CODEWITHSUNDEEP
pythonanacondagisgeopandasrasterio
Ayesha
Ayesha

Reputation: 11

Error while clip raster using geojson file

I am trying to clip the raster file with geojson file. both files have same coordinates system and overlap. here is my code and i am getting the error ( TypeError: can't multiply sequence by non-int of type 'Affine')

shape_geojson=path of jeojson
input_file= path of geotiff filesenter code here



with open(shape_geojson) as data_file:    
    geoms= json.load(data_file)

geoms = [{'type': 'Polygon', 'coordinates': [[[372187.1496072686, 5068258.904150674, 0.0], 
[373626.0661280414, 5068286.477440579, 0.0], [373548.73687596567, 5069922.161341429, 0.0], 
[372152.12599016255, 5069847.257662993, 0.0], [372187.1496072686, 5068258.904150674, 0.0]]]}]


for file in os.listdir(input_file):
        with rio.open(os.path.join(input_file, file)) as src:
        out_image, out_transform = rio.mask.mask(src, geoms, crop=True)
        out_meta = src.meta.copy()
        out_meta.update({"driver": "GTiff",
        "height": out_image.shape[1],
        "width": out_image.shape[2],
        "transform": out_transform})

Error

TypeError: can't multiply sequence by non-int of type 'Affine'

Upvotes: 1

Views: 781

Answers (2)

Roger Lew
Roger Lew

Reputation: 328

Rob's answer got me close, but my coords were nested.

Ended up having to do this:

def is_mappable_of_floats(x):
    try:
        float(x[0])
        return True
    except:
        return False

def coords_3d_to_2d(coords):
    _coords = []
    for coord in coords:
        if is_mappable_of_floats(coord):
            _coords.append((coord[0], coord[1]))
        else:
            _coords.append(coords_3d_to_2d(coord))
    return _coords


geoms = [
    {
        "type": g["type"],
        "coordinates": coords_3d_to_2d(g["coordinates"]),
    }
    for g in geoms
]

Upvotes: 0

Rob Raymond
Rob Raymond

Reputation: 31156

Very unusually your polygons are defined in 3D coordinates. Change them to 2D and the error goes away. Provide the URL of your raster and a full MWE can be shown.

geoms2d = [
    {
        "type": g["type"],
        "coordinates": [[xyz[0:2] for xyz in p] for p in g["coordinates"]],
    }
    for g in geoms
]

out_image, out_transform = rasterio.mask.mask(src, geoms2d, crop=True)

Upvotes: 1

Related Questions