Feature 1000 cycloneplotter

[bikegeek]

Pull Request Testing

• Describe testing already performed for these changes:
  
• generate plots for use case using the CyclonePlotter ExtraTC config file at:

https://github.com/dtcenter/METplus/blob/feature_1000_cycloneplotter/parm/use_cases/model_applications/tc_and_extra_tc/CyclonePlotter_fcstGFS_obsGFS_UserScript_ExtraTC.conf

to verify that the extraneous lines are absent.

• also run the use case at:
  https://github.com/dtcenter/METplus/blob/feature_1000_cycloneplotter/parm/use_cases/model_applications/tc_and_extra_tc/Plotter_fcstGFS_obsGFS_ExtraTC.conf

This will produce more tracks as the original code was not correctly retrieving all the available tracks.

• Recommend testing for the reviewer(s) to perform, including the location of input datasets, and any additional instructions:
  

• re-create the plot using the same two use cases above.

• Do these changes include sufficient documentation updates, ensuring that no errors or warnings exist in the build of the documentation? [Yes ]

• Do these changes include sufficient testing updates? [NA]

• Will this PR result in changes to the test suite? [Yes ]
  

Different images produced than the original implementation, resulting in failed checks (2, because I added a test for the same thing early on in the testing cycle):
_

ERROR: Some differences were found
file not found (new output)
A:
B:/data/output/CyclonePlotter/metplus_final.conf

file not found (new output)
A:
B:/data/output/CyclonePlotter/cyclone/20150301.png

file not found (new output)
A:
B:/data/output/CyclonePlotter/cyclone/20150301.csv

Finished comparing directories

_

• Please complete this pull request review by [end of Week of October 12].
  

Pull Request Checklist

See the METplus Workflow for details.

• Complete the PR definition above.
• Ensure the PR title matches the feature or bugfix branch name.
• Define the PR metadata, as permissions allow.
  Select: Reviewer(s)
  Select: Organization level software support Project or Repository level development cycle Project
  Select: Milestone as the version that will include these changes
• After submitting the PR, select Linked issues with the original issue number.
• After the PR is approved, merge your changes. If permissions do not allow this, request that the reviewer do the merge.
• Close the linked issue and delete your feature or bugfix branch from GitHub.

[bikegeek]

Now some use case tests are failing because previous plots don't match the new ones.

[j-opatz]

I ran both use cases indicated in the PR and was successful in creating output that reflected the desired update: storm lines no longer cross the entire globe when they go across the PM. In this respect, the code updates were successful. Not only that, but there was update to the size of the legend box, so it no longer overlaps on the map area and the size of the font looks more appropriate for the image size. The only errors in the automated checks are explainable by new output, which @bikegeek noted in the PR.
However, I did want to note one thing: in an earlier test, the output showed one of the issues, the overlap of the 180 E/180 W text, had been resolved:

My current run shows that these values are again overlapping:


This is not listed in the issue so I'll approve the PR, but would it be possible to address this issue in this code update? If not, we can open a new issue for this.

[bikegeek]

Requires cartopy 0.18 to remove the overlapping 180/-180 longitude labels when the central longitude=180. Use the conda env cycloneplotter_carto_0_18 on 'seneca' in /home/met_test/.conda/envs/cycloneplotter_carto_0_18

[j-opatz]

As @bikegeek indicated in the last comment, the bug of write-over text on the longitudes at the international dateline has been solved; it requires a new cartopy version to be used. I reran the 2 use cases and confirmed that this fix works as intended with no other undesired effects.
I suggest discussing the inclusion of the new cartopy version in the existing cyclone Conda environment with @georgemccabe, since I don't think we want/need a completely new conda environment to support.