Martin Yeo comments

Results 406 comments of


                                            Martin Yeo

Unpin tox<4

> @trexfeathers However, there is [tox-dev/tox-conda#163](https://github.com/tox-dev/tox-conda/pull/163), which is an open PoC PR ... > > Still needs some effort, but looks promising 🤞 I know, I referenced it in my...

Support for ESMF Format offline weights

Perhaps the simplest thing would be to write a converter to work with the existing [`io`](https://github.com/SciTools-incubator/iris-esmf-regrid/blob/main/esmf_regrid/experimental/io.py) module, which would avoid us needing to write any further ingestion code.

Support for ESMF Format offline weights

I didn't mean to close this though! We'll still need to work on the converter

Lockfile PR's can't trigger further GHA's (e.g. the test suite)

Closed by #418 (see https://github.com/SciTools/iris-esmf-regrid/pull/414#issuecomment-2391377430)

Raise exceptions instead of using `assert`

[assert (S101) - Ruff (astral.sh)](https://docs.astral.sh/ruff/rules/assert/)

`use_src_mask` and `use_tgt_mask` with nearest regridding scheme do not produce correct results

Closed by #276

Make modifying `Cube` `CoordSystem`s easier

> This might be a naive question, but is there a use case for having different grid mappings for different coordinates? The CF data model actually attaches grid mappings to...

Make modifying `Cube` `CoordSystem`s easier

During the work on #6536, @ukmo-ccbunney and I worked out that it is not appropriate to have this as a `CoordSystem` property - in the file it is controlled on...

Performance Shift(s): `de4843c4`

> This is real + repeatable, but I think for unknown cause following lockfile update. Not due to project code changes, so presumably some change a dependency. I think we...

Performance Shift(s): `de4843c4`

I agree with @stephenworsley that it might be worth seeing what happens with a `(2, 2, 10000)` benchmark - is that still 1.5 times slower?