Visualization Module

Requires the viz extra (pip install "implicit-word-network[viz]").

implicit_word_network.visualization.plot_network

plot_network(network: ImplicitNetwork | Graph, *, min_weight: float = 0.0, top_k: int | None = 100, labels: Sequence[str] | None = None, label_order: Sequence[str] | None = None, min_component_size: int = 1, ax: Any | None = None, layout: str = 'spring', seed: int = 42, node_size: tuple[float, float] = (90.0, 1100.0), edge_width: tuple[float, float] = (0.6, 5.5), with_labels: bool = True, max_labels: int | None = None, font_size: float = 8.5, legend: bool = True, palette: Sequence[str] = CATEGORICAL_PALETTE, title: str | None = None, output_path: str | None = None, show: bool = False) -> Any

Draw the entity layer of a network with matplotlib.

Node area scales with the number of mentions, edge width and opacity with the edge weight, and node colour encodes the entity type (fixed hue per type, listed in a legend). Labels are set in text ink with a surface halo and nudged apart so they do not overlap. Isolated entities are not drawn.

PARAMETER	DESCRIPTION
network	An ImplicitNetwork or a graph produced by to_networkx (node attributes text, label, count; edge weight). TYPE: ImplicitNetwork | Graph
min_weight	Drop edges lighter than this (networks only). TYPE: float DEFAULT: 0.0
top_k	Keep only the heaviest top_k edges (networks only; None keeps all). TYPE: int | None DEFAULT: 100
labels	Restrict to these entity types (networks only). TYPE: Sequence[str] | None DEFAULT: None
label_order	Entity types in the order they take palette slots. Defaults to types sorted by total mentions (most frequent first). TYPE: Sequence[str] | None DEFAULT: None
min_component_size	Drop connected components with fewer nodes, so small satellites do not squeeze the main network into a corner. TYPE: int DEFAULT: 1
ax	Matplotlib axes to draw on (a new figure is created otherwise). TYPE: Any | None DEFAULT: None
layout	"spring", "kamada_kawai" or "circular". TYPE: str DEFAULT: 'spring'
seed	Random seed of the spring layout. TYPE: int DEFAULT: 42
node_size	Minimum and maximum node area in points². TYPE: tuple[float, float] DEFAULT: (90.0, 1100.0)
edge_width	Minimum and maximum edge width in points. TYPE: tuple[float, float] DEFAULT: (0.6, 5.5)
with_labels	Draw entity names. TYPE: bool DEFAULT: True
max_labels	Label only the max_labels most mentioned entities. TYPE: int | None DEFAULT: None
font_size	Label font size in points. TYPE: float DEFAULT: 8.5
legend	Show the entity-type legend. TYPE: bool DEFAULT: True
palette	Categorical colours, assigned to types in label_order. TYPE: Sequence[str] DEFAULT: CATEGORICAL_PALETTE
title	Optional figure title. TYPE: str | None DEFAULT: None
output_path	Save the figure to this path when given. TYPE: str | None DEFAULT: None
show	Call plt.show(). TYPE: bool DEFAULT: False

RETURNS	DESCRIPTION
Any	The matplotlib axes.