Open rachel3834 opened 1 year ago
Following these instructions on setting up a token and using it through a Python 3.10 venv under Mac OS, I get the following:
` In [1]: from lsst.rsp import get_tap_service, retrieve_query
In [2]: service = get_tap_service()
OK the authentication error was simply due to not having the environment variables named according to the convention required. The two parameters need to be called: EXTERNAL_TAP_URL and ACCESS_TOKEN
Here's a wireframe of a v1.0 implementation of a tool to search the RSP source catalogs as an example.
The documentation includes some notebooks which give good examples of the essential search capabilities that I would expect a TOM to provide, such as this one for searching the point source catalogs.
The workflow that I would expect for this would be:
The LSST Data Products Definition Document gives a full description of the source catalog tables and terminology.
This epic will be broken into 2 main sections:
This epic will be broken into 2 main sections:
Using this thread to share my notes from the team's meeting with Leanne Guy at the Rubin Project and Community Workshop, 2023 in Tucson, AZ:
https://data.lsst.cloud/ - RSP URL Github login or ORCID most common CILogon used for OAuth workflow, handles data rights.
TOM permissions must have at least one person with data rights. They will need to be an admin user.
https://ds0-2.lsst.io/ - documentation for user login
https://dp0-2.lsst.io/data-access-analysis-tools/nb-intro.html
Documentation on configuring security tokens for external applications: https://nb.lsst.io/environment/tokens.html
RSP platform -> User profile -> create token -> Store token in TOM/Topcat etc.
Set an environment variable with the access token, and one with the external TAP URL.
Once configured, can use LSST software within application, e.g. Python
This requires pip install lsst-rsp package.
Documentation describing import and use of service from Python.
Service URL is https://data.lsst.cloud/api/tap Also one for image cutouts
Need to warn users not that high traffic calls may be throttled. Include this in documentation. Consider bulk-queries at the institute level.