Review MODE implementation to look for more computational efficiencies

[JohnHalleyGotway]

Describe the Enhancement

Issue #2724 and corresponding PR #2726 enhance MODE in MET-12.0.0-beta2 by improving its convolution algorithm, adding OpenMP, and minimizing avoidable memory allocation. This issue to conduct a careful review of the rest of the MODE application and associated library code to look for additional potential efficiencies.

Some ideas to consider are copied below from this issue comment:

• The ShapeData::n_objects() function calls split which loops over the entire grid and can be slow. Could eliminate calls to n_objects() or remove that function entirely to prevent its use.
• Most of the time MODE loops over a grid that is very sparse, often only containing a single object. Consider adding bounding box information to the ShapeData class to keep track of the extent of non-zero object data. Update member functions to loop over that bounding box rather than the entire domain.
• Review each call to DataPlane::get(x,y) which calls two_to_one() which performs a bounds check. Can they replaced with direct operation on the vector of Data? Or can they be replaced with a new accessor function that skips the bounds checking? Running @DanielAdriaansen test case with 75 tiny objects through a profiler, 50% of the ~10 min runtime is spent accessing data.
• Data is read in twice in multivar_frontend.cc, with calls to ModeFrontEnd::create_multivar_simple_objects() and ModeFrontend::create_multivar_merge_objects(). Instead the data should be kept in memory when creating the simple objects and accessed when creating the merge objects.

Time Estimate

2 days?

Sub-Issues

Consider breaking the enhancement down into sub-issues.

• Add a checkbox for each sub-issue here.

Relevant Deadlines

List relevant project deadlines here or state NONE.

Funding Source

Define the source of funding and account keys here or state NONE.

Define the Metadata

Assignee

• Select engineer(s) or no engineer required
• Select scientist(s) or no scientist required

Labels

• Review default alert labels
• Select component(s)
• Select priority
• Select requestor(s)

Milestone and Projects

• Select Milestone as the next official version or Backlog of Development Ideas
• For the next official version, select the MET-X.Y.Z Development project

Define Related Issue(s)

Consider the impact to the other METplus components.

• METplus, MET, METdataio, METviewer, METexpress, METcalcpy, METplotpy
  Expect no impacts.

Enhancement Checklist

See the METplus Workflow for details.

• Complete the issue definition above, including the Time Estimate and Funding Source.
• Fork this repository or create a branch of develop.
  Branch name: feature_<Issue Number>_<Description>
• Complete the development and test your changes.
• Add/update log messages for easier debugging.
• Add/update unit tests.
• Add/update documentation.
• Push local changes to GitHub.
• Submit a pull request to merge into develop.
  Pull request: feature <Issue Number> <Description>
• Define the pull request metadata, as permissions allow.
  Select: Reviewer(s) and Development issue
  Select: Milestone as the next official version
  Select: MET-X.Y.Z Development project for development toward the next official release
• Iterate until the reviewer(s) accept and merge your changes.
• Delete your fork or branch.
• Close this issue.

[davidalbo]

The subtask of reading data in twice was resolved in this issue: https://github.com/dtcenter/MET/issues/2745