predict-example.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c1c8697e-ac8e-4e6a-bd84-ab7e28eaeeee",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import rioxarray\n",
    "import rasterio\n",
    "import glob\n",
    "import xarray as xr\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "4cde6c22-dee7-48b7-b193-d20979074469",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import geomod\n",
    "from geomod.predictor import Predictor\n",
    "from geomod import utils"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65615dcc-cb1f-4bd2-8a24-eda628bd2d14",
   "metadata": {},
   "source": [
    "### Preparing the Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "697d622a-18e1-4e58-87ba-9ec9ca995e49",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "data_directory = '/home/jovyan/PlanetaryComputerExamples/adv-python-final/geomod/example-notebooks/data/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "2ea4c9e3-aa40-413f-aaf2-477dcc20c5c7",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "driver_map_names = ['elevation01', 'geology01', 'protection01']\n",
    "driver_map_paths = [data_directory+name+'.rst' for name in driver_map_names]\n",
    "driver_map_xr_list = [rioxarray.open_rasterio(map, masked=True) for map in driver_map_paths]\n",
    "all_driver_maps_xr = xr.concat(driver_map_xr_list, dim='band').assign_coords(band=driver_map_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "7905babc-6e6a-4a27-a16e-e34909924dac",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/srv/conda/envs/notebook/lib/python3.11/site-packages/xarray/core/duck_array_ops.py:188: RuntimeWarning: invalid value encountered in cast\n",
      "  return data.astype(dtype, **kwargs)\n"
     ]
    }
   ],
   "source": [
    "# round the driver maps to the nearest integer to get categorical data.\n",
    "# a future implementation will include the option to categorize driver maps according to user input\n",
    "\n",
    "all_driver_maps_xr = all_driver_maps_xr.round().astype('int')\n",
    "\n",
    "# this is necessary due to rounding not working correctly with nan values\n",
    "all_driver_maps_xr = xr.where(all_driver_maps_xr>-1000000, all_driver_maps_xr, np.nan)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "3b272502-b583-403a-ab2d-bf740dfa2d19",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# import the 1985 classification map, our starting map\n",
    "input_classification_map_xr = rioxarray.open_rasterio(data_directory+'1985anderson01.rst', masked=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "ec7ab074-24f5-4e7c-899e-9365e9ce6528",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# get the metadata from the .RDC that comes with every .rst terrset raster\n",
    "\n",
    "def get_metadata(rdc_file):\n",
    "# Open the text file for reading\n",
    "    metadata_dict = {}\n",
    "    with open(rdc_file, \"r\") as file:\n",
    "    # Read each line in the file\n",
    "        for line in file:\n",
    "            # Split each line into key and value using ':'\n",
    "            key, value = line.strip().split(':')[0:2]\n",
    "            # Store the key-value pair in the dictionary\n",
    "            metadata_dict[key.strip()] = value.strip()\n",
    "    return metadata_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "e1d775a5-085b-4e3e-9ef7-84889ddbd7e9",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'file format': 'IDRISI Raster A.1', 'file title': '1985 Anderson Land Category', 'data type': 'byte', 'file type': 'binary', 'columns': '512', 'rows': '512', 'ref. system': 'spc83ma1', 'ref. units': 'm', 'unit dist.': '1.0000000', 'min. X': '242160.0000000', 'max. X': '257520.0000000', 'min. Y': '923060.0000000', 'max. Y': '938420.0000000', \"pos'n error\": 'unknown', 'resolution': '30', 'min. value': '0', 'max. value': '7', 'display min': '0', 'display max': '7', 'value units': 'categories', 'value error': 'unknown', 'flag value': '0', \"flag def'n\": 'background', 'legend cats': '7', 'code      1': 'Built', 'code      2': 'Agriculture', 'code      3': 'Range', 'code      4': 'Forest', 'code      5': 'Water', 'code      6': 'Wetland', 'code      7': 'Barren', 'lineage': '3*C'}\n"
     ]
    }
   ],
   "source": [
    "# get the metadata for our starting classification map\n",
    "print(get_metadata(data_directory+'1985anderson01.RDC'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "90205c94-8aea-486a-9e9b-d7bd9ff7f4b6",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# reclassify the land cover map - we know from the metadata that the class of interest is 1 for built\n",
    "land_cover_map = utils.reclassify_landcover_map(input_classification_map_xr, class_of_interest=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "5f73124a-d40b-4dec-853b-c15c74860db8",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# plot the land cover map, we see the x and y coordinate scale, built areas as 1, all others as 0, and the mask as nan\n",
    "#land_cover_map.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "33116b9a-f913-446e-a27d-a0391bc33e75",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# we do the same with the validation map for the ending time\n",
    "validation_class_map_xr = rioxarray.open_rasterio(data_directory+'1999anderson01.rst', masked=True)\n",
    "validation_map = utils.reclassify_landcover_map(validation_class_map_xr, class_of_interest=1)\n",
    "#validation_map.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "389ba92c-2aec-4a97-887c-65093d2df6fc",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# users may define weights for each of the driver maps, or use no wieghts for an evenly weighted suitability map\n",
    "weights = {'elevation01':.8, 'geology01':.1, 'protection01':.1}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "6d2beb89-f2ea-40f2-8ac5-967876b68d13",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Here we initialize a model using the Predictor class and the weights\n",
    "landchange_model01 = Predictor(\n",
    "    land_cover_map = land_cover_map,\n",
    "    beginning_time = 1985,\n",
    "    ending_time = 1999,\n",
    "    time_step = 1,\n",
    "    suitability_map = None,\n",
    "    driver_maps = all_driver_maps_xr,\n",
    "    validation_map = validation_map,\n",
    "    pixel_quantities=None,\n",
    "    mask_image=None,\n",
    "    strata_map=None,\n",
    "    constrain_to_neighborhood=None,\n",
    "    driver_map_weights=weights)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "e1294ac3-6d45-4275-a056-7ddbbe41bd97",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "target number of change pixels: 3830\n",
      "Not using neighborhood constraint\n",
      "The number of pixels that can be changed: 75717\n",
      "Number of pixels above the change threshold: 2763\n",
      "There are 1320 tied pixels, of which 1067 must be selected arbitrarily.\n",
      "If this is unacceptable, consider reducing tied pixels by using more categorical variables to generate your suitability map\n"
     ]
    }
   ],
   "source": [
    "# run predict() on the model to get the predicted map\n",
    "landchange_model01.predict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "4a9472cb-6c56-4d0d-9ad4-2a12d72627b1",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3830\n",
      "0\n",
      "27473\n",
      "71887\n",
      "3830\n",
      "671\n",
      "26802\n",
      "71887\n",
      "31303.0\n",
      "30632.0\n"
     ]
    }
   ],
   "source": [
    "# access the predicted map as part of the model class\n",
    "print(((landchange_model01.predicted_map == 1) & (landchange_model01.land_cover_map == 0)).sum().item())\n",
    "print(((landchange_model01.predicted_map == 0) & (landchange_model01.land_cover_map == 1)).sum().item())\n",
    "print(((landchange_model01.predicted_map == 1) & (landchange_model01.land_cover_map == 1)).sum().item())\n",
    "print(((landchange_model01.predicted_map == 0) & (landchange_model01.land_cover_map == 0)).sum().item())\n",
    "\n",
    "print(((landchange_model01.validation_map == 1) & (landchange_model01.land_cover_map == 0)).sum().item())\n",
    "print(((landchange_model01.validation_map == 0) & (landchange_model01.land_cover_map == 1)).sum().item())\n",
    "print(((landchange_model01.validation_map == 1) & (landchange_model01.land_cover_map == 1)).sum().item())\n",
    "print(((landchange_model01.validation_map == 0) & (landchange_model01.land_cover_map == 0)).sum().item())\n",
    "\n",
    "print(landchange_model01.predicted_map.sum().item())\n",
    "print(landchange_model01.validation_map.sum().item())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "8026ade3-abbd-455b-9303-0a2e15a7848a",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'True Positives': 27197,\n",
       " 'True Negatives': 68452,\n",
       " 'False Positives': 4106,\n",
       " 'False Negatives': 3435,\n",
       " 'Precision': 0.8688304635338466,\n",
       " 'Recall': 0.8878623661530426,\n",
       " 'F1 Score': 0.8782433196092678}"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# call validate to get the CROSSTAB table\n",
    "landchange_model01.validate()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "58996887-4711-49a9-a622-0b236602fe66",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Here we initialize a model using the Predictor class\n",
    "# we pass a 3,3 kernel to constrain_to_neighborhood\n",
    "landchange_model02 = Predictor(\n",
    "    land_cover_map = land_cover_map,\n",
    "    beginning_time = 1985,\n",
    "    ending_time = 1999,\n",
    "    time_step = 1,\n",
    "    suitability_map = None,\n",
    "    driver_maps = all_driver_maps_xr,\n",
    "    validation_map = validation_map,\n",
    "    pixel_quantities=None,\n",
    "    mask_image=None,\n",
    "    strata_map=None,\n",
    "    constrain_to_neighborhood=(3,3),\n",
    "    driver_map_weights=weights)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "1b5a23b2-0a4d-4809-84f7-9dfb21aa6b5a",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "target number of change pixels: 3830\n",
      "Final neighborhood constraint: [3, 3]\n",
      "The number of pixels that can be changed: 15269\n",
      "Number of pixels above the change threshold: 3804\n",
      "There are 171 tied pixels, of which 26 must be selected arbitrarily.\n",
      "If this is unacceptable, consider reducing tied pixels by using more categorical variables to generate your suitability map\n"
     ]
    }
   ],
   "source": [
    "# run predict() on the model to get the predicted map\n",
    "landchange_model02.predict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "aa280910-f47d-41de-96b9-b7712ddf1129",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.QuadMesh at 0x7db633f0bc90>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# access the predicted map as part of the model class\n",
    "landchange_model02.predicted_map.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "e4b58a3b-afa5-498b-b3dc-811e0ebcccfc",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'True Positives': 27145,\n",
       " 'True Negatives': 68400,\n",
       " 'False Positives': 4158,\n",
       " 'False Negatives': 3487,\n",
       " 'Precision': 0.8671692808995943,\n",
       " 'Recall': 0.8861647949856359,\n",
       " 'F1 Score': 0.876564139824009}"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# call validate to get the CROSSTAB table\n",
    "landchange_model02.validate()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e5dbc2ee-73d7-4e2f-90a0-de793e7b18c7",
   "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}