Skip to content

Autory v3.1.0

RunMultiple is much more flexible

autory run multiple can now read more than just workbooks:

  • YAML and JSON files.
  • Inline YAML and JSON with --config-json or --config-yaml.
  • YAML or JSON on stdin.

A new command autory convert run-multiple was added to convert RunMultiple configurations between workbook, YAML, and JSON formats.

This makes it much easier to keep RunMultiple configurations in version control, generate them from scripts, or move them in and out of Excel without rebuilding it by hand.

See Run multiple Autory models for the full schema, examples, and workbook format.

More details shown in terminals by default

Before, the default log level for terminal output was still set to WARNING instead of INFO in some places. The default is now INFO everywhere. A normal run therefore shows more useful progress messages without any extra setup.

This can still be changed with the --log-level command line option or the LogLevel RunControl option.

See how to view run logs and run control for more info.

Better audit trail for model files

Autory now saves a copy of the original model file next to the per-run .AutoryLog file in the run's logs folder. This makes it easier to see exactly which model was used for a past run.

Better memory usage and isolation

Each model run now starts in a fresh Python process. This improves isolation between runs and prevents memory leaks.

If you reuse results from previous runs, links back to precedent workbooks are now more reliable, especially when precedent paths are absolute or only partly available.

More vigilant about creating valid Excel files

If your model uses names that are forbidden in Excel, this is now caught earlier, which helps surface the issue sooner.

Documentation improvements