Create subset files for V18 Circle CI

[cansavvy]

Purpose/implementation Section

This PR is for creating the subset files for circle CI testing for v18 release.
Closes #871

What was your approach?

This branch was created from jharenza:v18-files.
I made the changes to analyses/create-subset-files/create_subset_files.sh as described by #871 in 9510df5 :

• Update RELEASE to v18
• Add copy over step for pbta-histologies-base.tsv file.

With those changes I ran this on AWS as described by #871 and then in f14ca50 I added biospecimen_ids_for_subset.RDS file to this branch.

I have testing_v18.zip it seems reasonable (82 files, md5sum shows 80 files because no release notes or md5sum). I can upload whenever we are ready.

What GitHub issue does your pull request address?

Closes #871

[jaclyn-taroni]

Hmmm, the instructions on #871 are out of date for what we need now which is to:

1. Start merging the early PBTA histologies PRs to facilitate review of the later PRs.
2. Review the later PRs
3. Merge the release add v18 release #857

Which is to say that we don't want to add the changes from the #857 branch yet. Do we want to cherry pick 9510df5 onto a new branch?

[jaclyn-taroni]

Make that 9510df5 and f14ca50.

[cansavvy]

Which is to say that we don't want to add the changes from the #857 branch yet. Do we want to cherry pick 9510df5 onto a new branch?

Ah. Right. Forgot about that aspect. Yes so we'd like a PR with only the changes here and not the v18 changes from jharenza:v18-files?

And how should this relate to the order of me adding the actual subset files to S3?

[jaclyn-taroni]

Yes so we'd like a PR with only the changes here and not the v18 changes from jharenza:v18-files?

Yep!

And how should this relate to the order of me adding the actual subset files to S3?

I think we were holding off on merging the PBTA Histologies PRs so that they can be tested and the "depth" of that stack of stacked PRs can make it a little tricky to do a thorough review. I believe we also sometimes have testing files on CI "early" or ahead of the release itself so we can find and remedy any breaking changes. Which is to say - let's get them on S3 and see what happens!

[cansavvy]

v18 subset files have been uploaded to S3 bucket.

[cansavvy]

This reverting the unwanted commits was not as cleaned as I hoped. Going to close this PR and only copy over the actual wanted changes from 9510df5 and f14ca50 as mentioned previously. #873 (comment)

[cansavvy]

Closing this PR in favor of #888