feat(dis_properties): add grid method to get discretization properties - #2795
Conversation
* replace numpy .shape = with .reshape() in Raster * replace .utcfromtimestamp with fromtimestamp(timestamp, timezone.utc) in ModelTime
* numpy will be deprecating `.shape =` reshaping, migrating to reshape convention to prevent DeprecationWarnings
* returns dictionary of keyword arguments to build DIS, DISV, and DISU depending on grid type * added cell `.area` calculation via shoelace algorithm to `Grid` * remove deprecated flopy.mf6.utils/reference.py which housed "pre-modelgrid" spatial reference support for MF6 models
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## develop #2795 +/- ##
===========================================
+ Coverage 55.5% 73.3% +17.7%
===========================================
Files 644 659 +15
Lines 124135 132157 +8022
===========================================
+ Hits 68947 96913 +27966
+ Misses 55188 35244 -19944
🚀 New features to boost your workflow:
|
jdhughes-dev
left a comment
There was a problem hiding this comment.
The grid methods and the cell area calculation look good. I checked area against areas I could work out by hand and it is right for a structured grid with variable row and column spacing and for a vertex grid holding a square and a triangle.
Removing flopy/mf6/utils/reference.py is overdue. It was deprecated in 3.9 for removal in 3.10+, we are at 3.11, and nothing in flopy imports it.
Two things before this goes in:
.docs/code.rststill lists./source/flopy.mf6.utils.reference.rst. The files undersource/are written bysphinx-apidocat build time, so once the module is gone that line points at a document that is never generated. Read the Docs is not set to fail on warnings and the docs are not built on a pull request, so this will not turn up until after the merge.- Patch coverage is 55%, mostly the new
get_dis*_kwargspaths.
Should this go in the release that accompanies MODFLOW 6 6.8.0, or wait for the one after?
|
I'll work on some additional testing and clean up the docs for the removed There isn't a big rush to get this one in, so feel free to move forward with the release if this PR isn't completed in time. |
|
@jlarsen-usgs CI errors are unrelated and have been fixed separately, don't let them keep you from merging. In any case we won't be ready for the release for another few days yet. |
.areacalculation via shoelace algorithm toGridMirrors support outlined in removed PR #2700