Solution

pybasin.solution.Solution

Solution: Represents the time integration result for a single initial condition.

This class stores:

• The initial condition used for integration.
• The time series result from integration.
• Features extracted from the trajectory.
• Optional labels/classification for each trajectory.
• Optional model parameters that were used in the integration.
• Optional orbit data for orbit diagram plotting.

Attributes:

Name	Type	Description
initial_condition	Tensor	The initial condition used for integration (shape: B, S).
time	Tensor	Time points of the solution (shape: N).
y	Tensor	State values over time (shape: N, B, S).
features	Tensor | None	Filtered features used for classification.
extracted_features	Tensor | None	Original extracted features before filtering.
extracted_feature_names	list[str] | None	Names of extracted features.
feature_names	list[str] | None	Names of filtered features.
labels	ndarray | None	Labels assigned to each solution in the batch.
model_params	dict[str, Any] | None	Parameters of the ODE model.
orbit_data	OrbitData | None	Peak amplitude data for orbit diagram plotting.

Functions

__init__

__init__(
    initial_condition: Tensor,
    time: Tensor,
    y: Tensor,
    features: Tensor | None = None,
    labels: ndarray | None = None,
    model_params: dict[str, float] | None = None,
)

Initialize the Solution object.

Parameters:

Name	Type	Description	Default
initial_condition	Tensor	shape: (B, S) => B batches, S state variables	required
time	Tensor	shape: (N,) => N time points	required
y	Tensor	shape: (N, B, S) => N time points, B batches, S state variables	required
features	Tensor | None	Optional features describing the trajectory.	None
labels	ndarray | None	Optional classification labels for the solutions.	None
model_params	dict[str, float] | None	Optional dictionary of model parameters used in the simulation.	None

set_labels

set_labels(labels: ndarray)

Assign a label to this solution.

Parameters:

Name	Type	Description	Default
labels	ndarray	The label to assign from the classification results.	required

set_extracted_features

set_extracted_features(features: Tensor, names: list[str])

Store extracted features before filtering.

Parameters:

Name	Type	Description	Default
features	Tensor	Extracted feature tensor.	required
names	list[str]	List of feature names.	required

set_features

set_features(
    features: Tensor, names: list[str] | None = None
)

Store features extracted from the trajectory.

Parameters:

Name	Type	Description	Default
features	Tensor	A feature vector describing the solution (typically filtered).	required
names	list[str] | None	Optional list of feature names (for filtered features).	None

get_summary

get_summary() -> dict[str, Any]

Return a summary of the solution.

Returns:

Type	Description
dict[str, Any]	Dictionary with key information about the solution.

pybasin.utils.OrbitData dataclass

Data structure for orbit diagram plotting.

Stores peak amplitude information from steady-state trajectories, organized by degree of freedom (DOF). For period-N orbits, trajectories will have N distinct peak amplitude levels.

Attributes:

Name	Type	Description
peak_values	Tensor	Peak amplitudes for each DOF. Shape (max_peaks, B, D) where D is number of DOFs. Padded with NaN for trajectories with fewer peaks.
peak_counts	Tensor	Number of peaks per trajectory per DOF. Shape (B, D).
dof_indices	list[int]	Which state indices were analyzed.
time_steady	float	Time threshold used for steady-state filtering.