Update count_if_e, transform_reduce_e, and transform_e to support edge masking #4001
Conversation
seunghwak
added
feature request
|
@jnke2016 FYI |
…to fea_transform_e_mask
seunghwak
changed the title
|
/ok to test |
…to fea_transform_e_mask
…nto fea_transform_e_mask
…to fea_transform_e_mask
…to fea_transform_e_mask
| edge_mask_view_ = edge_mask_view; | ||
| } | ||
|
|
||
| void clear_edge_mask() { edge_mask_view_ = std::nullopt; } |
There was a problem hiding this comment.
I have no preference. Is there a reason to prefer this syntax over edge_mask_view_.reset()?
There was a problem hiding this comment.
I assume they are same. I guess mainly personal preference.
According to
https://stackoverflow.com/questions/47441623/how-to-assign-nothing-to-stdoptionalt
It seems like = {} or = std::nullopt were initially recommended and .reset() is added later.
There was a problem hiding this comment.
But consistency matters, so if we are mainly using .reset() elsewhere in our codebase, I am OK to make a change.
There was a problem hiding this comment.
That was my assessment. A quick check in the code, I see a few calls to .reset() and a bunch more assignments to std::nullopt.
I'm going to merge this PR. We should probably pick a way and create a PR to update things to be consistent.
|
/merge |
Update transform_e to support graphs with edge masking. This is necessary for K-truss.