Does anyone know how to get the affine transformation from a geopandas dataframe?
I need this for rasterising geospatial vector data with rasterio.
shapes = ((geom,value) for geom, value in zip(geodataframe.geometry, len(geodataframe)*[255]))
numpy_array = features.rasterize(
shapes,
out_shape=self.tile_size,
transform=geodataframe.transform???,
all_touched=True)
I am however still missing the transform part.
The solution I found to work for anybody else searching is to create the transform yourself.
import affine
transform = affine.Affine(res, 0, minx, 0, -res, maxy)
shapes = ((geom,value) for geom, value in zip(dataset.geometry, len(dataset)*[val]))
raster = rasterio.features.rasterize(shapes,
out_shape=tile_size,
transform=transform)
whereby res is the spatial resolution (ie 1pixel = 2m). minx and maxy correspond to the geographic position and the y axis is chosen negative to match the different coordinate system of the another raster I am using.