|
|
|
EEGLAB restingIAF Übergabeparameter funktionieren nicht |
|
| jens Eymann |
Gast
|
 |
Beiträge: ---
|
|
|
|
Anmeldedatum: ---
|
|
|
|
Wohnort: ---
|
|
|
|
Version: ---
|
|
|
|
 |
|
Verfasst am: 08.07.2021, 11:18
Titel: EEGLAB restingIAF Übergabeparameter funktionieren nicht
|
 |
| |
 |
 |
Hallo Zusammen,
ich werte EEG Ableitungen mit Matlabund EEGLAB aus, und bin dabei jetzt eine wichtige Peakfrequenzen zu sonderien (PAF oder IAF). Dabei muss ich variablen an das script restingIAF (restingIAF-master) übergeben.
In Zeile 111 bricht das tutorial script immer wieder ab, so dass kein Zugriff auf die Ergebnisse möglich ist.
Wer kann mir da helfen, herauszufinden, wie ich die entsprechenden Datasets übergeben muss, damit das Programm diese erkennt.
Fehlermeldung
>> restingIAF('data', 6, 3, [1 40], 250, [7 13], 11, 5)
Error using restingIAF (line 111)
The value of 'data' is invalid. Expected input to be one of these types:
double, single, uint8, uint16, uint32, uint64, int8, int16, int32, int64
Code mit Fehlermeldung (bitte nach Fehler suchen, damit die Zeile lokalisiert werden kann, Fehler ist drangeschrieben)
function [pSum, pChans, f] = restingIAF(data, nchan, cmin, fRange, Fs, w, Fw, k, varargin)
% Primary function for running `restingIAF` analysis routine for estimating
% two indices of individual alpha frequency (IAF): Peak alpha frequency
% (PAF) and the alpha centre of gravity (CoG) or mean frequency.
%
% Calls on the Signal Processing Toolbox function `pwelch` to derive power
% spectral density estimates of one or more vectors of EEG channel data,
% which are subsequently smoothed by `sgfDiff` (depends on `sgolay`).
% Alpha peak activity is parameterised by `peakBounds` and CoG estimated by
% `chanGravs`. Channel-wise peak and gravity estimates are averaged across
% channels by `chanMeans`.
%
% This function and all custom-designed dependencies are part of the
% `restingIAF` package, (c) Andrew W. Corcoran, 2016-2018.
%
% Please consult our methods paper for a more detailed exposition of the
% analysis routine, factors to consider when selecting parameter settings,
% and a study of its performance on empirical and simulated EEG signals:
%
% Corcoran, A.W., Alday, P.M., Schlesewsky, M., & Bornkessel-Schlesewsky,
% I. (2018). Toward a reliable, automated method of individual alpha
% frequency (IAF) quantification. Psychophysiology, 55(7), e13064.
% doi: 10.1111/psyp.13064.
%
% Each release version is also citable and archived on GitHub, making it
% easier for others to fully replicate your analysis (see README.md).
%
% Visit github.com/corcorana/restingIAF for further info on licencing and
% updates on package development.
%
%% Outputs:
% pSum = structure containing summary statistics of alpha-band parameters
% pChans = structure containing channel-wise spectral and alpha parameter data
% f = trimmed vector of frequency bins resolved by `pwelch`
%
%% Required inputs:
% data = vector or matrix containing continuous EEG channel data
% (matrix rows = channels, cols = sample points)
% nchan = number of channels in data array
% cmin = minimum number of channel estimtes that must be resolved in
% order to calculate average PAF/CoG estimates
% fRange = frequency range to be included in analysis (e.g., [1, 40] Hz)
% Fs = EEG sampling rate
% w = bounds of alpha peak search window (e.g., [7 13])
% Fw = frame width, Savitzky-Golay filter (corresponds to number of
% freq. bins spanned by filter; must be odd)
% k = polynomial order, Savitzky-Golay filter (must be < Fw)
%
%% Optional inputs:
%
% mpow = error bound (s.d.) used to determine threshold differentiating
% substantive peaks from background spectral noise (default = 1)
% mdiff = minimal height difference distinguishing a primary peak from
% competing peaks (default = 0.20; i.e. 20% peak height)
% taper = taper window function applied by `pwelch` (default = 'hamming')
% tlen = length of taper window applied by `pwelch` (default = 4 sec)
% tover = length of taper window overlap in samples (default = 50% window length)
% nfft = specify number of FFT points used to calculate PSD (default =
% next power of 2 above window length)
% norm = normalise power spectra (default = true)
%
%% setup inputParser
p = inputParser;
p.addRequired('data',...
@(x) validateattributes(x, {'numeric'}, ...
{'2d', 'nonempty'}));
p.addRequired('nchan',...
@(x) validateattributes(x, {'numeric'}, ...
{'scalar', 'integer', 'positive'}));
p.addRequired('cmin',...
@(x) validateattributes(x, {'numeric'}, ...
{'scalar', 'integer', 'positive', '<=', size(data, 1)}));
p.addRequired('fRange',...
@(x) validateattributes(x, {'numeric'}, ...
{'integer', 'nonnegative', 'increasing', 'size', [1,2]}));
p.addRequired('Fs',...
@(x) validateattributes(x, {'numeric'}, ...
{'scalar', 'integer', 'positive', '>=', 2*fRange(2)}));
p.addRequired('w',...
@(x) validateattributes(x, {'numeric'}, ...
{'nonnegative', 'increasing', 'size', [1,2], '>', fRange(1), '<', fRange(2)}));
p.addRequired('Fw',...
@(x) validateattributes(x, {'numeric'}, ...
{'scalar', 'integer', 'positive', 'odd'}));
p.addRequired('k',...
@(x) validateattributes(x, {'numeric'}, ...
{'scalar', 'integer', 'positive', '<', Fw }));
p.addOptional('mpow', 1,...
@(x) validateattributes(x, {'numeric'}, ...
{'scalar', 'positive'}));
p.addOptional('mdiff', .20,...
@(x) validateattributes(x, {'numeric'}, ...
{'scalar', '>=', 0, '<=', 1}));
p.addOptional('taper', 'hamming',...
@(x) validateattributes(x, {'char'}, ...
{}));
p.addOptional('tlen', (Fs*4),...
@(x) validateattributes(x, {'numeric'}, ...
{'scalar', 'integer', 'positive'}));
p.addOptional('tover', [],...
@(x) validateattributes(x, {'numeric'}, ...
{'integer', 'nonnegative'}));
p.addOptional('nfft', [],...
@(x) validateattributes(x, {'numeric'}, ...
{'integer', 'positive'}));
p.addOptional('norm', true,...
@(x) validateattributes(x, {'logical'}, ...
{'scalar'}));
p.parse(data, nchan, cmin, fRange, Fs, w, Fw, k, varargin{:}) % Hiertritt der fehler auf
mpow = p.Results.mpow;
mdiff = p.Results.mdiff;
taper = p.Results.taper;
tlen = p.Results.tlen;
tover = p.Results.tover;
nfft = p.Results.nfft;
norm = p.Results.norm;
%%
% struct for channel data (PSD estimates & derivatives, some additional info)
pChans = struct('pxx', [], 'minPow', [], 'd0', [], 'd1', [], 'd2', [],...
'peaks', [], 'pos1', [], 'pos2', [], 'f1', [], 'f2', [], 'inf1', [], 'inf2', [],...
'Q', [],'Qf', [], 'gravs', [], 'selP', [], 'selG', [] );
for kx = 1:nchan
if sum(isnan(data(kx, ))==0 % ensure no channel NaNs
% perform pwelch routine to extract PSD estimates by channel
fh = str2func(taper);
[pxx, f] = pwelch(data(kx,, fh(tlen), tover, nfft, Fs);
% delimit range of freq bins to be included in analysis
frex = dsearchn(f, fRange(1)):dsearchn(f, fRange(2));
f = f(frex);
pxx = pxx(frex); % truncate PSD to frex range
% normalise truncated PSD
if norm == true
pChans(kx).pxx = pxx / mean(pxx);
else
pChans(kx).pxx = pxx;
end
% calculate minPower vector
[pfit, sig] = polyfit(f, log10(pChans(kx).pxx), 1); % fit 1st order poly (regression line) to normalised spectra (log-scaled)
[yval, del] = polyval(pfit, f, sig); % derive yval coefficients of fitted polynomial and delta (std dev) error estimate
pChans(kx).minPow = yval + (mpow * del); % takes [minPowThresh * Std dev] as upper error bound on background spectral noise
% apply Savitzky-Golay filter to fit curves to spectra & estimate 1st and 2nd derivatives
[pChans(kx).d0, pChans(kx).d1, pChans(kx).d2] = sgfDiff(pChans(kx).pxx, Fw, k, Fs, tlen);
% calculate frequency resolution
exp_tlen = nextpow2(tlen);
fres = Fs/2.^exp_tlen;
% take derivatives, find peak(s) and boundaries of alpha band
[pChans(kx).peaks, pChans(kx).pos1, pChans(kx).pos2, pChans(kx).f1, pChans(kx).f2,...
pChans(kx).inf1, pChans(kx).inf2, pChans(kx).Q, pChans(kx).Qf]...
= peakBounds(pChans(kx).d0, pChans(kx).d1, pChans(kx).d2, f, w,...
pChans(kx).minPow, mdiff, fres);
else
warning('Row #%s contains NaNs, skipping channel...', num2str(kx))
nchan = nchan-1; % trim nchans for cx loop later on
end
end
% estimate gravities for smoothed spectra (average IAF window across channels)
[ gravs, selG, iaw ] = chanGravs([pChans(.d0], f, [pChans(.f1], [pChans(.f2] );
% calculate average pt estimates/spectra across k-th channels for each j-th recording
[ selP, pSum ] = chanMeans(gravs, selG, [pChans(.peaks], [pChans(.d0], [pChans(.Qf], cmin);
% retain gravity estimates and selected channels in channel data struct
% (only loop through trimmed channels)
for cx = 1:nchan
pChans(cx).gravs = gravs(cx);
pChans(cx).selP = selP(cx);
pChans(cx).selG = selG(cx);
end
% get total number of chans that contributed to PAF/CoG estimation
pSum.pSel = sum(selP);
pSum.gSel = sum(selG);
pSum.iaw = iaw;
end
| Beschreibung: |
|
Download |
| Dateiname: |
restingIAF-master.zip |
| Dateigröße: |
6.37 MB |
| Heruntergeladen: |
251 mal |
|
|
|
|
|
|
| Harald |
Forum-Meister
|
|
Beiträge: 24.448
|
|
|
|
Anmeldedatum: 26.03.09
|
|
|
|
Wohnort: Nähe München
|
|
|
|
Version: ab 2017b
|
|
|
|
|
|
|
Verfasst am: 09.07.2021, 12:19
Titel:
|
|
Hallo,
ich habe das .zip File entpackt und zudem EEGLAB heruntergeladen und beide Verzeichnisse mit Unterverzeichnissen auf den MATLAB-Pfad gelegt. Dann habe ich tutorial.m aus dem tutorial - Verzeichnis ausgeführt. Ich habe zwar keine Ahnung, was das gemacht hat, aber es lief fehlerfrei durch.
Du schreibst, dass das Tutorial in Zeile 111 abbricht - das ist aber doch eine Leerzeile?
Wo kommt denn die Zeile
her? Was sind die beteiligten Variablen? Ist es wirklich sinnvoll, ' ' um data zu setzen?
Bei Fragen zu einem speziellen Tool kann es generell sinnvoll sein, die Autoren oder andere Nutzer dieses speziellen Tools zu kontaktieren. Ich weiß nicht, wieviele Anwender von EEGLAB sich hier aufhalten.
Grüße,
Harald
_________________
1.) Ask MATLAB Documentation
2.) Search gomatlab.de, google.de or MATLAB Answers
3.) Ask Technical Support of MathWorks
4.) Go mad, your problem is unsolvable ;)
|
|
|
|
| Jens Antwort |
Gast
|
|
Beiträge: ---
|
|
|
|
Anmeldedatum: ---
|
|
|
|
Wohnort: ---
|
|
|
|
Version: ---
|
|
|
|
|
|
|
Verfasst am: 12.07.2021, 10:25
Titel: Danke Dir
|
|
Lieber Harald, danke dir, ich habe vermutlich nicht alle Pfade richtig zusammengelinkt, und den titel von eeglab2021 auf eeglab geändert. Es läuft jetzt bei mir auch.
Vielen Dank,
Jens
|
|
|
|
|
|
|
Einstellungen und Berechtigungen
|
|
Du kannst Beiträge in dieses Forum schreiben.
Du kannst auf Beiträge in diesem Forum antworten.
Du kannst deine Beiträge in diesem Forum nicht bearbeiten.
Du kannst deine Beiträge in diesem Forum nicht löschen.
Du kannst an Umfragen in diesem Forum nicht mitmachen.
Du kannst Dateien in diesem Forum posten
Du kannst Dateien in diesem Forum herunterladen
|
|
Impressum
| Nutzungsbedingungen
| Datenschutz
| FAQ
|  RSS
Hosted by:
Copyright © 2007 - 2024
goMatlab.de | Dies ist keine offizielle Website der Firma The Mathworks
MATLAB, Simulink, Stateflow, Handle Graphics, Real-Time Workshop, SimBiology, SimHydraulics, SimEvents, and xPC TargetBox are registered trademarks and The MathWorks, the L-shaped membrane logo, and Embedded MATLAB are trademarks of The MathWorks, Inc.
|
|
