Basic processing workflows

The Pipeline Manager builds and runs a reproducible sequence of analysis functions on the active recording. Open a .dat or .umt file with an accessible Save Folder first.

The Pipeline tab opens the manager and summarizes the current input, pipeline state, and most recent result.

The app lists available functions on the left, shows the processing graph in the center, and displays the selected step and its parameters on the right. The main file, data, and run controls are along the top.

1. Open the Pipeline tab and choose Pipeline Manager.
2. Search the function list, select a compatible function, and choose Add.
3. If prompted, choose the source for each required input.
4. Select a graph step and review its options in the Parameters panel on the right; change them and choose Apply.
5. Add later steps in the desired order. Use Add as New Branch when two paths should process files independently. Zoom and pan the graph with the mouse wheel and drag, or choose Fit.
6. Choose Run, review the execution plan, and confirm.
7. At the end, the Pipeline Run dialog reports the result; choose OK, or See full summary for the details. Select an image result to open it in DataViewer. Save any temporary result needed later.

Running the same pipeline again skips the steps that are already done. To process several recordings at once, use File > Batch pipeline... (see Batch mode).

The current image file is the starting point shown in the graph; analysis functions are added from the searchable list.

What is saved?
Image results returned to DataViewer are temporary by default. Save any result that must remain in the Save Folder. Permanent outputs and the processing records in dataHistory.mat and pipeLog.mat are written by Pipeline Manager.

Review destructive functions.
applyRegistrationTformOnFolder rewrites every .dat file in place and is effectively irreversible.

Continue with the Pipeline Manager reference.