Resuming Experiments

You can resume a previously created experiment to continue analysis or rerun steps with different parameters.

How to resume

1. Set the Work Directory to the parent folder containing your experiment.
2. Select an existing Experiment Name — one whose folder contains a *_config.yml or config.yml file.
3. The button switches to Resume Experiment. Click it to load the experiment.

What gets loaded

When you resume, the GUI loads the experiment's config file and restores:

• Data directory and work directory paths
• Config template reference (or uses the experiment config if not set)
• Cores setting
• Segmentation values
• Script location

Each analysis tab also updates its state:

Tab	On Resume
Mesh Generation	Reads mesh settings from config. Checks for existing .ply, .surface.vtp, or .xyz files to determine completion status.
PyCurv	Resets to ready state. Repopulates the VTP file list by scanning the experiment directory.
Distance & Orientation	Resets to ready state. Loads distance and orientation measurement settings from config.

Rerunning jobs

When you run a job on an experiment that already has results, you're prompted to choose:

• Overwrite — Deletes old result files and runs fresh.
• Archive — Moves old files to a timestamped folder (e.g., results/archive_20241208_120000/) with a config snapshot, then runs fresh.
• Cancel — Aborts the run.

This prevents accidental data loss when iterating on parameters.