flownet.ert._create_ert_setup module

flownet.ert._create_ert_setup._create_ert_parameter_file(parameters: List[flownet.parameters._base_parameter.Parameter], output_folder: pathlib.Path) → None

Takes in the parameters prior distribution as a dataframe, and outputs them in an ert parameter definition file

Parameters

  • parameters – List with Parameters

  • output_folder – Path to the output_folder

Returns

Nothing

flownet.ert._create_ert_setup.create_ert_setup(args: argparse.Namespace, network, schedule: flownet.realization._schedule.Schedule, config: <property object at 0x7f15b0f78830>, parameters=None, training_set_fraction: float = 1, prediction_setup: bool = False)

Create a ready-to-go ERT setup in output_folder.

Parameters

  • schedule – FlowNet Schedule instance to create ERT setup from

  • args – Arguments given to FlowNet at execution

  • network – FlowNet network instance

  • config – Information from the FlowNet config yaml

  • parameters – List with parameters (default = None)

  • training_set_fraction – Fraction of observations to be used for model training (default = 1)

  • prediction_setup – Set to true if it is a prediction run (default = False)

Returns

Nothing

flownet.ert._create_ert_setup.create_observation_file(schedule: flownet.realization._schedule.Schedule, obs_file: pathlib.Path, config: <property object at 0x7f15b0f78830>, training_set_fraction: float = 1, yaml: bool = False)

Creates an ERT observation file from a given schedule instance. It has not yet been decided if schedule is to be given at creation or should be part of self.

Parameters

  • schedule – FlowNet Schedule instance to create observations from.

  • obs_file – Path to store the observation file.

  • config – Information from the FlowNet config yaml

  • training_set_fraction – Fraction of observations in schedule to use in training set

  • yaml – Flag to indicate whether a yaml observation file is to be stored. False means ertobs.

Returns

Nothing

flownet.ert._create_ert_setup.resample_schedule_dates(schedule: flownet.realization._schedule.Schedule, resampling: Optional[str]) → List

Resampling dates based on requested frequency without introducing interpolated dates, i.e., keeps nearest existing date

Parameters

  • schedule – FlowNet Schedule instance to create observations from.

  • resampling – Pandas resampling frequency (str)

Returns

list of resampled dates

Return type

dates