SuperCDMS DID Finder

[BenGalewsky]

As an analyzer I want to resolve SuperCDMS Datasets So I can extract data from that experiment

Assumptions

1. Based on the ServiceX_DID_Finder_lib
2. Will use the SuperCDMS:// DID Finder Scheme (https://gitlab.com/supercdms/slaclab-datacat)

Acceptance Criteria

• Query data, load them and plot a pulse like in https://gitlab.com/supercdms/Analysis/tutorials/-/blob/master/tutorials/Introduction/Basic%20Raw%20Data%20Access.ipynb

[zonca]

@BenGalewsky what is the simplest DID finder implementation that I can look for an example?

[BenGalewsky]

The CERN OpenData DID Finder is quite simple. @Michael-D-Johnson is working on a requests based DID Finder for yt which will be very very similar to the SuperCDMS implementation

[zonca]

thanks @BenGalewsky

@Michael-D-Johnson do you already have some code for the yt DID finder I can look at? It would be really nice for me to understand how to implement one.

[zonca]

@BenGalewsky what about using https://github.com/ssl-hep/ServiceX_DID_Finder_Demo?

[Michael-D-Johnson]

@zonca I don't have code yet to share as it's untested, but I did start with the the DID Finder Demo you found. My main addition was to the find_files function in src/demo_did.py where I wrote some logic to pass the url's I needed as the file_path variable in the yield statement. I'll share the repo once its done.

[Michael-D-Johnson]

@zonca There is an initial DID finder for yt/girder here: https://github.com/pondd-project/ServiceX_DID_Finder_Girder. If you can't access it let me know.

[zonca]

thanks @Michael-D-Johnson , is the selection attribute in https://github.com/pondd-project/ServiceX_DID_Finder_Girder/blob/main/tests/yt.json filtering the YT hub data and returning a subset of it?

[Michael-D-Johnson]

@zonca the did finder is ignoring the selection line. The did finder only needs the first line (did: ). It uses that to find the dataset on YT hub. I am beginning work on a YT hub transformer which will allow you to make a selection/subset/filter on the YT hub data. The other lines currently in the json (are needed for servicex to work) are I believe an example I got using opencerndata and so will be irrelevant to the YT transformer. Once I have the YT hub transformer working, the selection, result-format, etc. will change, and I will update the json as needed. But for testing the did finder all you need is a valid collection id (did).