Re: [Erppcatoolkit-support] Problem loading/appending data and question

[Andreas Widmann <widmann@un...>]

Dear Joe,

> Sorry for long delay!  I came down with a really bad cold.
Oh, I hope you have fully recovered!

> 1) Could you send me a copy of your file so I can test it out?  I can’t assume that all such files would trigger the error.
I will send two files in a separate mail without copy to the list. I hope this is ok.

It is well possible (likely? ;) that I use the function in an incorrect or unintended way (and it worked in previous versions by accident). I try to add a second (within subject) condition to an already loaded grand-average dataset by appending a cell. Reading the first dataset (e.g., gavr-d.set, order doesn’t matter) from my EEGLAB grand-average works flawlessly. Trying to Edit -> Cells -> Append the second dataset (gavr-s.set) gives the reported error (because EPdataIn.trialSpecs is empty).

Should I add the condition/cell in a different way? Should I store both conditions in a single grand-average file? How to encode/store condition information then?

> I’d don’t remember exactly why I made the change but expect it was to fix a bug of some sort.
> I’ll see what I can do to come up with a better solution.
Thanks a lot!

> It’s not possible to use the Single File Mode since it is designed for files where each file is a single waveform and the Widmann variant by definition has multiple waveforms in it.
Ok.

> 2) I’ll take a closer look at it.
> 
> 3) I think that approach makes sense.  I do much the same for the function where it provides a test of the lateralization of dipole pairs.
Very good! Thank you!

Best,
Andreas

> Cheers!
> 
> Joe
> 
>> On Mar 31, 2015, at 13:11, Andreas Widmann <widmann@...> wrote:
>> 
>> Dear Joe and list!
>> 
>> Thank you for providing your toolbox (and supporting the „Widmann variant“ of the EEGLAB file format; which I didn’t even know that it exists :)))! May I ask three short questions?
>> 
>> (1) I have a problem appending cells with the above mentioned file format in 2.47 and 2.48 but not in 2.45:
>> The error is:
>> 
>> Conversion to double from cell is not possible.
>> Error in ep_addData (line 287)
>>                EPdataOut.trialSpecs(end+1:end+numAdded,:)=cell(numAdded,size(EPdataIn.trialSpecs,2));
>> Error in ep_editData (line 2666)
>>            EPoverview.workData=ep_addData(EPoverview.workData,newData,'cells'); 
>> Error while evaluating uicontrol Callback
>> 
>> If I add a check in ep_addData whether EPdataIn.trialSpecs is empty (as it was in previous versions) it works:
>> if ~isempty(EPdataIn.trialSpecs)
>>                EPdataOut.trialSpecs(end+1:end+numAdded,:)=cell(numAdded,size(EPdataIn.trialSpecs,2));
>> end
>> 
>> Would you expect any unexpected side effects from this workaround? Why did you remove the check? (Btw: Is there a way to use the single cell file mode for reading this file format? Couldn’t get that to work. Most probably due to my ignorance.)
>> 
>> (2) Only short suggestion: In ep_addEloc, line 134 ...
>>        if ~isempty(eloc(i).labels) && ~isempty(eloc(i).theta) && isempty(eloc(i).type)
>> … it could be helpful to check whether theta is numeric before wiping. EEGLAB pop_chanedit always ignores type when importing ced-files even if it is contained in the file. This practically disables importing electrode locations if one does not know that the type field has to be filled manually.
>> 
>> (3) My main question: I would like to compare the oddball P3a component latency with the latency of some other external non-ERP psychophysical measurement. Usually, we would do this using jackknifing with some relative peak amplitude criterion. However, I would prefer doing this with a „true" PCA-P3a component and not the ERP-P3a. I cannot separate N1/P2/P3a components with spatial PCA as they are too similar. Now, I hope my question is not too stupid: Do you think it would be a valid approach to do so some kind of jackknifing temporal PCA? That is, performing several temporal PCAs omitting one subject and estimating jackknifing latency from each of the resulting PCA-P3a component time courses?
>> 
>> Thank you! Best regards,
>> Andreas
>> 
>> 
>> ------------------------------------------------------------------------------
>> Dive into the World of Parallel Programming The Go Parallel Website, sponsored
>> by Intel and developed in partnership with Slashdot Media, is your hub for all
>> things parallel software development, from weekly thought leadership blogs to
>> news, videos, case studies, tutorials and more. Take a look and join the 
>> conversation now. http://goparallel.sourceforge.net/
>> _______________________________________________
>> Erppcatoolkit-support mailing list
>> Erppcatoolkit-support@...
>> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support
> 
> 

Re: [Erppcatoolkit-support] Problem loading/appending data and question

[Joseph Dien <jdien07@ma...>]

Sorry for long delay!  I came down with a really bad cold.

Sure thing, happy to support innovations like yours!

1) Could you send me a copy of your file so I can test it out?  I can’t assume that all such files would trigger the error.

I’d don’t remember exactly why I made the change but expect it was to fix a bug of some sort.

I’ll see what I can do to come up with a better solution.

It’s not possible to use the Single File Mode since it is designed for files where each file is a single waveform and the Widmann variant by definition has multiple waveforms in it.

2) I’ll take a closer look at it.

3) I think that approach makes sense.  I do much the same for the function where it provides a test of the lateralization of dipole pairs.

Cheers!

Joe

> On Mar 31, 2015, at 13:11, Andreas Widmann <widmann@...> wrote:
> 
> Dear Joe and list!
> 
> Thank you for providing your toolbox (and supporting the „Widmann variant“ of the EEGLAB file format; which I didn’t even know that it exists :)))! May I ask three short questions?
> 
> (1) I have a problem appending cells with the above mentioned file format in 2.47 and 2.48 but not in 2.45:
> The error is:
> 
> Conversion to double from cell is not possible.
> Error in ep_addData (line 287)
>                EPdataOut.trialSpecs(end+1:end+numAdded,:)=cell(numAdded,size(EPdataIn.trialSpecs,2));
> Error in ep_editData (line 2666)
>            EPoverview.workData=ep_addData(EPoverview.workData,newData,'cells'); 
> Error while evaluating uicontrol Callback
> 
> If I add a check in ep_addData whether EPdataIn.trialSpecs is empty (as it was in previous versions) it works:
> if ~isempty(EPdataIn.trialSpecs)
>                EPdataOut.trialSpecs(end+1:end+numAdded,:)=cell(numAdded,size(EPdataIn.trialSpecs,2));
> end
> 
> Would you expect any unexpected side effects from this workaround? Why did you remove the check? (Btw: Is there a way to use the single cell file mode for reading this file format? Couldn’t get that to work. Most probably due to my ignorance.)
> 
> (2) Only short suggestion: In ep_addEloc, line 134 ...
>        if ~isempty(eloc(i).labels) && ~isempty(eloc(i).theta) && isempty(eloc(i).type)
> … it could be helpful to check whether theta is numeric before wiping. EEGLAB pop_chanedit always ignores type when importing ced-files even if it is contained in the file. This practically disables importing electrode locations if one does not know that the type field has to be filled manually.
> 
> (3) My main question: I would like to compare the oddball P3a component latency with the latency of some other external non-ERP psychophysical measurement. Usually, we would do this using jackknifing with some relative peak amplitude criterion. However, I would prefer doing this with a „true" PCA-P3a component and not the ERP-P3a. I cannot separate N1/P2/P3a components with spatial PCA as they are too similar. Now, I hope my question is not too stupid: Do you think it would be a valid approach to do so some kind of jackknifing temporal PCA? That is, performing several temporal PCAs omitting one subject and estimating jackknifing latency from each of the resulting PCA-P3a component time courses?
> 
> Thank you! Best regards,
> Andreas
> 
> 
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website, sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for all
> things parallel software development, from weekly thought leadership blogs to
> news, videos, case studies, tutorials and more. Take a look and join the 
> conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> Erppcatoolkit-support mailing list
> Erppcatoolkit-support@...
> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support

[Erppcatoolkit-support] Problem loading/appending data and question

[Andreas Widmann <widmann@un...>]

Dear Joe and list!

Thank you for providing your toolbox (and supporting the „Widmann variant“ of the EEGLAB file format; which I didn’t even know that it exists :)))! May I ask three short questions?

(1) I have a problem appending cells with the above mentioned file format in 2.47 and 2.48 but not in 2.45:
The error is:

Conversion to double from cell is not possible.
Error in ep_addData (line 287)
                EPdataOut.trialSpecs(end+1:end+numAdded,:)=cell(numAdded,size(EPdataIn.trialSpecs,2));
Error in ep_editData (line 2666)
            EPoverview.workData=ep_addData(EPoverview.workData,newData,'cells'); 
Error while evaluating uicontrol Callback

If I add a check in ep_addData whether EPdataIn.trialSpecs is empty (as it was in previous versions) it works:
if ~isempty(EPdataIn.trialSpecs)
                EPdataOut.trialSpecs(end+1:end+numAdded,:)=cell(numAdded,size(EPdataIn.trialSpecs,2));
end

Would you expect any unexpected side effects from this workaround? Why did you remove the check? (Btw: Is there a way to use the single cell file mode for reading this file format? Couldn’t get that to work. Most probably due to my ignorance.)

(2) Only short suggestion: In ep_addEloc, line 134 ...
        if ~isempty(eloc(i).labels) && ~isempty(eloc(i).theta) && isempty(eloc(i).type)
… it could be helpful to check whether theta is numeric before wiping. EEGLAB pop_chanedit always ignores type when importing ced-files even if it is contained in the file. This practically disables importing electrode locations if one does not know that the type field has to be filled manually.

(3) My main question: I would like to compare the oddball P3a component latency with the latency of some other external non-ERP psychophysical measurement. Usually, we would do this using jackknifing with some relative peak amplitude criterion. However, I would prefer doing this with a „true" PCA-P3a component and not the ERP-P3a. I cannot separate N1/P2/P3a components with spatial PCA as they are too similar. Now, I hope my question is not too stupid: Do you think it would be a valid approach to do so some kind of jackknifing temporal PCA? That is, performing several temporal PCAs omitting one subject and estimating jackknifing latency from each of the resulting PCA-P3a component time courses?

Thank you! Best regards,
Andreas

[Erppcatoolkit-support] EP Toolkit 2.48 released

[Joseph Dien <jdien07@me...>]

EP Toolkit version 2.48 has been released:
https://sourceforge.net/projects/erppcatoolkit/

This is a minor bug fix release.

1) Fixed crash in Save function when name of dataset to be saved is changed.
2) Fixed crash when FontSize not provided as in reading .study files.
3) Fixed crash when channels selected in the Chans tab and then the working copy was saved.
4) Fixed peak chans in Topoplots sometimes not being computed correctly for voltage data (was reporting peak positive peak rather than absolute value peak).
5) Fixed crash when reading in multiple data files as in Single File mode.

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

Joseph Dien, PhD
Research Associate
Cognitive Neurology
The Johns Hopkins University School of Medicine


 

Re: [Erppcatoolkit-support] EP Toolkit 2.46 released

[Joseph Dien <jdien07@me...>]

Yes

Sent from my iPad

> On Dec 30, 2014, at 00:36, Arild Hestvik <hestvik@...> wrote:
> 
> Will .ept files always be backwards compatible?
> 
>> On Tue, Dec 30, 2014 at 7:53 AM, Joseph Dien <jdien07@...> wrote:
>> EP Toolkit version 2.46 has been released:
>> https://sourceforge.net/projects/erppcatoolkit/
>> 
>> Matlab 2014b onwards has made major changes to the graphics system.  EP Toolkit will not currently work on 2014b or newer on any platform.  This EP Toolkit update (2.46) will be the last one that will run under the older Matlab versions (prior to 2014b).  The next will be a 2.50 update which will only run under Matlab 2014b or later.
>> 
>> Aside from the usual bug fixes and refinements, notable upgrades include the addition of a batch convert option for the Save function, enhancements to the segmentation functionality, and enhancements to the blink/saccade template function.
>> 
>> 1) Workaround for NetStation bug where EGIS average files have number of trials in cell in wrong spot in the cell header.  
>> 2) Fixed crash in Segment function and others for files where event values are all numbers rather than strings.  
>> 3) Added support for saving event info to a text file for file formats that don’t contain event information and support for reading in such event text files when present, which also provides a mechanism for easily modifying event information.  
>> 4) Added support for bdf recording stop events (by marking them as 'boundary' events).  
>> 5) Added field to Transform panel indicating how much of a delay has been added to ERP latencies by one-pass filters.  
>> 6) Added delay field to the Segment function to allow for correction for timing offsets.  
>> 7) Added check to update cache if the current one was generated by an older version of the EP Toolkit, avoiding crashes.  
>> 8) Added Type event field to the list of possible criteria for Segment function.  
>> 9) Fixed crash in average function when bad channels resulted in not being able to compute covariance matrix.  
>> 10) Fixed appending subjects using Edit function not working.  
>> 11) Appending cells using Edit function now works even when they have different trial spec fields.  
>> 12) Support for SMI and text event files now for any function loading in a file, not just preprocessing.  
>> 13) Fixed crash in ANOVA when analysis only has between-group factors, no within-group factors.  
>> 14) Fixed preferences for SMI and text event file suffixes not being changeable.  
>> 15) Further improvements in SMI eye-tracker output support.  
>> 16) In Template function, added click 2D topos to expand into 3D view.  
>> 17) In Template function, added EOG channel controls.  
>> 18) Fixed crash when selecting mff files on a PC.  
>> 19) Since this was causing confusion, one now selects one of the mff files in the directory and then is given the opportunity to select the other files as well (unless it was the only mff file present).  
>> 20) Fixed crash when filtering when there is a boundary event at the start of the file.  
>> 21) Fixed crash in sampTest function when there are bad trials present in single-trial data.  
>> 22) Fixed crash when adding ced electrode coordinates info using Edit function.  
>> 23) Fixed crash when using Transform pane and rereference set to "none.”  
>> 24) Fixed crash when selecting points in continuous data where the number of points is not an even multiple of one second.  
>> 25) Eliminated requirement that segmented epochs not overlap with each other.  
>> 26) Added sixth spec for segmenting.  
>> 27) Fixed no time marker for blink and saccades when forming templates.  
>> 28) Added vertical saccade correction.  
>> 29) Fixed crash in SampleTest when no suitable data in working set.  
>> 30) Fixed crash in windowing and autoPCA functions.  
>> 31) When appending two sets of single-trial data, ensure that each cell has uniqe set of trial names.  
>> 32) Fixed crash in Scan function when the datasets have different numbers of channels (including regional ones).  
>> 33) Fixed channels being left out of Waves or Scan figures when the last dataset has fewer channels (including regional) than earlier datasets.  
>> 34) Added Import Events button to the Events sub pane.  
>> 35) Fixed crash in template function when there are non-EEG channels present.  
>> 36) Fixed crash in sampleTest function when using temporal PCA template with Woody PCA option.  
>> 37) In Edit, fixed crash when changing channel type and not staying changed when changing cell type.  
>> 38) Fixed crash when frequency transforming and there are non EEG channels present.  
>> 39) Fixed crash when saving factor files with more than one factor in simple binary file format.  
>> 40) Fixed not handling event times correctly when saving files in simple binary form and event time samples are decimals (warning was being displayed by Matlab).  
>> 41) Added convert option to Save function.
>> 
>> -------------------------------------------------------------------------------- 
>> 
>> Joseph Dien,
>> Senior Research Scientist
>> Maryland Neuroimaging Center
>> University of Maryland 
>> 
>> E-mail: jdien07@...
>> http://joedien.com
>>  
>> 
>>  
>> 
>> 
>> 
>> ------------------------------------------------------------------------------
>> Dive into the World of Parallel Programming! The Go Parallel Website,
>> sponsored by Intel and developed in partnership with Slashdot Media, is your
>> hub for all things parallel software development, from weekly thought
>> leadership blogs to news, videos, case studies, tutorials and more. Take a
>> look and join the conversation now. http://goparallel.sourceforge.net
>> _______________________________________________
>> Erppcatoolkit-support mailing list
>> Erppcatoolkit-support@...
>> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support
> 
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming! The Go Parallel Website,
> sponsored by Intel and developed in partnership with Slashdot Media, is your
> hub for all things parallel software development, from weekly thought
> leadership blogs to news, videos, case studies, tutorials and more. Take a
> look and join the conversation now. http://goparallel.sourceforge.net
> _______________________________________________
> Erppcatoolkit-support mailing list
> Erppcatoolkit-support@...
> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support

Re: [Erppcatoolkit-support] EP Toolkit 2.46 released

[Arild Hestvik <hestvik@ud...>]

Will .ept files always be backwards compatible?

On Tue, Dec 30, 2014 at 7:53 AM, Joseph Dien <jdien07@...> wrote:

> EP Toolkit version 2.46 has been released:
> https://sourceforge.net/projects/erppcatoolkit/
>
> Matlab 2014b onwards has made major changes to the graphics system.  EP
> Toolkit will not currently work on 2014b or newer on any platform.  This EP
> Toolkit update (2.46) will be the last one that will run under the older
> Matlab versions (prior to 2014b).  The next will be a 2.50 update which
> will only run under Matlab 2014b or later.
>
> Aside from the usual bug fixes and refinements, notable upgrades include
> the addition of a batch convert option for the Save function, enhancements
> to the segmentation functionality, and enhancements to the blink/saccade
> template function.
>
> 1) Workaround for NetStation bug where EGIS average files have number of
> trials in cell in wrong spot in the cell header.
> 2) Fixed crash in Segment function and others for files where event values
> are all numbers rather than strings.
> 3) Added support for saving event info to a text file for file formats
> that don’t contain event information and support for reading in such event
> text files when present, which also provides a mechanism for easily
> modifying event information.
> 4) Added support for bdf recording stop events (by marking them as
> 'boundary' events).
> 5) Added field to Transform panel indicating how much of a delay
> has been added to ERP latencies by one-pass filters.
> 6) Added delay field to the Segment function to allow for correction for
> timing offsets.
> 7) Added check to update cache if the current one was generated by
> an older version of the EP Toolkit, avoiding crashes.
> 8) Added Type event field to the list of possible criteria for Segment
> function.
> 9) Fixed crash in average function when bad channels resulted in not being
> able to compute covariance matrix.
> 10) Fixed appending subjects using Edit function not working.
> 11) Appending cells using Edit function now works even when they have
> different trial spec fields.
> 12) Support for SMI and text event files now for any function loading in a
> file, not just preprocessing.
> 13) Fixed crash in ANOVA when analysis only has between-group factors, no
> within-group factors.
> 14) Fixed preferences for SMI and text event file suffixes not being
> changeable.
> 15) Further improvements in SMI eye-tracker output support.
> 16) In Template function, added click 2D topos to expand into 3D view.
> 17) In Template function, added EOG channel controls.
> 18) Fixed crash when selecting mff files on a PC.
> 19) Since this was causing confusion, one now selects one of the mff files
> in the directory and then is given the opportunity to select the
> other files as well (unless it was the only mff file present).
> 20) Fixed crash when filtering when there is a boundary event at the start
> of the file.
> 21) Fixed crash in sampTest function when there are bad trials present in
> single-trial data.
> 22) Fixed crash when adding ced electrode coordinates info using Edit
> function.
> 23) Fixed crash when using Transform pane and rereference set to "none.”
> 24) Fixed crash when selecting points in continuous data where the number
> of points is not an even multiple of one second.
> 25) Eliminated requirement that segmented epochs not overlap with each
> other.
> 26) Added sixth spec for segmenting.
> 27) Fixed no time marker for blink and saccades when forming templates.
> 28) Added vertical saccade correction.
> 29) Fixed crash in SampleTest when no suitable data in working set.
> 30) Fixed crash in windowing and autoPCA functions.
> 31) When appending two sets of single-trial data, ensure that each cell
> has uniqe set of trial names.
> 32) Fixed crash in Scan function when the datasets have different numbers
> of channels (including regional ones).
> 33) Fixed channels being left out of Waves or Scan figures when the last
> dataset has fewer channels (including regional) than earlier datasets.
> 34) Added Import Events button to the Events sub pane.
> 35) Fixed crash in template function when there are non-EEG channels
> present.
> 36) Fixed crash in sampleTest function when using temporal PCA template
> with Woody PCA option.
> 37) In Edit, fixed crash when changing channel type and not staying
> changed when changing cell type.
> 38) Fixed crash when frequency transforming and there are non EEG
> channels present.
> 39) Fixed crash when saving factor files with more than one factor
> in simple binary file format.
> 40) Fixed not handling event times correctly when saving files in simple
> binary form and event time samples are decimals (warning was being
> displayed by Matlab).
> 41) Added convert option to Save function.
>
>
> --------------------------------------------------------------------------------
>
> Joseph Dien,
> Senior Research Scientist
> Maryland Neuroimaging Center
> University of Maryland
>
> E-mail: jdien07@...
> http://joedien.com
>
>
>
>
>
>
>
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming! The Go Parallel Website,
> sponsored by Intel and developed in partnership with Slashdot Media, is
> your
> hub for all things parallel software development, from weekly thought
> leadership blogs to news, videos, case studies, tutorials and more. Take a
> look and join the conversation now. http://goparallel.sourceforge.net
> _______________________________________________
> Erppcatoolkit-support mailing list
> Erppcatoolkit-support@...
> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support
>
>

[Erppcatoolkit-support] EP Toolkit 2.46 released

[Joseph Dien <jdien07@me...>]

EP Toolkit version 2.46 has been released:
https://sourceforge.net/projects/erppcatoolkit/

Matlab 2014b onwards has made major changes to the graphics system.  EP Toolkit will not currently work on 2014b or newer on any platform.  This EP Toolkit update (2.46) will be the last one that will run under the older Matlab versions (prior to 2014b).  The next will be a 2.50 update which will only run under Matlab 2014b or later.

Aside from the usual bug fixes and refinements, notable upgrades include the addition of a batch convert option for the Save function, enhancements to the segmentation functionality, and enhancements to the blink/saccade template function.

1) Workaround for NetStation bug where EGIS average files have number of trials in cell in wrong spot in the cell header.  
2) Fixed crash in Segment function and others for files where event values are all numbers rather than strings.  
3) Added support for saving event info to a text file for file formats that don’t contain event information and support for reading in such event text files when present, which also provides a mechanism for easily modifying event information.  
4) Added support for bdf recording stop events (by marking them as 'boundary' events).  
5) Added field to Transform panel indicating how much of a delay has been added to ERP latencies by one-pass filters.  
6) Added delay field to the Segment function to allow for correction for timing offsets.  
7) Added check to update cache if the current one was generated by an older version of the EP Toolkit, avoiding crashes.  
8) Added Type event field to the list of possible criteria for Segment function.  
9) Fixed crash in average function when bad channels resulted in not being able to compute covariance matrix.  
10) Fixed appending subjects using Edit function not working.  
11) Appending cells using Edit function now works even when they have different trial spec fields.  
12) Support for SMI and text event files now for any function loading in a file, not just preprocessing.  
13) Fixed crash in ANOVA when analysis only has between-group factors, no within-group factors.  
14) Fixed preferences for SMI and text event file suffixes not being changeable.  
15) Further improvements in SMI eye-tracker output support.  
16) In Template function, added click 2D topos to expand into 3D view.  
17) In Template function, added EOG channel controls.  
18) Fixed crash when selecting mff files on a PC.  
19) Since this was causing confusion, one now selects one of the mff files in the directory and then is given the opportunity to select the other files as well (unless it was the only mff file present).  
20) Fixed crash when filtering when there is a boundary event at the start of the file.  
21) Fixed crash in sampTest function when there are bad trials present in single-trial data.  
22) Fixed crash when adding ced electrode coordinates info using Edit function.  
23) Fixed crash when using Transform pane and rereference set to "none.”  
24) Fixed crash when selecting points in continuous data where the number of points is not an even multiple of one second.  
25) Eliminated requirement that segmented epochs not overlap with each other.  
26) Added sixth spec for segmenting.  
27) Fixed no time marker for blink and saccades when forming templates.  
28) Added vertical saccade correction.  
29) Fixed crash in SampleTest when no suitable data in working set.  
30) Fixed crash in windowing and autoPCA functions.  
31) When appending two sets of single-trial data, ensure that each cell has uniqe set of trial names.  
32) Fixed crash in Scan function when the datasets have different numbers of channels (including regional ones).  
33) Fixed channels being left out of Waves or Scan figures when the last dataset has fewer channels (including regional) than earlier datasets.  
34) Added Import Events button to the Events sub pane.  
35) Fixed crash in template function when there are non-EEG channels present.  
36) Fixed crash in sampleTest function when using temporal PCA template with Woody PCA option.  
37) In Edit, fixed crash when changing channel type and not staying changed when changing cell type.  
38) Fixed crash when frequency transforming and there are non EEG channels present.  
39) Fixed crash when saving factor files with more than one factor in simple binary file format.  
40) Fixed not handling event times correctly when saving files in simple binary form and event time samples are decimals (warning was being displayed by Matlab).  
41) Added convert option to Save function.

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

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

E-mail: jdien07@...
http://joedien.com
 

 

[Erppcatoolkit-support] OS X 10.10 update (Yosemite)

[Joseph Dien <jdien07@ma...>]

For Mac users, please note that the newest OS X update (Yosemite) will greatly affect your ability to use Matlab.  Only the newest version (2014b) will start normally.  This version of Matlab has greatly changed the graphics routines and so the current version of EP Toolkit will not run under it (I’m not sure when the necessary revisions will be completed).  Matlab 2013b and 2014a can run under Yosemite if the following steps are taken:

If there is a white bar across the Matlab icon and clicking on it gives you a "You can't use this version of the application” message, then you need to follow the instructions on this following webpage: http://www.mathworks.com/support/bugreports/1116607 <http://www.mathworks.com/support/bugreports/1116607>;

The webpage claims that Apple has fixed this problem but I’m not seeing it.

If following these instructions results in a very long java error message and Matlab crashes, then you need to also follow the instructions at: http://www.mathworks.com/support/bugreports/1098655 <http://www.mathworks.com/support/bugreports/1098655>;

Older versions of Matlab will not run at all under Yosemite.  I’m therefore going to stop trying to support anything earlier than Matlab 2013b.

I should have a minor EP Toolkit update (2.46) finished soon, which will be the last one that will run under the older Matlab versions (prior to 2014b).
After that, the next will be a 2.50 update which will only run under Matlab 2014b or later.

Joe
--------------------------------------------------------------------------------

Joseph Dien, PhD
Research Associate
Cognitive Neurology
The Johns Hopkins University School of Medicine

[Erppcatoolkit-support] Matlab 2014b update note

[Joseph Dien <jdien07@ma...>]

Please note that there have been significant revisions in Matlab 2014b’s graphic routines and so the EP Toolkit cannot yet be run under it.  A computer can have different versions of Matlab installed so just be sure to use an earlier version of Matlab with it until I’ve been able to make the necessary revisions.  A new bugfix revision will be coming out soon but will still need to run under 2013a.

Joe
--------------------------------------------------------------------------------

Joseph Dien, PhD
Research Associate
Cognitive Neurology
The Johns Hopkins University School of Medicine

[Erppcatoolkit-support] combining single subject data

[Arild Hestvik <hestvik@ud...>]

Joe,
how can I I combine two separate sessions / recordings of the same subject
into one single average file? I mean, I can average the two files (after
preprocessing), but the average tool creates a file that says it's 2
subjects, rather than 1. ?
Arild

Re: [Erppcatoolkit-support] ERPLab files

[Benikos N.P. <N.P.Benikos@so...>]

Hi Joe,

Thank you for getting back to me! I'm afraid this time it was a case of user error. Because I was using .avg files that had already been processed I didn't go through the processing part of the manual (where it explains in more detail the read function), but just start the section after. I missed the bit about uniform filenames denoting the subject number and condition. Thank you for the great toolkit.

Kind Regards,
Nick


--------------------------------------
Nicholas Benikos

Post Doctoral Research Fellow/Lab Manager
Developmental Brain Behaviour Laboratory 
Psychology Academic Unit
Room 3091
Shackleton Building
University of Southampton
Highfield Campus M35
Southampton SO17 1BJ, UK

Tel.: +44 0 2380 594593 (Ext. 24593)

-----Original Message-----
From: Joseph Dien [mailto:jdien07@...] 
Sent: 29 August 2014 02:06
To: Jack Shelley-Tremblay
Cc: erppcatoolkit-support@...
Subject: Re: [Erppcatoolkit-support] ERPLab files

Hi, welcome!

It shouldn't need any kind of configuration.

Am I to understand it's not working for you?

If not, could you describe what is happening?

Joe


On Aug 28, 2014, at 4:14 PM, Jack Shelley-Tremblay <jstremblay@...> wrote:

> Hello,
> 
> First post here so forgive my ignorance!
> 
> I have an ERPLab .erp file with 8 subjects that I wish now to perform PCA on using EP Toolkit.
> 
> I can load it in with the Read function, but I cannot find any 
> instructions on how to configure the subjects and cells correctly so
> 
> that EP Toolkit can understand it correctly.
> 
> Is there a tutorial anywhere that begins with a fully processed ERPLab file and goes on through PCA?
> 
> Thank you,
> 
> 
> Jack Shelley-Tremblay
> Professor of Psychology
> Assistant Director: University Honors Program University of South 
> Alabama
> 251-460-6883
> 
> ----------------------------------------------------------------------
> --------
> Slashdot TV.  
> Video for Nerds.  Stuff that matters.
> http://tv.slashdot.org/_______________________________________________
> Erppcatoolkit-support mailing list
> Erppcatoolkit-support@...
> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support


------------------------------------------------------------------------------
Slashdot TV.  
Video for Nerds.  Stuff that matters.
http://tv.slashdot.org/
_______________________________________________
Erppcatoolkit-support mailing list
Erppcatoolkit-support@...
https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support

Re: [Erppcatoolkit-support] ERPLab files

[Joseph Dien <jdien07@ma...>]

Hi, welcome!

It shouldn’t need any kind of configuration.

Am I to understand it’s not working for you?

If not, could you describe what is happening?

Joe


On Aug 28, 2014, at 4:14 PM, Jack Shelley-Tremblay <jstremblay@...> wrote:

> Hello,
> 
> First post here so forgive my ignorance!
> 
> I have an ERPLab .erp file with 8 subjects that I wish now to perform PCA on using EP Toolkit.
> 
> I can load it in with the Read function, but I cannot find any instructions on how to configure the subjects and cells correctly so 
> 
> that EP Toolkit can understand it correctly.
> 
> Is there a tutorial anywhere that begins with a fully processed ERPLab file and goes on through PCA?
> 
> Thank you,
> 
> 
> Jack Shelley-Tremblay
> Professor of Psychology
> Assistant Director: University Honors Program
> University of South Alabama
> 251-460-6883
> 
> ------------------------------------------------------------------------------
> Slashdot TV.  
> Video for Nerds.  Stuff that matters.
> http://tv.slashdot.org/_______________________________________________
> Erppcatoolkit-support mailing list
> Erppcatoolkit-support@...
> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support

[Erppcatoolkit-support] ERPLab files

[Jack Shelley-Tremblay <jstremblay@so...>]

Hello,

First post here so forgive my ignorance!

I have an ERPLab .erp file with 8 subjects that I wish now to perform PCA
on using EP Toolkit.

I can load it in with the Read function, but I cannot find any instructions
on how to configure the subjects and cells correctly so

that EP Toolkit can understand it correctly.

Is there a tutorial anywhere that begins with a fully processed ERPLab file
and goes on through PCA?

Thank you,


Jack Shelley-Tremblay
Professor of Psychology
Assistant Director: University Honors Program
University of South Alabama
251-460-6883

[Erppcatoolkit-support] ERP PCA Toolkit 2.45 release

[Joseph Dien <jdien07@ma...>]

New release of EP Toolkit.  Some highlights are: upgrades to the filtering functionality, support for BrainVision EEG files, sample-by-sample analysis, latency-locked and jitter-correction averaging, improved EEGlab .set file format support, improved mff file format support, added option to explore rereferencing effects in Topos view, added PARE-corrected average referencing to Transform function.

1) Fixed crash when performing combination of subjects with file containing .cov information.

2) Added -all- and -erpimage- options to the Factors list in View.

3) Fixed keys field of events not being added when missing, resulting in EP files created by older versions of Toolkit not being readable.

4) Fixed crash when there are multiple global and trialwise bad channels.  Apparently Mathworks changed something again in recent versions of Matlab.

5) Fixed crash when running artifact correction on an average file with multiple subjects.

6) Fixed problem where under some circumstances the files after the first of a batch could not be found by the Transform function.

7) Fixed not putting factor variance information in correct location when loading PCA .ept files, resulting in "There are 0 factors that meet the minimum variance criterion" error messages when trying to autoPCA them.

8) Fixed AutoPCA generating nothing but missing data values when maxcentroid and mincentroid measures chosen.

9) Fixed crash in View when overplotting two sets of data where one has regional channels not present in the other.

10) -all- and -erpimage- options in View leave out adds if lower levels are present (e.g., for subjects, leaves out grand averages if individual subjects are present).

11) Fixed weighting not correct when computing difference waves that do not sum to zero (not sure why anyone might do this but fixed just in case).

12) Fixed calculation of the SubNum field (number of subjects going into averages) when combining cells.

13) Based on additional information from the MNE community, made some corrections to how the covariance matrix and the effective sample size are calculated, for when generating FIFF files.

14) in Topos function, for finding peak channels and points and Hz, uses only the first dataset as the index for this.

15) Fixed crash when replacing bad channels in an average file and a cell is bad.

16) Allows choosing either Hanning or multi-taper methods for spectral measures.

17) Eliminated upper bound to smoothing in multi-taper frequency measures as no longer seems to apply.

18) Fixed 'all' option for Window Data leaving out last channel.

19) Repair referencing after performing artifact correction to support performing analyses on unreferenced data, as in spectral analyses.

20) Added conversion of REF channel type to EEG for older EP files, which can otherwise result in various reference channel problems.

21) Added ability to perform PCA on datasets with bad data.

22) FacScr observations are now always arranged with permutations in order of the seven data dimensions.  PCA files made prior to

2.44 cannot be used for two-step PCA.  The first step will need to be rerun.

23) Fixed minimum and maximum values for spectral PCA datasets not being calculated corrected when using View functions.

24) Fixed not able to load in EP files with frequency PCA data.

25) Fixed crash when changing pages in Topos view with frequency data.

26) Fixed 2D expanded head plots for Topos view for frequency data.


27) Fixed jack-knife test possibly conducted on wrong factor or just crashing in Topos view.

28) Fixed rescale option for frequency data in Topos view.

29) Fixed when spectral range changed in View function and dB or psd options are on, the  values are immediately further transformed.

30) Fixed crash on last page of topos for frequency PCA data.

31) Fixed crash when saving edits in Scan function.

32) In the Scan function, fixed all subjects shown as having a bad cell if the first subject in an average dataset has a bad cell.

33) Fixed peak latency of factors expressed as one sample too late in the Edit function.

34) Added legend to wave plots.

35) When using factors to set channel areas in the Window function, can now specify whether to use largest absolute, negative, or positive loadings.

36) Will no longer keep resetting the factor loading threshold back to the original number in the Window function.

37) Fixed crash when running an ANOVA on a windowed text file generated by autoPCA and the adds option is on.

38) Fixed minimum and maximum voltages in View pane not reflecting correct values for single trial data.

39) Added support for reading BrainVision EEG files.

40) Fixed crash in blink template function when frequency or factor files are present by excluding them entirely.

41) Fixed blink template waveform plot not adjusting to new length when switching datasets.

42) For blink template function, fixed marker for blink and saccade not placed at correct latency after switching between datasets with the same number of samples but different baselines.

43) Fixed crash when loading an ept file with no theta values for the electrode coordinates.

44) Fixed crash when there is more than one bad channel and the reference is mean mastoid.

45) Fixed crash in saccade correction due to bug introduced in last version.

46) Fixed crash when the ced file has no type field.

47) Fixed crash in topoplots when there are electrodes without coordinates.

48) Allow for manual windowing of PCA files rather than just autoPCA.

49) Fixed list of trials for single-trial data in View pane not correct.

50) Added simple averaging of trials in Edit function's cell subpane for single-trial data.

51) In preprocessing, if current reference not specified, default to original reference.

52)  Fixed blink and saccade correction in Preprocessing function could issue false error that template had wrong number of electrodes.

53) Fixed crash in Preprocessing function when correcting blink or saccade in continuous file.

54) Fixed crash in Preprocessing when there are multiple chunks and some have blink factors and some do not.

55) Fixed crash in Preprocessing function when replacing bad channels in continuous data.

56) Transforms can be applied to non-EEG channel types and which data mode can be specified.

57) Fixed crash when filtering continuous data and there are no boundary events present.

58) Fixed blank keys event fields being created without a .key after the keys level.

59) Added "starts", "ends", "contains", and "follows" keywords to Segment function.

60) Fixed Edit function reordering cells in single-trial data when anything clicked or changed in the Cells table and they were not already in alphabetical order.

61) Added support for sample-by-sample, CWT, and Woody Filter PCA comparisons using both non-parametric t-tests and jack-knife tests (including addition of STS chanType).

62) In Edit function, fixed deleting wrong cell of single-trial data when cells were not in alphabetical order.

63) Added support for SMI eye-tracking data to be merged with EEG data.

64) Added latency-locked and jitter-correction averaging options.

65) Fixed cov.Nq field (used for FIFF file format support) not being formed correctly when combining subjects, resulting in crashes later on.

66) Fixed subject selection not being applied to .cov.Nq field (used for FIFF file format support), resulting in crashes down the line for combined subject average files.

67) Added fix for crash when trying to read mff files that erroneously label their COM channel as being a reference channel.

68) Fixed crash when saving file with an empty .keys field in EEGlab .set format.

69) Added additional filtering options, including one-pass causal filters and order settings.  Also providing graphs of effects of filtering on the data in the frequency and temporal domains.

70) Simplified event keys field structure.

71) Fixed labels for REG channels being blank when saving EEGlab .set format files.

72) For single_trial data, the event latency values now conform to "for epoched datasets the event latencies are also encoded in sample points with respect to the beginning of the data (as if the data were continuous)". http://sccn.ucsd.edu/wiki/Chapter_03:_Event_Processing rather than being in terms of the beginning of the original continuous data.

73) Fixed adding 'trigger' events if they are already present.

74) Fixed epoch event field in saved .set files referring to trial number from complete dataset rather than in terms of the EEG file (the one condition).

75) Fixed urevent event field in saved .set files reflecting event numbering of full dataset rather than just the one condition in the EEG file.

76) Fixed sometimes adding too many epoch entries when exporting single_trial .set files, resulting in aborted export process.

77) Handles better situation where a field of an event is empty.

78) Fixed bug where windowing of spatial PCAs of ERPs and temporal and spatial PCAs of spectral data started with first sample of the epoch rather than the sample specified.

79) Fixed crash under some conditions when selecting points from a continuous file, as when using the trimData function.

80) Fixed only first event sample being updated when points selected from continuous data, as in trimming data.

81) Fixed boundary events falling on edge of selected time range not being deleted.

82) Fixed edits in trimData function not being saved when clicking on "keep".

83) Fixed crash in blink Template function when the dataset contains a boundary event.

84) Further improved mff file format support.

85) Adds, like blink and saccade channels, are stripped out of preprocessed data if the output file format is not ept.

86) Fixed crash in Edit function when reordering subjects.

87) Fixed .cov.Nq not being updated when stripping off subject adds, resulting in error messages.

88) Fixed CSD option of Transform crashing.

89) Fixed crash when median averaging frequency-domain data.

90) Fixed Topos function not displaying frequency-domain data.

91) Added option to explore effects of rereferencing to View Topos function.

92) In transform function, setting std field to empty rather than trying to rereference.

93) Added PARE-corrected average reference option to the Transform function.

Cheers!

Joe




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

Joseph Dien,
Senior Research Scientist
University of Maryland 

E-mail: jdien07@...
Phone: 301-226-8848
Fax: 301-226-8811
http://joedien.com//

Re: [Erppcatoolkit-support] Error in ep_readData

[Joseph Dien <jdien07@ma...>]

Sorry about that!  I’ve got a fix for it but it’s embedded in a major revision I’m finishing up.  It’s hard for me to paste the fix into the current release without breaking other things since so much has been changed.  I hopefully will have the revision finished and tested sometime next week if that works for you.  Or I can send you a link to my current build, although it’s strictly alpha (which is to say I’m still working on it).

Joe


On Jun 19, 2014, at 12:51 PM, Katerina Nikolakopoulou <anikolak13@...> wrote:

> Hallo,
> 
> I have run into a problem while trying to use ep toolkit. I have processed and cleaned my EEG data with eeglab and then created the average .ept files for every eeg dataset with ep toolkit. However when I try to read the .ept files I run into this error :
> 
> Undefined function or variable "hdr".
> 
> Error in ep_readData (line 3161)
>     [eloc,chanNames,chanTypes,sixDdata,noise,stanDev,badChans,reference,facVecS]=ep_addEloc(ced,eloc,fileFormat,dataType,chanNames,timeNames,cellNames,subNames,freqNames,chanTypes,sixDdata,noise,stanDev,covMatrix,badChans,reference,facVecS,hdr);
>  
> Error in ep (line 2581)
>                 EPdata=ep_readData(inArg);
> 
>  Error while evaluating uicontrol Callback
> 
> Its the first time I am using ep toolkit and any help would be great!
> 
> Thank you in advance,
> 
> Best regards,
> Katerina
> ------------------------------------------------------------------------------
> HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions
> Find What Matters Most in Your Big Data with HPCC Systems
> Open Source. Fast. Scalable. Simple. Ideal for Dirty Data.
> Leverages Graph Analysis for Fast Processing & Easy Data Exploration
> http://p.sf.net/sfu/hpccsystems_______________________________________________
> Erppcatoolkit-support mailing list
> Erppcatoolkit-support@...
> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support


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

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

E-mail: jdien07@...
Phone: 202-297-8117
http://joedien.com

[Erppcatoolkit-support] Error in ep_readData

[Katerina Nikolakopoulou <anikolak13@gm...>]

Hallo,

I have run into a problem while trying to use ep toolkit. I have processed
and cleaned my EEG data with eeglab and then created the average .ept files
for every eeg dataset with ep toolkit. However when I try to read the .ept
files I run into this error :

Undefined function or variable "hdr".

Error in ep_readData (line 3161)

[eloc,chanNames,chanTypes,sixDdata,noise,stanDev,badChans,reference,facVecS]=ep_addEloc(ced,eloc,fileFormat,dataType,chanNames,timeNames,cellNames,subNames,freqNames,chanTypes,sixDdata,noise,stanDev,covMatrix,badChans,reference,facVecS,hdr);

Error in ep (line 2581)
                EPdata=ep_readData(inArg);

 Error while evaluating uicontrol Callback

Its the first time I am using ep toolkit and any help would be great!

Thank you in advance,

Best regards,
Katerina

Re: [Erppcatoolkit-support] channel numbers in topoplot

[Joseph Dien <jdien07@ma...>]

I don’t currently have a way implemented to provide labels in the topoplots.  I can add that to the to-do list.  For now, the easiest way to accomplish what you’d like to do (maybe a better way) is to go to the channels function and tell it to set up a channel region based on a factor loading threshold.  That should get you all the channels in the isograph.

Cheers!

Joe


On Apr 11, 2014, at 11:00 AM, Arild Hestvik <hestvik@...> wrote:

> Hi Joe,
> how can I add channel labels to the temporospatial plot top of interest, e.g. fig 4 in the attached screenshot. I'd like to use a subset of electrodes, e.g the arbitrarily chosen inner "isograph" and use that as a PCA-motivated "region of interest" in analyzing the undecomposed voltage data, so seeing the channel numbers would help. I know I can look at the facpatST values, but a topoplot is more direct.
> 
> Arild Hestvik, PhD
> Department of Linguistics and Cognitive Science
> University of Delaware
> http://udel.edu/~hestvik
> <Untitled.png>


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

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

E-mail: jdien07@...
Phone: 202-297-8117
http://joedien.com

[Erppcatoolkit-support] channel numbers in topoplot

[Arild Hestvik <hestvik@ud...>]

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

[Joseph Dien <jdien07@ma...>]

Thinking further, I suppose one way the congruency effect could appear to get split off like that is if some subjects’ N400 have a congruency effect and some don’t and they have different scalp topographies, so that might be worth looking at.  I also don’t meant to claim that PCA always works, just that when it doesn’t it generally takes the form of not successfully separating ERP components.  The part of the N400 variance with the congruency effect should not normally go wandering off into a distinct factor by itself unless (for a spatial PCA) it has a distinct spatial distribution from the rest of the N400 variance, in which case it is indeed arguably a different ERP component.  Keep in mind that, within a given subject, a different scalp topography means that a different part of the brain generated it, although the reverse cannot necessarily be claimed.

Cheers!

Joe

On Apr 10, 2014, at 8:57 PM, Joseph Dien <jdien07@...> wrote:

> 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 <petter.kallioinen@...> 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
>> Erppcatoolkit-support@...
>> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support
> 
> 
> --------------------------------------------------------------------------------
> 
> Joseph Dien,
> Senior Research Scientist
> Maryland Neuroimaging Center
> University of Maryland 
> 
> E-mail: jdien07@...
> Phone: 202-297-8117
> http://joedien.com
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 


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

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

E-mail: jdien07@...
Phone: 202-297-8117
http://joedien.com

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

[Joseph Dien <jdien07@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 <petter.kallioinen@...> 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
> Erppcatoolkit-support@...
> https://lists.sourceforge.net/lists/listinfo/erppcatoolkit-support


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

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

E-mail: jdien07@...
Phone: 202-297-8117
http://joedien.com

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

[Petter Kallioinen <petter.kallioinen@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 Dien <jdien07@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 <mparvaz@...> 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 <jdien07@...> 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 <mparvaz@...> 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: jdien07@...
> Phone: 202-297-8117
> http://joedien.com
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 


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

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

E-mail: jdien07@...
Phone: 202-297-8117
http://joedien.com

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

[Muhammad Adeel Parvaz <mparvaz@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 <jdien07@...> 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 <mparvaz@...>
> 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: jdien07@...
> Phone: 202-297-8117
> http://joedien.com
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>

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

[Joseph Dien <jdien07@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 <mparvaz@...> 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: jdien07@...
Phone: 202-297-8117
http://joedien.com

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

[Muhammad Adeel Parvaz <mparvaz@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