Sample Tracer Module

The sampletracer class is the in-memory event store. It manages metadata, events, graphs, and selections organised into container objects (one per ROOT file/label pair), and provides the bridge from fully compiled event objects to the dataloader’s graph tensor dataset.

Class: `sampletracer`

Header: <generators/sampletracer.h>

Inheritance: tools, notification

Public Fields

Field	Type	Description
`output_path`	`std::string*`	Pointer to the output directory string (set by `analysis`).

Public Methods

Signature	Description
`bool add_meta_data(meta* meta_, std::string filename)`	Associates the `meta` object meta_ with filename. Returns `false` if an entry for filename already exists.
`meta* get_meta_data(std::string filename)`	Returns the `meta` associated with filename*, or `nullptr`.
`std::vector<event_template*> get_events(std::string label)`	Returns all events stored under dataset label label.
`bool add_event(event_template* ev, std::string label)`	Adds a fully compiled event to the container for label. Returns `false` if the event already exists.
`bool add_graph(graph_template* gr, std::string label)`	Adds a compiled graph to the container for label. Returns `false` if it already exists.
`bool add_selection(selection_template* sel)`	Adds a selection result. Returns `false` if it already exists.
`void fill_selections(std::map<std::string, selection_template> inpt)`	Populates inpt from each `container`’s stored selections.
`void populate_dataloader(dataloader* dl)`	Transfers all compiled `graph_t` objects to the dataloader’s internal dataset for training.
`void compile_objects(int threads, int intrath)`	Compiles events, graphs, and selections across threads parallel workers with intrath intra-op threads each.