We need to test CFS for complex model intercomparison use cases. A good example of that will be RAMIP, where all the data is currently on JASMIN.
Steps anticipated are
[x] Write code to parse RAMIP (and hopefully any CMIP6 class) experiment to lists of files. This is done, see ramip.py in the ramip branch.
[ ] Write code to aggregate RAMIP filelists into CFA files one per table. This is part done, see mip2cfa.py, but it needs testing, and in particular, the base pointers for the filenames needs to be got right.
[ ] We then need to run a few of these for a few tables from several models and investigate how this works in the GUI.
[ ] We can then run all of RAMIP to see how it scales.
I have got two RAMIP cfa files, for AMon from two models in the data directory. Next test would be to write some upload code for those, and investigate.
We need to test CFS for complex model intercomparison use cases. A good example of that will be RAMIP, where all the data is currently on JASMIN.
Steps anticipated are