Repository navigation
Rework Predicates and document intersects and nearest with cpp:function - #1430
Conversation
97af3dc to
0af11a8
Compare
| the pages above document helper functions that generate the whole set at once, | ||
| e.g., ``ArborX::Experimental::make_intersects(geometries)``. | ||
| An equivalent ``ArborX::Experimental::make_ordered_intersects(geometries)`` | ||
| exists for ordered spatial predicates. |
There was a problem hiding this comment.
Either include make_nearest too, or drop make_ordered_intersects. Also, consider dropping ArborX:: namespace, we don't seem to use much in the documentation.
| Defined in header `<ArborX_Predicates.hpp> <https://github.com/arborx/ArborX/blob/master/src/spatial/detail/ArborX_Predicates.hpp>`_ | ||
|
|
||
| .. cpp:function:: template <typename Geometry> \ | ||
| KOKKOS_INLINE_FUNCTION auto intersects(Geometry const &geometry) |
There was a problem hiding this comment.
This alignment is weird. What should be aligned is the rendered text.
There was a problem hiding this comment.
Meaning you don't want the break line \ here?
There was a problem hiding this comment.
I do want to break the line, but I don't like that it renders like
instead of being flushed left.
It's not specific to this PR, so not blocking.
| .. cpp:function:: template <typename Primitives> \ | ||
| auto Experimental::make_intersects(Primitives const &primitives) |
|
|
||
| This is equivalent to creating the set of predicates | ||
| ``intersects(ArborX::Sphere(AccessTraits<Primitives>::get(primitives, i), r))``. | ||
|
|
There was a problem hiding this comment.
I'm surprised but this is correct. We don't explicitly restrict to point in the code, but it will break otherwise. Could consider using approx_expand_by_radius.
| * :doc:`Predicates </API-Reference/Predicates>` | ||
| * :cpp:func:`nearest` | ||
| * :doc:`Callbacks </API-Reference/Callbacks>` | ||
| * :doc:`ArborX::BVH::query() </API-Reference/ArborX%3A%3ABoundingVolumeHierarchy%3A%3Aquery>` |
There was a problem hiding this comment.
Why just BVH? Or why have BVH?
There was a problem hiding this comment.
You mean we should just say query()?
There was a problem hiding this comment.
No, I wonder why we link to BVH::query at all, and not to the other queries (e.g., BruteForce), and whether we should mention it here at all.
Move intersects and nearest, together with their make_* helpers, under API-Reference/predicates/ as native cpp:function declarations, and rework the Predicates landing page. Fix the nearest signature (int k) and reference the predicates with :cpp:func: from the query pages. Assisted-By: Claude:aws-gov.claude-sonnet-5-5
Signed-off-by: Damien L-G <dalg24@gmail.com>
Without a blank line before the nested list, the predicate names were rendered inline in the preceding list item instead of as a list. Assisted-By: Claude:aws-gov.claude-sonnet-5-5
0af11a8 to
4642236
Compare
Add the blank line before the nested list so that the predicate names render as a list instead of inline in the preceding list item, as was done for the current query() pages. Also separate the items of the list in the DistributedTree page.
|
OK, going to ignore alignments and #1369. It's good enough. |
Move intersects and nearest, together with their make_* helpers, under API-Reference/predicates/ as native cpp:function declarations, and rework the Predicates landing page. Fix the nearest signature (int k) and reference the predicates with :cpp:func: from the query pages.
Assisted-By: Claude:aws-gov.claude-sonnet-5-5