Add to WCS tutorial: How to place a scale bar on an image with WCS

[eblur]

A colleague of mine asked me this simple question, and I thought it would make a good addition to the WCS tutorial. He ended up finding a solution in the source code for AplPy, but I'm certain we could make it simpler. (i.e., we don't need to the if statements to know if our inputs are an Astropy Quantity versus a Unit)

## Scale length from my colleague is calculated for a cosmology case
#a2261 is at z = 0.22331. For Omega_M = 0.3, Omega_Lambda = 0.7, this corresponds to 1 arcsec = 3.593 kpc
scale_length = 100.0 / 3.593 * u.arcsec # = 100 kpc

## This code plots that scale length on the image
# The following is taken from AplPy overlays
if isinstance(scale_length, u.Quantity):
    barlength = scale_length.to(u.degree).value
elif isinstance(scale_length, u.Unit):
    barlength = length.to(u.degree)
if wcs.is_celestial:
    pix_scale = wcsutils.proj_plane_pixel_scales(wcs)
else:
    print("you can't do that in this coordinate system")
    pix_scale = [1.0, 1.0]
degrees_per_pixel = np.sqrt(pix_scale[0] * pix_scale[1])
barlength = barlength / degrees_per_pixel

[kakirastern]

@eblur Yeah, I agree that it would be a good idea!

[kakirastern]

Have used matplotlib's axes.arrow for the plotting of a scale bar on an image in the Celestial Coordinates 1 (or celestial-coords1) notebook on Google Colab, with a wcs transform thrown in so that I can directly enter the degrees for the attributes for plotting. Maybe this simplified version would be more suitable for our purposes.

[keflavich]

Even though this was closed... I don't see this material in the tutorials. This is the snippet I use. Should we consider adding to core?

def make_scalebar(ax, left_side, length, color='w', linestyle='-', label='',
                  fontsize=12, text_offset=0.1*u.arcsec):
    axlims = ax.axis()
    lines = ax.plot(u.Quantity([left_side.ra, left_side.ra-length]),
                    u.Quantity([left_side.dec]*2),
                    color=color, linestyle=linestyle, marker=None,
                    transform=ax.get_transform('fk5'),
                   )
    txt = ax.text((left_side.ra-length/2).to(u.deg).value,
                  (left_side.dec+text_offset).to(u.deg).value,
                  label,
                  verticalalignment='bottom',
                  horizontalalignment='center',
                  transform=ax.get_transform('fk5'),
                  color=color,
                  fontsize=fontsize,
                 )
    ax.axis(axlims)
    return lines,txt

[kelle]

There is an implementation of a scale bar in Section 3 of Working with Celestial Coords with WCS I. So, as far as I'm concerned, from a Tutorial perspective, this is addressed.

If you want to propose to add it to core -- which makes total sense -- I assume you'd want to open that issue/PR in a different repo.

fig = plt.figure(figsize=(10, 10), frameon=False)
ax = plt.subplot(projection=wcs_helix)
ax.arrow(337, -21.2, 0, 0.1,
         head_width=0, head_length=0,
         fc='white', ec='white', width=0.003,
         transform=ax.get_transform('icrs'))
plt.text(337.05, -21.18, '0.1 deg',
         color='white', rotation=90,
         transform=ax.get_transform('icrs'))
plt.imshow(image, origin='lower', cmap='cividis', aspect='equal')
plt.xlabel(r'RA')
plt.ylabel(r'Dec')