SpatialUncertainty

Provides comprehensive uncertainty quantification methods for spatial predictions including epistemic, aleatoric, and total uncertainty calculation, confidence interval generation, and uncertainty visualization capabilities.

Details

The SpatialUncertainty class provides:

• Multiple uncertainty types (epistemic, aleatoric, total)

• Confidence interval generation with various methods

• Spatial uncertainty mapping and visualization

• Uncertainty calibration and validation metrics

Public fields

uncertainty_results

Stored uncertainty calculation results

visualization_config

Configuration for uncertainty visualizations

calibration_metrics

Uncertainty calibration assessment results

Active bindings

uncertainty_results

Stored uncertainty calculation results

visualization_config

Configuration for uncertainty visualizations

calibration_metrics

Uncertainty calibration assessment results

Methods

Public methods

• SpatialUncertainty$new()

• SpatialUncertainty$calculate_uncertainties()

• SpatialUncertainty$generate_confidence_intervals()

• SpatialUncertainty$create_uncertainty_maps()

• SpatialUncertainty$visualize_uncertainty()

• SpatialUncertainty$assess_calibration()

• SpatialUncertainty$calculate_coverage_probability()

• SpatialUncertainty$clone()

---

Method new()

Usage

SpatialUncertainty$new(config = list())

Arguments

config

Optional configuration list for uncertainty parameters

Examples

\dontrun{
uncertainty <- SpatialUncertainty$new()
}

---

Method calculate_uncertainties()

Usage

SpatialUncertainty$calculate_uncertainties(
  predictions,
  method = "bdl",
  uncertainty_types = c("epistemic", "aleatoric", "total")
)

Arguments

predictions

Prediction results from ML models (BDL, RF, etc.)

method

Method used for predictions ("bdl", "rf", "ensemble")

uncertainty_types

Types of uncertainty to calculate

Returns

UncertaintyResults object with all uncertainty estimates

Examples

\dontrun{
uncertainties <- uncertainty$calculate_uncertainties(
  predictions = bdl_result,
  method = "bdl",
  uncertainty_types = c("epistemic", "aleatoric", "total")
)
}

---

Method generate_confidence_intervals()

Usage

SpatialUncertainty$generate_confidence_intervals(
  predictions,
  confidence_level = 0.95,
  method = "normal"
)

Arguments

predictions

Prediction results with uncertainty estimates

confidence_level

Confidence level (e.g., 0.95 for 95% CI)

method

Method for interval calculation ("normal", "bootstrap", "quantile")

Returns

List with lower and upper confidence bounds

Examples

\dontrun{
intervals <- uncertainty$generate_confidence_intervals(
  predictions = predictions,
  confidence_level = 0.95,
  method = "normal"
)
}

---

Method create_uncertainty_maps()

Usage

SpatialUncertainty$create_uncertainty_maps(
  uncertainty_results,
  field_data,
  map_types = c("epistemic", "aleatoric", "total"),
  resolution = NULL
)

Arguments

uncertainty_results

Uncertainty calculation results

field_data

Spatial field data for mapping

map_types

Types of uncertainty maps to create

resolution

Spatial resolution for mapping

Returns

List of uncertainty raster maps

Examples

\dontrun{
maps <- uncertainty$create_uncertainty_maps(
  uncertainty_results = uncertainties,
  field_data = field_data,
  map_types = c("epistemic", "total")
)
}

---

Method visualize_uncertainty()

Usage

SpatialUncertainty$visualize_uncertainty(
  uncertainty_results,
  plot_type = "all",
  interactive = FALSE
)

Arguments

uncertainty_results

Uncertainty calculation results

plot_type

Type of visualization ("maps", "histograms", "scatter", "all")

interactive

Whether to create interactive plots

Returns

List of uncertainty visualization plots

Examples

\dontrun{
plots <- uncertainty$visualize_uncertainty(
  uncertainty_results = uncertainties,
  plot_type = "all",
  interactive = TRUE
)
}

---

Method assess_calibration()

Usage

SpatialUncertainty$assess_calibration(
  predictions,
  observations,
  calibration_method = "reliability_diagram"
)

Arguments

predictions

Predictions with uncertainty estimates

observations

True observed values for validation

calibration_method

Method for calibration assessment

Returns

Calibration assessment results

Examples

\dontrun{
calibration <- uncertainty$assess_calibration(
  predictions = predictions,
  observations = true_values,
  calibration_method = "reliability_diagram"
)
}

---

Method calculate_coverage_probability()

Usage

SpatialUncertainty$calculate_coverage_probability(
  predictions,
  observations,
  confidence_level = 0.95
)

Arguments

predictions

Predictions with confidence intervals

observations

True observed values

confidence_level

Expected confidence level

Returns

Coverage probability assessment

Examples

\dontrun{
coverage <- uncertainty$calculate_coverage_probability(
  predictions = predictions_with_ci,
  observations = true_values,
  confidence_level = 0.95
)
}

---

Method clone()

The objects of this class are cloneable with this method.

Usage

SpatialUncertainty$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Examples

if (FALSE) { # \dontrun{
# Initialize uncertainty quantification
uncertainty <- SpatialUncertainty$new()

# Calculate uncertainties from BDL predictions
uncertainties <- uncertainty$calculate_uncertainties(bdl_predictions)

# Generate confidence intervals
intervals <- uncertainty$generate_confidence_intervals(predictions, level = 0.95)

# Create uncertainty maps
maps <- uncertainty$create_uncertainty_maps(uncertainties, field_data)
} # }


## ------------------------------------------------
## Method `SpatialUncertainty$new`
## ------------------------------------------------

if (FALSE) { # \dontrun{
uncertainty <- SpatialUncertainty$new()
} # }


## ------------------------------------------------
## Method `SpatialUncertainty$calculate_uncertainties`
## ------------------------------------------------

if (FALSE) { # \dontrun{
uncertainties <- uncertainty$calculate_uncertainties(
  predictions = bdl_result,
  method = "bdl",
  uncertainty_types = c("epistemic", "aleatoric", "total")
)
} # }


## ------------------------------------------------
## Method `SpatialUncertainty$generate_confidence_intervals`
## ------------------------------------------------

if (FALSE) { # \dontrun{
intervals <- uncertainty$generate_confidence_intervals(
  predictions = predictions,
  confidence_level = 0.95,
  method = "normal"
)
} # }


## ------------------------------------------------
## Method `SpatialUncertainty$create_uncertainty_maps`
## ------------------------------------------------

if (FALSE) { # \dontrun{
maps <- uncertainty$create_uncertainty_maps(
  uncertainty_results = uncertainties,
  field_data = field_data,
  map_types = c("epistemic", "total")
)
} # }


## ------------------------------------------------
## Method `SpatialUncertainty$visualize_uncertainty`
## ------------------------------------------------

if (FALSE) { # \dontrun{
plots <- uncertainty$visualize_uncertainty(
  uncertainty_results = uncertainties,
  plot_type = "all",
  interactive = TRUE
)
} # }


## ------------------------------------------------
## Method `SpatialUncertainty$assess_calibration`
## ------------------------------------------------

if (FALSE) { # \dontrun{
calibration <- uncertainty$assess_calibration(
  predictions = predictions,
  observations = true_values,
  calibration_method = "reliability_diagram"
)
} # }


## ------------------------------------------------
## Method `SpatialUncertainty$calculate_coverage_probability`
## ------------------------------------------------

if (FALSE) { # \dontrun{
coverage <- uncertainty$calculate_coverage_probability(
  predictions = predictions_with_ci,
  observations = true_values,
  confidence_level = 0.95
)
} # }