Notes on Scripting, Batching and Tasking for SPM5
- A large part of SPM code is complex housekeeping related to files and directories
- Tailoring requires attention to these housekeeping matters
Have I got the right files?
- Need to create novel suite of analyses out of a number of core modules
- Intercommunication via the SPM.mat structure
- Why Batch?
- Need to perform same or similar analyses on several datasets
- How to access them
- What they do
Query: how to get jobs saves a .m or .xml files as well as a .mat file? - These options could be useful as these are text files and easier to edit.
- Job/task manager can be extended
To be clarified: there is some barely undocumented functionality for linking up tasks in sequence so that output of one task become the input to the next. E.g. if you add a normalisation after creating a realignment you will see the option of choosing r* files for normalizing despite the fact that they do not yet exist!
- Other approaches