Back to the main page.

Bug 1427 - allow egi_mff to be read when one of the files in the directory/package is given as filename

Reported 2012-04-16 11:06:00 +0200
Modified 2014-02-24 10:56:24 +0100
Product: FieldTrip
Component: fileio
Version: unspecified
Hardware: PC
Operating System: Mac OS
Importance: P4 normal
Assigned to: Robert Oostenveld
Depends on:
See also:

Robert Oostenveld - 2012-04-16 11:06:20 +0200

This is requested as solution for EEGLAB, which expects the user to click on a filename in the GUI. Solution: ft_read_header and ft_read_data both call dataset2files. If dataset2files knows how to deal with either the *.mff directory name or with the files inside that directory, it can return the correctly specified mff directory name.

Robert Oostenveld - 2012-04-16 11:19:53 +0200

dataset2files relies on ft_filetype to detect the file. So step 1 is to make that function smarter and detect most (or all) of the possible files in the mff package. I have started with a test script (test_bug1427) that goes over the desired implementation steps.

Robert Oostenveld - 2012-04-16 12:25:49 +0200

manzana> svn commit fileio/private/dataset2files.m fileio/ft_filetype.m test/test_bug1427.m Sending fileio/ft_filetype.m Sending fileio/private/dataset2files.m Adding test/test_bug1427.m Transmitting file data ... Committed revision 5629. The change involves ft_filetype from also detecting the files in the mff directory as egi_mff. If the user specifies one of the files in the mff directory, dataset2files will return the mff directory as header/datafile instead. TODO: - test the latest implementation with EEGLAB

Robert Oostenveld - 2013-09-24 14:34:03 +0200

I have updated the test script to use another mff dataset, as the one initially used contains multiple epochs. These epochs were initially not visible, but recent improvements to the egi_mff_v1 implementation made them visible (and explicitly unsupported). With another (more sane) mff dataset the test script runs fine, and ft_read_data/header can be used on any file in the mff directory. I take it that EEGLAB will also just work with it.

Robert Oostenveld - 2014-02-24 10:56:24 +0100

I closed several bugs at once that all have been resolved for some time. If you disagree, please reopen.