DEP: change hipscat-->hats

[bsipocz]

There has been a name change and thus new releases, we should update the dependencies in the image, too.

(The full switchover should happen once @troyraen gives the green light that all our hipscat holdings have been replaced by HATS ones)

[troyraen]

IRSA has not released any catalogs in hipscat format. The only hipscat catalog used in Fornax notebooks is Panstarrs (MAST holding) in the Light Curves notebook here: https://github.com/nasa-fornax/fornax-demo-notebooks/blob/main/light_curves/code_src/panstarrs_functions.py.

[zoghbi-a]

We didn't update this because the data in s3://irsa-mast-tike-spitzer-data/data/ZTF/dr20 which is being used for benchmarking is still in hipscat format. Is this going to change.

We'll wait until that is updated and the notebooks are updated too.

[bsipocz]

Can we have both during the transitioning period?

[troyraen]

I think the only hipscat products we care about are the Panstarrs ones in s3://stpubdata/panstarrs/ps1/public/hipscat/ because they're used in the Light Curves notebook. AFAIK, all the benchmarking that used data in s3://irsa-mast-tike-spitzer-data is done -- I hope to remove the Hipscat products from that bucket this month. I don't know when MAST is planning on updating the Panstarrs products to HATS. In the meantime, I agree that it would be nice to have both libraries preinstalled on Fornax.

[zoghbi-a]

I think Zak from MAST was using the ZTF data in irsa-mast-tike-spitzer-data.

We cannot have both. lsdb version 0.3 and 0.4 are incompatible. The former uses hipscat and the latter uses hats. Are you suggesting having them in separate conda environments?

[troyraen]

I think Zak from MAST was using the ZTF data in irsa-mast-tike-spitzer-data.

Good to know he's still using that, thank you. I'm planning to migrate that bucket to a different account, so either the bucket name will change or there will be a period when the data is unavailable. I'll reach out to individuals and also post generally on Slack before I start that.

lsdb version 0.3 and 0.4 are incompatible

Oh right, thanks. I was only thinking about hipscat and hats being separate libraries, but of course lsdb is the main library here. In that case, I think your original thinking makes sense to leave it as-is until there are actual HATS products available.

[bsipocz]

We cannot have both. lsdb version 0.3 and 0.4 are incompatible. The former uses hipscat and the latter uses hats. Are you suggesting having them in separate conda environments?

Good point, my thinking stopped at the point of hipscat and hats and wasn't thinking about the incompatibility in lsdb. Full agree on staying put for now.