def write_shapefile(self, cells, out_shp, data=False):
driver = ogr.GetDriverByName('Esri Shapefile')
ds = driver.CreateDataSource(out_shp)
layer = ds.CreateLayer('', None, ogr.wkbPolygon)
layer.CreateField(ogr.FieldDefn('id', ogr.OFTString))
if data:
layer.CreateField(ogr.FieldDefn('data', ogr.OFSTJSON))
defn = layer.GetLayerDefn()
for cell in cells:
feat = ogr.Feature(defn)
feat.SetField('id', cell['id'])
geom = ogr.CreateGeometryFromWkt(cell['cell_poly'])
feat.SetGeometry(geom)
if data:
if cell['id'] in cell['data']:
feat.SetField('data', json.dumps(cell['data'][cell['id']]))
else:
feat.SetField('data', json.dumps(cell['data']))
layer.CreateFeature(feat)
feat = geom = None # destroy these
ds = layer = feat = geom = None
def save_polygons(poly, output_folder, fname, meta=None):
driver = ogr.GetDriverByName('Esri Shapefile')
ds = driver.CreateDataSource(output_folder + '{}.shp'.format(fname))
srs = osr.SpatialReference()
srs.ImportFromEPSG(4326)
layer = ds.CreateLayer('', srs, ogr.wkbMultiPolygon)
# Add one attribute
layer.CreateField(ogr.FieldDefn('id', ogr.OFTInteger))
defn = layer.GetLayerDefn()
## If there are multiple geometries, put the "for" loop here
# Create a new feature (attribute and geometry)
feat = ogr.Feature(defn)
feat.SetField('id', 123)
# Make a geometry, from Shapely object
geom = ogr.CreateGeometryFromWkb(poly.wkb)
feat.SetGeometry(geom)
layer.CreateFeature(feat)
feat = geom = None # destroy these
# Save and close everything
ds = layer = feat = geom = None
def create_index_fld(input_shp,
class_names,
output_name='training_indexed.shp'):
"""
Creates an extra field in shp with unique index value for each polygon
input_shp - input shapefile
class_names - list of unique class names to classify
output name - output shapefile name
"""
data_shp = input_shp
driver_shp = ogr.GetDriverByName('ESRI Shapefile')
vector = driver_shp.Open(data_shp, 1)
lyr = vector.GetLayer()
directory_out = os.getcwd()
# if file with given name exists delete
if output_name + '.shp' in os.listdir(directory_out):
driver_shp.DeleteDataSource(output_name + '.shp')
print('created file', output_name)
out_ds = driver_shp.CreateDataSource(directory_out)
lyr_copy = out_ds.CopyLayer(lyr, output_name)
fieldDefn = ogr.FieldDefn('indeks', ogr.OFTInteger)
fieldDefn.SetWidth(1)
lyr_copy.CreateField(fieldDefn)
for nb, f in enumerate(lyr_copy):
f.SetField('indeks', nb)
lyr_copy.SetFeature(f)
fieldDefn = ogr.FieldDefn('kod', ogr.OFTInteger)
fieldDefn.SetWidth(10)
lyr_copy.CreateField(fieldDefn)
code = 1
for a in class_names:
print(class_names[code - 1])
lyr_copy.SetAttributeFilter("klasa = '{0}'".format(class_names[code -
1]))
for f in lyr_copy:
f.SetField('kod', code)
lyr_copy.SetFeature(f)
code += 1
print('created')
return output_name + '.shp'
def vectorize_data():
setup_env()
temp_path = os.getenv("TEMP_PATH")
filelist = create_filelist()
print("Starting vectorization...")
for file in tqdm(filelist, unit=" file"):
file_split = file.split("/")
date_time_obj = datetime.strptime(
file_split[len(file_split)-1], 'RW_%Y%m%d-%H%M.asc')
filename_input = temp_path + "/cropped/{}".format(
date_time_obj.strftime("%Y%m%d-%H%M"))
filename_output = temp_path + "/vectorized/{}".format(
date_time_obj.strftime("%Y%m%d-%H%M"))
source = gdal.Open(filename_input + ".tif")
band = source.GetRasterBand(1)
_ = band.ReadAsArray()
driver = ogr.GetDriverByName("ESRI Shapefile")
if os.path.exists(filename_output + ".shp"):
driver.DeleteDataSource(filename_output + ".shp")
target = driver.CreateDataSource(filename_output + ".shp")
srs = osr.SpatialReference()
srs.ImportFromProj4(
"+proj=stere +lon_0=10.0 +lat_0=90.0 +lat_ts=60.0 +a=6370040 +b=6370040 +units=m")
targetLayer = target.CreateLayer("radolan", srs=srs)
targetField = ogr.FieldDefn("rain", ogr.OFTInteger)
targetLayer.CreateField(targetField)
gdal.Polygonize(band, None, targetLayer, 0, [], callback=None)
target.Destroy()
source = None
_ = None
print("Vectorization complete.")
# set up the shapefile driver
driver = ogr.GetDriverByName("ESRI Shapefile")
# create the data source
data_source = driver.CreateDataSource(csvname + ".shp")
# create the spatial reference, WGS84
srs = osr.SpatialReference()
srs.ImportFromEPSG(4326)
# create the layer
layer = data_source.CreateLayer(csvname + "_Aggregations", srs, ogr.wkbPoint)
# Add the fields we're interested in
field_name = ogr.FieldDefn("Name", ogr.OFTString)
field_name.SetWidth(24)
layer.CreateField(field_name)
layer.CreateField(ogr.FieldDefn("Latitude", ogr.OFTReal))
layer.CreateField(ogr.FieldDefn("Longitude", ogr.OFTReal))
with open(csvfile, 'r') as csvf:
reader = csv.DictReader(csvf)
# Process the text file and add the attributes and features to the shapefile
for row in reader:
# create the feature
feature = ogr.Feature(layer.GetLayerDefn())
# Set the attributes using the values from the delimited text file
feature.SetField("Name", row['geohash'])
feature.SetField("Latitude", row['c_y'])
feature.SetField("Longitude", row['c_x'])
# https://github.com/GeospatialPython/Learn/raw/master/islands.zip
import gdal
from gdal import ogr, osr
# Thresholded input raster name
src = "islands_classified.tiff"
# Output shapefile name
tgt = "extract.shp"
# OGR layer name
tgtLayer = "extract"
# Open the input raster
srcDS = gdal.Open(src)
# Grab the first band
band = srcDS.GetRasterBand(1)
# Force gdal to use the band as a mask
mask = band
# Set up the output shapefile
driver = ogr.GetDriverByName("ESRI Shapefile")
shp = driver.CreateDataSource(tgt)
# Copy the spatial reference
srs = osr.SpatialReference()
srs.ImportFromWkt(srcDS.GetProjectionRef())
layer = shp.CreateLayer(tgtLayer, srs=srs)
# Set up the dbf file
fd = ogr.FieldDefn("DN", ogr.OFTInteger)
layer.CreateField(fd)
dst_field = 0
# Automatically extract features from an image!
extract = gdal.Polygonize(band, mask, layer, dst_field, [], None)
def _tiger_to_tract(self, infile):
""" Converts collection of Census Tiger files into a geopandas.GeoDataFrame of census tracts
Modified from original at
https://svn.osgeo.org/gdal/tags/1.4.3/gdal/pymod/samples/tigerpoly.py
"""
class Module(object):
def __init__(mod):
mod.lines = {}
mod.poly_line_links = {}
outfile = 'tracts.shp'
# Open the datasource to operate on.
ds = ogr.Open(infile, update=0)
poly_layer = ds.GetLayerByName('Polygon')
# Create output file for the composed polygons.
nad83 = osr.SpatialReference()
nad83.SetFromUserInput('NAD83')
shp_driver = ogr.GetDriverByName('ESRI Shapefile')
shp_driver.DeleteDataSource(outfile)
shp_ds = shp_driver.CreateDataSource(outfile)
shp_layer = shp_ds.CreateLayer(
'out', geom_type=ogr.wkbPolygon, srs=nad83)
src_defn = poly_layer.GetLayerDefn()
poly_field_count = src_defn.GetFieldCount()
for fld_index in range(poly_field_count):
src_fd = src_defn.GetFieldDefn(fld_index)
fd = ogr.FieldDefn(src_fd.GetName(), src_fd.GetType())
fd.SetWidth(src_fd.GetWidth())
fd.SetPrecision(src_fd.GetPrecision())
shp_layer.CreateField(fd)
# Read all features in the line layer, holding just the geometry in a hash
# for fast lookup by TLID.
line_layer = ds.GetLayerByName('CompleteChain')
line_count = 0
modules_hash = {}
feat = line_layer.GetNextFeature()
geom_id_field = feat.GetFieldIndex('TLID')
tile_ref_field = feat.GetFieldIndex('MODULE')
while feat is not None:
geom_id = feat.GetField(geom_id_field)
tile_ref = feat.GetField(tile_ref_field)
try:
module = modules_hash[tile_ref]
except:
module = Module()
modules_hash[tile_ref] = module
module.lines[geom_id] = feat.GetGeometryRef().Clone()
line_count = line_count + 1
feat.Destroy()
feat = line_layer.GetNextFeature()
# Read all polygon/chain links and build a hash keyed by POLY_ID listing
# the chains (by TLID) attached to it.
link_layer = ds.GetLayerByName('PolyChainLink')
feat = link_layer.GetNextFeature()
geom_id_field = feat.GetFieldIndex('TLID')
tile_ref_field = feat.GetFieldIndex('MODULE')
lpoly_field = feat.GetFieldIndex('POLYIDL')
rpoly_field = feat.GetFieldIndex('POLYIDR')
link_count = 0
while feat is not None:
module = modules_hash[feat.GetField(tile_ref_field)]
tlid = feat.GetField(geom_id_field)
lpoly_id = feat.GetField(lpoly_field)
rpoly_id = feat.GetField(rpoly_field)
if lpoly_id == rpoly_id:
feat.Destroy()
feat = link_layer.GetNextFeature()
continue
try:
module.poly_line_links[lpoly_id].append(tlid)
except:
module.poly_line_links[lpoly_id] = [tlid]
try:
module.poly_line_links[rpoly_id].append(tlid)
except:
module.poly_line_links[rpoly_id] = [tlid]
link_count = link_count + 1
feat.Destroy()
feat = link_layer.GetNextFeature()
# Process all polygon features.
feat = poly_layer.GetNextFeature()
tile_ref_field = feat.GetFieldIndex('MODULE')
polyid_field = feat.GetFieldIndex('POLYID')
degenerate_count = 0
while feat is not None:
module = modules_hash[feat.GetField(tile_ref_field)]
polyid = feat.GetField(polyid_field)
tlid_list = module.poly_line_links[polyid]
link_coll = ogr.Geometry(type=ogr.wkbGeometryCollection)
for tlid in tlid_list:
geom = module.lines[tlid]
link_coll.AddGeometry(geom)
try:
poly = ogr.BuildPolygonFromEdges(link_coll)
if poly.GetGeometryRef(0).GetPointCount() < 4:
degenerate_count = degenerate_count + 1
poly.Destroy()
feat.Destroy()
feat = poly_layer.GetNextFeature()
continue
feat2 = ogr.Feature(feature_def=shp_layer.GetLayerDefn())
for fld_index in range(poly_field_count):
feat2.SetField(fld_index, feat.GetField(fld_index))
feat2.SetGeometryDirectly(poly)
shp_layer.CreateFeature(feat2)
feat2.Destroy()
except:
warn('BuildPolygonFromEdges failed.')
feat.Destroy()
feat = poly_layer.GetNextFeature()
if degenerate_count:
warn('Discarded %d degenerate polygons.' % degenerate_count)
# Cleanup
shp_ds.Destroy()
shp_ds = None
ds.Destroy()
ds = None
# build a fully-qualified fips code and dissolve on it to create tract geographies
gdf = gpd.read_file(outfile)
if "CTBNA90" in gdf.columns:
gdf = gdf.rename(columns={"CTBNA90": 'TRACT', "BLK90": "BLOCK"})
gdf['STATE'] = gdf['STATE'].astype(str).str.rjust(2, "0")
gdf['COUNTY'] = gdf['COUNTY'].astype(str).str.rjust(3, "0")
gdf['TRACT'] = gdf['TRACT'].astype(str).str.rjust(6, "0")
gdf['BLOCK'] = gdf['BLOCK'].astype(str).str.rjust(4, "0")
gdf['fips'] = gdf.STATE + gdf.COUNTY + gdf.TRACT
if self.geom == 'block':
gdf['fips'] += gdf.BLOCK
gdf = gdf.dropna(subset=['fips'])
gdf.geometry = gdf.buffer(0)
gdf = gdf.dissolve(by='fips')
gdf.reset_index(inplace=True)
shp_driver.DeleteDataSource(outfile)
return gdf
band_dict[i + 1] = file_only
i += 1
class_file = os.path.join(training_folder, sub, r'classification.tif')
f.write(class_file)
file_only = os.path.basename(os.path.normpath(class_file))
band_dict[i + 1] = file_only
# take the vector shapefile and convert the alphabetic codes to numeric codes
print('Mapping codes')
vector_classes = os.path.join(train_folder, r'classification.shp')
new_field_name = 'Classvalue'
driver = ogr.GetDriverByName("ESRI Shapefile")
dataSource = driver.Open(vector_classes, 1)
layer = dataSource.GetLayer()
new_field = ogr.FieldDefn(new_field_name, ogr.OFTInteger)
field_names = [field.name for field in layer.schema]
if new_field_name in field_names:
layer.DeleteField(field_names.index(new_field_name))
layer.CreateField(new_field)
for feature in layer:
alpha_code = feature.GetField('Classname')
num_code = code_dict[alpha_code]
feature.SetField(new_field_name, num_code)
layer.SetFeature(feature)
dataSource = None
