[x] Rewrite test (after Ag deleted the changes!) so it works again
[x] Get slurm interface working properly
[x] Make sure all files are being written to outputs
[x] Make sure that outputs are visible in the UI
[x] DO NOT implement e-mailing the user (yet) - AS to check if we already have a way to do this generically in the PyWPS
[x] Do the rest of the things on Alan's list:
(should make a ticket but whatever...)
make sure that the plot has on it:
correct text in correct places (example ini file agrees with Andrew's shell script but the
plot title ends up with longitude in an unexpected place in the title)
sensible number of decimal places for text representations of the lon, lat - see the
_util.py
a single metafile to contain both the plots and the main model output, rather than having these
split into two separate metafiles (see the base class)
plots for the different run types also - but may need more examples from Andrew to do this
copy of some notes with various locations of stuff...
NAME model runner:
/gws/smf/j04/cedaproc/cedawps/swallow/files/20220406_1/model/run_name.sh
various tests:
/gws/smf/j04/cedaproc/cedawps/swallow/alan-scratch/
Example shell scripts
/gws/nopw/j04/name/examples/WebInterfaceRuns_v8_3/TrajectoryRun
$ cd /gws/smf/j04/cedaproc/cedawps/adaq/src
$ source ./setup-env.sh
$ python adaqscripts/aq_plot.py
changes to adaqpython: in adaqcode/cube_statistics.py need "import iris.analysis" (after "import iris")
decide with Ag whether to point to adaqpython under the versioned copied files (in which case the
_plot_output.py should import the function from _run_name_model.py needed to find the relevant directory),
or if not then probably adaq should be removed from the script which takes the snapshot of this content
(but if it is, then the above "import iris.analysis" patch will need to be applied - probably via a
general means of adding patches)
(should make a ticket but whatever...)
make sure that the plot has on it:
correct text in correct places (example ini file agrees with Andrew's shell script but the plot title ends up with longitude in an unexpected place in the title)
sensible number of decimal places for text representations of the lon, lat - see the _util.py
a single metafile to contain both the plots and the main model output, rather than having these split into two separate metafiles (see the base class)
plots for the different run types also - but may need more examples from Andrew to do this
=====================================================================================================
copy of some notes with various locations of stuff...
NAME model runner: /gws/smf/j04/cedaproc/cedawps/swallow/files/20220406_1/model/run_name.sh
various tests: /gws/smf/j04/cedaproc/cedawps/swallow/alan-scratch/
Example shell scripts /gws/nopw/j04/name/examples/WebInterfaceRuns_v8_3/TrajectoryRun
$ cd /gws/smf/j04/cedaproc/cedawps/adaq/src $ source ./setup-env.sh $ python adaqscripts/aq_plot.py
changes to adaqpython: in adaqcode/cube_statistics.py need "import iris.analysis" (after "import iris")
(but if it is, then the above "import iris.analysis" patch will need to be applied - probably via a general means of adding patches)
==============================
link to the cartopy no binary issue, though probably not needed as it works with conda... https://github.com/SciTools/cartopy/issues/837