Add PRT support

[wpbonelli]

Add basic support for MF6 PRT. #1940 and MODFLOW-USGS/modflow6#1389 depend on this. Some tests are skipped/left unimplemented until PRT classes are merged.

Features

• to_coords() and to_prp() methods for ParticleData, LRCParticleData, NodeParticleData
  • get global coordinates and particle release point (PRP) package data, respectively
  • useful for converting MP7 release positions & models to PRT
  • limited to structured and quad-patch/-tree vertex grids
• to_prt_pathlines(), to_mp7_pathlines(), to_mp7_endpoints() adapters
  • convert PRT pathlines to MP7 pathlines or endpoints and vice versa
  • useful for comparing and plotting MP7 and PRT results
• support PRT pathline/endpoint plots for map view and cross section

Future directions

• create_prtsim() function?
  • pattern after Modpath6.create_mpsim(), generate identical PRT simulation
  • not strictly necessary but convenient to start particles in boundary locns
• TrackFile in binaryfile.pywith get_[pathlines/endpoints]() methods?
  • necessary? straightforward to read CSV or binary data with pandas or numpy, almost as easy to get endpoints e.g. sort_values("t").groupby(["imdl", "iprp", "irpt", "trelease"]).tail(1)
• model-agnostic internal format for particle tracks instead of using mp7 as standard?
  • would need more extensive changes to plot modules

[codecov]

Codecov Report

Merging #1753 (b49c185) into develop (c63bfca) will increase coverage by 0.1%.
The diff coverage is 86.8%.

Additional details and impacted files

@@            Coverage Diff            @@
##           develop   #1753     +/-   ##
=========================================
+ Coverage     72.8%   72.9%   +0.1%     
=========================================
  Files          259     259             
  Lines        59040   59282    +242     
=========================================
+ Hits         43004   43261    +257     
+ Misses       16036   16021     -15     

Files	Coverage Δ
flopy/utils/binaryfile.py	83.1% <100.0%> (ø)
flopy/utils/voronoi.py	70.0% <95.6%> (+2.7%)	⬆️
flopy/plot/map.py	83.6% <76.9%> (-0.4%)	⬇️
flopy/plot/plotutil.py	55.2% <92.3%> (+2.0%)	⬆️
flopy/plot/crosssection.py	66.4% <50.0%> (-0.3%)	⬇️
flopy/modpath/mp7particledata.py	86.0% <87.7%> (+4.0%)	⬆️

... and 7 files with indirect coverage changes

[langevin-usgs]

Looks like this is all cleanly set up to go in. Will be good to get @jlarsen-usgs buy in.

[wpbonelli]

This just mimics a modpath sequencenumber — can't guarantee particles are indexed in the same order so this should be ignored when comparing prt and modpath results

It's possible to preserve particle identity by running modpath first, then setting prt PRP release point numbers or boundnames to modpath sequencenumbers

Could maybe add switches name_as_sequencenumber/irpt_as_sequencenumber or maybe simpler sequencenumber="irpt" (or "name") to set sequencenumber to that property

[wpbonelli]

correction unrelated to PRT, lrcregions needs to be 2D

[wpbonelli]

@langevin-usgs suggested adding functions to convert mp7 to PRT results (currently only the other way is supported) but otherwise that full correspondence between result formats is not necessary until there is a demonstrated need

edit: added conversion functions

[jlarsen-usgs]

Overall, this PR look goods.

Is there anything else that needs to be added/modified for PRT export support (e.g., shapefiles, vtk)? If so, will that be included here or will there be a follow up PR?

[wpbonelli]

Yes export functionality is still to come, planned in a separate PR after MF6 package classes are in