-
Notifications
You must be signed in to change notification settings - Fork 5
Batch Mode
The batch mode is activated from SEV's main menubar.
Select 'Batch>Enter batch mode' or Ctrl+B.
This brings up the SEV batch mode dialog box. The first section allows you to select the EDF directory to load the EDF batch. The "EDF Directory" button allows you to select the pathname. "Output Settings" brings up another dialog box for the naming convention, classifier format, summaries, and target output directory for output of the batch process.
The Create Channels section loads channels. You can add as many channels as you desire by pressing the "Add Channel" button. "Config" (Rules for Synthesis) brings up the Filter Toolbox, described in the Single User Mode instructions. IMPORTANT: if you wish to export created images, check the "export" box next to the channel.
Event Detection and Artifact Detection allows selection of event/artifact detectors, similar to the event toolbox. Settings brings up the screen pictured below. The START column indicates the starting setting for the row parameter, the STOP is the endpoint, and NUM is the number of steps. For example, START of 1, STOP of 5, NUM of 3 would generate three output files per EDF with settings 1, 3, and 5. Here lies the power of the batch mode.
Power Spectrum allows selection of PSD or MUSIC power spectrum.
The batch mode will create a log file that outputs to the root pathname you supplied in the Output Settings above. It will be named '_log.txt' plus the date and time created. Information on number files the batch process was run on, event detectors run, artifact detectors run, and power spectral density run plus parameters for all. It then reports on successfully processed EDFs or errors encountered, and time needed to complete processes.
BE ADVISED that the SEV batch processor will automatically overwrite output files of the same type. For example, if you run the electrode pop artifact detector on a set of files on March 12, and run it again on on the same set of files on March 13, you will lose your March 12 data set.
Two ways to work around this: always designate a new root output directory in Output Settings. Or use the database feature.
A tab separated text file is generated for every EDF processed. Variables measured depend on the artifact/event detector used, but typically, start time of events, duration of events, epoch, and sleep stage are recorded. See M-file template section for specific fields returned.