time_bnds no longer auto-generated (was in 0.5.0)

[durack1]

Question criteria

• I added a descriptive title here.
• I searched the xCDAT GitHub Discussions to find a similar question and didn't find it.
• I searched the xCDAT documentation and/or Xarray documentation.

Describe your question

I recently created a mamba env, which installed cmor and xcdat (along with a bunch of other libraries). Weirdly, this pinned the xcdat version to 0.5.0, which I was blissfully using and enjoying the fact that time_bnds was autocreated during the xc.open_dataset(file) call - wonderful.

It seems somewhere between 0.5.0 and the latest 0.7.3 this auto time_bnds creation was either lost or turned off. Is this expected? I just took a read through the https://github.com/xCDAT/xcdat/releases notes and this wasn't obviously stated.

I've not generated any MVCE, log or environment info below, but if useful I can easily backfill this intel.

Are there are any possible answers you came across?

No response

Minimal Complete Verifiable Example (MVCE)

No response

Relevant log output

No response

Environment

No response

Anything else we need to know?

No response

[pochedls]

This was an intentional change. Up until v6.0 the functionality to create time bounds was overly simplistic, which could lead to incorrect time bounds. In v5.0, if a dataset had a monthly time axis with an un-centered date of 2000-01-01, we would (likely incorrectly) infer that the time bounds were ["1999-12-16", "2000-16-01"] (or something like that).

In v6.0, this was addressed, but there was a recognition that it is not always possible to infer the correct time bounds (e.g., for some sub-daily data or pentad data). We didn't want to automatically add time bounds that could be wrong. The user can elect to add time bounds by specifying xc.open_mfdataset(..., add_bounds=["T"]) or after open with ds = ds.bounds.add_missing_bounds().

[durack1]

@pochedls thanks for the info backfill. I was a little surprised that such an old version of xCDAT was pinned, not sure what caused it, but thought it was worth raising a query to understand - will close