nirs4all.controllers.charts.exclusion module
ExclusionChartController - Visualizes excluded vs included samples.
This controller creates 2D scatter plots showing which samples have been marked as excluded by sample filtering operations. Useful for understanding filtering decisions and identifying patterns in excluded data.
- class nirs4all.controllers.charts.exclusion.ExclusionChartController[source]
Bases:
OperatorControllerController for visualizing sample exclusions.
Creates 2D scatter plots using PCA to show the relationship between included and excluded samples. Supports coloring by: - Exclusion status (included vs excluded) - Target values (y) - Exclusion reason
- Pipeline syntax:
“exclusion_chart” # Basic exclusion visualization
{“exclusion_chart”: {“color_by”: “y”}} # Color by target values
{“exclusion_chart”: {“color_by”: “reason”}} # Color by exclusion reason
- {“exclusion_chart”: {
“n_components”: 3, # Use 3D PCA “show_legend”: True, “title”: “Custom Title”
}}
- execute(step_info: ParsedStep, dataset: SpectroDataset, context: ExecutionContext, runtime_context: RuntimeContext, source: int = -1, mode: str = 'train', loaded_binaries: List[Tuple[str, Any]] | None = None, prediction_store: Any | None = None) Tuple[ExecutionContext, Any][source]
Execute exclusion visualization.
Creates a 2D (or 3D) scatter plot showing included vs excluded samples using PCA for dimensionality reduction.
- Parameters:
step_info – Parsed step containing operator and configuration
dataset – Dataset to visualize
context – Pipeline execution context
runtime_context – Runtime infrastructure context
source – Data source index (unused)
mode – Execution mode
loaded_binaries – Pre-loaded binaries (unused)
prediction_store – External prediction store (unused)
- Returns:
Tuple of (context, StepOutput with chart image)