Homec4science
Diffusion solar_potential (aw_solar) Browse

solarPOT/libs/geo_ops.py
b2056863329baw_solar

Last Change
Raw File

geo_ops.py
View Options

import numpy as np
import pandas as pd
import geopandas as gpd
import time
def create_buffer( input_gdf,
buff_dist,
columns = None,
cap_style = 1,
join_style = 1,
tolerance = 0.001,
simplify = True):
# function which creates a buffer around a given input polyon
# Inputs:
# - input_gdf: geodataframe with polygons around which to create a buffer
# - buff_dist: width of buffer (for inwards buffering give negative distance)
# - columns: columns of input_gdf to be kept in output dataframe - if none is given, all columns are kept
# - cap_style, join_style: shape of buffer vertices (see https://shapely.readthedocs.io/en/latest/manual.html#object.buffer)
# - tolerance for simplifying polygons (default value 1mm, i.e. very small)
# Output: GeoDataFrame with buffered polygons and columns as given
print("Creating buffer with distance %f" %buff_dist)
if columns is None:
columns = input_gdf.columns
# extract relevant columns and make a copy (to not change original dataframe)
buffer_shp = input_gdf.loc[ :, columns]
# simplify buffer shapes (remove unnecessary vertices to avoid wrong buffer shapes)
if simplify: buffer_shp['geometry'] = buffer_shp.simplify(tolerance)
# create buffer around polygons
buffer_shp['geometry'] = buffer_shp.geometry.buffer(buff_dist, cap_style = cap_style, join_style = join_style)
# delete all rows with empty polygons
buffer_non_null = buffer_shp[~buffer_shp.is_empty]
return buffer_non_null
def get_polygon_corners(input_gdf, ID_field, attribute_df = None):
if attribute_df is None: attribute_df = input_gdf
# explode dataframe (divide multi-part polygons into individual polygons)
unique_gdf = input_gdf.explode().reset_index().drop(['level_0', 'level_1'], axis = 1)
# extract point coordinates for each polygon
points = []
for idx, geometry in enumerate(unique_gdf.geometry.exterior):
points.append(get_points(geometry, ID = unique_gdf[ID_field][idx], ID_name = ID_field))
# convert list to pandas dataframe
point_coords = pd.concat(points).reset_index().drop('index', axis = 1)
# add attribute information to dataframe
point_coords_with_info = pd.merge(point_coords, attribute_df, how = 'left', on = ID_field)
# remove duplicates and reset index
return point_coords_with_info.drop_duplicates().reset_index().drop('index', axis = 1)
def rotate_by_field(input_gdf, angle_field, origin = 'center', use_radians = 'False', backwards = False):
# create a copy of the input frame and get unique rotation angles
rotated_gdf = input_gdf.copy()
unique_rotation_angles = rotated_gdf[angle_field].unique()
# rotate each angle seperately
for angle in unique_rotation_angles:
if backwards: rotation_angle = - angle
else: rotation_angle = angle
# get rows that have the current angle and rotate
curr_idx = rotated_gdf[angle_field] == angle
rotated_gdf.loc[curr_idx, 'geometry'] = rotated_gdf[curr_idx].rotate(rotation_angle, origin = origin, use_radians = False)
return rotated_gdf
## ================================================== ##
############# AUXILARY FUNCTIONS #######################
def get_points(x, ID, ID_name, xcoord = 'x', ycoord = 'y'):
# function to create pandas dataframe with polygon corners
# Inputs: x - linear ring or similar geometry; ID - field identifier (and desired name of field)
# convert list of coordinate tuples into numpy array
coords = np.asarray(x)
# create dataframe and add ID field
new_df = pd.DataFrame(coords, columns = [ xcoord, ycoord ])
new_df[ ID_name ] = ID
return new_df
c4science · Help