Evaluate type OTHER dimensions support

Assess the actual support of openEO processes in case of datacubes that have a OTHER-typed dimension.

As a test case, multi-model multi-scenario ensembles of climate indices are a good candidate for stressing the capabilities of the openEO Python-based backend. A sample is available in the rasdaman catalogue, eg.: http://saocompute.eurac.edu/rasdaman/ows?&SERVICE=WCS&VERSION=2.1.0&REQUEST=DescribeCoverage&COVERAGEID=amt_19712100_tnst_1x1km_year_cordexadj_qdm&outputType=GeneralGridCoverage

Which is the datacube corresponding to the actual files in: /mnt/CEPH_PROJECTS/FACT_CLIMAX/CORDEX-Adjust/INDICES/amt

The functionalities to be tested should be: PUBLISHING use raster2stac to create a datacube with OTHER-typed dimension FILTERING usual spatial/temporal filters; along the OTHER dimension (filter_labels); select a band/model (filter_bands) AGGREGATION reduce_dimension: e.g. get the XY map of a given time interval (IPCC: near term = 2021–2040, mid-term = 2041–2060, long term = 2081–2100) whose "pixels" are the average between the scenarios (OTHER average) of the ensemble median value (BANDS median). INTRA-COMPARISON compute the difference between 2 identical selections of the different scenarios (OTHER dimension), eg. map of differences between 2050 ensemble-max values over South Tyrol (this one pretty much relies on the filter_labels filtering to work) The output can be a Jupyter notebook.