stilt.Simulation

class stilt.Simulation(receptor: 'Receptor', config: 'VariantConfig', *, met: 'MetStream | None' = None, parent: 'Simulation | None' = None, directory: 'str | Path | None' = None, store: 'Store | None' = None)

One receptor run under one variant.

A simulation runs HYSPLIT for its receptor and calculates a footprint from the particles. A derived simulation (a from: variant) runs no HYSPLIT and calculates its footprint from its parent’s particles. The simulation knows where its output files are and whether they all exist (is_complete()).

You rarely build one yourself. Get it from a model instead, as in model.simulations[receptor_id, "hrrr"].

Parameters

receptorReceptor

Where and when particles are released.

configVariantConfig

Settings of the variant. config.name is the variant name.

metMetStream, optional

Meteorology for the HYSPLIT run. Required unless parent is given.

parentSimulation, optional

Simulation whose particles this one uses. Such a simulation never runs HYSPLIT.

directorystr or Path, optional

Working directory where HYSPLIT runs and outputs are written. A temporary directory is used when omitted. Nothing is created until an output is written.

storeStore, optional

Project store. Outputs are copied there by publish() and read from there when they are not in directory.

Methods

clear_empty_footprint_marker()	Remove the empty-footprint marker.
delete()	Delete this simulation's outputs from the store and its working directory.
generate_footprint([config, write, ...])	Calculate the footprint from the particles.
is_complete()	Return whether every expected output exists.
key(path)	Return the store key of one of this simulation's files.
publish()	Copy this simulation's outputs from the working directory to the store.
resolve(path)	Return a local path to an output file, from the working directory or the store, or None.
run_trajectories([timeout, rm_dat, write])	Run HYSPLIT and keep the particles as trajectories.
transform_context()	Return the TransformContext passed to this simulation's transforms.
write_empty_footprint_marker(reason)	Write the empty-footprint marker holding reason and return its path.

Attributes

empty_footprint_path	Path of the marker written instead of the NetCDF when the footprint is empty.
empty_reason	Why the footprint is empty, or None when it is not.
footprint	The footprint, or None if it does not exist yet.
footprint_path	Path of the footprint NetCDF file in the working directory.
has_footprint	Whether the footprint file or the empty-footprint marker exists.
has_trajectory	Whether the trajectory file exists (the parent's for a derived simulation).
is_backward	Whether particles run backward in time (n_hours < 0).
is_derived	Whether this simulation uses another simulation's particles.
log	Text of the HYSPLIT log.
log_path	Path of the HYSPLIT log in the working directory.
makes_footprint	Whether this simulation makes a footprint, which it does when its variant has a grid.
met_dir	Directory where meteorology files are staged for HYSPLIT.
met_files	Meteorology files staged in met_dir for HYSPLIT.
outcome	How this simulation ended, read from its outputs and log.
outputs	Paths of the log, footprint, and empty-marker files, and the trajectory file unless derived.
plot	Plotting methods, such as sim.plot.map().
source_met_files	Meteorology files in the archive that cover this simulation's period.
time_range	Start and end of the period the particles cover.
trajectories	Particle trajectories, or None if they do not exist yet.
trajectories_path	Path of the trajectory Parquet file (the parent's for a derived simulation).
variant	Variant name.