nmf.ffb.DOT: Generate a Graphviz DOT Diagram for an NMF-FFB Model

In nmfkc: Non-Negative Matrix Factorization with Kernel Covariates

Generate a Graphviz DOT Diagram for an NMF-FFB Model

Description

Creates a Graphviz DOT script that visualizes the structural network estimated by nmf.sem. The resulting diagram displays:

• endogenous observed variables (Y_1),

• exogenous observed variables (Y_2),

• latent factors (F_1, ..., F_Q),

together with the non-negative path coefficients whose magnitudes exceed a user-specified threshold.

Directed edges represent estimated relationships:

• Y_2 \rightarrow F_q: entries of C2 (exogenous loadings),

• F_q \rightarrow Y_1: rows of X (factor-to-endogenous mappings),

• Y_1 \rightarrow F_q: entries of C1 (feedback paths).

Edge widths are scaled by coefficient magnitude, and nodes are placed in optional visual clusters. Only variables participating in edges above the threshold are displayed, while latent factors are always shown.

Usage

nmf.ffb.DOT(
  result,
  weight_scale = 5,
  weight_scale_c2 = weight_scale,
  weight_scale_x1 = weight_scale,
  weight_scale_feedback = weight_scale,
  threshold = 0.01,
  sig.level = 0.1,
  rankdir = "LR",
  fill = TRUE,
  cluster.box = c("normal", "faint", "invisible", "none"),
  cluster.labels = NULL,
  hide.isolated = TRUE,
  ...
)

Arguments

result	A list returned by nmf.sem, containing matrices X, C1, and C2.
weight_scale	Base scaling factor for edge widths.
weight_scale_c2	Scaling factor for edges Y_2 \rightarrow F_q (C2 matrix). Defaults to weight_scale.
weight_scale_x1	Scaling factor for edges F_q \rightarrow Y_1 (X matrix). Defaults to weight_scale.
weight_scale_feedback	Scaling factor for feedback edges Y_1 \rightarrow F_q (C1 matrix). Defaults to weight_scale.
threshold	Minimum coefficient value needed for an edge to be drawn.
sig.level	Significance level for filtering structural edges (C_1 feedback and C_2 exogenous loadings) when inference results are present. If result contains a coefficients data frame from nmf.sem.inference, only edges with p_value < sig.level are drawn, with significance stars (* ** ***) appended to the edge label. The X (factor-to-Y_1) edges are never starred since the basis is not the inference target. Set to NULL to disable significance filtering and fall back to the threshold magnitude filter for both C_1 and C_2. Default is 0.1.
rankdir	Graphviz rank direction (e.g., "LR", "TB").
fill	Logical; whether to use filled node shapes.
cluster.box	Character string controlling the visibility and style of cluster frames around Y2, factors, and Y1 blocks. One of "normal", "faint", "invisible", "none".
cluster.labels	Optional character vector of length 3 giving custom labels for the Y2, factor, and Y1 clusters.
hide.isolated	Logical. If TRUE (default), Y1 and Y2 nodes that have no edges at or above threshold are excluded from the graph.
...	For backward compatibility: accepts deprecated names weight_scale_y2f (use weight_scale_c2) and weight_scale_fy1 (use weight_scale_x1).

Value

A character string representing a valid Graphviz DOT script.

See Also

nmf.ffb, nmf.ffb.inference, plot.nmfkc.DOT

Examples

Y <- t(iris[, -5])
Y1 <- Y[1:2, ]
Y2 <- Y[3:4, ]
result <- nmf.ffb(Y1, Y2, rank = 2, maxit = 500)
dot <- nmf.ffb.DOT(result)
cat(dot)

nmfkc documentation built on July 14, 2026, 1:07 a.m.