Change map smooth widths to match Astropy

docs/image/survey_example.py

@@ -7,7 +7,7 @@
 filename = "$GAMMAPY_EXTRA/datasets/fermi_survey/all.fits.gz"
 survey_map = Map.read(filename, hdu="counts")
 survey_map.data = survey_map.data.astype("float")
-smoothed_map = survey_map.smooth(radius=Angle(0.2, unit="deg"))
+smoothed_map = survey_map.smooth(width=Angle(0.1, unit="deg"))
 
 fig = plt.figure(figsize=(15, 8))
 xlim = Angle([70, 262], unit="deg")

gammapy/data/event_list.py

@@ -659,7 +659,7 @@ def _counts_image(self):
         }
         m = WcsNDMap.create(**opts)
         m.fill_by_coord(self.radec)
-        m = m.smooth(radius=1)
+        m = m.smooth(width=1)
         return m
 
     def plot_image(self):

gammapy/maps/tests/test_wcsnd.py

@@ -489,8 +489,8 @@ def test_convolve_vs_smooth():
     m = WcsNDMap.create(binsz=binsz, width=1.05 * u.deg, axes=axes)
     m.data[:, :, 10, 10] = 1.
 
-    desired = m.smooth(kernel="gauss", radius=0.5 * u.deg, mode="constant")
-    gauss = Gaussian2DKernel(5).array
+    desired = m.smooth(kernel="gauss", width=0.5 * u.deg, mode="constant")
+    gauss = Gaussian2DKernel(10).array
     actual = m.convolve(kernel=gauss)
     assert_allclose(actual.data, desired.data, rtol=1e-3)
 

gammapy/maps/wcsnd.py

@@ -491,19 +491,20 @@ def _plot_format_allsky(self, ax):
         lat.grid(alpha=0.2, linestyle="solid", color="w")
         return ax
 
-    def smooth(self, radius, kernel="gauss", **kwargs):
+    def smooth(self, width, kernel="gauss", **kwargs):
         """
         Smooth the image (works on a 2D image and returns a copy).
 
-        The definition of the smoothing parameter radius is equivalent to the
-        one that is used in ds9 (see `ds9 smoothing <http://ds9.si.edu/doc/ref/how.html#Smoothing>`_).
 
         Parameters
         ----------
-        radius : `~astropy.units.Quantity` or float
+        width : `~astropy.units.Quantity` or float
             Smoothing width given as quantity or float. If a float is given it
             interpreted as smoothing width in pixels. If an (angular) quantity
             is given it converted to pixels using ``geom.wcs.wcs.cdelt``.
+            It corresponds to the standard deviation in case of a Gaussian kernel,
+            the radius in case of a disk kernel, and the side length in case
+            of a box kernel.
         kernel : {'gauss', 'disk', 'box'}
             Kernel shape
         kwargs : dict
@@ -518,22 +519,19 @@ def smooth(self, radius, kernel="gauss", **kwargs):
         """
         from scipy.ndimage import gaussian_filter, uniform_filter, convolve
 
-        if isinstance(radius, u.Quantity):
-            radius = (radius.to("deg") / self.geom.pixel_scales.mean()).value
+        if isinstance(width, u.Quantity):
+            width = (width.to("deg") / self.geom.pixel_scales.mean()).value
 
         smoothed_data = np.empty_like(self.data)
 
         for img, idx in self.iter_by_image():
             if kernel == "gauss":
-                width = radius / 2.
                 data = gaussian_filter(img, width, **kwargs)
             elif kernel == "disk":
-                width = 2 * radius + 1
                 disk = Tophat2DKernel(width)
                 disk.normalize("integral")
                 data = convolve(img, disk.array, **kwargs)
             elif kernel == "box":
-                width = 2 * radius + 1
                 data = uniform_filter(img, width, **kwargs)
             else:
                 raise ValueError("Invalid kernel: {!r}".format(kernel))