Querying Summit Logs#

Three summit logging services can be queried through rubin_nights:

  • Night report — a structured summary of the observing night, one per telescope.

  • Narrative log — a running record of observer comments and observatory state changes.

  • Exposure log — annotations attached to specific exposures.

Clients for all three are included in the dictionary returned by get_clients().

Setting a time window#

from astropy.time import Time, TimeDelta

day_obs = "2025-04-15"
t_start = Time(f"{day_obs}T12:00:00", format="isot", scale="utc")
t_end = t_start + TimeDelta(1, format="jd")

Night report#

query_night_report() returns the night report for a given day_obs and telescope as a dict. Passing return_html=True also returns a rendered HTML string suitable for display in a notebook.

from IPython.display import display, HTML

report, html = endpoints["night_report"].query_night_report(
    day_obs=day_obs,
    telescope="Simonyi",
    return_html=True,
)
display(HTML(html))

The telescope argument accepts "Simonyi" or "AuxTel".

Narrative log#

query_log() returns log messages between two times as a DataFrame indexed by timestamp:

log = endpoints["narrative_log"].query_log(t_start, t_end)
log[["component", "message_text", "user_id"]].head()

The timestamp index makes it straightforward to join narrative entries with EFD data or visit tables for contextual analysis.

For more granular control, the underlying query method accepts the full set of parameters supported by the narrativelog REST API.

messages = endpoints["narrative_log"].query(
    min_date_begin=t_start.isot,
    max_date_begin=t_end.isot,
    components=["ATMCS", "ATPneumatics"],
)

Exposure log#

query_log() retrieves exposure annotations. This log is typically sparse; many nights will return an empty DataFrame.

exposure_log = endpoints["exposure_log"].query_log(t_start, t_end)

Exposure log entries reference an exposure ID rather than a wall-clock time, so joining with visit data requires matching on obs_id or exposure_id.

The full set of query parameters is documented at the exposurelog REST API.

Async queries#

All three log clients expose both synchronous and asynchronous query methods. The query method is synchronous; async_query is its async counterpart for use in async contexts or event-loop-aware frameworks.

messages = await endpoints["narrative_log"].async_query(
    min_date_begin=t_start.isot,
    max_date_begin=t_end.isot,
)