How to specify colorbar keyword arguments when using quickplot.pcolormesh?

[krikru]

When using iris.quickplot.pcolormesh, a colorbar is created automatically. How can I specify keywords arguments for that colorbar, like I can do when I create a colorbar manually and add it to a plot created with iris.plot.pcolormesh?

[krikru]

Or is there some better way to solve it in than the way I’m trying to solve it in while still getting the same behavior?

[rcomer]

Which keywords do you want to pass? Some properties of the Colorbar can be set afterwards via its set_... methods, e.g.

import iris.quickplot as qplt

mesh = qplt.pcolormesh(my_data)
mesh.colorbar.set_label('stuff')

For full control, I think the right way is to use iris.plot.pcolormesh and add the colorbar manually. I think quickplot was only really designed for a "first look" at your data and not with plot configuration in mind.

[krikru]

Thanks for your reply!

The function I want to pass the keyword arguments in just takes an argument colorbar_kwargs as an argument which I had hoped that I could pass on. But using iris.plot.pcolormesh instead and creating a colorbar manually may indeed be the most proper way to go about it.

So in order to get the same behavior with iris.plot.pcolormesh as I would get with iris.quickplot.pcolormesh (which I want to for consistency purposes), I have noticed that there are a couple of things that I need to do. First I need to create and add a colorbar, which I know how to do. Then I also need to add a label, since iris.quickplot.pcolormesh seems to add the units of the cube as a label to the colorbar. Where in the code does that happen, so I can look at how a label is created from the units? Looking at the pieces of code for iris.plot.pcolormesh and iris.quickplot.pcolormesh respectively, though, there seems like they are quite different. Are there more differences between them that I should know of, except for that that the quicklplot version adds a colorbar and a label for the colorbar?

[trexfeathers]

Is this what you are looking for?

iris/lib/iris/quickplot.py

Lines 100 to 116 in 3e1e80e

	def _get_titles(u_object, v_object):
	if u_object is None:
	u_object = iplt._u_object_from_v_object(v_object)
	xlabel = _title(u_object, with_units=True)
	ylabel = _title(v_object, with_units=True)
	title = ""
	if u_object is None:
	title = _title(v_object, with_units=False)
	elif isinstance(u_object, iris.cube.Cube) and not isinstance(
	v_object, iris.cube.Cube
	):
	title = _title(u_object, with_units=False)
	elif isinstance(v_object, iris.cube.Cube) and not isinstance(
	u_object, iris.cube.Cube
	):
	title = _title(v_object, with_units=False)
	return xlabel, ylabel, title

[rcomer]

The above snippet gives the plot title and the labels for the x- and y-axis, which quickplot also adds automatically (although the x- and y- labels are skipped for maps). The colorbar label is defined here:

iris/lib/iris/quickplot.py

Lines 66 to 69 in 3e1e80e

	if _use_symbol(cube.units):
	bar.set_label(cube.units.symbol)
	else:
	bar.set_label(cube.units)

[trexfeathers]

#5970

[krikru]

The above snippet gives the plot title and the labels for the x- and y-axis, which quickplot also adds automatically (although the x- and y- labels are skipped for maps). The colorbar label is defined here:

iris/lib/iris/quickplot.py

Lines 66 to 69 in 3e1e80e

	if _use_symbol(cube.units):
	bar.set_label(cube.units.symbol)
	else:
	bar.set_label(cube.units)

Thanks; this is probably what I was looking for since I'm plotting maps, but the code for adding x- and y-labels could definitely be useful too.

How does Iris know when a map is being plotted? Does is just check whether the two axis coordinates are spatial coordinates?

[krikru]

Thank you both anyway for your answers. I think I have my code the way want it now.