1 · delivery type
How events are delivered — this decides what you can compute or include below.
Slack renders its own message — a
generic-webhook function won't be used.
3 · compute & filter (optional)
Runs for each matching event on a dynamical.org data product.
Write only the body of handler. ds is the triggering product's
dataset, already opened as an xarray Dataset. The return value (JSON or a scalar) is
placed on the result key of the webhook body; return None to
skip (filter out) the delivery. To send an image, base64-encode it and return the string.
numpy, pandas, xarray and matplotlib (as plt) are imported for you.
See examples ↗ ·
copied
See the dataset's dims, coordinates, and variables — e.g. whether
latitude is a 1-D axis or a 2-D projected coordinate (HRRR).
3 · forecast (optional)
Your delivery URL (step 4) must be a Slack incoming-webhook URL. Optionally
include the latest forecast for a location — one matplotlib chart per variable, rendered
from the just-completed run of each dynamical.org data product target
(ensemble datasets show every member).
Add a complete trigger to a
target to include a forecast — the run's data isn't available at in_flight or
progress.
