Provide ability to "unset" certain elements of the FlorisModel
When keyword arguments are set on the instantiated FlorisModel using the set() method, they are remembered. To override a previously set variable, one most override that variable with a new call to set(). However, for some arguments, in particular, heterogeneous_inflow_config, it would be nice to be able to "unset" or forget a previous setting. FlorisModel currently has no way to do this, since passing None (e.g. fmodel.set(heterogeneous_inflow_config=None)) simply means using the previously set case.
Proposed solution
A possible solution would be to allow a special (unambiguous) argument to be passed to keywords that could be "unset" in order to trigger the unsetting action. For example, fmodel.set(heterogeneous_inflow_config=-1) or fmodel.set(heterogeneous_inflow_config=False).
Alternatives considered
A second alternative is to define new methods on FlorisModel for the unsetting procedure. This is actually something that we already have for unsetting previously set control actions: FlorisModel.reset_operation(). However, if we go down this route, we'll need a new method for each non-critical variable, which seems a little annoying (especially if these changes are to be reflected in UncertainFlorisModel and ParallelFlorisModel).
Current workaround
For now, users can "forget" things by simply reinstantiating their FlorisModel from scratch.
Additional context
I'd like a general solution here, which requires identifying which of the arguments to FlorisModel.set() are valid candidates for "unsetting". I would think at least the following:
heterogeneous_inflow_config- Control setpoints (
yaw_angles, power_setpoints, awc_modes, awc_amplitudes, awc_frequencies, disable_turbines)
- possibly
wind_data, turbine_type and turbine_library_path, solver_settings
Provide ability to "unset" certain elements of the
FlorisModelWhen keyword arguments are set on the instantiated
FlorisModelusing theset()method, they are remembered. To override a previously set variable, one most override that variable with a new call toset(). However, for some arguments, in particular,heterogeneous_inflow_config, it would be nice to be able to "unset" or forget a previous setting.FlorisModelcurrently has no way to do this, since passingNone(e.g.fmodel.set(heterogeneous_inflow_config=None)) simply means using the previously set case.Proposed solution
A possible solution would be to allow a special (unambiguous) argument to be passed to keywords that could be "unset" in order to trigger the unsetting action. For example,
fmodel.set(heterogeneous_inflow_config=-1)orfmodel.set(heterogeneous_inflow_config=False).Alternatives considered
A second alternative is to define new methods on
FlorisModelfor the unsetting procedure. This is actually something that we already have for unsetting previously set control actions:FlorisModel.reset_operation(). However, if we go down this route, we'll need a new method for each non-critical variable, which seems a little annoying (especially if these changes are to be reflected inUncertainFlorisModelandParallelFlorisModel).Current workaround
For now, users can "forget" things by simply reinstantiating their
FlorisModelfrom scratch.Additional context
I'd like a general solution here, which requires identifying which of the arguments to
FlorisModel.set()are valid candidates for "unsetting". I would think at least the following:heterogeneous_inflow_configyaw_angles,power_setpoints,awc_modes,awc_amplitudes,awc_frequencies,disable_turbines)wind_data,turbine_typeandturbine_library_path,solver_settings