Print Functions API

The groups below separate graph/model printers from inference-result and WLS printers used during interactive inspection.

Model and Graph

FactorGraph.printModel — Function

printModel(graph::GaussianFactorGraph)

Print a compact summary of variables and factors.

Arguments

graph: Gaussian factor graph or Gaussian tree view.

Notes

This helper is intended for interactive inspection.

Example

x1 = GaussianVariable(:x1, 1)
f1 = GaussianFactor(:x1, 0.0, 1.0, 0.1; label = "f1")

graph = factorGraph([x1], [f1])

printModel(graph)

source

FactorGraph.printGraph — Function

printGraph(graph::AbstractFactorGraph)

Print variables, factors, and edge connectivity as tables.

Arguments

graph: Gaussian or discrete factor graph, or tree view.

Notes

The output shows each variable with connected factor labels and edge IDs, each factor with connected variable labels and edge IDs, and finally the edge list. This function is intended for interactive inspection.

Example

x1 = GaussianVariable(:x1, 1)
f1 = GaussianFactor(:x1, 0.0, 1.0, 0.1; label = "f1")

graph = factorGraph([x1], [f1])

printGraph(graph)

source

FactorGraph.printEdges — Function

printEdges(graph::AbstractFactorGraph)

Print one line per graph edge.

Arguments

graph: Gaussian or discrete factor graph, or tree view.

Notes

Each line has the form edge_id: variable <-> factor.

Example

x1 = GaussianVariable(:x1, 1)
f1 = GaussianFactor(:x1, 0.0, 1.0, 0.1; label = "f1")

graph = factorGraph([x1], [f1])

printEdges(graph)

source

Inference

FactorGraph.printMessages — Function

printMessages(
    graph::AbstractFactorGraph, inference::AbstractInference;
    variable = nothing, factor = nothing
)

Print messages using exactly one keyword filter:

Arguments

graph: Gaussian or discrete factor graph, or tree view.
inference: Matching inference object.

Keywords

variable: Variable ID or label for variable-to-factor messages.
factor: Factor index or label for factor-to-variable messages.

Notes

Use exactly one keyword filter. variable = label prints messages from the variable node to connected factor nodes. factor = label prints messages from the factor node to connected variable nodes.

The printed fields depend on the inference form: moment form prints mean and covariance, canonical form prints information and precision, and min-sum form prints costs or quadratic-message data.

Example

x1 = GaussianVariable(:x1, 1)
f1 = GaussianFactor(:x1, 0.0, 1.0, 0.1; label = "f1")

graph = factorGraph([x1], [f1])
inference = moment(graph)

messages!(graph, inference)
printMessages(graph, inference; variable = :x1)

source

FactorGraph.printMarginal — Function

printMarginal(
    graph::AbstractFactorGraph, inference::AbstractSumProductInference;
    variable = nothing
)

Print variable marginals. If variable is omitted, all variable marginals are printed.

Arguments

graph: Gaussian or discrete factor graph, or tree view.
inference: Matching inference object.

Keywords

variable: Optional variable ID or label.

Notes

Moment-form marginals print mean and covariance. Canonical-form marginals also print information and precision.

Example

x1 = GaussianVariable(:x1, 1)
f1 = GaussianFactor(:x1, 0.0, 1.0, 0.1; label = "f1")

graph = factorGraph([x1], [f1])
inference = moment(graph)
gbp!(graph, inference)

printMarginal(graph, inference; variable = :x1)

source

FactorGraph.printEstimate — Function

printEstimate(
    graph::AbstractFactorGraph, inference::AbstractMinSumInference;
    variable = nothing
)

Print min-sum MAP estimates. If variable is omitted, all variable estimates are printed.

Arguments

graph: Gaussian or discrete factor graph, or tree view.
inference: Matching min-sum inference object.

Keywords

variable: Optional variable ID or label.

Example

x1 = GaussianVariable(:x1, 1)
f1 = GaussianFactor(:x1, 0.0, 1.0, 0.1; label = "f1")

graph = factorGraph([x1], [f1])
inference = minsum(graph)
gbp!(graph, inference)

printEstimate(graph, inference; variable = :x1)

source

Weighted Least Squares

FactorGraph.printWLS — Function

printWLS(
    graph::Union{GaussianFactorGraph, TreeFactorGraph{GaussianFactorGraph}},
    result::WeightedLeastSquaresResult
)

Print a weighted least-squares result.

Arguments

graph: Gaussian factor graph or tree view.
result: Weighted least-squares result.

Notes

Prints one mean and covariance block per Gaussian variable.

Example

x1 = GaussianVariable(:x1, 1)
f1 = GaussianFactor(:x1, 0.0, 1.0, 0.1; label = "f1")

graph = factorGraph([x1], [f1])
result = solveWLS(graph)

printWLS(graph, result)

source