Tips for performance

Select only fields you need

If columns is not specified, we transfer all available data fields (original LSST fields and Fink science module outputs) to you! This is easily over 100 fields, and I doubt you need all of them. After a phase of debug to learn which fields are useful, it is wiser to select only a subset of the fields:

# select only time, flux, position
r = requests.post(
    "https://api.lsst.fink-portal.org/api/v1/sso",
    json={
        "n_or_d": "8467",
        "columns": "r:midpointMjdTai,r:psfFlux,r:psfFluxErr,r:ra,r:dec",
    }
)

Note that the fields should be comma-separated, and unknown field names are ignored. See the definition page for the list of available fields per endpoint.

Playing with the cache

We cache query results at the level of the database. When you hit a timeout (120 seconds), if you relaunch the query, it will progress faster as it will re-use data already fetched. Do not abuse, but relaunching a few times sometimes does not harm :-)