Fix iris (>2) warnings about esmvalcore

[schlunma]

Right now I'm using numpy 1.16.4 and iris 2.2.1, but I suppose this has changed since I initially created this issue.

[stefsmeets]

Is this still relevant when we are now working on iris3? #378 #819

[zklaus]

Cheers for the heads-up, @bjlittle

[valeriupredoi]

here's one more

WARNING /home/users/valeriu/anaconda3Feb19/envs/esmvaltool/lib/python3.7/site-packages/ESMValTool-2.0a1-py3.7.egg/esmvaltool/preprocessor/_regrid.py:355: FutureWarning: Conversion of the second argument of issubdtype from `int` to `np.signedinteger` is deprecated. In future, it will be treated as `np.int64 == np.dtype(int).type`.
  extrapolation=extrapolation)

[bascrezee]

/numpy/ma/core.py:3174: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use arr[tuple(seq)] instead of arr[seq]. In the future this will be interpreted as an array index, arr[np.array(seq)], which will result either in an error or a different result.

Some additional information on this warning, that you reported already. I also came across it while working on: ESMValGroup/ESMValTool#1200
This warning can be reproduced easily like this (taken from Stackoverflow )

x = np.zeros((4,2))
x[slice(None),1]    # works
x[[slice(None),1]]  # raises Warning

It should be fixed in latest Iris, see SciTools/iris#3212.

[bjlittle]

@schlunma What versions of iris and numpy are you guys using?

I'm looking to re-haul the amount of warnings that iris generates, which is over verbose as far as I'm concerned.

[bascrezee]

Warnings like this appear very often in the preprocessor. Most likely it's because of the cell_measures attribute in the raw cubes (e.g. tas:cell_measures = "area: areacella"). Would it break something if we removed this attribute?

@bouweandela what do you think? Removing this attribute? Or maybe filtering out the warnings at the top level (is that possible?), since according to @schlunma ESMValTool is designed like this on purpose and the warning is therefore not relevant at any level within ESMValTool. Is that correct?

[bouweandela]

@bjlittle is planning to build support for it into iris, see SciTools/iris#3329.

If that works out, the solution should be as simple as loading both files at the same time:

>>> import iris
>>> iris.load(["/badc/cmip5/data/cmip5/output1/BCC/bcc-csm1-1/historical/fx/atmos/fx/r0i0p0/latest/areacella/areacella_fx_bcc-csm1-1_historical_r0i0p0.nc", "/badc/cmip5/data/cmip5/output1/BCC/bcc-csm1-1/historical/mon/atmos/Amon/r1i1p1/latest/tas/tas_Amon_bcc-csm1-1_historical_r1i1p1_185001-201212.nc"])
/group_workspaces/jasmin2/cp4cds1/magic-test/tmp-bandela/version2_development/conda/envs/esmvaltool/lib/python3.6/site-packages/iris/fileformats/cf.py:798: UserWarning: Missing CF-netCDF measure variable 'areacella', referenced by netCDF variable 'tas'
  warnings.warn(message % (variable_name, nc_var_name))
[<iris 'Cube' of air_temperature / (K) (time: 1956; latitude: 64; longitude: 128)>,
<iris 'Cube' of cell_area / (m2) (latitude: 64; longitude: 128)>]

(note that this currently does not work yet).

Maybe we could filter the warning until support is implemented?

[zklaus]

I think this issue is obsolete now; closing. @schlunma, please reopen if you feel otherwise.

[bjlittle]

BTW we're going to be moving to logging for iris 3.1, which will silence, finally, all/most of the warning messages 👍