Averaging applied auxiliary coordinates - why mean of bounds?

[Goodwin Gibbins]

I've noticed that when a cube is averaged by, for example, cube.collapsed('time', iris.analysis.MEAN) or cube.aggregated_by('year', , iris.analysis.MEAN), the averaging which is applied to the auxiliary coordinates is mean(min(aux_coord), max(aux_coord)), which seems odd to me. Is this a bug or a feature? Why? I couldn't find mention of it in the documentation, but may have missed it.

A very minimal working example:

import iris

import numpy as np

# Load an example cube.

filename = iris.sample_data_path('ostia_monthly.nc')

cube = iris.load_cube(filename)

# Add a simple auxiliary coordinate.

addx = iris.coords.AuxCoord(1.0 + 0* cube.coord('time').points, long_name = 'x')

cube.add_aux_coord(addx, 0)

cube.coord('x').points[2] = 1000

col_cube = cube.collapsed('time', iris.analysis.MEAN)

print(col_cube.coord('x').points) # = 500.5

print(np.mean(cube.coord('x').points)) # = 19.5

[RuthC]

I think it was just done that way for simplicity.  There is an open pull request to change this for the collapsed method, but I'm not sure of the status as it hasn't been updated in a while:

https://github.com/SciTools/iris/pull/3029

Ruth

[Klaus Zimmermann]

Hello,

to me, this seems reasonable.

What don't you like about it? What behavior would you prefer?

Cheers
Klaus

> --
> You received this message because you are subscribed to the Google
> Groups "SciTools (iris, cartopy, cf_units, etc.) -
> https://github.com/scitools" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to scitools-iri...@googlegroups.com
> <mailto:scitools-iri...@googlegroups.com>.
> To view this discussion on the web, visit
> https://groups.google.com/d/msgid/scitools-iris/2f1cc368-4a45-4e4a-8bbb-1515ef64e20ao%40googlegroups.com
> <https://groups.google.com/d/msgid/scitools-iris/2f1cc368-4a45-4e4a-8bbb-1515ef64e20ao%40googlegroups.com?utm_medium=email&utm_source=footer>.