erppcatoolkit-support — support list

Re: [Erppcatoolkit-support] ERP-components and PCA-components

[Joseph D. <jd...@ma...>]

The spatial step shouldn’t be splitting apart the effects in the way you describe.  I certainly don’t see anything like that in my N400 datasets, although from time to time I do get datasets where the PCA results are noisy, probably due to the dataset itself being too noisy (too few trials/subjects).  I’d have to see the data myself to say more.  For example, it may be that there are ERP components involved that you’re not familiar with.  But anyway, statistically the portion of the ERP component associated with the semantic effect should have the identical spatial distribution as the portion that does not and so they should end up being described by the same spatial factor.  I can’t think of any statistical mechanism by which they might be split off like that.

Regarding the PARE, when I use NetStation to preprocess, I do use it.  I’ve been meaning to add it to the Toolkit.  It’s a low priority because my sense is that it really doesn’t make much of a substantive difference to the results.  Nonetheless, I do intend to implement it at some point.

Cheers!

Joe


On Apr 10, 2014, at 3:40 AM, Petter Kallioinen <pet...@gm...> wrote:

> Hi Joe!
> I use your toolkit all the time. I like the structured 
> 
> I find that the temporal PCA-components usually have nice correspondance to candidate ERP-components (as you know N400 is my primary interest). However in the spatial decomposition things get problematic: Very often the first spatial component is the best in terms of topography and amplitude, but hardly ever show differences between conditions. Condition differences is usually I usually find in 2 to 4th spatial components, with topographies that does not look like typical ERP-components. (Later spatial components usually captures noise, remaining artifacts etc which is fine). So the mapping between the two-step PCAs and ERP components does not seem to be direct. N400-incongruency effects does not appear, in my data, as a centroparietal negativity. The typical N400-negativity is caputerd in the first component while, incongruency-effect is captured in components with less typical topographies. Is this stuff you recognize Joe? Should I evaluate topographical components as related
> subcomponents of the same underlying ERP-component rather than as independent ones? And, if the matching between PCA-components and ERP-components is more nested than direct (a unrecognized spatial PC evaluated as part of known and predicted temporal PC) do you have suggestions about how to should do the bonferroni correction?
> 
> A second question: What are your views on PARE-correction? I've seen you using it in articles, but it is not a part of the toolkit. 
> 
> Hope to have you in Stockholm again in the future!
> 
> Best regards!
> /Petter Kallioinen
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> ------------------------------------------------------------------------------
> Put Bad Developers to Shame
> Dominate Development with Jenkins Continuous Integration
> Continuously Automate Build, Test & Deployment 
> Start a new project now. Try Jenkins in the cloud.
> http://p.sf.net/sfu/13600_Cloudbees
> _______________________________________________
> Erppcatoolkit-support mailing list
> Erp...@li...
> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support


--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
Maryland Neuroimaging Center
University of Maryland 

E-mail: jd...@ma...
Phone: 202-297-8117
http://joedien.com

[Erppcatoolkit-support] ERP-components and PCA-components

[Petter K. <pet...@gm...>]

Hi Joe!
I use your toolkit all the time. I like the structured 

I find that the temporal PCA-components usually have nice correspondance to candidate ERP-components (as you know N400 is my primary interest). However in the spatial decomposition things get problematic: Very often the first spatial component is the best in terms of topography and amplitude, but hardly ever show differences between conditions. Condition differences is usually I usually find in 2 to 4th spatial components, with topographies that does not look like typical ERP-components. (Later spatial components usually captures noise, remaining artifacts etc which is fine). So the mapping between the two-step PCAs and ERP components does not seem to be direct. N400-incongruency effects does not appear, in my data, as a centroparietal negativity. The typical N400-negativity is caputerd in the first component while, incongruency-effect is captured in components with less typical topographies. Is this stuff you recognize Joe? Should I evaluate topographical components as related subcomponents of the same underlying ERP-component rather than as independent ones? And, if the matching between PCA-components and ERP-components is more nested than direct (a unrecognized spatial PC evaluated as part of known and predicted temporal PC) do you have suggestions about how to should do the bonferroni correction?

A second question: What are your views on PARE-correction? I've seen you using it in articles, but it is not a part of the toolkit. 

Hope to have you in Stockholm again in the future!

Best regards!
/Petter Kallioinen

Re: [Erppcatoolkit-support] Addressing a reviewer's comment on analyzing latencies

[Joseph D. <jd...@ma...>]

That’s right!  Of course it’ll only work if your two-step PCA can verify that it won’t be confounding multiple ERP components that have similar scalp distributions and if the spatial PCA is able to adequately separate the ERP components (it doesn’t work as well as a temporal PCA, let alone a two-step PCA).  This is further complicated by the fact that if there is sufficient latency jitter and sufficient factors retained, a temporal PCA will split an ERP component into multiple factors to account for the different time courses.  So you’ll have to try things out and then determine what conclusions you can justify.

On Apr 9, 2014, at 12:38 PM, Muhammad Adeel Parvaz <mp...@gm...> wrote:

> Thanks Joe. Just to clarify, by 'the latency analysis using windowed or spatial PCA data' you mean that we can do only a spatial PCA on our ERP waveform and identify a N200-like frontal negativity and record the latency of that factor for each subject?
> 
> Muhammad
>   
> 
> 
> On Wed, Apr 9, 2014 at 12:01 PM, Joseph Dien <jd...@ma...> wrote:
> As you say, it’s not possible to get latency effects from a temporospatial factor.  Now that you’ve isolated the ERP component, you could use the information about peak channel and latency to perform a latency analysis using windowed or spatial PCA data is one possibility.  You could also use the factor as a template for performing a Woody filter analysis (not a feature of the EP Toolkit at this point), although that has a number of caveats.  What would work best depends on the features of your data.
> 
> Cheers!
> 
> Joe
> 
> 
> On Apr 9, 2014, at 11:32 AM, Muhammad Adeel Parvaz <mp...@gm...> wrote:
> 
>> Dear Joe and other EPtoolkiters,
>> 
>> We are using EPtoolkit to identify frontal N200 and P300 components and we analyze the amplitudes both between and within groups.
>> However, a reviewer is now asking why we don't analyze latencies when many previous studies have shown latency differences. Since we have isolated temporospatial components, we don't have individual subjects' latencies. What would be the most effective way to address the reviewers' comment?
>> 
>> Thanks
>> Muhammad
>>   
> 
> 
> --------------------------------------------------------------------------------
> 
> Joseph Dien,
> Senior Research Scientist
> Maryland Neuroimaging Center
> University of Maryland 
> 
> E-mail: jd...@ma...
> Phone: 202-297-8117
> http://joedien.com
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 


--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
Maryland Neuroimaging Center
University of Maryland 

E-mail: jd...@ma...
Phone: 202-297-8117
http://joedien.com

Re: [Erppcatoolkit-support] Addressing a reviewer's comment on analyzing latencies

[Muhammad A. P. <mp...@gm...>]

Thanks Joe. Just to clarify, by 'the latency analysis using windowed or
spatial PCA data' you mean that we can do only a spatial PCA on our ERP
waveform and identify a N200-like frontal negativity and record the latency
of that factor for each subject?

Muhammad



On Wed, Apr 9, 2014 at 12:01 PM, Joseph Dien <jd...@ma...> wrote:

> As you say, it's not possible to get latency effects from a temporospatial
> factor.  Now that you've isolated the ERP component, you could use the
> information about peak channel and latency to perform a latency analysis
> using windowed or spatial PCA data is one possibility.  You could also use
> the factor as a template for performing a Woody filter analysis (not a
> feature of the EP Toolkit at this point), although that has a number of
> caveats.  What would work best depends on the features of your data.
>
> Cheers!
>
> Joe
>
>
> On Apr 9, 2014, at 11:32 AM, Muhammad Adeel Parvaz <mp...@gm...>
> wrote:
>
> Dear Joe and other EPtoolkiters,
>
> We are using EPtoolkit to identify frontal N200 and P300 components and we
> analyze the amplitudes both between and within groups.
> However, a reviewer is now asking why we don't analyze latencies when many
> previous studies have shown latency differences. Since we have isolated
> temporospatial components, we don't have individual subjects' latencies.
> What would be the most effective way to address the reviewers' comment?
>
> Thanks
> Muhammad
>
>
>
>
>
> --------------------------------------------------------------------------------
>
> Joseph Dien,
> Senior Research Scientist
> Maryland Neuroimaging Center
> University of Maryland
>
> E-mail: jd...@ma...
> Phone: 202-297-8117
> http://joedien.com
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>

Re: [Erppcatoolkit-support] Addressing a reviewer's comment on analyzing latencies

[Joseph D. <jd...@ma...>]

As you say, it’s not possible to get latency effects from a temporospatial factor.  Now that you’ve isolated the ERP component, you could use the information about peak channel and latency to perform a latency analysis using windowed or spatial PCA data is one possibility.  You could also use the factor as a template for performing a Woody filter analysis (not a feature of the EP Toolkit at this point), although that has a number of caveats.  What would work best depends on the features of your data.

Cheers!

Joe


On Apr 9, 2014, at 11:32 AM, Muhammad Adeel Parvaz <mp...@gm...> wrote:

> Dear Joe and other EPtoolkiters,
> 
> We are using EPtoolkit to identify frontal N200 and P300 components and we analyze the amplitudes both between and within groups.
> However, a reviewer is now asking why we don't analyze latencies when many previous studies have shown latency differences. Since we have isolated temporospatial components, we don't have individual subjects' latencies. What would be the most effective way to address the reviewers' comment?
> 
> Thanks
> Muhammad
>   


--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
Maryland Neuroimaging Center
University of Maryland 

E-mail: jd...@ma...
Phone: 202-297-8117
http://joedien.com

[Erppcatoolkit-support] Addressing a reviewer's comment on analyzing latencies

[Muhammad A. P. <mp...@gm...>]

Dear Joe and other EPtoolkiters,

We are using EPtoolkit to identify frontal N200 and P300 components and we
analyze the amplitudes both between and within groups.
However, a reviewer is now asking why we don't analyze latencies when many
previous studies have shown latency differences. Since we have isolated
temporospatial components, we don't have individual subjects' latencies.
What would be the most effective way to address the reviewers' comment?

Thanks
Muhammad

[Erppcatoolkit-support] ERP PCA Toolkit 2.43 release

[Joseph D. <jd...@ma...>]

Some stragglers that missed the 2.42 release.  I’m about to make some substantial feature additions and wanted to get these in before I start making more changes to the internal architecture of the Toolkit.

Cheers!

Joe

1) Adds mne toolbox to the path to avoid crashes when reading or writing FIFF format files. 
2) Fixed crash when loading file type that has internal channel names (like Neuroscan) and the only mismatch between it and the ced file is a single implicit REF channel or there is no mismatch and there is a single explicit REF channel.
3) Added .cov field to hold the channel covariance matrix generated during averaging for later inspection regarding channel quality.
4) Fixed when loading in a new file that had the same name as multiple existing files, appending dashed number to prior dashed number instead of replacing it (e.g., "name-1-2”).
5) Fixed crash when running an ANOVA on a PCA dataset results in trimmed cell means being added to it.
6) Fixed crash when selecting time points and there are empty event cells.
7) When the dataset name has been changed on the Overview page, Edit will ask if the unedited dataset should be kept in addition to the edited version when Done is pressed.
8) Further improvements to writing out FIFF format files.


--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
University of Maryland 

E-mail: jd...@ma...
Phone: 301-226-8848
Fax: 301-226-8811
http://joedien.com//












------------------------------------------------------------------------------
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/13534_NeoTech
_______________________________________________
Erppcatoolkit-support mailing list
Erp...@li...
https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support

[Erppcatoolkit-support] ERP PCA Toolkit 2.42 release

[Joseph D. <jd...@ma...>]

Mostly bugfixes to squash bugs caused by recent changes to the internal plumbing of the Toolkit and further refinements of existing features.  Also, improved support for reading MFF files, initial support for reading and writing FIFF files, new trimData function to graph and delete unwanted portions of continuous data prior to segmentation, new option to overplot all the trials or all the subject averages using the View Waves function, and new option to plot all the trials or all the subject averages using erpimages. 

1) In Scan function, fixed showing bad channel markings for first cell regardless of the currently displayed cell. 
2) Fixed crash in saccade correction when there is more than one bad channel. 
3) Fixed crash when reading mff file with more than one subject field. 
4) Added workaround for Matlab bug which has been causing screen size to periodically register as being zero. 
5) Fixed crash when stripping off single factors from combination factors, as in the Save function. 
6) Fixed crash when trying to topoplot data where electrode coordinate information is present but all coordinates are missing. 
7) Added option in Edit function’s Overview subpane to load in new electrode coordinates. 
8) When reading in boundary events in mff files, duration field contains the length of the recording pause. 
9) Added Trim Data option to the Segment Data function. 
10) Fixed crash when non-Mac user tries to open mff files by selecting a directory that does not contain any mff files or selects the mff file itself. 
11) Fixed crash when running a two-step PCA on a PCA file generated prior to version 2.40. 
12) Fixed aborting PCA when factor loadings slightly over 1 due to rounding errors. 
13) Fixed crash when reading ced file with REF channel type indicated. 
14) Fixed crash when using View with TFT data. 
15) Fixed crash when trying to plot frequency data in Topos with dB scaling where the power equals zero. 
16) Added View function option to plot or erpimage all trials and all subjects. 
17) pca no longer optional field for EP file format. No fields are optional. 
18) EP file fields now have standardized ordering to fix crash in Scan function when looking at files with different orderings. 
19) Fixed crash, as in scree plot, when a file has different types of adds in a category (e.g., GAV and AVG in subjects) 
20) Fixed PCA not correctly recognizing that a file has already undergone two-step PCA. 
21) Fixed subjects subpane of Edit function specifying average subject type as being AVE rather than AVG. 
22) Fix for not properly adding new blinks to the template for single trial data. 
23) Fixed REF channel type not being changed to EEG for files with one reference channel, resulting in some continuous files not being made available for forming blink templates. 
24) Fixed crash for EGIS average files with custom cell header lengths. 
25) Fixed when using single file mode to read in single-trial files, all the resulting files are identical to the very first subject. 
26) Handling decimal sampling rates more gracefully. 
27) Changed MEG channel type to MGA (axial gradiometer) and MGP (planar gradiometer) and MGM (magnetometer).  MEG support still not ready.
28) Fixed crash when reading mff file with subject field where the field was left blank. 
29) ced label for electrode coordinates provided by file (e.g., eeglab, MFF, FIFF formats) is "internal”. 
30) Added support for reading FIFF files. It's a pretty complex file format so this is only a first-pass implementation. 
31) Eliminated file type check for EGIS files since NetStation generates average EGIS headers that are incorrectly marked as being session files. 
32) Uses internal electrode coordinates provided by MFF and FIFF files and added preferences to automatically rotate such electrode coordinates to face upwards if needed. 
33) Improved saving data in text format so will check for overwriting and for FFT data will save as freq by chan. 
34) Added support for writing Neuromag FIFF file format. It's a pretty complex file format so this is only a first-pass implementation. 
35) Changed uses of "temp" as a variable name to "tempVar" due to other Matlab programmers often using it as a function name, resulting in collisions. 
36) Fixed crash when using "all" channels option in Windowing function and no channel groups have been defined yet. 
37) Added ability to window single-trial data. 
38) Fixed crash when editing cells using the Edit function. 
38) For PCA files, fixed recTime field not including space for the 'all' cell, resulting in crashes when edited. 
39) Combining of factors in Edit function now done as simple addition rather than as averaging. 
40) Fixed error if adding a factor combination to a PCA file with no combined factors. 
41) Windowing outputs cells in actual order rather than alphabetical order. 
42) Eliminated noTable option for old versions of Matlab. 
43) Fixed crash when segmenting or previewing. 
44) Fixed segmenting table + button mirroring the first line of the table rather than the current settings above the table. 
45) Fixed segmenting table - button deleting all but second to last line rather than just the last line. 
46) The events 'SESS','CELL','TRSP','bgin' are not excluded from being displayed for continuous files in the Display and the Scan functions when the evt option is checked, only for single-trial and average files. 
47) Fixed crash when appending channels in the Edit function. 
48) Fixed all but one channel is flat for grand average combined factors, as in the "all" factor from PCAs. 
49) Fixed all but one channel is flat for combined cells if one already has a combined factor, as when one uses the Edit function to combined cells on a factor cell containing an "all" cell. 
50) Fixed waveforms added during ANOVAs to correspond to trimmed cell means being all flat.
51) Average numbers, trial specs, and events carried over to the PCA file.

I’ve also committed changes to the FieldTrip I/O to fix mff files generated by NetStation 4.5.4 as they have nanosecond timing rather than microsecond timing and to fix a crash when there are no events in the user markup track. I’ve also made a fix to FieldTrip to fix a crash when loading in an EEGlab .set file with no events, so those wishing to read mff or EEGlab .set files should update to 1/31/14 or later.

--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
University of Maryland 

E-mail: jd...@ma...
Phone: 301-226-8848
Fax: 301-226-8811
http://joedien.com//

Re: [Erppcatoolkit-support] problem with EP Toolkit V 2.41

[Joseph D. <jd...@ma...>]

For some reason your computer thinks that “temp” is a function.
Could you type the following in your command window?

which('temp')


On Feb 5, 2014, at 7:36 AM, Haiobo Zhou <se...@gm...> wrote:

> Dear all:
> I get the error message type the  " ep" in matlab command 2009b under win7:
>        ??? Error: File: ep.m Line: 462 Column: 23
>        A function name ("temp") cannot be indexed with ".".
>  
> but it's OK in Version 2.23(11-Apr-11)
> what's the problem?
> thanks  in advance.
> ------------------------------------------------------------------------------
> Managing the Performance of Cloud-Based Applications
> Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
> Read the Whitepaper.
> http://pubads.g.doubleclick.net/gampad/clk?id=121051231&iu=/4140/ostg.clktrk_______________________________________________
> Erppcatoolkit-support mailing list
> Erp...@li...
> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support


--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
Maryland Neuroimaging Center
University of Maryland 

E-mail: jd...@ma...
Phone: 202-297-8117
http://joedien.com

[Erppcatoolkit-support] problem with EP Toolkit V 2.41

[Haiobo Z. <se...@gm...>]

Dear all:
I get the error message type the  " ep" in matlab command 2009b under win7:
       ??? Error: File: ep.m Line: 462 Column: 23
       A function name ("temp") cannot be indexed with ".".

but it's OK in Version 2.23(11-Apr-11)
what's the problem?
thanks  in advance.

[Erppcatoolkit-support] EP Toolkit 2.41 released

[Joseph D. <jd...@me...>]

EP Toolkit version 2.41 has been released:
https://sourceforge.net/projects/erppcatoolkit/

Aside from the usual bug fixes and refinements, notable upgrades include the addition of MRI artifact correction routines to preprocessing, a fix for the font sizes on Windows computers, and a segmentation function.  A notable bug fix was for the robust ANOVA function for between-group analyses.  My usual practice is to sort the between group identifier column alphabetically.  It turned out that when it was not sorted alphabetically, the subjects were not being assigned to the correct group, usually resulting in null effects.  My apologies for this bug.  I’m continuing to work with EGI to improve the support for mff files but they are still a work in progress.  Please read the notes at the end of this list if you are an OS X user.

**I’d also like to note that I am interested in opportunities for collaborating on grant submissions.**

Changes include:
1) When saving simple binary files with events, fractional sample durations rounded up.

2) Dropped setting bad one second epochs to 1000µv when saving simple binary continuous files.

3) Fixed crash when reading file with TRSP events that was not in mff format.

4) Fixed crash when translating single-trial EP file to EEGlab format and there are no events.

5) "boundary" in .type as well as .value fields for boundary events.

6) Added .keys field to events field to hold additional information, especially for mff files.

7) Added subject specs support for mff files.

8) Fixed EEGlab .set files not including study name in file name.

9) Fixes font sizes on Windows.

10) Fixed “About EP Toolkit” spawning new main window.

11) Fixed not checking to see if file name already exists when saving an EEGLab .study file.

12) For continuous data in Waves and Scan functions, baseline correct each channel by entire one second epoch so that waves will be visible.

13) + and - change scale buttons added to Scan function.  

14) Scan and Waves functions can now present event markings.

14) Fixed crash in view waves function when superimposing two datasets where one is missing channel coordinates

15) When channel type is changed using Edit function to REG, ANS, or ECG, electrode coordinates are set to missing.

16) Added Segment Data function.

17) Filtering will not cross boundaries.

18) .analysis edit fields updated when points dropped from a continuous file.  Boundary events added as needed.

19) Added fMRI artifact correction option to Preprocess data function

20) Fixed crash in blink correction when there is more than one bad channel.

21) Adds blink artifact channel and event marking peak latency of the blink in each epoch.

22) Blink correction checks for semi-singular data matrix, as from bridged channels, and drops dimensions as needed to improve quality of solution.

23) Averaging keeps all the events from the averaged data.

23) Adds saccade artifact channel and event marking onset latency of the saccade in each epoch.

24) In Scan function, added toggle to center data and a toggle to switch between clicking to mark bad channels and clicking to expand the waveform into a separate window.

25) Fixed frequency domain transform of continuous data not generating proper recTime field, resulting in later crashes.

26) Fixed crash in Topos function when trying to change scaling of frequency-domain data.

27) Preprocessing does not write over existing files.

28) Fixed noise and std fields set same as the data when combining cells or subjects.

29) Fixed crashes in preprocessing when there are multiple bad channels.

30) Fixed in Edit function added new subject or cell duplicating existing name resulted in corrupted file.

31) Fixed crash when regenerating cache and there is a spatial PCA in the active set.

32) Added windowing function option to window 'all' the channels.  

33) Fixed crash in Preprocessing when there are multiple global and trialwise bad channels when running under a version of Matlab earlier than 2013a.

34) Fixed crash when reading .set files using single file mode.

35) Fixed between group ANOVAs not being calculated correctly when the between group variable is not sorted alphabetically.

36) Fixed crash in Edit function when adding factors together.

37) Fixed crashes in Template function when there is a non EEG channel present.

38) Fixed crash in Template function when switching between datasets with different electrode montages or epoch length.

39) Fixed crash introduced in 2.40 when performing two-step PCA.

I’ve also committed changes to the FieldTrip I/O to allow user-added custom EEGlab event fields to be read so the FieldTrip copy should be updated to 11/1/13 or later.  Also, when I upgraded to OS X 10.9 (Mavericks), I started getting a lot of momentary freezing of the Matlab user interface. It appears that a number of third-party packages are incompatible.  My own issues were solved using the Console utility, which identified RSSbot as having an endless stream of error messages, and the Activity Monitor utility, which identified Safari as having a lot of hung subprocesses, which I in turn tracked to the Flash Player plugin.  Also, upgrading XQuartz to 2.75 (which the Mavericks installer does not do for you) made a big difference for Matlab specifically.


-------------------------------------------------------------------------------- 

Joseph Dien,
Senior Research Scientist
Maryland Neuroimaging Center
University of Maryland 

E-mail: jd...@ma...
http://joedien.com
 

 

Re: [Erppcatoolkit-support] different average conditions per participant

[Joseph D. <jd...@ma...>]

Oops!  I’ll have a fix out to you tomorrow.  I’m in the final stages of posting a new release and will include the fix as part of it.  Thanks for the report!

Joe

On Dec 28, 2013, at 4:33 AM, Maria Serena Panasiti <m.s...@gm...> wrote:

> Dear Joe,
> 
> it worked perfectly, thank you very much. Unfurtunaltely I am having a new problem: I run the first step temporal PCA and everything goes fine, when I then try to run the spatial pca on the temporal PCA factors I get this error:
> 
> ??? Reference to non-existent field 'recTime'.
> 
> Error in ==> ep_doPCA at 879
>     FactorResults.recTime=theData.recTime;
> 
> Error in ==> ep_doPCAst at 339
>     [stFactorResults] = ep_doPCA('asis', ROTATION, ROTOPT, MAT_TYPE, NUM_FAC, theData, LOADING);
> 
> Error in ==> ep>pickPCAdata at 5593
>         FactorResults = ep_doPCAst(theData.pca, 'UNRT', EPmain.pca.rotopt, PCArel, 1, PCAloading,PCAmode); %scree test only
>         needs unrotated solution
> 
> ??? Error while evaluating uicontrol Callback
> 
> 
> If I put a comment before this part of the ep_doPCA script:
> 
>  %FactorResults.recTime=theData.recTime;
> 
> then the the scree test works, and the spatial PCA (Infomax rotation) actually starts but at the very end I get this error:
> 
> Done with reconstructing PCA waveforms...
> ??? Index exceeds matrix dimensions.
> 
> Error in ==> ep_addToEPworkCache at 123
>                 minFac=min(min(min(EPdata.data(EEGchans,:,theCell,theSub,theFactor,:))));
> 
> Error in ==> ep_saveEPdataset at 75
> newDataset=ep_addToEPworkCache(EPdata);
> 
> Error in ==> ep>pickPCAdata at 5730
>     EPdataset=ep_saveEPdataset(EPdataset,PCAoutput,length(EPdataset.dataset)+1,'no');
> 
> ??? Error while evaluating uicontrol Callback
> 
> The same thing happens if i try to run the spatial PCA on the average data.
> 
> I am using Matlab 2007b but I have also try Matlab 2011
>                    Eeglab 8_3
>                    Fieldtrip lite last version available
> 
> I thank you in advance for your help and apologize if the question is too basic, I've tried to solve the problem by reading the tutorial but I couldn't figure it out.
> 
> Best wishes,
> 
> Serena
> 
> 
> 
> On 15 December 2013 07:51, Joseph Dien <jd...@ma...> wrote:
> I’m going to assume that you are using a file format like Neuroscan .AVG where each condition of each subject is a separate file.  If so, use the Single File option of the Read function to import them (see documentation).  It’ll form a single file in which each subject and each condition is separate but are all held in a single file for convenience sake.
> 
> Cheers!
> 
> Joe
> 
> 
> On Dec 14, 2013, at 4:56 AM, Maria Serena Panasiti <m.s...@gm...> wrote:
> 
>> Dear All,
>> 
>> I have just downloaded the toolkit to perform PCA on my averaged data. I am struggling to understand how I can keep my experimental conditions separeted when I read my data on the toolkit.
>> 
>> I have two experimental conditions for each subject. I glued all of the averages together in the edit window (append function); I gave the same subject name to the two experimental conditions averages and assigned them a weight of :
>> 
>> weight   name
>> 1            sub001   (This would be: AVg condtion1 subject1)
>> 1            sub001   (This would be: AVg condtion2 subject1)
>> 1            sub002    (This would be: AVg condtion1 subject2)
>> 1            sub002   (This would be: AVg condtion2 subject2)
>> 1            sub003   (This would be: AVg condtion1 subject3)
>> 1            sub003   (This would be: AVg condtion2 subject3)
>> 
>> 
>> 
>> When I run the PCA, the toolkit treats each raw as a single participant (in the command window it would count up to six in this case)
>> So I was wondering if this was the right way to feed the program with different averages per conditions.
>> 
>> As I understood PCA I should need to keep conditions separated but please tell me if I am wrong.
>> 
>> Any help would be very much appreciated.
>> 
>> Best wishes,
>> 
>> Serena
>> 
>> 
>> 
>> -- 
>> Maria Serena Panasiti, Ph.D
>> 
>> Cognitive Social and Affective Neuroscience Lab
>> Department of Psychology.
>> University of Rome "La Sapienza".
>> Via dei Marsi 78 - 00185 - Roma.
>> Phone: (+39) 06-49917635. Fax: (+39) 06-49917635
>> 
>> School of Psychology & Clinical Language Sciences
>> University of Reading 
>> Reading, United Kingdom 
>> 
>> ------------------------------------------------------------------------------
>> Rapidly troubleshoot problems before they affect your business. Most IT 
>> organizations don't have a clear picture of how application performance 
>> affects their revenue. With AppDynamics, you get 100% visibility into your 
>> Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
>> http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk_______________________________________________
>> Erppcatoolkit-support mailing list
>> Erp...@li...
>> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support
> 
> 
> --------------------------------------------------------------------------------
> 
> Joseph Dien,
> Senior Research Scientist
> Maryland Neuroimaging Center
> University of Maryland 
> 
> E-mail: jd...@ma...
> Phone: 202-297-8117
> http://joedien.com
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> -- 
> Maria Serena Panasiti, Ph.D
> 
> Cognitive Social and Affective Neuroscience Lab
> Department of Psychology.
> University of Rome "La Sapienza".
> Via dei Marsi 78 - 00185 - Roma.
> Phone: (+39) 06-49917635. Fax: (+39) 06-49917635
> 
> School of Psychology & Clinical Language Sciences
> University of Reading 
> Reading, United Kingdom 

Re: [Erppcatoolkit-support] different average conditions per participant

[Maria S. P. <m.s...@gm...>]

*Dear Joe,*

*it worked perfectly, thank you very much. Unfurtunaltely I am having a new
problem: I run the first step temporal PCA and everything goes fine, when I
then try to run the spatial pca on the temporal PCA factors I get this
error:*

??? Reference to non-existent field 'recTime'.

Error in ==> ep_doPCA at 879
    FactorResults.recTime=theData.recTime;

Error in ==> ep_doPCAst at 339
    [stFactorResults] = ep_doPCA('asis', ROTATION, ROTOPT, MAT_TYPE,
NUM_FAC, theData, LOADING);

Error in ==> ep>pickPCAdata at 5593
        FactorResults = ep_doPCAst(theData.pca, 'UNRT', EPmain.pca.rotopt,
PCArel, 1, PCAloading,PCAmode); %scree test only
        needs unrotated solution

??? Error while evaluating uicontrol Callback


*If I put a comment before this part of the ep_doPCA script:*

 %FactorResults.recTime=theData.recTime;

*then the the scree test works, and the spatial PCA (Infomax rotation)
actually starts but at the very end I get this error:*

Done with reconstructing PCA waveforms...
??? Index exceeds matrix dimensions.

Error in ==> ep_addToEPworkCache at 123

minFac=min(min(min(EPdata.data(EEGchans,:,theCell,theSub,theFactor,:))));

Error in ==> ep_saveEPdataset at 75
newDataset=ep_addToEPworkCache(EPdata);

Error in ==> ep>pickPCAdata at 5730

EPdataset=ep_saveEPdataset(EPdataset,PCAoutput,length(EPdataset.dataset)+1,'no');

??? Error while evaluating uicontrol Callback

*The same thing happens if i try to run the spatial PCA on the average
data.*

*I am using Matlab 2007b but I have also try Matlab 2011*
*                   Eeglab 8_3*
*                   Fieldtrip lite last version available*

*I thank you in advance for your help and apologize if the question is too
basic, I've tried to solve the problem by reading the tutorial but I
couldn't figure it out.*

*Best wishes,*

*Serena*



On 15 December 2013 07:51, Joseph Dien <jd...@ma...> wrote:

> I’m going to assume that you are using a file format like Neuroscan .AVG
> where each condition of each subject is a separate file.  If so, use the
> Single File option of the Read function to import them (see documentation).
>  It’ll form a single file in which each subject and each condition is
> separate but are all held in a single file for convenience sake.
>
> Cheers!
>
> Joe
>
>
> On Dec 14, 2013, at 4:56 AM, Maria Serena Panasiti <
> m.s...@gm...> wrote:
>
> Dear All,
>
> I have just downloaded the toolkit to perform PCA on my averaged data. I
> am struggling to understand how I can keep my experimental conditions
> separeted when I read my data on the toolkit.
>
> I have two experimental conditions for each subject. I glued all of the
> averages together in the edit window (append function); I gave the same
> subject name to the two experimental conditions averages and assigned them
> a weight of :
>
> weight   name
> 1            sub001   (This would be: AVg condtion1 subject1)
> 1            sub001   (This would be: AVg condtion2 subject1)
> 1            sub002    (This would be: AVg condtion1 subject2)
> 1            sub002   (This would be: AVg condtion2 subject2)
> 1            sub003   (This would be: AVg condtion1 subject3)
> 1            sub003   (This would be: AVg condtion2 subject3)
>
>
>
> When I run the PCA, the toolkit treats each raw as a single participant
> (in the command window it would count up to six in this case)
> So I was wondering if this was the right way to feed the program with
> different averages per conditions.
>
> As I understood PCA I should need to keep conditions separated but please
> tell me if I am wrong.
>
> Any help would be very much appreciated.
>
> Best wishes,
>
> Serena
>
>
>
> --
> Maria Serena Panasiti, Ph.D
>
> Cognitive Social and Affective Neuroscience Lab
> Department of Psychology.
> University of Rome "La Sapienza".
> Via dei Marsi 78 - 00185 - Roma.
> Phone: (+39) 06-49917635. Fax: (+39) 06-49917635
>
> School of Psychology & Clinical Language Sciences
> University of Reading
> Reading, United Kingdom
>
> ------------------------------------------------------------------------------
> Rapidly troubleshoot problems before they affect your business. Most IT
> organizations don't have a clear picture of how application performance
> affects their revenue. With AppDynamics, you get 100% visibility into your
> Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics
> Pro!
>
> http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk_______________________________________________
> Erppcatoolkit-support mailing list
> Erp...@li...
> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support
>
>
>
>
> --------------------------------------------------------------------------------
>
> Joseph Dien,
> Senior Research Scientist
> Maryland Neuroimaging Center
> University of Maryland
>
> E-mail: jd...@ma...
> Phone: 202-297-8117
> http://joedien.com
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>


-- 
Maria Serena Panasiti, Ph.D

Cognitive Social and Affective Neuroscience Lab
Department of Psychology.
University of Rome "La Sapienza".
Via dei Marsi 78 - 00185 - Roma.
Phone: (+39) 06-49917635. Fax: (+39) 06-49917635

School of Psychology & Clinical Language Sciences
University of Reading
Reading, United Kingdom

Re: [Erppcatoolkit-support] different average conditions per participant

[Joseph D. <jd...@ma...>]

I’m going to assume that you are using a file format like Neuroscan .AVG where each condition of each subject is a separate file.  If so, use the Single File option of the Read function to import them (see documentation).  It’ll form a single file in which each subject and each condition is separate but are all held in a single file for convenience sake.

Cheers!

Joe


On Dec 14, 2013, at 4:56 AM, Maria Serena Panasiti <m.s...@gm...> wrote:

> Dear All,
> 
> I have just downloaded the toolkit to perform PCA on my averaged data. I am struggling to understand how I can keep my experimental conditions separeted when I read my data on the toolkit.
> 
> I have two experimental conditions for each subject. I glued all of the averages together in the edit window (append function); I gave the same subject name to the two experimental conditions averages and assigned them a weight of :
> 
> weight   name
> 1            sub001   (This would be: AVg condtion1 subject1)
> 1            sub001   (This would be: AVg condtion2 subject1)
> 1            sub002    (This would be: AVg condtion1 subject2)
> 1            sub002   (This would be: AVg condtion2 subject2)
> 1            sub003   (This would be: AVg condtion1 subject3)
> 1            sub003   (This would be: AVg condtion2 subject3)
> 
> 
> 
> When I run the PCA, the toolkit treats each raw as a single participant (in the command window it would count up to six in this case)
> So I was wondering if this was the right way to feed the program with different averages per conditions.
> 
> As I understood PCA I should need to keep conditions separated but please tell me if I am wrong.
> 
> Any help would be very much appreciated.
> 
> Best wishes,
> 
> Serena
> 
> 
> 
> -- 
> Maria Serena Panasiti, Ph.D
> 
> Cognitive Social and Affective Neuroscience Lab
> Department of Psychology.
> University of Rome "La Sapienza".
> Via dei Marsi 78 - 00185 - Roma.
> Phone: (+39) 06-49917635. Fax: (+39) 06-49917635
> 
> School of Psychology & Clinical Language Sciences
> University of Reading 
> Reading, United Kingdom 
> 
> ------------------------------------------------------------------------------
> Rapidly troubleshoot problems before they affect your business. Most IT 
> organizations don't have a clear picture of how application performance 
> affects their revenue. With AppDynamics, you get 100% visibility into your 
> Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
> http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk_______________________________________________
> Erppcatoolkit-support mailing list
> Erp...@li...
> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support


--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
Maryland Neuroimaging Center
University of Maryland 

E-mail: jd...@ma...
Phone: 202-297-8117
http://joedien.com

[Erppcatoolkit-support] different average conditions per participant

[Maria S. P. <m.s...@gm...>]

Dear All,

I have just downloaded the toolkit to perform PCA on my averaged data. I am
struggling to understand how I can keep my experimental conditions
separeted when I read my data on the toolkit.

I have two experimental conditions for each subject. I glued all of the
averages together in the edit window (append function); I gave the same
subject name to the two experimental conditions averages and assigned them
a weight of :

weight   name
1            sub001   (This would be: AVg condtion1 subject1)
1            sub001   (This would be: AVg condtion2 subject1)
1            sub002    (This would be: AVg condtion1 subject2)
1            sub002   (This would be: AVg condtion2 subject2)
1            sub003   (This would be: AVg condtion1 subject3)
1            sub003   (This would be: AVg condtion2 subject3)



When I run the PCA, the toolkit treats each raw as a single participant (in
the command window it would count up to six in this case)
So I was wondering if this was the right way to feed the program with
different averages per conditions.

As I understood PCA I should need to keep conditions separated but please
tell me if I am wrong.

Any help would be very much appreciated.

Best wishes,

Serena



-- 
Maria Serena Panasiti, Ph.D

Cognitive Social and Affective Neuroscience Lab
Department of Psychology.
University of Rome "La Sapienza".
Via dei Marsi 78 - 00185 - Roma.
Phone: (+39) 06-49917635. Fax: (+39) 06-49917635

School of Psychology & Clinical Language Sciences
University of Reading
Reading, United Kingdom

[Erppcatoolkit-support] ERP PCA Toolkit 2.40 release

[Joseph D. <jd...@ma...>]

New release of EP Toolkit.  Mostly a maintenance release.  It fixes a number of bugs, especially in the Channel Group function for performing windowing.  Note especially the fix for the feature of using PCA results to choose channels for windowing.  It also expands support for EEGlab, ERPlab, BDF, and MFF file formats.  The MFF file format is still a work in progress and I am working with EGI and FieldTrip folks to get full support working in Matlab.  There was also a lot of work on plumbing behind the scenes to get ready for upcoming features.  Please be sure to let me know if any of these revisions has broken a function or if you have any feature requests.

Cheers!

Joe

1) Fixed dipole analysis in Topos function crashing due to changes in FieldTrip.
2) Fixed crash in Read pane when segmented simple binary session file is incorrectly specified to be an average file by the user.
3) Fixed combining channels into Regional Channel in Edit function or from windowing results in flat waveform.
4) Fixed View crashing if none of the data have electrode coordinates available.
5) Fixed Change Work Directory menu function not working.
6) Initializes mff file support at startup so that it doesn’t crash the toolkit down the line as workaround for Matlab bug.
7) Modified so panes fit on screens with less than 700 vertical pixels.
8) Fixed overview subpane of Edit function graying out number of factors.
9) Fixed not making factor data available for setting channel groups in windowing function when first in analysis set had different number of channels.
10) Fixed not asking for channel group name in windowing function when creating a new one via factor loadings and is currently the initial blank one.
11) Fixed not updating the display of the area names when using factor loadings to define the channel groups in the windowing function.
12) Fixed when using factor loadings to define channel groups in the windowing function, area names not reflecting factors defining them.
13) Fixed when using factor loadings to define channel groups, wrong channels could be chosen.
14) Worked around crash when reading mff files under Matlab 2013b by using alternative mff file format function.
15) When reading file, fixed incorrect inference of reference scheme when a single ref channel is designated.
16) When reading a file, fixed channel type not changed from REF to EEG for flexible channel order file formats.
17) Fixed REF channels assumed to be last for fixed order channel file formats.
18) For reading files, better support for MEG and ANS chan types, and BAD CED code.
19) Added support for mff files with recording stops and with segments.
20) Added support for boundary events (for continuous data where recording was stopped and then restarted).
21) Workaround for EEGlab issue where if a CED file has just the label and the type filled out, the type info migrates over to the theta column for some reason.
22) Changed event sample to count from start of epoch rather than FieldTrip convention of start of recording.
23) Improved detection of Simple Binary files with scrambled cells.
24) Fixed preprocessing function not finding files past the first when batched and they are not in the active directory.
25) Fixed PCA of continuous data generating error message.
26) Full support for ECG channel type.
27) Fixed Read function rejecting files if sampling rate and time names different past three decimals due to rounding errors.
28) Fixed crash when reading EEGlab file if cell names are a mix of numbers and strings.
29) Fixed crash when reading file if type field from CED file contains numbers for some reason.
30) Added recTime field to keep track of  recording time of an epoch with respect to the start of the recording session.
31) No longer rearranging single trial data to group by cell.
32) Eliminated offset field from events structure.
33) Fixed topoplot not showing the correct peak channel for non-factor data.
34) Better updating of history field.
35) Added display of trial specs to Scan function.
36) Fixed one-second epochs displayed with View function one sample longer than intended for continuous data.
37) Made a number of fixes to exporting of single-trial data in EEGlab format as it wasn’t working.
38) When reading EEGlab single-trial data, rejected channel and trial information is included.
39) When reading and writing ERPlab files, nTrials information is supported.
40) When reading ERPlab files, channel location information is supported.
41) Ensure that power field comes after analysis field to avoid crash in functions like scan when looking at multiple EP data files.
 
I’ve also committed changes to the FieldTrip I/O to fix problems with trying to load simple binary files (unsegmented files being treated as segmented, crash when simple binary file has no event track, and crash when there are event tracks but no events marked) and mff files (Matlab bug causes global variables to be erased crashing the toolkit and not computing event times correctly and not registering multiple segments correctly and not including events only within period of the epoch and not separating the trials of segmented data and adding better support for segmented files) so the FieldTrip copy should be updated to 10/15/13 or later.  Currently, mff average files cannot yet be read.
 
I’ve also reported a bug in EEGlab (still present as of 12_0_2b5) that has been affecting CED files.  CED files created with pop_editChans, as described in the tutorial, are defective (the columns are scrambled).  If you use this function to create a CED file, check the resulting file using a program like Excel to make sure that the columns are in the right place with respect to the headers.  If they do not match up, then move the columns so that they do.  The order of the column headers is correct so do not move the column headers, just everything below the headers.

--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
University of Maryland 

E-mail: jd...@ma...
Phone: 301-226-8848
Fax: 301-226-8811
http://joedien.com//

[Erppcatoolkit-support] ERP PCA Toolkit 2.38 release

[Joseph D. <jd...@ma...>]

New release of EP Toolkit due to serious bug affecting two-step PCA since 2.30.  It shouldn't affect ANOVA results but will affect topographic maps or waveforms depending on whether the second step was spatial or temporal respectively, amplitudes of effects (but not statistical significance), and source analyses.

My apologies.  Thanks to Hiroshi Nittono for reporting the problem.

Joe



1) Fixed bug introduced in 2.30 wherein variable standard deviations for two-step PCAs are applied incorrectly, resulting in inaccurate PCA reconstructions.  ANOVA results should be fine but topographic maps and source analyses for PCAs where spatial was the second step will be affected, as will waveform plots when the second step was temporal.  Thanks to Hiroshi Nittono for reporting this problem. 

2) Fixed windowing pane crashing when Herz bins changed.

--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
University of Maryland 

E-mail: jd...@ma...
Phone: 301-226-8848
Fax: 301-226-8811
http://joedien.com//

[Erppcatoolkit-support] ERP PCA Toolkit 2.38 release

[Joseph D. <jd...@ma...>]

New release of EP Toolkit.  Mostly bug fixes for better supporting Neuroscan files and for a number of crashing bugs in the Windowing function.  Note especially fix for windowing temporal PCA data as it would result in the measures being for the wrong channel!  I'm not sure how that one got past me so apologies for any trouble it may have caused.  Spatial and temporo-spatial (and spatio-temporal) PCAs are unaffected.

Cheers!

Joe



1) Fixed Single File Mode in Preprocessing pane crashing when values entered. 
2) Fixed Single File Mode in Preprocessing pane crashing. 
3) Fixed crashing when merging fixed channel files (like Neuroscan) where there are channels in the data that are not in the CED file. 
4) Fixed single file mode single-trial data files not being merged successfully. 
5) Added option to eliminate unwanted channels, as in a GFP channel, when reading in the data by marking the channel type as BAD in the CED file. 
6) When merging files, as in Single File Mode, channels not present in the initial file simply dropped from succeeding files rather than aborting the run. 
7) Single-Trial Files from multiple subjects can now be selected using the Read pane's Single File Mode and read in as separate files.  
8) Fixed crash when loading in a channel group file for the Window function and there currently isn't any channel group defined. 
9) Fixed crash when going into channel group subpane, cancel out without defining a channel group, and then return to the subpane. 
10) Fixed crash when windowing using the minpeak or maxpeak measures with the adjoining samples option and the peak latency was at the upper end of the window. 
11) Fixed crash when windowing and there are multiple channels in the area of a channel group. 
12) Fixed crash during windowing when factors were used to define the areas of a channel group. 
13) During windowing, fixed minpeak and maxpeak measures yielding missing data numbers when the window size was less than three samples, as in the autoPCA mode. 
14) Fixed Export button on Edit’s Factors subpane not working. 
15) When reading in text data files, multiple delimiters between values (as in space-space) now treated as a single delimiter.  
16) Fixed output of windowing being for channel 1 rather than the intended channel for non-spatial PCAs (spatial and temporo-spatial PCAs not affected).

--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
University of Maryland 

E-mail: jd...@ma...
Phone: 301-226-8848
Fax: 301-226-8811
http://joedien.com//

[Erppcatoolkit-support] ERP PCA Toolkit 2.37 release

[Joseph D. <jd...@ma...>]

New release of EP Toolkit.  Mostly improvements to reading data files, especially simple binary files (including important bug fixes to handling of average files).  Also enhancements to windowing with peak measures and to manual editing of bad channels in the Scan function.

Cheers!

Joe


1) Scaling of topos now obeys the values on the View pane.  Also, fixed manual changes to plotting range no longer working.
2) Fixed Simple Binary average files being scrambled when read!  EP Toolkit now accommodates loose Simple Binary file format specification rather than make assumptions about its internal structure (there is no formal documentation on it).
3) Better handles ced files where there are channel types other than EEG, FID, and REF.
4) Fixed choosing "auto" as Edit Mode setting in Preprocessing pane resulting in error message.
5) Added option to average together samples around a peak to minpeak and maxpeak measures.
6) Implemented Luck (2005) suggestion to only count as a dip/peak a sample where both neighboring samples are higher/lower.
7) When there is missing data (more likely now with new peak measure code), the NaNs are converted to missing data code specified in preferences setting.  8) Fixed Read function where if baseline was zero ms then instead it was being reported as being 4 ms.
9) Fixed detrend button in Preprocess Data pane not working.
10) Fixed warning when trying to read in Neuroscan files with two physically linked explicit reference sites.
11) Added option to the Trials subpane of the Edit function to load a text file to rename the cell names of all the trials.
12) Added support for reading EGI's epoch-marked simple binary format for session files.
13) Added table to Scan function which lists %age of bad channels and allows channels to be marked globally good or bad.
14) Fixed Scan crashing if the scaling is from zero to zero, as with a bad trial.
15) Fixed automatic global bad channel detection being performed even when editMode set to Manual.


--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
University of Maryland 

E-mail: jd...@ma...
Phone: 301-226-8848
Fax: 301-226-8811
http://joedien.com//

[Erppcatoolkit-support] ERP PCA Toolkit 2.36 release

[Joseph D. <jd...@ma...>]

New release of EP Toolkit.  Mostly bug fixes and refinements.  Note especially fix to calculation of difference and combination waves.  Enhancement to Topos view of non-PCA files is helpful for looking for effects.

Cheers!

Joe



1) In Topos function, fixed peak samples of ERP data being identified by amplitude rather than absolute amplitude.
2) In Topos function, added display of topos at every 50 ms for ERP and TFT data and every Hz for FFT data for non-factor data.
3) Fixed Topos not allowing two datasets to be shown in parallel when they have different regional channels.
4) Fixed Topos crashing when trying to display factor and non-factor data side by side.
5) In Edit function, fixed combining of subjects and chans not correct when weights not the same (as in difference wave).
6) In Edit function, fixed weighting of difference waves for cells and chans and subjects incorrect (waves too small).
7) In Window function, fixed windowed files being labeled as being in dB even when voltage data.
8) In Channels subfunction of the Window Data function, fixed crash when applying factor loadings and there is more than one PCA dataset in the working set.
9) Wave plots can now accommodate datasets with different sets of channels.
10) In Scan function of View pane, fixed crash when using secondary datasets with differing fields, such as PCA and not.
11) In Scan function of View pane, fixed crash when right shifting the cell and a secondary dataset is already at the maximum cell.
12) Improved the controls for the Windowing pane.
13) Improved handling of channel groups in the Channels function of the Windowing pane.
14) In Topos function, added peak point/Hz line to expanded waveform windows.
15) Markers in waveform plots can be set at zero ms.
16) For saved files in ERPlab format, fixed fields ERP.ntrials.rejected and ERP.ntrials.invalid to be vectors of zeros rather than empty set (due Joseph Orr).
17) Accommodated change in Matlab 2013a’s userpath function that, at least on a Mac, can cause it to lose track of the existing EPwork directory.

--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
University of Maryland 

E-mail: jd...@ma...
Phone: 301-226-8848
Fax: 301-226-8811
http://joedien.com//

[Erppcatoolkit-support] ERP PCA Toolkit 2.35 release

[Joseph D. <jd...@ma...>]

Just realized it's pretty easy to have the EP Toolkit automatically check for updates whenever it is started so just added this functionality.

--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
University of Maryland 

E-mail: jd...@ma...
Phone: 301-226-8848
Fax: 301-226-8811
http://joedien.com//

[Erppcatoolkit-support] ERP PCA Toolkit 2.34 release

[Joseph D. <jd...@ma...>]

Bugfix release.  Nothing major unless you ran into one of them.

1) Fixed error when setting file format preferences to ERPlab files.
2) Fixed failure to save preferences when save preferences button clicked.
3) Changed ms window information provided in header of windowing text files so it ranges from onset of sample to offset of sample rather than onset to onset (e.g., 0-4 ms rather than 0-0 ms for first sample).
4) Fixed ms window information on Window pane so that it ranges from onset of sample to offset of sample rather than offset to offset (e.g., 4-4 ms rather than 0-0 ms for first sample) after changes to the windowing settings.
5) Fixed running baseline correction even on frequency domain data even though baseline fields are grayed out.
6) Fixed time names not being calculated correctly when sampling rate changed using Edit function.
7) Fixed crash in View waves or when setting channel groups in Window function when all channels are along midline or center line.
8) Fixed crashes after Edit function used to trim the range of frequencies due to not applying trimming to std or FacVecF fields.
9) Allow View Scan function to operate on average files.
10) Now checking installation configuration only when EP first started.
11) Fixed peak channels not being identified correctly in Topos view.
12) In Edit function, fixed %age of blink, saccade, move, and bad trials in QC subpane being computed incorrectly for average data.
13) Fixed power field not being included in merged files, such as from Single File Mode, causing crashes elsewhere. 
14) Fixed crash when loading in .study average file where the .set files have no prestimulus period.
15) In Topos view, fixed crash when displaying data with a regional channel.
16) Fixed crash when data file has no event information.
17) Fixed crash in Topos view when displaying frequency data in dB scaling and the maximum value is negative.    

--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
University of Maryland 

E-mail: jd...@ma...
Phone: 301-226-8848
Fax: 301-226-8811
http://joedien.com//

Re: [Erppcatoolkit-support] ERP PCA Toolkit question

[Joseph D. <jd...@ma...>]

Hi,
   Thanks!  It is possible.  You should present the factor loadings in microvolt scaling though.  Anyway, to get the factor loadings, go to the Edit function.  Click on PCA.  Click on FacPat.  Click on Export.  You'll get a text file with what you need.  If you want to scale to microvolts, multiply each factor loading by the standard deviation of the variable.  You can get the microvolt scaled waveforms by typing load('PCA.ept','-mat') in the command window and getting the numbers from EPdata.facVecT (for temporal).  facVecS for spatial.  I'll see about adding an option to the Edit function so you don't need to mess with the command window.

Cheers!

Joe


On Feb 14, 2013, at 8:13 AM, Joseph Dien <jd...@ca...> wrote:

>  
>  
> From: Alberto Gonzalez [mailto:vil...@ho...] 
> Sent: Wednesday, February 13, 2013 7:34 AM
> To: Joseph Dien
> Subject: ERP PCA Toolkit question
>  
> Hello Mr. Dien,
> 
> I am currently using your ERP PCA Toolkit (which is great), and i need to obtain a picture of the loadings of the extracted factors (similar to the attached document). Has the toolkit any option to obtain such data?.
> 
> Thanks for your time.
> 
> Yours sincerely.
> 
> Alberto González.
> <Captura de pantalla 2013-02-13 a la(s) 13.28.50.png>


--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
University of Maryland 

E-mail: jd...@ma...
Phone: 301-226-8848
Fax: 301-226-8811
http://joedien.com//

[Erppcatoolkit-support] ERP PCA Toolkit 2.33 release

[Joseph D. <jd...@ma...>]

New release of EP Toolkit.  Major feature additions include addition of filtering options, support for manual editing of bad channels and trials, and support for ERPlab file format.  Continuing improvements in support of spectral data and ongoing bug fixes and refinements to existing features.

Cheers!

Joe



1) Added option to set last row to be imported in text files.
2) Fixed crash when changing the sampling rate in the Samples subpane of the Edit function.
3) Fixed subNames, subTypes, cellNames, cellTypes, facNames, and facType fields generated by single file mode and by PCA output and by Transform data not necessarily being column vectors, resulting in crashes in other parts of the Toolkit.
4) Fixed crash when invoking 2D plots in Topos view.
5) Fixed crash when using combining cells or chans for spatial PCA data.
6) Fixed Transform not updating reference type when rereferencing data.
7) Fixed missing markers in Wave plots when data range is less than one.
8) Fixed crash under certain circumstances when windowing data with multiple channels in a channel group.
9) Added option to do internal calculations of frequency data in either amplitude or power form.
10) Fixed erroneous error message and crash when trying to PCA average file where bad channels were dropped rather than replaced.
11) Fixed error message when expanding channel for waveform plot for which the data are no longer available.
12) Fixed not changing prestimulus period in Transform function when reference set to "none."
13) Added frequency filtering (low pass, high pass, bandpass, bandstop, notch) to Transform function.
14) Fixed subNames and subTypes field not being a column vector when averaging multiple files, resulting in no average file being saved.
15) Detects when PCA yields complex numbers due to rounding errors causing relationship matrix to be unsymmetric and makes the relationship matrix symmetric.
16) Since eigenvalue decomposition during PCA is not always in ascending order, sort them first.
17) Handles situation where FFT data has negative values (e.g., due to imprecision in PCA results) and transforming to dB would result in complex numbers, by taking absolute value first.
18) Events assigned to wrong subject (off by one) when reading in average files.
19) Added support for reading .set files generated by Widmann's pop_grandaverage function.
20) Added support for reading .erp files generated by ERPlab.
21) Fixed erroneous "labels" error message when trying to load .study file.
22) Fixed error message when applying PCA to data with only one cell.
23) Fixed problem that loading EP files with frequency PCA data results in damaged data file and error messages.
24) Fixed problem where information for expanding channels in waveform plot is lost under some circumstances.
25) Added option to contextual menu in Topos view to rescale figures according to selected topo map.
26) Fixed bug in reconstruction of PCA data from frequency-domain PCAs.
27) Added markers and expanding window to waveform figures in Topos.
28) Added manual scanning for bad channels and trials.
29) If baseline correction chosen, will occur even if blink correction not performed and it will be applied prior to global bad channel and trial detection.  Detrending now performed prior to global bad channel and trial correction and baseline correction.
30) Fixed crash when trimming low end Hz of spectral data using Edit function.
31) Fixed identifying all EGIS files as Hydrocel-128P under OS X 10.8 when using Satimage osx.
32) Added frequency-domain peak measures to windowing function.
33) Fixed crash when combining channels for factor data under certain circumstances (presence of facData due to adds).
34) When saving a dataset, name of dataset changes if new name is chosen for saved dataset.
35) Fixed a variety of issues with the automatic scaling in the topomap figures, especially for spectral data.
36) Fixed error message when generating PCA data from a single subject average.
37) Fixed crash in View pane under when changing a dataset under certain circumstances.
38) Marker fields in View pane no longer reset to blank whenever a change is made in the settings.
39) Clearing volt, hz, and sample parameters in View pane no longer crashes.  If value is manually set, will not change until a new dataset is chosen or until the value is cleared (in which case it will be replaced with automatic value).
40) Added writing ERPlab format files.


--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
University of Maryland 

E-mail: jd...@ma...
Phone: 301-226-8848
Fax: 301-226-8811
http://joedien.com//

[Erppcatoolkit-support] ERP PCA Toolkit 2.31 release

[Joseph D. <jd...@ma...>]

New mostly bugfix release.  Note especially #13, which fixed a bug that caused data corruption for continuous (unsegmented) data files when performing spectral analysis.  Also, if you're interested in learning more about PCA and how to use the ERP PCA Toolkit, I'm holding an all-day workshop on the topic at the Society for Psychophysiological Research conference in New Orleans in October.

1) Average function can now merge together files with different sets of cells.
2) Added option to combine cells in Edit function weighted by number of trials in each average.
3) In windowing function, fixed bad channel handling referring only to first channel in channel area.
4)  Fixed missing data numbers being transformed to psd and dB for frequency data.
5) Allow continuous files that have been frequency transformed to be averaged.
6) Fixed overwriting existing file with Transform function if the output file format differed from the input file format.
7) For the Edit function, fixed crash when examining std table of the QC subpane for frequency data.
8) Fixed edit's add cells trial weighting option not working correctly when the cells are not a consecutive series starting with the first.
9) For topo plots, fixed minimum voltage scaling being set to -1000 whenever minimum voltages are below 1000. 
10) Fixed crash when trying to add subject ANOVA waveform after computing robust ANOVA and subjects have been trimmed.
11) Fixed crash when trying to add subject ANOVA waveform after computing robust ANOVA and data is not from PCA output.
12) Fixed crash when loading older EP format files with factor data.
13) Fixed selecting incorrect time points when running spectral analysis on continuous data files (data corruption bug!!!!).
14) Improved ability to figure out the cell names of EEGlab files.
15) Fixed crash when combining multiple session files into one average file.
16) Fixed blink templates being saved as text files.
17) For windowing, fixed channel numbers appearing black on black background on windows computers.    

--------------------------------------------------------------------------------

Joseph Dien,
Senior Research Scientist
University of Maryland 

E-mail: jd...@ma...
Phone: 301-226-8848
Fax: 301-226-8811
http://joedien.com//