Fix centering of full-symmetry cartesian latticeMaps #346
Conversation
Asciimaps assume everything starts at (0,0), so grid blueprints need to offset indices when reading in full geometry inputs. Also, when writing asciimaps, gridContents need to be offset in the other direction for things to work out correctly. This also fixes a bug related to asciimaps writing cartesian maps at all.
|
Fixes #332 |
|
The code changes look good. Can you post a screenshot of the before and after for the full core Cartesian case with the |
There was a problem hiding this comment.
Can we improve some of the grids documentation to include the images that you provided me with where the origin points are for the different grid types and symmetries? I think that this will be very useful going forward as a visual aid for users and developers.
| gridDesign2 = self.grids["sfp"] | ||
| _grid = gridDesign2.construct() | ||
| self.assertEqual(gridDesign2.gridContents[1, 1], "3") | ||
| self.assertEqual(gridDesign2.gridContents[1, 1], "1") | ||
| self.assertEqual(gridDesign2.gridContents[0, 0], "3") | ||
| self.assertEqual(gridDesign2.gridContents[-1, -1], "3") |
There was a problem hiding this comment.
Can you add unit tests to cover the following scenarios:
- Cartesian, Full Core, Even
- Cartesian, Quarter Core, Odd
- Cartesian, Quarter Core, Even
The SFP test is nice, but let's make sure that we have the other variations tested too.
There was a problem hiding this comment.
Added quarter and even cases
|
Before: After: |
|
Also, switching to 1/4 symmetry in the blueprints works as expected: |
Asciimaps assume everything starts at (0,0), so grid blueprints need to
offset indices when reading in full geometry inputs. Also, when writing
asciimaps, gridContents need to be offset in the other direction for
things to work out correctly.
This also fixes a bug related to asciimaps writing cartesian maps at
all.