SpinupSettings

class rameau.core.SpinupSettings(cycles=0, starting_date=None, ending_date=None)[source]

Spinup settings.

Parameters:

cycles (int, optional) – Number of initialisation period cycles. Default is 0.
starting_date (datetime.datetime, optional) – Starting date of the initialisation period. Default is the starting date of the rainfall input data.
ending_date (datetime.datetime, optional) – Ending date of the initialisation period. An error will be returned if the provided date is sooner than the spinup starting date. Default is the date preceding the SimulationSettings.starting_date.

Returns:

SpinupSettings

Examples

Just a dummy example.

>>> data = pd.DataFrame(np.random.rand(25))
>>> model = rm.Model(
...     tree=dict(watersheds=[{}]),
...     inputs=dict(
            rainfall=data,
            pet=data,
            date=pd.date_range("2000-01-01", "2000-01-25")),
        simulation_settings=dict(
            starting_date=datetime.datetime(2000, 1, 4)
            spinup_settings=dict(
                starting_date=datetime.datetime(2000, 1, 1)
                ending_date=datetime.datetime(2000, 1, 3),
                cycles=2
            )
        )
... )
>>> for key, val in model.simulation_setttings.spinup_settings.item():
...     print(key, ":", val)
cycles = 2
starting_date = 2000-01-01 00:00:00
ending_date = 2000-01-03 00:00:00

Let’s define a new model but without simulation settings.

>>> model = rm.Model(
...     tree=dict(watersheds=[{}]),
...     inputs=dict(
            rainfall=data,
            pet=data,
            date=pd.date_range("2000-01-01", "2000-01-25")),
... )
>>> for key, val in model.simulation_setttings.spinup_settings.item():
...     print(key, ":", val)
cycles = 0
starting_date = None
ending_date = None

The starting date and ending date of the spinup settings attached to the model are not defined. If we run a simulation, these dates should be initialised.

>>> sim = model.run_simulation()
>>> for key, val in sim.simulation_setttings.spinup_settings.item():
...     print(key, ":", val)
cycles = 0
starting_date = None
ending_date = None

What’s wrong? Spinup dates are initiliased only when cycles > 0. Let’s change that:

>>> model.simulation_settings.spinup_settings.cycles = 1
>>> sim = model.run_simulation()
Traceback (most recent call last):
...
RuntimeError: Spinup ending date older than spinup starting date
>>> print(sim.simulation_settings.starting_date)
    2000-01-01 00:00:00

The method throws an error. Try defining a simulation starting date one day later.

>>> model.simulation_settings.starting_date = datetime.datetime(2000, 1, 2)
>>> sim = model.run_simulation()
>>> for key, val in sim.simulation_setttings.spinup_settings.item():
...     print(key, ":", val)
cycles = 1
starting_date = 2000-01-01 00:00:00
ending_date = 2000-01-01 00:00:00

Now it works. Default value for spinup starting date is 2000-01-01 (first date of rainfall data). If the simulation starting date is also equal to 2000-01-01, then the spinup ending date is 1999-12-31, which is sooner than the default spinup starting date. So we need to change the simulation starting date to 2000-01-02 for a working initialisation period of 1 day.

Attributes

`cycles`	Number of spinup cycles.
`ending_date`	Ending date of the initialisation period.
`starting_date`	Starting date of the initialisation period.