Closed nilshempelmann closed 5 years ago
candidate and target would be netCDF files holding different indices (annual mean temp, annual mean pr), and indices would be the name of the indices to perform the comparison. I'll try to come up with a test based on public data as well as a notebook example.
OK I got it. So the Process would be the following step after an indieces calculation process (from finch).
name of the indices to perform the comparison.
Means the variable name in the netCDF file, used to read the data?
The notebook can be stored in the notebooks
repository and linked into the documentation
Correct. Correct. Will do.
@cehbrecht checks failing, due to wokingdir missdeclaration. files are saved to curdir, not to the temp folder ... but self.workdir is set, can you help?
@nilshempelmann I have started to fix workir
... see wps_subset_continents.py
for example. But there is more to do. You don't need rename_complexinput
... should be fixed in pywps.
There is still something wrong with the test-suite ... haven't figured out. When I only run the test_wps_subset_continent
test then it works.
@cehbrecht the workdirs are set in all Processes.
`flyingpigeon.handler_common
is quite strange... please check this module. I also propose to move the only function there wfs_common
to eggshell.nc.nc_utils
may be even rename it to simply wfs
The output effects all Processes. So here is the issue: https://github.com/bird-house/flyingpigeon/issues/281
@nilshempelmann I have skipped the subset continent/countries tests. They work when tested alone:
pytest tests/test_wps_subset_continents.py
... but not when running the test suite:
pytest -v tests/
I suppose there is still an issue with the clipping
.
@cehbrecht
I suppose there is still an issue with the
clipping
.
OK will check it. Processes are running with good outputs. https://github.com/bird-house/flyingpigeon/issues/282
@huard leave it to you to wrap up with handler_common
https://github.com/bird-house/flyingpigeon/issues/281 issue and merge the PR and delete the branch
spatial analogs and map spatial anlogs slighly modified and integrated. tests passed.
I didn't made test runs, since I don't know how what to write for the argument 'indice'.
@huard are there a tutorial available may be based on public available datasets?
I also propose to rename
map_spatial_analog
toplot_spatial_analog
for consitent naminghttps://github.com/bird-house/flyingpigeon/issues/277