KGBN.vis¶

KGBN.vis.is_entire_rule_negated(rule)[source]¶

Check if the entire rule is a single negated expression like !(A | B).

Returns True for rules like:

!(A | B)
!(A & B & C)
! (A | B)

Returns False for rules like:

!(A) | B (negation is only partial)
A & !B (negation is on individual variable)
!A (no parentheses)

KGBN.vis.read_logic_rules(source)[source]¶

Reads logic rules from a file path or from a string containing rules.

Args:: source (str): Path to the file or the string containing logic rules.
Returns:: dict: Mapping from variable names to their logic rules.

KGBN.vis.extract_logic_rules_from_network(network)[source]¶

Extract logic rules from a BooleanNetwork or PBN object using stored equations.

Args:: network: BooleanNetwork or ProbabilisticBN object
Returns:: dict: Mapping from variable names to their logic rules (for BN) or list of rules (for PBN) dict: Mapping from edges to probabilities (for PBN) or empty dict (for BN)

KGBN.vis.build_igraph_pbn(logic_rules, edge_probabilities)[source]¶: Build igraph for PBN with multiple rules per node.

KGBN.vis.build_igraph(logic_rules)[source]¶: Build igraph for BN with single rule per node.

KGBN.vis.create_matplotlib_visualization(logic_rules, removed_nodes=None, removed_edges=None, measured_nodes=None, perturbed_nodes=None, color_node=None)[source]¶

Create a matplotlib-based visualization for Jupyter notebooks.

Args:: color_node (str): If provided, all nodes will use this color (overrides default coloring)

KGBN.vis.vis_network(source, output_html='network_graph.html', interactive=False, removed_nodes=None, removed_edges=None, measured_nodes=None, perturbed_nodes=None, color_node=None, color_edge=None, physics=True, figsize=(13, 8), title='Boolean Network', node_groups=None, layout='hierarchical', layout_kwargs=None, activation_color='#D73027', inhibition_color='#4575B4', node_size=1500, font_size=9, ax=None, return_fig=False, seed=42)[source]¶

Visualize the Boolean network.

When interactive=False (default), produces a publication-quality matplotlib figure with a configurable layout and clearly distinguished activation / inhibition edges.

When interactive=True, saves an interactive HTML file via PyVis.

Args:

source: Logic rules — dict, file path (str), network string (str),: BooleanNetwork, or ProbabilisticBN object.

output_html (str): Output HTML file name (interactive mode only). interactive (bool): If True, generate an interactive HTML file. removed_nodes (set): Nodes shown as removed (light grey). removed_edges (set): Edges shown as removed (interactive mode only). measured_nodes (set): Measured / readout nodes (orange). perturbed_nodes (set): Perturbed nodes (red). color_node (str): Uniform node colour override (interactive mode only). color_edge (str): Uniform edge colour override (interactive mode only). physics (bool): Enable physics simulation (interactive mode only).

figsize (tuple): Figure size (width, height) in inches: (static mode only).

title (str): Figure title (static mode only). node_groups (dict): Custom colour groups that override

connectivity-based defaults. Format:

{group_name: (node_set, color)}

Example:

{'Drug targets': ({'BCL2', 'SYK'}, '#E1F2D0'),
 'Mutations':    ({'NRAS'},         '#D783FF')}

layout (str): Node placement algorithm (static mode only). Options:

'hierarchical' (default) — layered top-down.

'spring' — force-directed (dense/cyclic nets).

'kamada_kawai' — stress-minimisation.

'circular' — evenly spaced on a circle.

'shell' — concentric shells.

'spectral' — eigenvector-based.

'dot' — Graphviz dot (requires pygraphviz or pydot).

layout_kwargs (dict): Extra keyword arguments for the layout: algorithm. For 'hierarchical', x_gap (default 3.0) and y_gap (default 2.5) control inter-node spacing.

activation_color (str): Edge colour for activation (static mode only). inhibition_color (str): Edge colour for inhibition (static mode only). node_size (int): Node marker size in matplotlib units (static mode only). font_size (int): Label font size in points (static mode only). ax: Existing matplotlib.axes.Axes to draw on; a new figure

is created when None (static mode only).

return_fig (bool): If True, return (fig, ax) instead of: calling plt.show() (static mode only).

seed (int): Random seed for reproducible layouts (static mode only).

Returns:

None, or (fig, ax) when return_fig is True and interactive=False.

KGBN.vis.vis_compression(original_network, compressed_network, compression_info, output_html='compression_comparison.html', interactive=False)[source]¶

Visualize the original network with removed/collapsed nodes highlighted.

Args:: original_network: Original BooleanNetwork or ProbabilisticBN compressed_network: Compressed network (not used for visualization) compression_info: Dictionary with compression information output_html (str): Output HTML file name interactive (bool): If True, return network visualization in interactive html file

KGBN.vis.vis_extension(original_network, extended_network, output_html='network+KG.html', interactive=True, color_node='#AED6F1', color_edge=None, extension_color_node='#FBE1BE', extension_color_edge=None, physics=True, figsize=(22, 9), title=None, node_groups=None, layout='hierarchical', layout_kwargs=None, edge_scores=None, activation_color='#D73027', inhibition_color='#4575B4', node_size=1500, font_size=9, return_fig=False, seed=9)[source]¶

Visualize the network with KG (Knowledge Graph) extension, highlighting new nodes and edges.

When interactive=True (default), saves an interactive HTML file via PyVis. When interactive=False, produces a side-by-side publication-quality matplotlib figure comparing the original and extended networks.

Args:

original_network: Original BooleanNetwork or ProbabilisticBN. extended_network: Extended network with additional nodes / edges. output_html (str): Output HTML file name (interactive mode only). interactive (bool): If True, generate an interactive HTML file. color_node (str): Fill colour for original-network nodes. color_edge (str): Uniform edge colour override (interactive mode only). extension_color_node (str): Fill colour for nodes added by the KG. extension_color_edge (str): Edge colour override for KG-added edges

(interactive mode only).

physics (bool): Enable physics simulation (interactive mode only).

figsize (tuple): Overall figure size (width, height) in inches: (static mode only).

title (str): Super-title above both panels (static mode only). node_groups (dict): Optional extra colour groups applied to both

panels, overriding color_node / extension_color_node for the listed nodes. Format:
{group_name: (node_set, color)}
Example:
{'Drug targets': ({'BCL2', 'SYK'}, '#FBE1BE'),
 'Mutations':    ({'NRAS'},         '#D783FF')}

layout (str): Node placement algorithm shared by both panels

(static mode only). See vis_network() for available options.

layout_kwargs (dict): Extra keyword arguments for the layout

algorithm (static mode only).

edge_scores (dict): Optional {(src, tgt): score} mapping used to

scale new-edge line widths by evidence score (static mode only). Build this from the all_relations list returned by load_signor_network():

_, relations = load_signor_network(genes, ...)
edge_scores = {
    (src, tgt): score
    for src, tgt, _, score in relations
    if score is not None
}

Width is mapped as 0.5 + score × 4.

activation_color (str): Colour for activation edges (static mode only). inhibition_color (str): Colour for inhibition edges (static mode only). node_size (int): Node marker size (static mode only). font_size (int): Label font size in points (static mode only). return_fig (bool): If True, return (fig, axes) instead of

calling plt.show() (static mode only).

seed (int): Random seed for reproducible layouts (static mode only).

Returns:

None, or (fig, axes) when return_fig is True and interactive=False.

KGBN.vis.create_matplotlib_extension_visualization(logic_rules, new_nodes, new_edges, color_node='lightblue', extension_color_node='#E1F2D0', extension_color_edge=None)[source]¶

Create a matplotlib-based visualization for extension comparison.

Args:: logic_rules: Dictionary of logic rules new_nodes: Set of new node names from KG extension new_edges: Set of new edge tuples from KG extension color_node (str): Color for nodes from the original network extension_color_node (str): Color for new nodes from KG extension_color_edge (str): Color for new edges from KG. If None, uses default color