Using Local Python Files

All tasks in this package can also be executed outside of Fractal using standard Python files. This is useful for:

• testing and debugging
• running tasks locally
• integrating tasks into custom scripts or pipelines
• development and verification

Local execution relies on small Python files that act as “task launchers,” each containing all parameters required for a given task.

---

Procedure

Below is the recommended workflow for running tasks locally.

1. Copy the Python File Templates

   Template files are stored in the repository and kept up to date with each release.
   Each template includes a complete list of parameters for a specific task.

   To avoid accidentally editing the original files, you should copy the templates before modifying them.
   This only needs to be done once unless you want to refresh them later.

   cd path/to/mesospim-fractal-tasks
   copy-run-templates
   

   Or for Windows:

   cd C:\path\to\mesospim-fractal-tasks
   copy-run-templates
   

   This will generate editable files in the current directory, following the naming pattern:

   run_<taskname>.py
   

2. Edit Template File

   Before running the task, you need to edit the template file to set the parameters specific to your dataset. This includes the path to the OME-Zarr image to process for example.

3. Run the task

   Once the template file is edited, you can run the task using the following command:

   python run_<taskname>.py
   

Tip

In case the list of parameters has been compromised, you can reiterate the copying step to retrieve the latest version of the template files.