Statistical display

At the end of each simulated year, SimPaths produces year-specific summary statistics for post-simulation analysis. A selected subset is also displayed graphically while a simulation is running.

Output handling

SimPathsCollector collects and persists simulation outputs. When the graphical interface is enabled, SimPathsObserver updates the live charts and monitoring views.

The collector has separate switches for person, benefit-unit and household records, and for summary statistics covering income and wealth, demographics, alignment, labour, health and wellbeing. A displayed chart is not a substitute for saving the data needed for analysis.

Choosing outputs

For a YAML-configured run, use the collector_args section. For example, this fragment enables CSV export while omitting the three large entity-level tables:

collector_args:
  exportToCSV: true
  exportToDatabase: false
  persistPersons: false
  persistBenefitUnits: false
  persistHouseholds: false

Merge these settings into the existing collector_args block rather than creating a second block. Summary-statistic switches remain at their configured values. The documented options are in config/default.yml; use Multiple Runs to run a configuration, including a single repetition.

Before a large experiment, run a small population and check that:

1. The expected files appear in the run's output/ directory.
2. The saved years and output frequency match the experiment.
3. The required variables and population groups are present. Some statistics, such as Gini coefficients, have a separate calculation switch.
4. Enough detail has been retained for validation and subgroup analysis.

Entity-level exports can be large and may contain sensitive microdata. Retain them only in an appropriate research environment. Population persistence for restarting a run is a separate facility from collector export.

See Running your first simulation for the main run components and Saving outputs for the underlying JAS-mine output facilities.