fix: use fabsf() consistently for float symmetry checks

[andrewwhitecdw]

Use fabsf() instead of abs() for the floating-point symmetry difference in all learner-facing locations.

abs() selects the integer overload, which truncates/rounds the float before comparing it against the 0.1 threshold. That can hide real asymmetry or report false positives.

The format specifier stays %d everywhere because temperature_grid_f uses cuda::std::dextents<int, 2>, so temp.extent(1) returns int, not size_t.

Files updated:

• Solutions/symmetry-check.cpp
• Sources/symmetry-check.cpp
• Solutions/row-symmetry-check.cpp
• Sources/row-symmetry-check.cpp
• Sources/row-symmetry-check-assertions.cpp
• Sources/row-symmetry-check-fixed.cpp
• 03.02.02-Exercise-Symmetry.ipynb
• 03.02.03-Exercise-Row-Symmetry.ipynb
• 03.02.04-Dev-Tools.ipynb

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[github-actions]

👋 Thank you for your contribution!

This pull request is from a forked repository so GitHub Actions will not be able to run CI. A maintainer will review your changes shortly and manually trigger the CI.

@maintainers Please review this PR when you have a chance and follow the instructions in the CONTRIBUTING.md file to trigger the CI.

[brycelelbach]

/ok to test a96f6f0

[github-actions]

❌ Git Signature Check Failed

🔗 View workflow run logs

Found 1 unsigned commit(s):

Unsigned commits

• 5a641d4: Fix float abs and size_t format specifier in symmetry-check solution (unsigned)

How to fix:

1. Configure commit signing (if not already done):

   # For GPG signing
   git config --global commit.gpgsign true
   
   # Or for SSH signing (Git 2.34+)
   git config --global gpg.format ssh
   git config --global user.signingkey ~/.ssh/id_ed25519.pub

2. Re-sign your commits:

   git fetch https://github.com/NVIDIA/accelerated-computing-hub.git main
   git rebase -i FETCH_HEAD --exec "git commit --amend --no-edit -S"
   git push --force-with-lease

📚 GitHub documentation on signing commits

[robobryce]

Changes requested for the format regression below. The floating-point change also needs to be applied consistently to the code the notebooks generate and to the closely related row-symmetry examples, rather than only to one linked solution.

Separately, the required git-signature check currently fails because commit 5a641d4 is unsigned.

[robobryce]

temperature_grid_f is declared as cuda::std::mdspan<float, cuda::std::dextents<int, 2>> in Sources/ach.cuh, so its index_type—and therefore the return type of temp.extent(1)—is int, not size_t. %zu expects a size_t; passing an int through variadic printf is a format/argument mismatch and undefined behavior. Please retain %d here (or explicitly cast to size_t, though that is unnecessary).

[robobryce]

Please apply this correctness fix to the code learners actually generate and to the related examples as well. 03.02.02-Exercise-Symmetry.ipynb still emits abs(...) in all three symmetry-check code cells, while Solutions/row-symmetry-check.cpp, Sources/row-symmetry-check-assertions.cpp, and the fixed-code cell in 03.02.04-Dev-Tools.ipynb retain the same expression. As written, the exercise continues teaching/executing the form this PR says is incorrect, and the linked solution silently differs. Keep the extent format as %d in those locations because the mdspan index type is int.

[github-actions]

👋 Thank you for your contribution!

This pull request is from a forked repository so GitHub Actions will not be able to run CI. A maintainer will review your changes shortly and manually trigger the CI.

@maintainers Please review this PR when you have a chance and follow the instructions in the CONTRIBUTING.md file to trigger the CI.