corrGraph package

Submodules

corrGraph.corrGraph module

class corrGraph.corrGraph.Config

Bases: object

Configuration class for setting default parameters for the correlation graph visualization.

AXIS_CONFIG = {'showgrid': False, 'showline': False, 'showticklabels': False, 'title': '', 'zeroline': False}

DEFAULT_EDGE_WIDTH = 0.01

DEFAULT_NODE_SIZE = 0.5

EDGE_WIDTH_SCALE = 1

FIGURE_SIZE = '800px'

LAYOUT_CONFIG = {'annotations': [{'showarrow': False, 'text': 'Visualization of the correlation graph', 'x': 0.005, 'xref': 'paper', 'y': -0.002, 'yref': 'paper'}], 'height': 800, 'hovermode': 'closest', 'margin': {'b': 20, 'l': 5, 'r': 5, 't': 40}, 'showlegend': False, 'title': '<br>Correlation Graph', 'titlefont_size': 16, 'width': 800, 'xaxis': {'showgrid': False, 'showline': False, 'showticklabels': False, 'title': '', 'zeroline': False}, 'yaxis': {'showgrid': False, 'showline': False, 'showticklabels': False, 'title': '', 'zeroline': False}}

NEGATIVE_COLOR = 'orange'

NODE_SIZE_SCALE = 20

POSITIVE_COLOR = 'blue'

class corrGraph.corrGraph.CorrGraph(corr_matrix: DataFrame, threshold: float = 0.5, use_correlations_as_weights: bool = True)

Bases: object

Class for creating and visualizing a correlation graph from a correlation matrix.

get_centrality(centrality_type: str = 'degree') → Dict[str, float]

Calculate and return the centrality of each node in the graph based on the specified centrality type.

Parameters:

centrality_type – A string to specify the type of centrality measure (‘degree’, ‘betweenness’, ‘closeness’, ‘eigenvector’).

Returns:

A dictionary where keys are node names and values are their centrality measures.

Raises:

ValueError – If an invalid centrality type is provided.

get_graph() → Graph

Get the created graph.

Returns:

A networkx Graph object representing the correlation graph.

update_node_weights(weights: Dict[str, float]) → None

Update the weights of the nodes in the graph.

Parameters:

weights – A dictionary where keys are node names and values are the weights to be assigned.

Raises:

• ValueError – If a node is not found in the graph.

• TypeError – If a weight is not numeric.

visualize_graph(visualization_type: str = 'plotly', node_weight_is_size: bool = True, edge_weight_is_size: bool = True, use_as_notebook: bool = True) → None

Visualize the graph using either Plotly or PyVis based on the visualization_type argument.

Parameters:

• visualization_type – A string to decide which visualization library to use (‘plotly’ or ‘pyvis’).

• node_weight_is_size – A boolean to decide if node weights should determine node sizes.

• edge_weight_is_size – A boolean to decide if edge weights should determine edge widths.

• use_as_notebook – A boolean to decide if the visualization should be displayed in a Jupyter notebook (only for PyVis).

visualize_graph_with_plotly(node_weight_is_size: bool = True, edge_weight_is_size: bool = True) → None

Visualize the graph using Plotly.

Parameters:

• node_weight_is_size – A boolean to decide if node weights should determine node sizes.

• edge_weight_is_size – A boolean to decide if edge weights should determine edge widths.

visualize_graph_with_pyvis(node_weight_is_size: bool = True, edge_weight_is_size: bool = True, use_as_notebook=True) → None

Visualize the graph using PyVis.

Parameters:

• node_weight_is_size – A boolean to decide if node weights should determine node sizes.

• edge_weight_is_size – A boolean to decide if edge weights should determine edge widths.

• use_as_notebook – A boolean to decide if the visualization should be displayed in a Jupyter notebook.

Module contents