online_cpd_plotter

Online Change-Point Detection Plotter.

This module provides the OnlineCpdPlotter class that coordinates multiple visualizers and components to create comprehensive visualizations for change-point detection analysis.

class pysatl_cpd.analysis.visualization.online.plotters.online_cpd_plotter.OnlineCpdPlotter(backend, data_provider=None, detection_trace=None, layout=None)[source]

Bases: Generic

Coordinator class for online change-point detection visualizations.

This class manages visualizers and components to create comprehensive visualizations. It provides a simplified API for setting up visualizations with configurable defaults.

Parameters:

backend (DrawBackend) – The plotting backend to use (MATPLOTLIB or PLOTLY).
data_provider (Optional[DataProvider[Any, TimeseriesAnnotation]]) – The data provider containing observations.
detection_trace (Optional[OnlineDetectionTrace[TypeVar(StateT, bound= OnlineAlgorithmState)]]) – The detection trace from online algorithm execution.
layout (ILayoutStrategy | str | None) –
Layout strategy to use. Can be:
- An ILayoutStrategy instance
- One of: “vertical”, “split”, “dashboard-lite”, “dashboard”
- None (defaults to “vertical”)

__init__(backend, data_provider=None, detection_trace=None, layout=None)[source]

Parameters:

backend (DrawBackend)
data_provider (DataProvider[Any, TimeseriesAnnotation] | None)
detection_trace (OnlineDetectionTrace | None)
layout (ILayoutStrategy | str | None)

Return type:

None

property backend: DrawBackend

Return the plotting backend used by this visualizer.

Returns:: Current backend (MATPLOTLIB or PLOTLY).
Return type:: DrawBackend

default_layout()[source]

Create figure and axes using the current layout strategy.

Returns:: Figure object and axes mapping for drawing.
Return type:: tuple[Figure | Figure, dict[str, Axes] | dict[str, tuple[int, int]]]

set_data_provider(data_provider)[source]

Set the data provider for the timeseries visualizer.

Parameters:: data_provider (DataProvider[Any, TimeseriesAnnotation]) – Data provider containing observations.
Returns:: Returns self to allow method chaining.
Return type:: Self

set_timeseries_visualizer(timeseries_visualizer)[source]

Replace the timeseries visualizer coordinated by this plotter.

Return type:: Self
Parameters:: timeseries_visualizer (ITimeseriesVisualizer[Any])

set_detection_trace(detection_trace)[source]

Set the detection trace for the trace visualizer.

Parameters:: detection_trace (OnlineDetectionTrace[TypeVar(StateT, bound= OnlineAlgorithmState)]) – Detection trace from online algorithm execution.
Returns:: Returns self to allow method chaining.
Return type:: Self

set_ground_truth(change_points, margin=0)[source]

Set ground truth change points and margin.

Parameters:

change_points (list[int]) – List of ground truth change point indices.
margin (int | tuple[int, int]) – Margin around change points. An integer creates a symmetric margin. A tuple (left, right) creates an asymmetric margin. If both values are 0, the margin is not drawn.

Returns:

Returns self to allow method chaining.

Return type:

Self

set_legend_axis(axis_name)[source]

Set the axis where legend should be displayed for detection components.

Parameters:: axis_name (str | None) – Name of the axis for legend display. None means auto (timeseries).
Returns:: Returns self to allow method chaining.
Return type:: Self

add_component(name, component, axes_names, show_legend=True)[source]

Add a visual component to the plotter.

Parameters:

name (str) – Name to identify the component.
component (IVisualComponent) – Component instance to add.
axes_names (list[str]) – Names of axes where the component should be drawn. Must be provided.
show_legend (bool) – Whether to show legend for this component.

Returns:

Returns self to allow method chaining.

Return type:

Self

remove_component(name)[source]

Remove a component from the plotter.

Parameters:: name (str) – Name of the component to remove.
Returns:: Returns self to allow method chaining.
Return type:: Self

get_component(name)[source]

Get a component by name.

Parameters:: name (str) – Name of the component to retrieve.
Returns:: The component instance or None if not found.
Return type:: IVisualComponent | None

draw(figure, axes)[source]

Coordinate drawing of all visualizers and components.

Parameters:

figure (Figure | Figure) – The figure to draw on (Matplotlib or Plotly).
axes (dict[str, Axes] | dict[str, tuple[int, int]]) – Mapping from subplot names to their axes objects or positions.

Returns:

The modified figure with all visual elements drawn.

Return type:

Figure | Figure

property required_axes: set[str]

Return the set of all axes names required by visualizers and components.

Returns:: Set of subplot names needed for drawing.
Return type:: set[str]

set_layout(layout)[source]

Set the layout strategy for this plotter.

Parameters:

layout (ILayoutStrategy | str) –

Layout strategy to use. Can be:

An ILayoutStrategy instance
One of: “vertical”, “split”, “dashboard-lite”, “dashboard”

Returns:

Returns self to allow method chaining.

Return type:

Self