EEGLAB version 9 revision history

From SCCN
Jump to: navigation, search

This version is currently no longer supported.

Important note: computing ERSP in STUDY seems to be causing problem. Use more recent version instead. This problem will not be fixed in this version.

Important note: ERP polarity in STUDY seems to be problematic sometimes. This problem was fixed in version 11 and 12 and will not be fixed for this version.

Contents

EEGLAB rev. 9.0.8.6 (stable)

  • Issue date: September 13, 2011
  • SVN revision 9394 branch https://sccn.ucsd.edu/svn/software/branches/eeglab9
  • External modules: Fieldtrip (partial) SVN version 1919, BIOSIG (partial) SVN version 2540
  • Changes compared to previous version:
    • pop_erpplot.m, fix plotting scalp topographies at the STUDY level; pop_readersp.m, fix subtracting common ERSP baseline at the STUDY level; std_plotcurve.m, minor fix for legends (SVN 9393 - Arno)
    • Update revision number (SVN 9394 - Arno)

EEGLAB rev. 9.0.7.6 (stable)

  • Issue date: August 27, 2011
  • SVN revision 9376 branch https://sccn.ucsd.edu/svn/software/branches/eeglab9
  • External modules: Fieldtrip (partial) SVN version 1919, BIOSIG (partial) SVN version 2540
  • Changes compared to previous version:
    • pop_loadstudy.m & std_ersp.m, fix ICA ERSP computation bug when component have been preselected (SVN 9374 - Arno)
    • pop_precom.m, fix computing files when channels are interpolated - only some designs affected (SVN 9375 - Arno)
    • Update revision number (SVN 9376 - Arno)

EEGLAB rev. 9.0.6.6 (stable)

  • Issue date: August 26, 2011
  • SVN revision 9371 branch https://sccn.ucsd.edu/svn/software/branches/eeglab9
  • External modules: Fieldtrip (partial) SVN version 1919, BIOSIG (partial) SVN version 2540
  • Changes compared to previous version:
    • std_readerp.m, std_readfile.m, fixing reading interpolated channels (SVN 9369-9373 - Arno)
    • Update revision number (SVN 9372 - Arno)

EEGLAB rev. 9.0.5.6 (stable)

  • Issue date: July 27, 2011
  • SVN revision 9170 branch https://sccn.ucsd.edu/svn/software/branches/eeglab9
  • External modules: Fieldtrip (partial) SVN version 1919, BIOSIG (partial) SVN version 2540
  • Changes compared to previous version:
    • std_readersp.m, std_readerp.m, fixing reading channels when they are not saved in the same order as the channel location structure. For more information see Interpolation_bug_July_2011 (SVN 9345 - Arno)
    • Update revision number (SVN 9346 - Arno)

EEGLAB rev. 9.0.4.6 (stable)

  • Issue date: March 18, 2010
  • SVN revision 9170 branch https://sccn.ucsd.edu/svn/software/branches/eeglab9
  • External modules: Fieldtrip (partial) SVN version 1919, BIOSIG (partial) SVN version 2540
  • Changes compared to previous version:
    • eeglab.m, fix call to pop_rejchan.m to avoid double pop-up of new data window (SVN 9170 - Arno)

EEGLAB rev. 9.0.4.5 (stable)

  • Issue date: December 26, 2010
  • SVN revision 9121 branch https://sccn.ucsd.edu/svn/software/branches/eeglab9
  • External modules: Fieldtrip (partial) SVN version 1919, BIOSIG (partial) SVN version 2540
  • Changes compared to previous version:
    • pop_studydesign.m, disabling the 'combine' push button because the combine feature is not fully functional (SVN 9118 - Arno)
    • std_plotcurve.m, fix minor bug plotting curve together for groups (SVN 9119 - Arno)
    • Contents.m, change revision number to 9.0.4.5 (SVN 9121 - Arno)

EEGLAB rev. 9.0.4.4 (stable)

  • Issue date: December 7, 2010
  • SVN revision 9114
  • External modules: Fieldtrip (partial) SVN version 1919, BIOSIG (partial) SVN version 2540
  • Changes compared to previous version:
    • std_readerp.m, std_erpplot.m, fix reading single-trial ERP from one or two subjects (SVN 9087 - Arno)
    • std_spec.m, fix computing spectrum for merged dataset and interpolated channels (SVN 9088 - Arno)
    • std_erp.m, fix computing ERP for interpolated channels (SVN 9089 - Arno)
    • importevent.m, fix reading file from the command line and setting 'init_time' and 'init_type' fields (SVN 9090 - Arno)
    • pop_chanedit.m, fix not remembering the plotting radius (SVN 9091 - Arno)
    • eeglab2fieldtrip.m, fix performing dipole fitting on subset of channels (SVN 9092 - Arno)
    • eeglab.m, leaving SIFT plugin menu activated when multiple datasets are selected (SVN 9093 - Arno)
    • std_getindvar.m, fix processing numerical independent variable (SVN 9094 - Arno)
    • std_checkset.m, fix converting old STUDY format for independent variable pairing (SVN 9095 - Arno)
    • std_plotcurve.m, use IIR filtering instead of FIR filtering to avoid baseline offset problem (SVN 9096 - Arno)
    • std_makedesign.m, fix old STUDY design "filebase" field issue where STUDY design with index of more than 1 could use the old file name legacy format (without the "design" prefix) (SVN 9097 - Arno)
    • pop_studydesign.m, GUI text fix (SVN 9098 - Arno)
    • std_checkset.m, added check for the presence of ICA components (in case a user compute ICA within a STUDY) (SVN 9099 - Arno)
    • std_preclust.m, fixed checking for a "complete" STUDY when doing pre-clustering (SVN 9100 - Arno)
    • ft_electroderealign.m, fixed assignment between dissimilar structures - this function (unlike other Fieldtrip function) is copied from the Fieldtrip repository, the function has also been fixed in Fieldtrip (SVN 9101 - Arno)
    • coregister.m, fix warping to match new Fieldtrip function calling convention, and put warning for fiducial alignment since it could not be fixed (SVN 9102 - Arno)
    • std_plotcurve.m, add warning message for plotting standard error (SVN 9103 - Arno)
    • pop_precomp.m, disabling channel selection because of some issues reported in bug 981 (SVN 9104 - Arno)
    • eeg_mergelocs.m, fix merging channel location structures with dissimilar fields - added function eeg_mergelocs_diffstruct (SVN 9105 - Arno)
    • plotcurve.m, remove Latex interpretation of legends (SVN 9106 - Arno)
    • epoch.m, take the lower sample of data epoch event (instead of round) for non integer latencies to avoid having the event outside the limit from the data epochs (SVN 9107 - Arno)
    • std_spec.m, remove warning when computing FFT with uneven epoch length (SVN 9108 - Arno)
    • std_checkset.m, fix the check for the presence of ICA components, added new warning for non uniform sampling rate and check for channels without labels bug 987 (SVN 9109 - Arno)
    • std_plotcurve.m, fix plotting single subject in topographic format (SVN 9111 - Arno)
    • Contents.m, change revision number (SVN 9112 - Arno)
    • loc_subsets.m, fix axis aspect ratio (SVN 9113 - Nima)
    • creating a branch for EEGLAB 9 stable release (SVN 9114 - Arno)

EEGLAB rev. 9.0.3.4b

  • Issue date: Nov 18th, 2010
  • SVN revision 1986
  • External modules: Fieldtrip (partial) SVN version 1919, BIOSIG (partial) SVN version 2540
  • Changes compared to previous version:
    • supergui.m, fixing figure aspect ratio for Mac intel 64-bits (SVN rev. 9082 - Arno)
    • newtimef.m, fixing [min max] baseline (SVN rev. 9083 - Arno)
    • std_ersp.m, fix missing indices for ERSP interpolated channels that was generating some crash - if no crash was observed the results were correct (SVN rev. 9084 - Arno)
    • Contents.m, update revision number before release (SVN 1985 - Arno)
    • std_selectdataset.m, fix processing numerical data (SVN 1986 - Arno)

EEGLAB rev. 9.0.3.3b

  • Issue date: Nov 15th, 2010
  • SVN revision 1981
  • External modules: Fieldtrip (partial) SVN version 1919, BIOSIG (partial) SVN version 2540
  • Changes compared to previous version:
    • eeglab.m, fix typo for detecting operating system (SVN rev. 9043 - Arno)
    • eeg_checkset, fix header typo (SVN 9044 - Arno)
    • gradmap, check if "imresize" from image processing toolbox is present (SVN 9045 - Arno)
    • updating the 4 Fieldtrip external functions that have to be updated manually (SVN 9046 to 9049 - Arno)
    • eegplot.m, fix problem when using very small scale (SVN rev. 9050 & 9051 - Arno)
    • pop_subcomp.m, fix rejecting default components when using pure command line call (SVN 9052 - Arno)
    • std_chaninds.m, better error message (SVN 9053 - Arno)
    • pop_loadstudy.m, automatically recompute STUDY trial information (SVN 9054 - Arno)
    • std_checkset.m, check trial information consistency across datasets and recompute if necessary (SVN 9055 - Arno)
    • std_erp.m and std_spec.m, fix inconsistent channel labels in data file when selecting a subset of channels (SVN 9056 - Arno)
    • biosig2eeglab.m, fix problem when status channel index not in channel location structure (SVN 9057 - Arno) bug 925
    • std_precomp.m, fix channel computation for ERSP design when channels are not interpolated (SVN 9058 - Arno) bug 924
    • std_spec.m, fix reading already precomputed channel data (obsolete functionality - for backward compatibility only) (SVN 9059 - Arno) bug 923
    • eeg_decodechan.m, fix decoding integer channel indices and auto-detection of integer channel labels (SVN 9060 - Arno) bug 926
    • pop_select.m, allow selecting time range in epoched dataset with no event and add warming message (SVN 9061 - Arno) bug 926
    • dipplot.m, fix minor coordintate auto-detection issue (SVN 9062 - Arno) bug 931
    • loadcnt.m, fix reading unrelevant data at the end of some file (SVN 9063 - Arno) bug 918
    • pophelp.m, removing debug text output (SVN 9065 - Arno) bug 917
    • tftopo.m, adding suport for 5 dimentional inputs thanks to Joshua Balsters (SVN 9066 - Arno) bug 914
    • compile_eeglab.m, fix to handle new external version of Fieldtrip (SVN 9067 - Arno) bug 916
    • timewarp.m, replacing function in timefreqfunc by the function in sigprocfunc and removing the first one (SVN 9068 (replace) & 9064 (remove) - Arno) bug 932
    • tftopo.m, undoing change in SVN 9066 (SVN 9069 - Arno) bug 914
    • std_readersp.m, fix inconsistent ERSP optional common baseline removal removing common baseline only for independent variables that have been paired (SVN 9070 - Arno)
    • pop_autorej.m, nogui set to 'on' by default so it does not pop up when calling the function from the command line (SVN 9071 - Arno)
    • std_createclust.m, fix name and index when creating new cluster (SVN 9072 - Arno)
    • eeg_getdatact.m, edit error message (SVN 9073 - Arno)
    • eeg_store, pop_saveset, fixed having to resave a data file twice when modifying it in a STUDY (SVN 9074 - Arno)
    • fix VidEd plugin for manual event insertion (SVN 1975-1978 - James Desjardin)
    • Contents.m, update revision number before release (SVN 1979 - Arno)
    • std_checkset.m, check that datasets with the same ICA decompositions have uniform dipole models (SVN 1980 - Arno)
    • std_editset.m, pop_study.m, fix the 'inbrain' option for dipole selection (SVN 1981 - Arno)

EEGLAB rev. 9.0.2.3b

  • Issue date: October 7th, 2010
  • SVN revision 9042
  • Major bug fixes compared to previous version:
    • mri3dplot, fix horizontal white line when plotting specific slices
    • std_precomp, allow STUDY to have non-uniform datasets with different epoch lengths (EEGLAB still enforces specific designs to have the same epoch length)
    • pop_newcrossf, fix coherence between ICs (bug introduced in version 9.0.0.1b)
    • pop_prop, does not require any more the field EEG.reject.gcompreject to be set
    • readegilocs, fix reading file GSN-HydroCel-257.sfp (was absent from repository)
    • eegplot, fix plotting memory mapped data when the mean is subtracted
    • pop_runica, fix merging datasets when running ICA from a STUDY (thanks Jan Kacper Kamiński)

EEGLAB rev. 9.0.2.2b

  • Issue date: September 23rd, 2010
  • SVN revision 9032
  • Major bug fixes compared to previous version:
    • pop_envtopo, fix accessing data when comparing 2 datasets
    • std_precomp, fix component indices for ICA cluster ERP and spetrum computation (some ICA component might have been missing but the computation was accurate)
    • std_erpplot and std_erspplot fix pairing bug leading comparisons to always use unpaired statistics. This was true for both ERP, ERSP, ITC and spectrum plotting. The command line information was indicating correctly which type of statistics was been used. This bug is specific to version 9 and was not present in version 8.
    • eeg_decodechan, fix selecting channel labels which are numerical
    • eeglab function, now add path to eeglab/functions for memmapdata object
  • Major improvements compared to previous version:
    • pop_readegi and readegilocs accept custom channel location file. pop_readegi allows to select a location file in the GUI.
    • pop_saveset, new option to allow saving big files in Matlab 7.3 file format

EEGLAB rev. 9.0.1.2b

  • Issue date: September 21th, 2010
  • SVN revision 9015
  • Major bug fixes compared to previous version:
    • erpimage, avoid rare error when statistics toolbox is missing
    • newtimef, enforce symmetry of ERSP axis by default
    • dipoledensity, fix the function folder detection and the associated and the Bug in Fieldtrip (Fieldtrip repository was updated accordingly).
    • pop_eegplot, fixed electrode range for ICA components leading to wrong component indices being displayed
    • std_readersp, fix common ERSP baseline subtraction (when selected in GUI)
    • pop_readegi, now saves the channel location filename in the "chaninfo" substructure
  • Major improvements compared to previous version:
    • std_plotcurve, added possibility to plot difference and standard error (command line only)
    • std_preclust, better detection of incomplete or inconsistent studies
    • added function std_dipoleclusters
    • allow numerical types in function std_getindvar
    • eegplot speedup by a factor of 3

EEGLAB rev. 9.0.0.2b

  • Issue date: September 7th, 2010
  • SVN revision 8991
  • Major difference with previous version:
    • pop_clustedit, fix scalp map button not enabled
    • topoplot, fix scalp map not plotted in color
    • minor header edits

EEGLAB rev. 9.0.0.1b

  • Issue date: September 1st, 2010
  • SVN revision 8989
  • Major difference with previous version:
    • pop_editset fix for reading binary data bug 826
    • scanfold.m, can now process "functions" folder (there was a name conflict with Matlab)
    • setfont.m, fix setting specificaly specific handles (labels or title)
    • rmart.m, fix header and processing channel labels bug 848
    • pop_precomp.m, std_precomp.m, std_erp.m, std_spec.m, std_ersp.m, fix removing component cluster from data
    • std_selcomp.m, fix selecting the component for plotting
    • runpca2.m, fix keeping return n principal components bug 850
    • pop_select.m, fix selecting channels by indices from the command line bug 878
    • eegplot.m, fix non functional zoom option for events bug 879
    • loadtxt.m, fix numerical conversion and reading data sheet with empty cells bug 828
    • biosig2eeglab.m, fix channel location conversion bug 829
    • pop_rejepoch.m, fixed problem when all trials are selected bug 832
    • pop_rmdat.m, now supports selecting multiple event types bug 833
    • finputcheck.m, fix processing empty argument when limits are set
    • gauss3d.m, fix 3-D standard deviation calculation crash bug 846
    • rejkurt.m, fix normalize option for 3-D data bug 852
    • pop_runica.m, now puts channel type selection in command history bug 837
    • fdr.m, fix minor numerical problem bug 838
    • pop_editset.m, fix read float 32 data bug 826
    • pop_iirfilt.m, now supported for STUDY (can filter all the datasets of a STUDY) bug 826
    • pop_plotdata.m, fix abnormal figure position arrangement
    • minor issues or help inconsistency fixed in rejkurt.m, gauss3d.m, histtoexe.m, remart.m, mapcorr.m, runpca2.m, rmbase.m, eegh.m, getkeyval.m, eegplotgold.m, eeg_lat2point.m, eeg_matchchans.m (bugs 841 to 861).
    • pop_chanedit.m, fixing "EEG.chaninfo" update for command line calls bug 838
    • eeg_epochformat.m, now can export event structure as a table. We will soon add a menu to EEGLAB to support this functionality bug 862
    • eeglab2fieldtrip.m, fix crash for datasets with no channel locations bug 872
    • runica.m, fix slow down because of a misplaced "drawnow" command bug 881
    • std_setcomps2cell, generate error (as an option) if cell in design do not intersect with any of the datasets given as input
    • eeg_checkset, fix missing ".ref" field for datasets from very old versions of EEGLAB
    • std_ersp, fix giving channel labels as input
    • std_precomp, enforcing to use a design that reaches all datasets to perform pre-clustering
    • std_makedesign, removing spurious character in subjects and independent variable values so they can be used in filenames
    • pop_rejchanspec, allow for upper and lower threshold
    • eeg_chaninds, allow ignoring missing channels
    • pop_rmbase, fixed problem when using empty data sample point range
    • VisEd plugin, fixed adding events manually to dataset with no pre-existing events
    • fmrib plugin, perform filtering in double precision and recompiled 64-bit mex files
    • std_spec, allow concatenating continuous datasets in STUDY and processing boundaries of the concatenated datasets
    • eeg_interp, do not reorder channels when interpolating a single channel and fix the history command
    • pop_interp, fix the history command
    • pop_resample, now checks event consistency after resampling
    • pop_chanedit, fix command line call problem when using code from history
    • spectopo, remove warning for not rounded boundaries
    • eeg-dipselect, fix call to new version of sourcedepth in Fieldtrip
    • std_preclust, fix error when processing continuous data and when processing a design with more than 1 condition
    • all admistrative STUDY function, allow processing independent variables with numerical values
    • about 15 EEGLAB pop function were updated to use the eeg_getdatact function to access ICA components
    • erpimage now accepts input of the size 1 x samples x trials
    • eeg_getdatact now accepts "samples" and "reshape" as input
    • eeglab function has updated help message and help menu
    • help functions now do not use the "text" variable any more (this was generating crashes in very specific circumstances).
    • eeg_getdatact does not crash for accessing ICA activities for a dataset that has not been saved
    • timefreq.m; Because of an error to speed up the time-frequency decomposition functions, FFT frequency have been slightly inaccurate from the period of June 2009 to this present release. This problem only occurred when selecting custom window size (not using the default window size which is a power of 2). Frequencies where overestimated by a factor of 0% to 50% (the exact number depending on the size of the window). You need not be concerned by this bug unless you selected manually a window size which was not a power of 2 (thanks to Wu Quian). bug 874
    • statcond.m; A problem remains with the bootstrap function for paired data. The permutation function should be used instead. This problems arises because of resampling method we used. It leads to overestimation of significance by a factor of about 2 compared to permutation. We have not found a solution yet for this problem but have placed a warning when you choose to use bootstrap in the graphic interface. Bootstrap for unpaired data is functional and recommended (thanks to Erik Muelle).