API Reference

Model code imports the modeling API from cimba.sim and a few top-level helpers from cimba. The pages below group the public surface by area.

A model is declared as a Model subclass whose annotated fields describe the shape of one trial. Process behaviour is written as ordinary Python functions registered with @model.process and @model.collect, and experiments are built with model.experiment(...) and run with Experiment.run().

• Models, Components and Experiments
  • Model declarations
  • Components
  • Per-process fields
  • Process graphs
  • Experiments
• Processes and Events
  • Process verbs
  • Dynamic processes
  • Low-level events
• Queues, Resources and Conditions
  • Queues and resources
  • Stores, priority queues and conditions
• Datasets, Summaries and Reporting
  • Datasets
  • Time series
• Random Draws
  • Parameter Conventions
  • Continuous Draws
  • Discrete Draws
  • Probability Vectors
  • Seeds And Reproducibility
• Logging and Signals
  • Logging
  • Signal constants and casts
• Data-driven Traces

Top-level package

cimba.use_threads(n)

Return the number of worker threads Cimba will use.

The upstream library always runs one worker thread per logical CPU core. The n argument is accepted for API compatibility (0 means all cores) but is not passed through to the C library yet.

Parameters:

n (int)

Return type:

int

cimba.version()

Return the cimba library version string.

Return type:

str