif_file, stats, col_name, n_jobs, verbose=10, timeout=10000):\n",
    "    gdf = gpd.read_parquet(vec_file)\n",
    "    if gdf.crs is None:\n",
    "        gdf = gdf.set_crs(\"EPSG:4326\")\n",
    "    gdf = gdf.rename(columns={\"geom\": \"geometry\"})\n",
    "    gdf = gdf.set_geometry(\"geometry\")\n",
    "    gdf = gdf[gdf[\"geometry\"].notna()].copy()\n",
    "\n",
    "    with rasterio.open(tif_file) as src:\n",
    "        raster_crs = src.crs\n",
    "        gdf = gdf.to_crs(raster_crs)  # Transform vector to raster CRS\n",
    "        \n",
    "        # CA bounding box + convert it to a polygon in raster CRS\n",
    "        california_polygon = box(*gdf.total_bounds)\n",
    "        \n",
    "        out_image, out_transform = mask(src, [california_polygon], crop=True, nodata=src.nodata)\n",
    "\n",
    "        # If raster is 3D, select the first band\n",
    "        if out_image.ndim == 3:\n",
    "            out_image = out_image[0]\n",
    "\n",
    "    # compute zonal statistics for each geometry slice\n",
    "    def get_stats(geom_slice):\n",
    "        geom = [geom_slice.geometry]\n",
    "        stats_result = zonal_stats(\n",
    "            geom, out_image, stats=stats, affine=out_transform, all_touched=True, nodata=src.nodata\n",
    "        )\n",
    "        return stats_result[0] if stats_result and stats_result[0].get(\"mean\") is not None else {'mean': None}\n",
    "\n",
    "    output = [get_stats(row) for row in gdf.itertuples()]\n",
    "    gdf[col_name] = [res['mean'] for res in output]\n",
    "\n",
    "    return gdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ce66bae6-bac5-4837-9b01-fde16a00c303",
   "metadata": {},
   "outputs": [],
   "source": [
    "# getting local copies of data \n",
    "# aws s3 cp s3://vizzuality/hfp-100/hfp_2021_100m_v1-2_cog.tif . --endpoint-url=https://data.source.coop\n",
    "# aws s3 cp s3://vizzuality/lg-land-carbon-data/natcrop_bii_100m_cog.tif . --endpoint-url=https://data.source.coop\n",
    "# aws s3 cp s3://vizzuality/lg-land-carbon-data/natcrop_fii_100m_cog.tif . --endpoint-url=https://data.source.coop\n",
    "# aws s3 cp s3://vizzuality/lg-land-carbon-data/natcrop_expansion_100m_cog.tif . --endpoint-url=https://data.source.coop\n",
    "# aws s3 cp s3://vizzuality/lg-land-carbon-data/natcrop_reduction_100m_cog.tif . --endpoint-url=https://data.source.coop\n",
    "# aws s3 cp s3://cboettig/carbon/cogs/irrecoverable_c_total_2018.tif . --endpoint-url=https://data.source.coop\n",
    "# aws s3 cp s3://cboettig/carbon/cogs/manageable_c_total_2018.tif . --endpoint-url=https://data.source.coop\n",
    "# ! aws s3 cp s3://cboettig/justice40/disadvantaged-communities.parquet . --endpoint-url=https://data.source.coop\n",
    "# minio/shared-biodiversity/redlist/cog/combined_sr_2022.tif\n",
    "# /home/rstudio/minio/shared-biodiversity/redlist/cog/combined_rwr_2022.tif\n",
    "# ! aws s3 cp s3://cboettig/social-vulnerability/svi2020_us_tract.parquet . --endpoint-url=https://data.source.coop\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "531e7f88-1ce1-4027-b0ab-aab597e9a2b2",
   "metadata": {},
   "source": [
    "# Biodiversity Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "66dec912-ad8a-41cf-a5c2-6ec9cc350984",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "tif_file = 'SpeciesRichness_All.tif'\n",
    "vec_file = \"/home/rstudio/github/ca-30x30/ca2024-30m.parquet\"\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'], col_name = \"richness\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b081ec1a-ea91-485e-95f9-12cd06c2002a",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "tif_file = 'RSR_All.tif'\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'],\n",
    "                      col_name = \"rsr\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d5133f36-404e-4f6a-a90b-eb5f098e6f06",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "tif_file = 'combined_sr_2022.tif'\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'],  col_name = \"all_species_richness\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2ce56a66-34e3-4f61-95ae-65d1f06bc468",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "tif_file = 'combined_rwr_2022.tif'\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'],  col_name = \"all_species_rwr\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6c129894-3775-4842-8767-f81a8f626d2c",
   "metadata": {},
   "source": [
    "# Carbon Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "19c3e402-8712-450f-b3dd-af9d0c01689c",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "tif_file = 'irrecoverable_c_total_2018.tif'\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'],  col_name = \"irrecoverable_carbon\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c55c777a-48ce-4403-a171-cfc0d2351df6",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "tif_file = 'manageable_c_total_2018.tif'\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'],  col_name = \"manageable_carbon\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "33ac0fb7-2cde-448d-a634-1973e34ac14f",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "tif_file = 'deforest_carbon_100m_cog.tif'\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'], \n",
    "                     col_name = \"deforest_carbon\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "096c00a8-57af-41d7-93cc-85d85414aa4f",
   "metadata": {},
   "source": [
    "# Human Impact Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d2a8c10f-e94b-4eef-940f-2af9599edee1",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "tif_file = 'natcrop_bii_100m_cog.tif'\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'], \n",
    "                     col_name = \"biodiversity_intactness_loss\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1c318f39-7ca0-4f3c-80fb-73f72202e4e0",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "tif_file = 'natcrop_fii_100m_cog.tif'\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'],\n",
    "                     col_name = \"forest_integrity_loss\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "aef9070a-c87a-463e-81b8-3cc6c5c9d484",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "tif_file = 'natcrop_expansion_100m_cog.tif'\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'],  col_name = \"crop_expansion\", n_jobs=threads, verbose=0)\n",
    "gpd.GeoDataFrame(df, geometry=\"geometry\").to_parquet(\"cpad-stats-temp.parquet\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d94f937b-b32c-4de1-b4ac-93ce33f0919f",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "tif_file = 'natcrop_reduction_100m_cog.tif'\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'],  col_name = \"crop_reduction\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6bdaba61-30c1-49d6-a4e6-db68f1daafa3",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "tif_file = 'hfp_2021_100m_v1-2_cog.tif'\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'],  col_name = \"human_impact\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f8e037d4-7a34-42bc-941f-0c09ee80ef3b",
   "metadata": {},
   "source": [
    "# Need to convert SVI & Justice40 files to tif"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c4a19013-65f1-4eef-be2d-0cf1be3d0f7f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import geopandas as gpd\n",
    "import numpy as np\n",
    "import rasterio\n",
    "from rasterio.features import rasterize\n",
    "from rasterio.transform import from_bounds\n",
    "\n",
    "def get_geotiff(gdf, output_file,col):\n",
    "    gdf = gdf.set_geometry(\"geometry\")\n",
    "    gdf = gdf.set_crs(\"EPSG:4326\")\n",
    "    print(gdf.crs)\n",
    "\n",
    "    # Set raster properties\n",
    "    minx, miny, maxx, maxy = gdf.total_bounds  # Get the bounds of the geometry\n",
    "    pixel_size = 0.01  # Define the pixel size in units of the CRS\n",
    "    width = int((maxx - minx) / pixel_size)\n",
    "    height = int((maxy - miny) / pixel_size)\n",
    "    transform = from_bounds(minx, miny, maxx, maxy, width, height)\n",
    "    \n",
    "    # Define rasterization with continuous values\n",
    "    shapes = ((geom, value) for geom, value in zip(gdf.geometry, gdf[col]))\n",
    "    raster = rasterize(\n",
    "        shapes,\n",
    "        out_shape=(height, width),\n",
    "        transform=transform,\n",
    "        fill=0.0,  # Background value for areas outside the geometry\n",
    "        dtype=\"float32\"  # Set data type to handle continuous values\n",
    "    )\n",
    "    print(\"Unique values in raster:\", np.unique(raster))\n",
    "\n",
    "    # Define GeoTIFF metadata\n",
    "    out_meta = {\n",
    "        \"driver\": \"GTiff\",\n",
    "        \"height\": height,\n",
    "        \"width\": width,\n",
    "        \"count\": 1,\n",
    "        \"dtype\": raster.dtype,\n",
    "        \"crs\": gdf.crs,\n",
    "        \"transform\": transform,\n",
    "        \"compress\": \"deflate\"  # Use compression to reduce file size\n",
    "    }\n",
    "    \n",
    "    # Write to a GeoTIFF file with COG options\n",
    "    with rasterio.open(output_file, \"w\", **out_meta) as dest:\n",
    "        dest.write(raster, 1)\n",
    "        dest.build_overviews([2, 4, 8, 16], rasterio.enums.Resampling.average)\n",
    "        dest.update_tags(1, TIFFTAG_RESOLUTION_UNIT=\"Meter\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4925a74-5ed2-49a4-845b-6a0f0398a43e",
   "metadata": {},
   "source": [
    "# SVI"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4e678f01-73af-4f99-a565-e9b7f04d9547",
   "metadata": {},
   "outputs": [],
   "source": [
    "# clean up SVI data\n",
    "svi_df =  (con\n",
    " .read_parquet(\"svi2020_us_tract.parquet\")\n",
    " .select(\"RPL_THEMES\",\"RPL_THEME1\",\"RPL_THEME2\",\"RPL_THEME3\",\"RPL_THEME4\",\"Shape\")\n",
    " .rename(SVI = \"RPL_THEMES\", socioeconomic = \"RPL_THEME1\", \n",
    "         household_char = \"RPL_THEME2\", racial_ethnic_minority =  \"RPL_THEME3\",\n",
    "         housing_transit = \"RPL_THEME4\", geometry = \"Shape\")\n",
    ".cast({\"geometry\":\"geometry\"})\n",
    ")\n",
    "svi_df.execute().to_parquet(\"svi2020_us_tract_clean.parquet\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c5046d6b-9798-46d3-a1bc-548e29414007",
   "metadata": {},
   "outputs": [],
   "source": [
    "gdf = gpd.read_parquet(\"svi2020_us_tract_clean.parquet\")\n",
    "svi = gdf[['SVI','geometry']]\n",
    "socio = gdf[['socioeconomic','geometry']]\n",
    "house = gdf[['household_char','geometry']]\n",
    "minority = gdf[['racial_ethnic_minority','geometry']]\n",
    "transit = gdf[['housing_transit','geometry']]\n",
    "\n",
    "#convert SVI parquet to tif\n",
    "get_geotiff(svi,\"svi.tif\",\"SVI\")\n",
    "get_geotiff(socio,\"svi_socioeconomic.tif\",\"socioeconomic\")\n",
    "get_geotiff(house,\"svi_household.tif\",\"household_char\")\n",
    "get_geotiff(minority,\"svi_minority.tif\",\"racial_ethnic_minority\")\n",
    "get_geotiff(transit,\"svi_transit.tif\",\"housing_transit\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6a36b77f-d0be-45bd-9318-da4b57eaf353",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "tif_file = 'svi.tif'\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'],  col_name = \"SVI\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "05ef74e2-3f23-4f69-8cd3-8862cb73a259",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "tif_file = 'svi_socioeconomic.tif'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'],  col_name = \"socioeconomic_status\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "23417a03-38c2-4b31-8340-f08ec8a11631",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "tif_file = 'svi_household.tif'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'],  col_name = \"household_char\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "de86d7f0-6cdc-4d05-bdee-d9803cbd83bd",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "tif_file = 'svi_minority.tif'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'],  col_name = \"racial_ethnic_minority\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0c49dd50-7dd3-4240-9af8-3e32ec656bc0",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "tif_file = 'svi_transit.tif'\n",
    "df = big_zonal_stats(vec_file, tif_file, stats = ['mean'],  col_name = \"housing_transit\", n_jobs=threads, verbose=0).to_parquet(\"cpad-stats-temp.parquet\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ff4b6604-9828-4882-90bd-554c21f5c6e6",
   "metadata": {},
   "source": [
    "# Justice40 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3678a91f-72f7-4339-a409-a97776cba043",
   "metadata": {},
   "outputs": [],
   "source": [
    "#clean up\n",
    "justice40 =  (con\n",
    "        .read_parquet(\"disadvantaged-communities.parquet\")\n",
    "        .rename(geometry = \"SHAPE\",justice40=\"Disadvan\")\n",
    "        .filter(_.StateName == \"California\")\n",
    "        .mutate(geometry = _.geometry.convert(\"ESRI:102039\",\"EPSG:4326\"))\n",
    "        .select(\"justice40\",\"geometry\")\n",
    "        )\n",
    "justice40.execute().to_parquet(\"ca_justice40.parquet\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8faa425f-6f9c-4189-a53a-24dd0250c539",
   "metadata": {},
   "outputs": [],
   "source": [
    "# #justice40 is either 0 or 1, so we want to get the percentage of polygon where justice40 = 1.  \n",
    "\n",
    "def big_zonal_stats_binary(vec_file, justice40_file, col_name,projected_crs=\"EPSG:3310\"):\n",
    "    # Read both vector files as GeoDataFrames\n",
    "    gdf = gpd.read_parquet(vec_file)\n",
    "    justice40_gdf = gpd.read_parquet(justice40_file)\n",
    "    \n",
    "    # Set CRS if not already set (assuming both should be in EPSG:4326, modify if needed)\n",
    "    if gdf.crs is None:\n",
    "        gdf = gdf.set_crs(\"EPSG:4326\")\n",
    "    if justice40_gdf.crs is None:\n",
    "        justice40_gdf = justice40_gdf.set_crs(\"EPSG:4326\")\n",
    "     # Ensure both GeoDataFrames are in the same CRS and reproject to a projected CRS for area calculations\n",
    "    gdf = gdf.to_crs(projected_crs)\n",
    "    justice40_gdf = justice40_gdf.to_crs(projected_crs)\n",
    "    \n",
    "    # Ensure both GeoDataFrames are in the same CRS\n",
    "    gdf = gdf.to_crs(justice40_gdf.crs)\n",
    "    \n",
    "    # Filter justice40 polygons where justice40 == 1\n",
    "    justice40_gdf = justice40_gdf[justice40_gdf['justice40'] == 1].copy()\n",
    "    \n",
    "    # Prepare a list to hold percentage of justice40 == 1 for each polygon\n",
    "    percentages = []\n",
    "    \n",
    "    # Iterate over each polygon in the main GeoDataFrame\n",
    "    for geom in gdf.geometry:\n",
    "        # Find intersecting justice40 polygons\n",
    "        justice40_intersections = justice40_gdf[justice40_gdf.intersects(geom)].copy()\n",
    "        \n",
    "        # Calculate the intersection area\n",
    "        if not justice40_intersections.empty:\n",
    "            justice40_intersections['intersection'] = justice40_intersections.intersection(geom)\n",
    "            total_intersection_area = justice40_intersections['intersection'].area.sum()\n",
    "            \n",
    "            # Calculate percentage based on original polygon's area\n",
    "            percentage_1 = (total_intersection_area / geom.area) \n",
    "        else:\n",
    "            percentage_1 = 0.0  # No intersection with justice40 == 1 polygons\n",
    "        \n",
    "        # Append result\n",
    "        percentages.append(percentage_1)\n",
    "    \n",
    "    # Add results to the original GeoDataFrame\n",
    "    gdf[col_name] = percentages\n",
    "    return gdf\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fe80fc28-73ce-4a26-9925-851c2798e467",
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "vec_file = './cpad-stats-temp.parquet'\n",
    "\n",
    "df = big_zonal_stats_binary(vec_file, \"ca_justice40.parquet\", col_name=\"percent_disadvantaged\")\n",
    "df.to_parquet(\"cpad-stats-temp.parquet\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3083b85-1322-4188-ac08-e73c2570978c",
   "metadata": {},
   "source": [
    "# Cleaning up"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2e4de199-82d4-4e2b-8572-6fe19b57d1ee",
   "metadata": {},
   "outputs": [],
   "source": [
    "## clean up\n",
    "con = ibis.duckdb.connect(extensions=[\"spatial\"])\n",
    "ca_geom = con.read_parquet(\"ca2024-30m.parquet\").cast({\"geom\":\"geometry\"}).select(\"id\",\"geom\")\n",
    "\n",
    "ca = (con\n",
    "        .read_parquet(\"cpad-stats-temp.parquet\")\n",
    "        .cast({\n",
    "                \"crop_expansion\": \"int64\",\n",
    "                \"crop_reduction\": \"int64\",\n",
    "                \"manageable_carbon\": \"int64\",\n",
    "                \"irrecoverable_carbon\": \"int64\"\n",
    "            })\n",
    "        .rename(svi = \"SVI\")\n",
    "        .mutate(\n",
    "            richness=_.richness.round(3),\n",
    "                rsr=_.rsr.round(3),\n",
    "                all_species_rwr=_.all_species_rwr.round(3),\n",
    "                all_species_richness=_.all_species_richness.round(3),\n",
    "                percent_disadvantaged=(_.percent_disadvantaged).round(3),\n",
    "                svi=_.svi.round(3),\n",
    "                svi_socioeconomic_status=_.socioeconomic_status.round(3),\n",
    "                svi_household_char=_.household_char.round(3),\n",
    "                svi_racial_ethnic_minority=_.racial_ethnic_minority.round(3),\n",
    "                svi_housing_transit=_.housing_transit.round(3),\n",
    "                human_impact=_.human_impact.round(3),\n",
    "                deforest_carbon=_.deforest_carbon.round(3),\n",
    "                biodiversity_intactness_loss=_.biodiversity_intactness_loss.round(3),\n",
    "                forest_integrity_loss=_.forest_integrity_loss.round(3),\n",
    "            )\n",
    "          .select('established', 'reGAP', 'name', 'access_type', 'manager', 'manager_type', 'Easement', 'Acres', 'id', 'type','richness', \n",
    "                 'rsr', 'all_species_rwr', 'all_species_richness','deforest_carbon', 'irrecoverable_carbon', 'manageable_carbon', 'human_impact',\n",
    "                 'percent_disadvantaged','svi', 'svi_socioeconomic_status', 'svi_household_char', \n",
    "                 'svi_racial_ethnic_minority','svi_housing_transit')\n",
    "          .join(ca_geom, \"id\", how=\"inner\")\n",
    "        )\n",
    "\n",
    "ca.head(5).execute()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3780de2c-3a68-442c-bb3b-64c792418979",
   "metadata": {},
   "source": [
    "# Save as PMTiles + Upload data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "05c791c9-888a-483a-9dbb-a2ba7eb1bce2",
   "metadata": {},
   "outputs": [],
   "source": [
    "import subprocess\n",
    "import os\n",
    "from huggingface_hub import HfApi, login\n",
    "import streamlit as st\n",
    "\n",
    "login(st.secrets[\"HF_TOKEN\"])\n",
    "# api = HfApi(add_to_git_credential=False)\n",
    "api = HfApi()\n",
    "\n",
    "def hf_upload(file, repo_id):\n",
    "    info = api.upload_file(\n",
    "            path_or_fileobj=file,\n",
    "            path_in_repo=file,\n",
    "            repo_id=repo_id,\n",
    "            repo_type=\"dataset\",\n",
    "        )\n",
    "def generate_pmtiles(input_file, output_file, max_zoom=12):\n",
    "    # Ensure Tippecanoe is installed\n",
    "    if subprocess.call([\"which\", \"tippecanoe\"], stdout=subprocess.DEVNULL) != 0:\n",
    "        raise RuntimeError(\"Tippecanoe is not installed or not in PATH\")\n",
    "\n",
    "    # Construct the Tippecanoe command\n",
    "    command = [\n",
    "        \"tippecanoe\",\n",
    "        \"-o\", output_file,\n",
    "        \"-zg\",\n",
    "        \"--extend-zooms-if-still-dropping\",\n",
    "        \"--force\",\n",
    "        \"--projection\", \"EPSG:4326\",  \n",
    "        \"-L\",\"layer:\"+input_file,\n",
    "    ]\n",
    "    # Run Tippecanoe\n",
    "    try:\n",
    "        subprocess.run(command, check=True)\n",
    "        print(f\"Successfully generated PMTiles file: {output_file}\")\n",
    "    except subprocess.CalledProcessError as e:\n",
    "        print(f\"Error running Tippecanoe: {e}\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1f2d179d-6d47-4e84-83c6-7cb3d969fc00",
   "metadata": {},
   "outputs": [],
   "source": [
    "gdf = ca.execute().set_crs(\"EPSG:4326\")\n",
    "gdf.to_file(\"cpad-stats.geojson\")\n",
    "\n",
    "generate_pmtiles(\"cpad-stats.geojson\", \"cpad-stats.pmtiles\")\n",
    "hf_upload(\"cpad-stats.pmtiles\", \"boettiger-lab/ca-30x30\")\n",
    "\n",
    "gdf.to_parquet(\"cpad-stats.parquet\")\n",
    "hf_upload(\"cpad-stats.parquet\", \"boettiger-lab/ca-30x30\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bf53bfe6-2175-4aa8-b3cb-be2c1db86b1b",
   "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": "3.12.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}