djura.slf

Storey loss function generation.

Storey-loss-function (SLF) Generator

The tool allows the automatic production of SLFs based on input fragility, consequence and quantity data.

Considerations for double counting should be done at the input level and the consequence function should mirror it.

SLF estimation procedure: Ramirez and Miranda 2009, CH.3 Storey-based building-specific loss estimation (p. 17) FEMA P-58 for fragilities: https://femap58.atcouncil.org/reports For consequence functions: https://femap58.atcouncil.org/reports

EDP: Engineering Demand Parameter DV: Decision Variable DS: Damage State

class djura.slf.slf.SLF(inventory, edp, correlations=None, edp_range=None, edp_bin=None, do_grouping=True, conversion=1.0, realizations=20, replacement_cost=1.0, regression='Weibull', storey=None, directionality=None, seed=None, max_psd=10, max_pfa=5, n_prev=0)

Bases: object

Storey-loss-function (SLF) Generator for Storey-Based Loss Assessment

The tool allows the automatic production of SLFs based on input fragility, consequence and quantity data.

Considerations for double counting should be done at the input level and the consequence function should mirror it.

SLF estimation procedure: Ramirez and Miranda 2009, CH.3 Storey-based building-specific loss estimation (p. 17) FEMA P-58 for fragilities: https://femap58.atcouncil.org/reports For consequence functions: https://femap58.atcouncil.org/reports

EDP: Engineering Demand Parameter DV: Decision Variable DS: Damage State

NEGLIGIBLE = 1e-08

fragility_function()

Derives fragility functions

Returns:

Fragility functions associated with each damage state and component

Return type:

tuple[FragilityModel, ndarray, ndarray]

perform_monte_carlo(fragilities)

Performs Monte Carlo simulations and simulates damage state(DS) for each engineering demand parameter (EDP) value

Parameters:

fragilities (FragilityModel) – Fragility functions of all components at all DSs

Returns:

Sampled damage states of each component for each simulation

Return type:

DamageStateModel

enforce_ds_dependent(damage_state)

Enforces new DS for each dependent component

Parameters:

damage_state (DamageStateModel) – Sampled damage states of each component for each simulation

Returns:

Sampled DS of each component for each simulation after enforcing DS for dependent components if a correlation matrix is provided

Return type:

DamageStateModel

calculate_costs(damage_state)

Evaluates the damage cost on the individual i-th component at each EDP level for each n-th simulation

Parameters:

damage_state (DamageStateModel) – Sampled damage states

Return type:

tuple[CostModel, CostModel, SimulationModel]

Returns:

• CostModel – Total replacement costs in absolute values

• CostModel – Total replacement costs as a ratio of replacement cost

• SimulationModel – Repair costs associated with each component and simulation

perform_regression(loss, loss_ratio, percentiles=None)

Performs regression and outputs final fitted results as storey-loss functions (SLFs)

Parameters:

• loss (CostModel) – Total loss for the floor segment in absolute values

• loss_ratio (CostModel) – Total loss for the floor segment as a ratio of replacement cost

• percentiles (List[float]) – Percentiles to estimate, by default [0.16, 0.50, 0.84], ‘mean’ is always included

Return type:

tuple[LossModel, FittedLossModel, FittingParametersModel]

Returns:

• LossModel – Loss quantiles in terms of both absolute values and ratio to replacement cost

• FittedLossModel – Fitted loss functions

• FittingParametersModel – Fitting parameters or each quantiles and mean

estimate_accuracy(y, yhat)

Estimate prediction accuracy

Parameters:

• y (ndarray) – Observations

• yhat (ndarray) – Predictions

Returns:

Maximum error in %, and Cumulative error in %

Return type:

tuple[float, float]

generate_slfs()

Genearte SLFs

Returns:

SLFs per each performance group

Return type:

Dict[str, SLFPGModel]

export_to_json(out, export_path)

Return type:

None

class djura.slf.models.ComponentDataModel(**data)

Bases: BaseModel

id: int

name: str

EDP: str

Component: str

Group: int | None

Quantity: float

damage_states: int

median_demand: List[float]

total_dispesion: List[float]

repair_cost: List[float]

cost_dispersion: List[float]

best_fit: List[str | None]

classmethod allow_none(v)

model_config: ClassVar[ConfigDict] = {}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class djura.slf.models.CorrelationTreeModel(**data)

Bases: BaseModel

id: int

dependent_on_item: str

min_ds: List[str]

model_config: ClassVar[ConfigDict] = {}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class djura.slf.models.ItemBase(root=PydanticUndefined, **data)

Bases: RootModel

root: Dict[str, ndarray]

model_config: ClassVar[ConfigDict] = {'arbitrary_types_allowed': True}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class djura.slf.models.ItemsModel(root=PydanticUndefined, **data)

Bases: RootModel

root: Dict[int, ItemBase]

model_config: ClassVar[ConfigDict] = {}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class djura.slf.models.FragilityModel(**data)

Bases: BaseModel

EDP: ndarray

ITEMs: ItemsModel

model_config: ClassVar[ConfigDict] = {'arbitrary_types_allowed': True}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class djura.slf.models.DamageStateModel(root=PydanticUndefined, **data)

Bases: RootModel

root: Dict[int, Dict[int, ndarray]]

model_config: ClassVar[ConfigDict] = {'arbitrary_types_allowed': True}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class djura.slf.models.CostModel(root=PydanticUndefined, **data)

Bases: RootModel

root: Dict[int, ndarray]

model_config: ClassVar[ConfigDict] = {'arbitrary_types_allowed': True}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class djura.slf.models.SimulationModel(root=PydanticUndefined, **data)

Bases: RootModel

root: Dict[int, CostModel]

model_config: ClassVar[ConfigDict] = {'arbitrary_types_allowed': True}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class djura.slf.models.FittingModelBase(**data)

Bases: BaseModel

popt: ndarray | List

pcov: ndarray | List

multiplier: float | None

model_config: ClassVar[ConfigDict] = {'arbitrary_types_allowed': True}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class djura.slf.models.FittingParametersModel(root=PydanticUndefined, **data)

Bases: RootModel

root: Dict[str, FittingModelBase]

model_config: ClassVar[ConfigDict] = {}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class djura.slf.models.FittedLossModel(root=PydanticUndefined, **data)

Bases: RootModel

root: Dict[str, ndarray]

model_config: ClassVar[ConfigDict] = {'arbitrary_types_allowed': True}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class djura.slf.models.LossModel(**data)

Bases: BaseModel

loss: Dict[int, Dict[int | str, float]]

loss_ratio: Dict[int, Dict[int | str, float]]

model_config: ClassVar[ConfigDict] = {}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class djura.slf.models.SLFModel(**data)

Bases: BaseModel

directionality: int | None

component_type: str

storey: int | List[int] | None

edp: str

edp_range: List[float]

slf: List[float]

fitting_parameters: Dict[str, FittingModelBase] | None

model_config: ClassVar[ConfigDict] = {}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class djura.slf.models.SLFPGModel(root=PydanticUndefined, **data)

Bases: RootModel

root: Dict[str, SLFModel]

model_config: ClassVar[ConfigDict] = {'arbitrary_types_allowed': True}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

djura.slf.regression_methods.weibull(x, popt)

djura.slf.regression_methods.papadopoulos(x, popt)

djura.slf.utilities.aggregate_values(row, columns, key)

djura.slf.utilities.convert_inv(df)

djura.slf.utilities.convert_corr(df)

djura.slf.utilities.slf_aggregator(*args)

Aggregates input SLFs into a list supported by the loss assessment module and creates interpolation functions for the EDP vs Loss relationships

Each argument must have a specific structure:

{
    'group name': {
        'Directionality': null,
        'Storey': null,
        'edp': 'edp_name',
        'edp_range': list(),
        'slf': list()
    }
}

Return type:

list