-
Notifications
You must be signed in to change notification settings - Fork 197
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Mdspanifying (currently tested) raft::matrix
#846
Mdspanifying (currently tested) raft::matrix
#846
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
EXCELLENT
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There's quite a few redundant element-wise functions in here. Can you create an issue to convert those to encapsulate any mdspan shape for later on?
/** | ||
* @brief Overload of `sort_keys_per_row` to help the | ||
* compiler find the above overload, in case users pass in | ||
* `std::nullopt` for one or both of the optional arguments. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I thought this overload was only needed when there are two optional arguments? @mhoemmen
* @param[in] scalar scalar value to fill matrix elements | ||
*/ | ||
template <typename math_t, typename idx_t, typename layout> | ||
void fill(const raft::handle_t& handle, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We should probably have another element wise function for every possible mdspan shape
I mentioned this offline but it's still unclear to me what exactly the final state of matrix vs linalg should look like after consolidation. For example, I'm not completely convinced that these element wise ops should be in linalg because they aren't linear algebra, yet because they are element wise ops they might not necessarily belong in the matrix namespace either. We need to think about it a bit more, I think. |
@gpucibot merge |
No description provided.