Visualizations.
Functions
execute_plot(**kwargs) -> Dict[str, Any]
Generate a visualization and return it as a saved PNG plus base64.
Backs the
tuiml_plot tool. Supported plot types: confusion_matrix, roc_curve, pr_curve, learning_curve, tree, feature_importance, cd_diagram, boxplot_comparison, heatmap, ranking_table.Parameters
plot_type
str
Which plot to produce (arrives via
**kwargs, like all parameters below).
title
str
= None
Custom plot title; each plot type has a sensible default.
model_id
str
= None
Trained model to plot from (model-based plot types).
model_path
str
= None
Explicit path to a serialized model file.
data
str
= None
Dataset to evaluate/plot against (dataset_id, path, or built-in name); required by data-driven plot types.
normalize
bool
= False
Normalize counts in the confusion matrix.
algorithm
str
= None
Algorithm name; required for
learning_curve.
benchmark_results
dict
= None
{algorithm_name: [scores...]} mapping; required for the comparison plots (cd_diagram, boxplot_comparison, heatmap, ranking_table).
Returns
result
dict
On success:
status ('success'), plot_type, description, path (saved PNG on disk), _image_base64 and _image_mime. On failure: status ('error'), error and optionally error_type / suggestion.