"""Whole published growth-and-yield systems for Sweden.
Each module here is one publication reproduced end to end: its own coefficients,
its own state, its own stepping rules. Eriksson (1976), Persson (1992), Petterson
(1955), Ekö (1985) and Elfving & Hägglund (1975) are not assemblies of
interchangeable equations — the parts were fitted together and only agree with
the printed yield tables when used together. Keeping each one whole is what makes
it checkable against its source.
That is the difference from :mod:`pyforestry.sweden.adapters`, and it runs in one
direction only: a system may carry coefficients *and* ship the
:class:`~pyforestry.base.simulation.growth_model.GrowthModel` that drives it,
because a self-contained system owns its own interface. An adapter may not carry
coefficients at all.
Individual published equations that stand on their own — a volume function, a
site-index curve, a bark thickness model — belong in the domain packages
(``sweden/volume``, ``sweden/siteindex``, ``sweden/bark``, …) instead.
"""
from .eko1985 import (
DominantHeightObservation,
Eko1985Cohort,
Eko1985Model,
Eko1985SiteContext,
Eko1985Stand,
RegionSE,
)
from .elfving_hagglund_1975 import ElfvingHagglundInitialStand
from .eriksson_1976 import (
Eriksson1976ManagementSchedule,
Eriksson1976Model,
Eriksson1976Stand,
SimulationResult,
StandInit,
ThinningProgram,
ThinningRequest,
estimate_initial_stand,
simulate,
)
from .nystrom_soderberg_1987 import NystromSoderberg1987
from .persson_1992 import (
Persson1992Model,
Persson1992Stand,
PerssonSimulationResult,
PerssonStandInit,
PerssonThinningProgram,
PerssonThinningRequest,
persson_estimate_initial_stand,
persson_simulate,
)
from .petterson_1955 import (
PETTERSON_VARIANTS,
Petterson1955Model,
Petterson1955Stand,
PettersonSimulationResult,
PettersonStandInit,
PettersonThinningProgram,
PettersonVariant,
PettersonYieldRow,
petterson_simulate,
)
#: The stateful whole-rotation runner each published system ships, by the name
#: its `GrowthModel` adapter is registered under. Elfving & Hägglund (1975) is
#: absent deliberately: it estimates a starting stand and does not step one.
_SYSTEM_RUNNERS: dict[str, type] = {
"eko_1985": Eko1985Stand,
"eriksson_1976": Eriksson1976Stand,
"persson_1992": Persson1992Stand,
"petterson_1955": Petterson1955Stand,
}
[docs]
def available_systems() -> dict[str, type]:
"""Return each published system's own runner, by name.
These project a stand end to end, exactly as
:func:`~pyforestry.sweden.simulation.presets.get_pipeline`'s composites do --
which ``available_pipelines()`` alone does not say, since it lists only the
two compositions pyforestry assembled itself. Asking what this package can
run means asking both.
The classes are returned rather than built, because there is no shared
constructor to build them with and inventing one would misrepresent them:
each takes the starting state and thinning programme its own publication
defines. Read the class to see what it wants.
Returns:
Name to runner class. The same names their ``GrowthModel`` adapters use
in :func:`~pyforestry.available_models`, so one system is one name
whichever way it is reached -- the adapter steps a stand you supply, the
runner builds and projects its own.
"""
return dict(_SYSTEM_RUNNERS)
__all__ = [
"available_systems",
"DominantHeightObservation",
"Eko1985Cohort",
"Eko1985Model",
"Eko1985SiteContext",
"Eko1985Stand",
"ElfvingHagglundInitialStand",
"Eriksson1976ManagementSchedule",
"Eriksson1976Model",
"Eriksson1976Stand",
"NystromSoderberg1987",
"PETTERSON_VARIANTS",
"Persson1992Model",
"Persson1992Stand",
"PerssonSimulationResult",
"PerssonStandInit",
"PerssonThinningProgram",
"PerssonThinningRequest",
"Petterson1955Model",
"Petterson1955Stand",
"PettersonSimulationResult",
"PettersonStandInit",
"PettersonThinningProgram",
"PettersonVariant",
"PettersonYieldRow",
"RegionSE",
"SimulationResult",
"StandInit",
"ThinningProgram",
"ThinningRequest",
"estimate_initial_stand",
"persson_estimate_initial_stand",
"persson_simulate",
"petterson_simulate",
"simulate",
]