07-02_LakeAnalysis_Z056_local.py

from landsattrend.lake_analysis import LakeMaker
import os, platform
import shutil

PROCESS_ROOT = os.getcwd()

def set_conda_gdal_paths():
    if platform.system() == 'Windows':
        os.environ['GDAL_BIN'] = os.path.join(os.environ['CONDA_PREFIX'], 'Library', 'bin')
        os.environ['GDAL_PATH'] = os.path.join(os.environ['CONDA_PREFIX'], 'Scripts')
    else:
        os.environ['GDAL_BIN'] = os.path.join(os.environ['CONDA_PREFIX'], 'bin')
        os.environ['GDAL_PATH'] = os.environ['GDAL_BIN']

process_dir = os.path.join(PROCESS_ROOT, 'process')
site_name = '32604'
CLASS_PERIOD = '2000-2020'
CLASS_MODEL = os.path.join(PROCESS_ROOT, 'models', 'PDG_6idx2feat_elslope_model_py38_sklearn0232_v04.z')
LAKE_FILTER_MODEL = os.path.join(PROCESS_ROOT, 'models', '20180820_lakefilter_12039samples_py3.z')

DEM_LOCATION = os.path.join(PROCESS_ROOT, r'aux_data', 'dem', 'DEM.vrt')
FOREST_LOCATION = os.path.join(PROCESS_ROOT, r'aux_data', 'forestfire', 'forestfire.vrt')

def main():
    set_conda_gdal_paths()
    tiles_directory = os.path.join(PROCESS_ROOT, 'data', site_name, CLASS_PERIOD, 'tiles')
    tif_files = os.listdir(tiles_directory)

    if '.DS_Store' in tif_files:
        tif_files.remove('.DS_Store')
    print('Available Images:\n')
    for t in tif_files:
        print(t)

    l = LakeMaker(site_name, os.path.join(process_dir, site_name), tiles_directory, classperiod=CLASS_PERIOD)
    print("\nStart Classification")
    l.classify(CLASS_MODEL)

    print("\nPreparing additional Data")
    l.prepare_aux_data(DEM_LOCATION, FOREST_LOCATION)

    print("\nCreating Masks")
    l.make_masks()

    print("\nCalculating Stats")
    l.make_stats()

    print("\nSaving DataFrame to Disk")
    l.save_df()

    # errors come somewhere here
    print("\nFiltering non-lake objects")
    l.filter_data(LAKE_FILTER_MODEL)
    print("\nSaving DataFrame to Disk")
    l.save_filtered_data()
    print("\nTransforming data to metric values")
    l.finalize_calculations()
    print("\nSaving DataFrame to Disk")
    l.save_results()
    print("\nSaving ResultGrid at 3km resolution")
    l.export_gridded_results([100, 250])


if __name__ == "__main__":
    main()