def test_rasterization_raise_error_for_too_big_image():
shape = Polygon([(0, 0), (1, 0), (1, -1), (0, -1)])
fcol = FeatureCollection([GeoFeature(GeoVector(shape), {})])
with pytest.raises(ScaleError) as excinfo:
fcol.rasterize(1e-50)
assert "Scale is too fine, increase it for a smaller image" in excinfo.exconly()
def test_rasterization_function_raises_error_if_no_dtype_is_given():
sq1 = GeoFeature(
GeoVector.from_bounds(xmin=0, ymin=2, xmax=1, ymax=3, crs=WGS84_CRS),
{'value': 1.0})
sq2 = GeoFeature(
GeoVector.from_bounds(xmin=1, ymin=0, xmax=3, ymax=2, crs=WGS84_CRS),
{'value': 2.0})
fc = FeatureCollection([sq1, sq2])
with pytest.raises(ValueError) as excinfo:
fc.rasterize(1.0, fill_value=lambda x: 1, crs=WGS84_CRS)
assert "dtype must be specified for multivalue rasterization" in excinfo.exconly(
)
def test_rasterization_function():
sq1 = GeoFeature(
GeoVector.from_bounds(xmin=0, ymin=2, xmax=1, ymax=3, crs=WGS84_CRS),
{'value': 1.0}
)
sq2 = GeoFeature(
GeoVector.from_bounds(xmin=1, ymin=0, xmax=3, ymax=2, crs=WGS84_CRS),
{'value': 2.0}
)
fc = FeatureCollection([sq1, sq2])
def func(feat):
return feat['value']
expected_image = np.ma.masked_array(
[[
[1.0, 0.0, 0.0],
[0.0, 2.0, 2.0],
[0.0, 2.0, 2.0]
]],
[
[False, True, True],
[True, False, False],
[True, False, False],
]
)
result = fc.rasterize(1.0, fill_value=func, crs=WGS84_CRS, dtype=np.float32)
assert_array_equal(result.image.mask, expected_image.mask)
assert_array_equal(result.image.data, expected_image.data)
def test_rasterization_of_line_has_correct_pixel_width(resolution):
xmax, ymax = 11, 5
pixels_width = 1
line = GeoFeature.from_shape(
LineString([(xmax / 2, 0), (xmax / 2, ymax * 4 / 5)]))
roi = GeoVector.from_bounds(xmin=0,
ymin=0,
xmax=xmax,
ymax=ymax,
crs=DEFAULT_CRS)
fc = FeatureCollection([line])
expected_image = np.zeros(
(int(ymax // resolution), int(xmax // resolution)), dtype=np.uint8)
expected_image[int(1 // resolution):, expected_image.shape[1] // 2] = 1
expected_affine = Affine(resolution, 0.0, 0.0, 0.0, -resolution, 5.0)
expected_crs = DEFAULT_CRS
expected_result = GeoRaster2(expected_image,
expected_affine,
expected_crs,
nodata=0)
result = fc.rasterize(resolution,
polygonize_width=pixels_width,
crs=DEFAULT_CRS,
bounds=roi)
assert result == expected_result
def test_rasterization_of_line_simple():
resolution = 1
pixels_width = 1
line = GeoFeature.from_shape(LineString([(2.5, 0), (2.5, 3)]))
roi = GeoVector.from_bounds(xmin=0,
ymin=0,
xmax=5,
ymax=5,
crs=DEFAULT_CRS)
fc = FeatureCollection([line])
expected_image = np.zeros((5, 5), dtype=np.uint8)
expected_image[2:, 2] = 1
expected_affine = Affine(1.0, 0.0, 0.0, 0.0, -1.0, 5.0)
expected_crs = DEFAULT_CRS
expected_result = GeoRaster2(expected_image,
expected_affine,
expected_crs,
nodata=0)
result = fc.rasterize(resolution,
polygonize_width=pixels_width,
crs=DEFAULT_CRS,
bounds=roi)
assert result == expected_result
def test_rasterization_has_expected_affine_and_crs():
shape = Polygon([(0, 0), (1, 0), (1, -1), (0, -1)])
crs = CRS({'init': 'epsg:32631'})
fcol = FeatureCollection([GeoFeature(GeoVector(shape, crs), {})])
expected_affine = ~Affine(1.0, 0.0, 0.0, 0.0, -1.0, 0.0)
georaster = fcol.rasterize(1, crs=crs)
assert georaster.crs == crs
assert georaster.affine.a == pytest.approx(expected_affine.a)
assert georaster.affine.b == pytest.approx(expected_affine.b, abs=1e-10)
assert georaster.affine.c == pytest.approx(expected_affine.c, abs=1e-10)
assert georaster.affine.d == pytest.approx(expected_affine.d, abs=1e-10)
assert georaster.affine.e == pytest.approx(expected_affine.e)
assert georaster.affine.f == pytest.approx(expected_affine.f, abs=1e-10)
def test_rasterization_function_with_empty_collection():
fc = FeatureCollection([])
def func(feat):
return feat['value']
expected_image = np.ma.masked_array(
[[[0.0, 0.0, 0.0], [0.0, 0.0, 0.0], [0.0, 0.0, 0.0]]], [
[True, True, True],
[True, True, True],
[True, True, True],
])
bounds = GeoVector.from_bounds(0.0, 0.0, 3.0, 3.0)
result = fc.rasterize(1.0,
bounds=bounds,
fill_value=func,
crs=WGS84_CRS,
dtype=np.float32)
assert_array_equal(result.image.mask, expected_image.mask)
assert_array_equal(result.image.data, expected_image.data)
