def run(self): lulc_nodata = raster_utils.get_nodata_from_uri(GLOBAL_LANDCOVER_URI) forest_lulc_codes = [1, 2, 3, 4, 5] mask_uri = os.path.join(OUTPUT_DIR, "forest_mask.tif") mask_nodata = 2 def mask_nonforest(lulc): """Takes in a numpy array of landcover values and returns 1s where they match forest codes and 0 otherwise""" mask = numpy.empty(lulc.shape, dtype=numpy.int8) mask[:] = 1 for lulc_code in forest_lulc_codes: mask[lulc == lulc_code] = 0 mask[lulc == lulc_nodata] = mask_nodata return mask cell_size = raster_utils.get_cell_size_from_uri(GLOBAL_LANDCOVER_URI) raster_utils.vectorize_datasets( [GLOBAL_LANDCOVER_URI,], mask_nonforest, mask_uri, gdal.GDT_Byte, mask_nodata, cell_size, 'intersection', dataset_to_align_index=0, dataset_to_bound_index=None, aoi_uri=None, assert_datasets_projected=True, process_pool=None, vectorize_op=False, datasets_are_pre_aligned=True) raster_utils.distance_transform_edt( mask_uri, FOREST_EDGE_DISTANCE_URI)
def run(self): raster_utils.vectorize_datasets( list(self.dataset_uri_list), self.dataset_pixel_op, self.dataset_out_uri, self.datatype_out, self.nodata_out, self.pixel_size_out, self.bounding_box_mode, dataset_to_align_index=self.dataset_to_align_index, vectorize_op=self.vectorize_op)
def run(self): nodata = raster_utils.get_nodata_from_uri(UNION_LANDCOVER_URI) cell_size = raster_utils.get_cell_size_from_uri(UNION_LANDCOVER_URI) raster_utils.vectorize_datasets( [UNION_LANDCOVER_URI, UNION_BIOMASS_URI], lambda x, y: x, GLOBAL_LANDCOVER_URI, gdal.GDT_Int16, nodata, cell_size, "intersection", dataset_to_align_index=0, vectorize_op=False)
def _align_raster_with_biomass(input_uri, output_uri): """Function to use internally to take an input and align it with the GLOBAL_BIOMASS_URI raster""" nodata = raster_utils.get_nodata_from_uri(input_uri) if nodata is None: nodata = -9999 cell_size = raster_utils.get_cell_size_from_uri(GLOBAL_BIOMASS_URI) raster_utils.vectorize_datasets( [input_uri, GLOBAL_BIOMASS_URI], lambda x, y: x, output_uri, gdal.GDT_Float32, nodata, cell_size, "dataset", dataset_to_bound_index=1, vectorize_op=False)
def run(self): def union_op(*array_list): """Given an array stack return an array that has a value defined in the stack that is not nodata. used for overlapping nodata stacks.""" output_array = array_list[0] for array in array_list[1:]: output_array = numpy.where( array != nodata, array, output_array) return output_array nodata = raster_utils.get_nodata_from_uri(self.dataset_uri_list[0]) cell_size = raster_utils.get_cell_size_from_uri( self.dataset_uri_list[0]) raster_utils.vectorize_datasets( list(self.dataset_uri_list), union_op, self.dataset_out_uri, gdal.GDT_Int16, nodata, cell_size, "union", dataset_to_align_index=0, vectorize_op=False)
def run(self): biomass_raster_list = [ "C:/Users/rpsharp/Dropbox_stanford/Dropbox/forest_edge_carbon/af_biov2ct1.tif", "C:/Users/rpsharp/Dropbox_stanford/Dropbox/forest_edge_carbon/am_biov2ct1.tif", "C:/Users/rpsharp/Dropbox_stanford/Dropbox/forest_edge_carbon/as_biov2ct1.tif", ] nodata = raster_utils.get_nodata_from_uri(biomass_raster_list[0]) cell_size = raster_utils.get_cell_size_from_uri(biomass_raster_list[0]) def union_op(*biomass_array_list): output_array = biomass_array_list[0] for biomass_array in biomass_array_list[1:]: output_array = numpy.where( biomass_array != nodata, biomass_array, output_array) return output_array raster_utils.create_directories([os.path.dirname(self.output_uri)]) raster_utils.vectorize_datasets( biomass_raster_list, union_op, self.output_uri, gdal.GDT_Int16, nodata, cell_size, 'union', dataset_to_align_index=0, vectorize_op=False)
def average_layers(): base_table_uri = "C:/Users/rich/Desktop/all_grid_results_100km_clean_v2.csv" base_table_file = open(base_table_uri, 'rU') table_header = base_table_file.readline() #need to mask the average layers to the biomass regions giant_layer_uri = "C:/Users/rich/Desktop/average_layers_projected/giant_layer.tif" af_uri = "C:/Users/rich/Desktop/af_biov2ct1.tif" am_uri = "C:/Users/rich/Desktop/am_biov2ct1.tif" as_uri = "C:/Users/rich/Desktop/as_biov2ct1.tif" cell_size = raster_utils.get_cell_size_from_uri(am_uri) #raster_utils.vectorize_datasets( # [af_uri, am_uri, as_uri], lambda x,y,z: x+y+z, giant_layer_uri, gdal.GDT_Float32, # -1, cell_size, 'union', vectorize_op=False) table_uri = base_table_uri table_file = open(table_uri, 'rU') table_header = table_file.readline().rstrip() lookup_table = raster_utils.get_lookup_from_csv(table_uri, 'ID100km') out_table_uri = "C:/Users/rich/Desktop/all_grid_results_100km_human_elevation.csv" out_table_file = codecs.open(out_table_uri, 'w', 'utf-8') average_raster_list = [ ("C:/Users/rich/Desktop/average_layers_projected/lighted_area_luminosity.tif", 'Lighted area density'), ("C:/Users/rich/Desktop/average_layers_projected/fi_average.tif", 'Fire densities'), ("C:/Users/rich/Desktop/average_layers_projected/glbctd1t0503m.tif", 'FAO_Cattle'), ("C:/Users/rich/Desktop/average_layers_projected/glbgtd1t0503m.tif", 'FAO_Goat'), ("C:/Users/rich/Desktop/average_layers_projected/glbpgd1t0503m.tif", 'FAO_Pig'), ("C:/Users/rich/Desktop/average_layers_projected/glbshd1t0503m.tif", 'FAO_Sheep'), ("C:/Users/rich/Desktop/average_layers_projected/glds00ag.tif", 'Human population density AG'), ("C:/Users/rich/Desktop/average_layers_projected/glds00g.tif", 'Human population density G'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_11.tif', '"11: Urban, Dense settlement"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_12.tif', '"12: Dense settlements, Dense settlements"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_22.tif', '"22: Irrigated villages, Villages"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_23.tif', '"23: Cropped & pastoral villages, Villages"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_24.tif', '"24: Pastoral villages, Villages"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_25.tif', '"25: Rainfed villages, Villages"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_26.tif', '"26: Rainfed mosaic villages, Villages"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_31.tif', '"31: Residential irrigated cropland, Croplands"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_32.tif', '"32: Residential rainfed mosaic, Croplands"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_33.tif', '"33: Populated irrigated cropland, Croplands"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_34.tif', '"34: Populated rainfed cropland, Croplands"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_35.tif', '"35: Remote croplands, Croplands"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_41.tif', '"41: Residential rangelands, Rangelands"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_42.tif', '"42: Populated rangelands, Rangelands"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_43.tif', '"43: Remote rangelands, Rangelands"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_51.tif', '"51: Populated forests, Forested"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_52.tif', '"52: Remote forests, Forested"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_61.tif', '"61: Wild forests, Wildlands"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_62.tif', '"62: Sparse trees, Wildlands"'), ('C:/Users/rich/Desktop/average_layers_projected/anthrome_63.tif', '"63: Barren, Wildlands"'), ("C:/Users/rich/Desktop/average_layers_projected/5km_global_pantropic_dem.tif", '"Average Elevation"'), ] clipped_raster_list = [] for average_raster_uri, header in average_raster_list: print 'clipping ' + average_raster_uri clipped_raster_uri = os.path.join(os.path.dirname(average_raster_uri), 'temp', os.path.basename(average_raster_uri)) cell_size = raster_utils.get_cell_size_from_uri(average_raster_uri) raster_utils.vectorize_datasets( [average_raster_uri, giant_layer_uri], lambda x,y: x, clipped_raster_uri, gdal.GDT_Float32, -1, cell_size, 'intersection', vectorize_op=False) clipped_raster_list.append((clipped_raster_uri, header)) dataset_list = [gdal.Open(uri) for uri, label in clipped_raster_list] band_list = [ds.GetRasterBand(1) for ds in dataset_list] nodata_list = [band.GetNoDataValue() for band in band_list] extended_table_headers = ','.join([header for _, header in average_raster_list]) def write_to_file(value): try: out_table_file.write(value) except UnicodeDecodeError as e: out_table_file.write(value.decode('latin-1')) write_to_file(table_header + ',' + extended_table_headers + '\n') #print table_header + ',' + extended_table_headers for line in table_file: split_line = line.rstrip().split(',') grid_id = split_line[2] #for grid_id in lookup_table: try: split_grid_id = grid_id.split('-') grid_row_index, grid_col_index = map(int, split_grid_id) except ValueError as e: month_to_number = { 'Jan': 1, 'Feb': 2, 'Mar': 3, 'Apr': 4, 'May': 5, 'Jun': 6, 'Jul': 7, 'Aug': 8, 'Sep': 9, 'Oct': 10, 'Nov': 11, 'Dec': 12, } grid_row_index, grid_col_index = month_to_number[split_grid_id[0]], int(split_grid_id[1]) print 'processing grid id ' + grid_id ds = dataset_list[0] base_srs = osr.SpatialReference(ds.GetProjection()) lat_lng_srs = base_srs.CloneGeogCS() coord_transform = osr.CoordinateTransformation( base_srs, lat_lng_srs) gt = ds.GetGeoTransform() grid_resolution = 100 #100km row_coord = grid_row_index * grid_resolution * 1000 + GLOBAL_UPPER_LEFT_ROW col_coord = grid_col_index * grid_resolution * 1000 + GLOBAL_UPPER_LEFT_COL lng_coord, lat_coord, _ = coord_transform.TransformPoint( col_coord, row_coord) write_to_file(','.join(split_line[0:2]) + ',%d-%d,' % (grid_row_index, grid_col_index) + ','.join(split_line[3:11]) +',%f,%f,' % (lat_coord, lng_coord)+','.join(split_line[13:])) for (_, header), band, ds, nodata in zip(clipped_raster_list, band_list, dataset_list, nodata_list): gt = ds.GetGeoTransform() n_rows = ds.RasterYSize n_cols = ds.RasterXSize xoff = int(grid_col_index * (grid_resolution * 1000.0) / (gt[1])) yoff = int(grid_row_index * (grid_resolution * 1000.0) / (-gt[5])) win_xsize = int((grid_resolution * 1000.0) / (gt[1])) win_ysize = int((grid_resolution * 1000.0) / (gt[1])) if xoff + win_xsize > n_cols: win_xsize = n_cols - xoff if yoff + win_ysize > n_rows: win_ysize = n_rows - yoff block = band.ReadAsArray( xoff=xoff, yoff=yoff, win_xsize=win_xsize, win_ysize=win_ysize) block_average = numpy.average(block[block != nodata]) write_to_file(',%f' % block_average) write_to_file('\n')
import os import numpy import gdal from invest_natcap import raster_utils anthrome_uri = "C:/Users/rich/Desktop/average_layers_projected/gl_anthrome.tif" anthrome_ds = gdal.Open(anthrome_uri) anthrome_band = anthrome_ds.GetRasterBand(1) anthrome_array = anthrome_band.ReadAsArray() unique_anthrome_ids = numpy.unique(anthrome_array) for anthrome_id in unique_anthrome_ids: output_uri = os.path.join(os.path.dirname(anthrome_uri), 'anthrome_%d.tif' % anthrome_id) out_pixel_size = raster_utils.get_cell_size_from_uri(anthrome_uri) raster_utils.vectorize_datasets( [anthrome_uri], lambda x: x==anthrome_id, output_uri, gdal.GDT_Byte, 127, out_pixel_size, "union", vectorize_op=False) print numpy.unique(anthrome_array).shape
def process_ecoregion(prefix): ecoregion_shapefile_uri = os.path.join( DATA_DIR, 'ecoregions', 'ecoregions_projected.shp') ecoregion_lookup = raster_utils.extract_datasource_table_by_key( ecoregion_shapefile_uri, 'ECO_ID_U') ecoregion_nodata = -1 ecoregion_lookup[ecoregion_nodata] = { 'ECO_NAME': 'UNKNOWN', 'ECODE_NAME': 'UNKNOWN', 'WWF_MHTNAM': 'UNKNOWN', } lulc_raw_uri = os.path.join(DATA_DIR, '%s%s' % (prefix, LULC_BASE)) biomass_raw_uri = os.path.join(DATA_DIR, '%s%s' % (prefix, BIOMASS_BASE)) cell_size = raster_utils.get_cell_size_from_uri(lulc_raw_uri) lulc_uri = os.path.join(OUTPUT_DIR, "%s_lulc_aligned.tif" % (prefix)) biomass_uri = os.path.join(OUTPUT_DIR, "%s_biomass_aligned.tif" % (prefix)) raster_utils.align_dataset_list( [lulc_raw_uri, biomass_raw_uri], [lulc_uri, biomass_uri], ['nearest']*2, cell_size, 'intersection', 0, dataset_to_bound_index=None, aoi_uri=None, assert_datasets_projected=True, process_pool=None) #create ecoregion id ecoregion_dataset_uri = os.path.join( OUTPUT_DIR, "%s_ecoregion_id.tif" % (prefix)) raster_utils.new_raster_from_base_uri( lulc_uri, ecoregion_dataset_uri, 'GTiff', ecoregion_nodata, gdal.GDT_Int16) raster_utils.rasterize_layer_uri( ecoregion_dataset_uri, ecoregion_shapefile_uri, option_list=["ATTRIBUTE=ECO_ID_U"]) lulc_nodata = raster_utils.get_nodata_from_uri(lulc_uri) forest_lulc_codes = [1, 2, 3, 4, 5] mask_uri = os.path.join(OUTPUT_DIR, "%s_mask.tif" % prefix) mask_nodata = 2 def mask_nonforest(lulc): mask = numpy.empty(lulc.shape, dtype=numpy.int8) mask[:] = 1 for lulc_code in forest_lulc_codes: mask[lulc == lulc_code] = 0 mask[lulc == lulc_nodata] = mask_nodata return mask raster_utils.vectorize_datasets( [lulc_uri,], mask_nonforest, mask_uri, gdal.GDT_Byte, mask_nodata, cell_size, 'intersection', dataset_to_align_index=0, dataset_to_bound_index=None, aoi_uri=None, assert_datasets_projected=True, process_pool=None, vectorize_op=False, datasets_are_pre_aligned=True) forest_edge_distance_uri = os.path.join(OUTPUT_DIR, "%s_forest_edge.tif" % prefix) raster_utils.distance_transform_edt(mask_uri, forest_edge_distance_uri) biomass_stats_uri = os.path.join(OUTPUT_DIR, "%s_biomass_stats.csv" % prefix) _aggregate_results(forest_edge_distance_uri, biomass_uri, ecoregion_dataset_uri, ecoregion_lookup, biomass_stats_uri)