Skip to content
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

New/improved Experts #19

Open
AlexandreMary opened this issue Feb 14, 2024 · 0 comments
Open

New/improved Experts #19

AlexandreMary opened this issue Feb 14, 2024 · 0 comments

Comments

@AlexandreMary
Copy link
Contributor

AlexandreMary commented Feb 14, 2024

Experts analyse the outputs of tests and compare them to a reference, according to various metrics depending on the type of test.
Some experts would benefit from improvements, some new experts addressing aspects not cheched at the moment should be developped:

To be improved:

  • pack2bin (gmkpack): retrieve path to gmkpack outputs when build fails ? to be printed in the stdout of the terminal ? (not sure it is possible)
  • fields: scatter plots or histogram of errors if normalized RMSE > threshold ? link with tecpaf/F.Suzat
  • fields: too slow on Arpege T798 for instance (cf. 3rd bullet in Make davai test suite faster #34)
  • DrHook: how to make it more reliable (at the moment way too volatile)

New experts:

  • DDH: epygram is able to read DDH files and could compare DDH outputs
  • EC_MEMINFO: in replacement of the RSS expert, get more information from EC_MEMINFO
  • Coding norms checker: for routines modified in branch, get norms checker outputs and set a "score": 0 if no improvement in norms checking violation, positive if some improvement, negative if introducing new violations. How to use the norms checker out of the box (Ryad) ?
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant