-
Notifications
You must be signed in to change notification settings - Fork 28
/
limo_central_tendency_and_ci.m
1179 lines (1099 loc) · 55.1 KB
/
limo_central_tendency_and_ci.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
function result=limo_central_tendency_and_ci(varargin)
% The function computes estimates of central tendency (mean, trimmed mean,
% Harell-Davis 0.5 decile, median) with 95% Bayesian Highest Density Intervals.
% INPUTS are either a data matrix, con files or LIMO.mat files.
% If you input LIMO files, estimates of the raw data for the
% categorical variables will be performed (it makes no sense to summarize continuous
% variables). Non overlap of 95% HDI shows univariate and 'non-corrected' significant
% differences. This can also be assessed directly using limo_plot_difference
%
% FORMAT
% limo_central_tendency_and_ci(varargin)
% result = limo_central_tendency_and_ci(varargin)
%
% INPUTS
% limo_central_tendency_and_ci(expected_chan_loc)
% expected_chan_loc is the name of the channel structure from EEGLAB
% this option calls the GUI
%
% limo_central_tendency_and_ci(data, 'Analysis_type',selected_channels,savename)
% data is a [channel * [freq/time] frames * trials/subjects] matrix
% Analysis_type should be 'Mean', 'Trimmed mean', 'HD' or 'Median'
% selected_channels can be [] for all brain or 1 or many channels (1 per trial/subject)
% savename (optional) name for saving files
%
% limo_central_tendency_and_ci(Files, parameters, expected_chan_loc, 'Estimator1', 'Estimator2', selected_channels,savename)
% Files are the full names (with paths) of LIMO.mat files or con files
% parameters are which part of the raw data to analyse based on the design matrix, e.g. [1 2];
% it can also be 'con_X' (x being the contrast number) meaning that the columns of the design matrix spanned by
% computed contrasts will be used (useful when the design change among subjects)
% if Files are contrast files, parameter must be 1
% expected_chan_loc is the channel structure from EEGLAB but for the group of subjects
% Estimator should be 'Mean', 'Weighted mean', 'Trimmed mean', 'HD' or 'Median' (doesn't matter for con files)
% Estimator 1 is applied to trials within-subjects
% Estimator 2 is applied across subjects
% selected_channels can be [] for all brain or 1 or many channels (=nb files)
% savename (optional) name for saving files
%
% OUTPUTS result = limo_central_tendency_and_ci()
% result is a structure with the fields 'subject' and 'central'
% if not called, the equivalent of the results fields are saved on the drive
% result.subjects returns the estimator1 computed per subject DIM [channel freq/time parameter subject]
% result.estimator2 returns the estimator 2 computed across subjects DIM [channel freq/time parameter 3]
% the last dim is 3 for low CI bound, estimator value, high CI bound
% estimator2 van be Median, Harrell_Davis, trimmed_mean, or mean
%
% if empty files are created on the drive (typically when called via GUI)
% for single subject the fomat is channel, freq/time, parameters, subjects
% for the group it's a structure data.estimator2 name and data.limo
%
% Examples Call the GUI
% ------------
% limo_central_tendency_and_ci('limo_gp_level_chanlocs.mat')
%
% Trimmed mean of Beta parameters
% -------------------------------
% data = load('Yr.mat'); for instance betas from a rep- measure ANOVAs
% % being in the ANOVA directory, it will also use LIMO.mat for extra info
% for condition = 1:9
% tmp = squeeze(data.Yr(:,:,:,condition));
% limo_central_tendency_and_ci(tmp, 'Trimmed mean',[],['Condition' num2str(condition)])
% end
%
% Weighed means ERP per subject + group level trimmed mean and 95% HDI
% ---------------------------------------------------------------------
% Files = fullfile('..derivatives/LIMO_studyname','LIMO_files_face_detection_all_Face_time_GLM_Channels_Time_WLS.txt')
% expected_chan_loc = fullfile('.../derivatives','limo_gp_level_chanlocs.mat')
% limo_central_tendency_and_ci(Files, [1 2 3], expected_chan_loc, 'Weighted mean', 'Trimmed mean', [], 'ERPs')
%
% Trimmed mean ERP + 95% HDI for a condition for a single subject from EEGLAB .daterp
% ------------------------------------------------------------------------------------
% data = load('sub-002.daterp','-mat'); % read single trials
% index = arrayfun(@(x) contains(x.type,'famous','IgnoreCase',true), data.trialinfo); % get condition of intetest
% FN = fieldnames(data); all_channels = find(contains(FN,'chan'));
% for channel = length(all_channels):-1:1
% data_matrix(channel,:,:) = data.(FN{channel})(:,index); % make a data matrix
% end
% limo_central_tendency_and_ci(data_matrix, 'Trimmed mean',[],'Famous_trimmed_mean')
%
% Guillaume Rousselet provided the initial code to do the stats
% Cyril Pernet made the interface, organize to suite EEG data etc - version 1. 18 May 2010
% June/July 2013 - Fixed some bugs CP / thx to Benedikt Ehinger
% Novembre 2013 - fixed further issues related to parameter selection CP / thx to Matt Craddock
% version 2 September 2015 - included within subject weighted mean + update for time frequency
% version 3 February 2016 - CP/GAR updated for Bayesian HDI
% version 4+ CP maintenance of inputs/arguments/beta or con/etc .. see gitlog
%
% see also limo_central_estimator.m limo_add_plots.m limo_plot_difference.m
% -------------------------------------------------------------------------
% Copyright (C) LIMO Team 2021
%% file selection and checkings
% -----------------------------
current_dir = pwd; warning off
result = []; % the output if requested
data = []; % the matrix of data to compute summary stats on
if nargin == 3 || nargin == 4
% ------------------------
data = varargin{1};
if ischar(data)
if ~exist(data,'file')
limo_errordlg('%s does not exist',data);
return
else
try
data = load(data);
data = data.(cell2mat(fieldnames(data)));
catch
limo_errordlg('%s is not a matrix',data);
return
end
end
end
if ndims(data)<3 || ndims(data) >4 %#ok<*ISMAT>
if ndims(data) == 2
disp('for 2D data, try using limo_central_estimator.m');
end
limo_errordlg('data in must be 3 or 4 dimensional: [1/all channels], [freq/time] frames, subjects');
return
elseif ndims(data) == 4
limo.Analysis = 'Time-Frequency';
if exist('LIMO.mat','file')
disp('updating data structure with local LIMO.mat')
LIMO = load('LIMO.mat');
limo.Level = LIMO.LIMO.Level;
limo.Type = LIMO.LIMO.Type;
limo.data.sampling_rate = LIMO.LIMO.data.sampling_rate;
limo.data.trim1 = LIMO.LIMO.data.trim1;
limo.data.trim2 = LIMO.LIMO.data.trim2;
limo.data.start = LIMO.LIMO.data.start;
limo.data.end = LIMO.LIMO.data.end;
limo.data.trim_lowf = LIMO.LIMO.data.trim_lowf;
limo.data.trim_highf = LIMO.LIMO.data.trim_highf;
limo.data.lowf = LIMO.LIMO.data.lowf;
limo.data.highf = LIMO.LIMO.data.highf;
limo.data.tf_times = LIMO.LIMO.data.tf_times;
limo.data.tf_freqs = LIMO.LIMO.data.tf_freqs;
if isfield(LIMO.LIMO.data, 'neighbouring_matrix')
limo.data.neighbouring_matrix = LIMO.LIMO.data.neighbouring_matrix;
end
if isfield(LIMO.LIMO.data, 'expected_chanlocs')
limo.data.expected_chanlocs = LIMO.LIMO.data.expected_chanlocs;
end
if isfield(LIMO.LIMO.data, 'chanlocs')
limo.data.expected_chanlocs = LIMO.LIMO.data.chanlocs;
end
end
else
if exist('LIMO.mat','file')
disp('updating data structure with local LIMO.mat')
LIMO = load('LIMO.mat');
limo.Level = LIMO.LIMO.Level;
limo.Analysis = LIMO.LIMO.Analysis;
limo.Type = LIMO.LIMO.Type;
limo.data.sampling_rate = LIMO.LIMO.data.sampling_rate;
limo.data.trim1 = LIMO.LIMO.data.trim1;
limo.data.trim2 = LIMO.LIMO.data.trim2;
limo.data.start = LIMO.LIMO.data.start;
limo.data.end = LIMO.LIMO.data.end;
if isfield(LIMO.LIMO.data, 'timevect')
limo.data.timevect = LIMO.LIMO.data.timevect;
end
if isfield(LIMO.LIMO.data, 'freqlist')
limo.data.expected_chanlocs = LIMO.LIMO.data.freqlist;
end
if isfield(LIMO.LIMO.data, 'neighbouring_matrix')
limo.data.neighbouring_matrix = LIMO.LIMO.data.neighbouring_matrix;
end
if isfield(LIMO.LIMO.data, 'expected_chanlocs')
limo.data.expected_chanlocs = LIMO.LIMO.data.expected_chanlocs;
end
if isfield(LIMO.LIMO.data, 'chanlocs')
limo.data.expected_chanlocs = LIMO.LIMO.data.chanlocs;
end
else
limo.Analysis = 'Time or Frequency';
end
end
Estimator2 = varargin{2};
if strcmpi(Estimator2,'Trimmed mean') || strcmpi(Estimator2,'HD') ...
|| strcmpi(Estimator2,'Median') || strcmpi(Estimator2,'Mean') ...
|| strcmpi(Estimator2,'All')
parameters = 1; %#ok<NASGU>
else
limo_errordlg('type of estimator not recognized');
return
end
selected_channels = varargin{3};
if ~isempty(selected_channels)
Analysis_type = '1 channel only';
if strcmpi(limo.Analysis,'Time-Frequency')
data = data(selected_channels,:,:,:);
else
data = data(selected_channels,:,:);
end
else
Analysis_type = 'Full brain analysis';
end
if nargin == 4
savename = varargin{4};
[p,f,ext]=fileparts(savename);
if strcmp(ext,'.mat')
savename=fullfile(p,f);
end
end
elseif nargin == 6 || nargin == 7
% ---------------------------
if exist(varargin{1},'file')
Files = varargin{1};
if size(Files,1) == 1 % select a txt file listing all files
[Names,Paths,Files] = limo_get_files([],[],[],Files);
end
else
limo_errordlg('input file not found');
return
end
parameters = varargin{2};
is_limo = zeros(1,size(Names,2));
is_con = zeros(1,size(Names,2));
for i=size(Names,2):-1:1
if strfind(Names{i},'LIMO'); is_limo(i) = 1;
elseif strfind(Names{i},'con'); is_con(i) = 1; end
end
if all(is_con) && parameters ~=1; parameters = 1;
warning on; warning('all con files in, parameter set to 1'); warning off
end
if exist(fullfile(pwd,'LIMO.mat'),'file')
disp('updating data structure with local LIMO.mat')
LIMO = load('LIMO.mat');
limo.Level = LIMO.LIMO.Level;
limo.Analysis = LIMO.LIMO.Analysis;
limo.data.sampling_rate = LIMO.LIMO.data.sampling_rate;
limo.data.trim1 = LIMO.LIMO.data.trim1;
limo.data.trim2 = LIMO.LIMO.data.trim2;
limo.data.start = LIMO.LIMO.data.start;
limo.data.end = LIMO.LIMO.data.end;
if isfield(LIMO.LIMO.data, 'timevect')
limo.data.timevect = LIMO.LIMO.data.timevect;
end
if isfield(LIMO.LIMO.data, 'freqlist')
limo.data.expected_chanlocs = LIMO.LIMO.data.freqlist;
end
if isfield(LIMO.LIMO.data, 'neighbouring_matrix')
limo.data.neighbouring_matrix = LIMO.LIMO.data.neighbouring_matrix;
end
if isfield(LIMO.LIMO.data, 'expected_chanlocs')
limo.data.expected_chanlocs = LIMO.LIMO.data.expected_chanlocs;
end
if isfield(LIMO.LIMO.data, 'chanlocs')
limo.data.expected_chanlocs = LIMO.LIMO.data.chanlocs;
end
else
limo.Analysis = 'Time or Frequency';
end
expected_chanlocs = varargin{3};
if ischar(expected_chanlocs)
expected_chanlocs = load(expected_chanlocs);
limo.data.neighbouring_matrix = expected_chanlocs.channeighbstructmat;
limo.data.expected_chanlocs = expected_chanlocs.expected_chanlocs;
expected_chanlocs = limo.data.expected_chanlocs;
else
if isfield(expected_chanlocs,'expected_chanlocs') && ...
isfield(expected_chanlocs,'channeighbstructmat')
limo.data.neighbouring_matrix = expected_chanlocs.channeighbstructmat;
limo.data.expected_chanlocs = expected_chanlocs.expected_chanlocs;
expected_chanlocs = limo.data.expected_chanlocs;
else
limo.data.expected_chanlocs = expected_chanlocs;
end
end
Estimator1 = varargin{4};
Estimator2 = varargin{5};
selected_channels = varargin{6};
if isempty(selected_channels)
Analysis_type = 'Full brain analysis';
else
Analysis_type = '1 channel only';
expected_chanlocs = expected_chanlocs(selected_channels);
end
% match frames
% -------------
[first_frame,last_frame,subj_chanlocs,limo] = limo_match_frames(Paths,limo);
% get data for all parameters dim [channel, frame, param, nb subjects
% ---------------------------------------------------------------------
disp('gathering data ...');
for i=size(Paths,2):-1:1 % for each subject
fprintf('processing subject %g\n',i);
LIMO = load(fullfile(Paths{i},'LIMO.mat'));
LIMO = LIMO.LIMO;
limo.Type{i} = LIMO.Type;
if all(is_limo)
Yr = load(fullfile(Paths{i},'Yr.mat'));
elseif all(is_con)
Yr = load(Files{i});
end
Yr = Yr.(cell2mat(fieldnames(Yr)));
if strcmpi(LIMO.Analysis,'Time-Frequency')
begins_at = fliplr((max(first_frame) - first_frame(i,:) + 1)); % returns time/freq/or freq-time
ends_at(1) = size(Yr,2) - (last_frame(i,2) - min(last_frame(:,2)));
ends_at(2) = size(Yr,3) - (last_frame(i,1) - min(last_frame(:,1)));
else
begins_at = max(first_frame) - first_frame(i) + 1;
ends_at = size(Yr,2) - (last_frame(i) - min(last_frame));
end
if max(parameters) <= sum(LIMO.design.nb_conditions+LIMO.design.nb_interactions) || ...
max(parameters) == size(LIMO.design.X,2) || ...% any categorial or the constant
any([contains(num2str(parameters),{'con'}) strcmp(num2str(parameters),'1')]) % or all con files or 1
if all(is_limo)
if isnumeric(parameters)
index = logical(sum(LIMO.design.X(:,parameters)==1,2));
else
if contains(parameters,'con')
tmp=find(LIMO.contrast{str2double(parameters(5:end))}.C);
index = logical(sum(LIMO.design.X(:,tmp)==1,2)); clear tmp
else
limo_error('unrecognized input parameter')
end
end
for channel=size(Yr,1):-1:1
if strcmpi(Estimator1,'Weighted Mean')
if strcmpi(LIMO.Analysis,'Time-Frequency')
for f=size(Yr,2):-1:1
fw(1,f,:,:) = squeeze(Yr(channel,f,:,index)).*repmat(squeeze(LIMO.design.weights(channel,f,index))',size(Yr,3),1);
end
tmp(channel,:,:) = limo_tf_4d_reshape(fw,LIMO.data.size3D);
clear fw;
else
tmp(channel,:,:) = squeeze(Yr(channel,:,index)).*repmat(LIMO.design.weights(channel,index),size(Yr,2),1);
end
else
tmp(channel,:,index) = squeeze(Yr(channel,:,index));
end
end
% 1st level analysis
% --------------------
if strcmpi(Estimator1,'Trimmed mean') % trim raw data @ 20%
tmp = limo_trimmed_mean(tmp,20);
elseif strcmpi(Estimator1,'Median') % median raw data
tmp = nanmedian(tmp,3);
elseif strcmpi(Estimator1,'HD') % mid-decile Harrell-Davis of raw data
tmp = limo_harrell_davis(tmp,0.5);
elseif strcmpi(Estimator1,'Mean') || strcmpi(Estimator1,'Weighted Mean') % mean of raw data
tmp = nanmean(tmp,3);
end
else
if strcmpi(LIMO.Analysis,'Time-Frequency')
tmp = limo_tf_4d_reshape(Yr,LIMO.data.size3D);
tmp = squeeze(tmp(:,:,1));
else
tmp = Yr;
tmp = squeeze(tmp(:,:,1));
end
end
clear Yr
if strcmpi(Analysis_type,'Full brain analysis') && size(subj_chanlocs(i).chanlocs,2) == size(tmp,1)
if strcmpi(LIMO.Analysis,'Time-Frequency')
data(:,:,:,i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,reshape(tmp,LIMO.data.size4D(1:3)));
else
data(:,:,i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,tmp);
end
elseif strcmpi(Analysis_type,'1 channel only') && length(subj_chanlocs(i).chanlocs) == size(tmp,1)
if strcmpi(LIMO.Analysis,'Time-Frequency')
if size(selected_channels,2) == 1
data(1,:,:,i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,reshape(tmp,LIMO.data.size4D(1:3)));
else
out = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,reshape(tmp,LIMO.data.size4D(1:3)));
data(1,:,:,i) = out(i,:,:);
end
else
if size(selected_channels,2) == 1
data(1,:,i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,tmp);
else
out = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,tmp); % out is for all expected chanlocs, ie across subjects
data(1,:,i) = out(i,:,:); % matches the expected chanloc of the subject
end
end
end
clear tmp
else
if max(parameters) > size(LIMO.design.X,2)
warning('subject %g, parameter %g not computed: \n the design only includes %g regressors plus the constant',Paths{i},size(LIMO.design.X,2));
else
warning('subject %g, \n parameter %g not computed - continuous regressor',Paths{i},max(parameters));
end
end
end
if nargout == 1
result.subjects = data;
end
if nargin == 7
savename = varargin{7};
[p,f,ext]=fileparts(savename);
if strcmp(ext,'.mat')
savename=fullfile(p,f);
end
end
if all(cellfun(@(x) strcmpi(x,limo.Type{1}), limo.Type))
limo.Type = limo.Type{1};
else
limo_errordlg('despite successful data aggregation, LIMO.Type differ?? channels/compomnents/sources - check your data')
return
end
elseif nargin == 1
% ---------------------------
% Expected_chanlocs
expected_chanlocs = load(varargin{1});
% check if Betas/Con
option = limo_questdlg('type of analysis','what data to analyse?','Raw Data','Betas','Con','Betas');
if isempty(option)
return
end
% -----------------------------
% ANALYSIS ON BETAS PARAMETERS
% -----------------------------
if strcmpi(option,'Betas') || strcmpi(option,'Con')
Estimator1 = option;
Estimator2 = limo_questdlg('Estimation option','which estimator?','Mean','Trimmed mean','HD/Median','Trimmed mean');
if strcmpi(Estimator2,'HD/Median')
Estimator2 = 'HD';
end
% get the data
% ------------
Names = {}; %#ok<NASGU>
[Names,Paths,Files] = limo_get_files([],{'*.mat;*.txt','matlab or text'},sprintf('Select %s files',option)); %#ok<ASGLU>
if isempty(Names)
return
elseif size(Names,2) < 3
limo_errordlg('LIMO cannot do group bootrap estimates - too few subjects')
return
end
% check type of files and returns which beta param to test
% -------------------------------------------------------
is_betas = [];
is_con = [];
for i=size(Names,2):-1:1
if strfind(Names{i},'Betas')
is_betas(i) = 1;
elseif strfind(Names{i},'con')
is_con(i) = 1;
end
end
if (isempty(is_betas)) == 0 && sum(is_betas) == size(Names,2)
if strcmpi(Estimator1,'Con')
limo_warndlg('you indicated computation for contrasts, but all files are beta parameters - still computing though',...
'selection warning');
Estimator1 = 'Betas';
end
parameters = limo_inputdlg('which parameters to test e.g [1:3]','parameters option');
if isempty(parameters)
return
else
parameters = cell2mat(parameters);
if ~strcmp(parameters(1),'[') && ~strcmp(parameters(end),']')
parameters = ['[' parameters ']'];
end
parameters = eval(parameters);
end
elseif (isempty(is_con)) == 0 && sum(is_con) == size(Names,2)
if strcmpi(Estimator1,'Betas')
limo_warndlg('you indicated computation for Betas, but all files are contrasts - still computing though',...
'selection warning')
Estimator1 = 'Con';
end
parameters = 1;
else
limo_errordlg('file selection failed, only Betas.mat files are supported');
return
end
% match frames
% ------------
limo.data.neighbouring_matrix = expected_chanlocs.channeighbstructmat;
limo.data.expected_chanlocs = expected_chanlocs.expected_chanlocs;
[first_frame,last_frame,subj_chanlocs,limo] = limo_match_frames(Paths,limo);
limo.Level = 2;
% match channels
% --------------
Analysis_type = limo_questdlg('Rdx option','type of analysis?','Full brain analysis','1 channel only','Full brain analysis');
if isempty(Analysis_type)
return
end
if strcmpi(Analysis_type,'1 channel only')
channel = limo_inputdlg('which channel to analyse [?]','channel option'); % can be 1 nb or a vector of channels (channel optimized analysis)
if isempty(cell2mat(channel))
[file,dirf,index] = uigetfile('*.mat','select your channel file');
if index == 0
return
else
channel_vector = load(fullfile(dirf,file));
channel_vector = channel_vector.cell2mat(fieldname(channel_vector));
% check the vector has the same length as the number of files
if length(channel_vector) ~= size(Names,2)
errordlg('the nb of channels does not match the number of subjects','channel error'); return;
end
% restric the channels
expected_chanlocs = limo.data.expected_chanlocs(channel_vector);
end
elseif size(eval(cell2mat(channel)),2) == 1 || size(eval(cell2mat(channel)),2) == size(Names,2)
selected_channels = eval(cell2mat(channel));
expected_chanlocs = limo.data.expected_chanlocs(selected_channels);
else
limo_errordlg('the nb of channels does not match the number of subjects','channel error');
return
end
else
expected_chanlocs = limo.data.expected_chanlocs;
end
% make one large matrix
disp('gathering data ...'); index = 1;
for i=size(Paths,2):-1:1 % for each subject
fprintf('processing subject %g\n',i);
% load file and store contend
LIMO = load([Paths{i} filesep 'LIMO.mat']); LIMO = LIMO.LIMO;
Yr = load([Paths{i} filesep Names{i}]);
Yr = Yr.(cell2mat(fieldnames(Yr)));
if strcmpi(LIMO.Analysis,'Time-Frequency')
begins_at = fliplr((max(first_frame) - first_frame(i,:) + 1)); % returns time/freq/or freq-time
ends_at(1) = size(Yr,2) - (last_frame(i,2) - min(last_frame(:,2)));
ends_at(2) = size(Yr,3) - (last_frame(i,1) - min(last_frame(:,1)));
else
begins_at = max(first_frame) - first_frame(i) + 1;
ends_at = size(Yr,2) - (last_frame(i) - min(last_frame));
end
if strcmpi(Analysis_type,'Full brain analysis')
if strcmpi(LIMO.Analysis,'Time-Frequency')
data(:,:,:,:,i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,squeeze(Yr(:,:,:,parameters)));
else
data(:,:,i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,squeeze(Yr(:,:,parameters)));
end
elseif strcmpi(Analysis_type,'1 channel only')
if size(selected_channels,2) == 1
if strcmpi(LIMO.Analysis,'Time-Frequency')
data(1,:,:,1:length(parameters),i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,squeeze(Yr(:,:,:,parameters)));
else
data(1,:,1:length(parameters),i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,squeeze(Yr(:,:,parameters)));
end
else % optimized channel
if strcmpi(LIMO.Analysis,'Time-Frequency')
out = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,squeeze(Yr(:,:,:,parameters))); % out is for all expected chanlocs, i.e. across subjects
data(1,:,:,:,i) = out(i,:,:,:); % matches the expected chanloc of the subject
else
out = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,squeeze(Yr(:,:,parameters))); % out is for all expected chanlocs, i.e. across subjects
data(1,:,:,i) = out(i,:,:); % matches the expected chanloc of the subject
end
end
end
clear tmp
end
else
% -------------------
% ANALYSIS ON ERPs
% -------------------
% select data
% -----------
[Names,Paths,Files] = limo_get_files([],{'*.mat;*.txt','matlab or text'},'Select LIMO files'); %#ok<ASGLU>
if isempty(Names)
return
elseif size(Names,2) < 3
limo_errordlg('LIMO cannot do group bootrap estimates - too few subjects')
return
end
% check it's LIMO.mat files and which param to test
% --------------------------------------------------
is_limo = [];
for i=size(Names,2):-1:1
if strcmpi(Names{i},'LIMO.mat')
is_limo(i) = 1;
end
end
if (isempty(is_limo)) == 0 && sum(is_limo) == size(Names,2)
Q = limo_questdlg('Type of merging','Options','Evaluate single conditions','Pool Conditions','Evaluate single conditions');
if strcmpi(Q,'Evaluate single conditions')
parameters = limo_inputdlg('which parameters to test e.g [1:3]','parameters option');
else
parameters = limo_inputdlg('which parameters to pool e.g [1 3 5]','parameters option');
end
if isempty(parameters)
return
else
parameters = eval(cell2mat(parameters));
if isnan(parameters)
parameters = str2double(cell2mat(parameters));
end
end
else
limo_errordlg('file selection failed, only LIMO.mat files are supported');
return
end
% check what type of analysis
% ---------------------------
Analysis_type = limo_questdlg('Rdx option','type of analysis?','Full brain analysis','1 channel only','Full brain analysis');
if isempty(Analysis_type)
return;
else
limo.Type = 'Channel';
end
limo.data.neighbouring_matrix = expected_chanlocs.channeighbstructmat;
if strcmpi(Analysis_type,'1 channel only')
channel = limo_inputdlg('which channel to analyse [?]','channel option'); % can be 1 nb or a vector of channels (channel optimized analysis)
if isempty(cell2mat(channel))
[file,dir,index] = uigetfile('*.mat','select your channel file');
if isempty(file)
return
else
cd(dir);
channel_vector = load(file);
channel_vector = channel_vector.getfield(channel_vector);
% check the vector has the same length as the number of files
if length(channel_vector) ~= length(Paths)
errordlg('the nb of channels does not match the number of subjects','channel error'); return;
end
selected_channels = channel_vector;
expected_chanlocs = expected_chanlocs.expected_chanlocs(selected_channels);
end
elseif size(eval(cell2mat(channel)),2) == 1 || size(eval(cell2mat(channel)),2) == size(Names,2)
selected_channels = eval(cell2mat(channel));
expected_chanlocs = expected_chanlocs.expected_chanlocs(selected_channels);
else
limo_errordlg('the nb of channels does not match the number of subjects','channel error');
return;
end
else
selected_channels = [];
expected_chanlocs = expected_chanlocs.expected_chanlocs;
end
limo.data.expected_chanlocs = expected_chanlocs;
% select method
% -------------
[Estimator1,Estimator2] = limo_central_tendency_questdlg;
if isempty(Estimator1) && isempty(Estimator2)
return
end
if strcmpi(Estimator1,'All') || strcmpi(Estimator1,'Mean')
weighted_mean = limo_questdlg('do you want to use weights to compute means?','saving option','yes','no','yes');
end
% match frames
% -------------
[first_frame,last_frame,subj_chanlocs,limo] = limo_match_frames(Paths,limo);
limo.Level = 2;
% get data for all parameters dim [channel, frame, param, nb subjects
% ---------------------------------------------------------------------
disp('gathering data ...');
for i=size(Paths,2):-1:1 % for each subject
fprintf('processing subject %g',i); disp(' ')
LIMO = load(fullfile(Paths{i},'LIMO.mat')); LIMO = LIMO.LIMO;
Yr = load(fullfile(Paths{i},'Yr.mat')); Yr = Yr.Yr;
if strcmpi(LIMO.Analysis,'Time-Frequency')
begins_at = fliplr((max(first_frame) - first_frame(i,:) + 1)); % returns time/freq/or freq-time
ends_at(1) = size(Yr,2) - (last_frame(i,2) - min(last_frame(:,2)));
ends_at(2) = size(Yr,3) - (last_frame(i,1) - min(last_frame(:,1)));
else
begins_at = max(first_frame) - first_frame(i) + 1;
ends_at = size(Yr,2) - (last_frame(i) - min(last_frame));
end
if strcmpi(Q,'Evaluate single conditions')
for j=length(parameters):-1:1
if parameters(j) <= sum(LIMO.design.nb_conditions+LIMO.design.nb_interactions) || ...
parameters(j) == size(LIMO.design.X,2)
index = LIMO.design.X(:,parameters(j))==1;
if strcmpi(weighted_mean,'yes')
for channel=1:size(Yr,1)
if strcmpi(LIMO.Analysis,'Time-Frequency')
for f=size(Yr,2):-1:1
fw(1,f,:,:) = squeeze(Yr(channel,f,:,index)).*repmat(squeeze(LIMO.design.weights(channel,f,index))',size(Yr,3),1);
end
tmp(channel,:,:) = limo_tf_4d_reshape(fw,LIMO.data.size3D);
clear fw;
else
tmp(channel,:,:) = squeeze(Yr(channel,:,index)).*repmat(LIMO.design.weights(channel,index),size(Yr,2),1);
end
end
else
tmp = squeeze(Yr(:,:,index)); % retain those trials only
end
% 1st level analysis
% --------------------
if strcmpi(Estimator1,'Trimmed mean') % trim raw data @ 20%
tmp = limo_trimmed_mean(tmp,20);
elseif strcmpi(Estimator1,'Median') % median raw data
tmp = nanmedian(tmp,3);
elseif strcmpi(Estimator1,'HD') % mid-decile Harrell-Davis of raw data
tmp = limo_harrell_davis(tmp,0.5);
elseif strcmpi(Estimator1,'Mean') % mean of raw or weighted data
tmp = nanmean(tmp,3);
end
if strcmpi(Analysis_type,'Full brain analysis') && length(subj_chanlocs(i).chanlocs) == size(tmp,1)
if strcmpi(LIMO.Analysis,'Time-Frequency')
data(:,:,:,j,i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,reshape(tmp,LIMO.data.size4D(1:3)));
else
data(:,:,j,i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,tmp);
end
elseif strcmpi(Analysis_type,'1 channel only') && length(subj_chanlocs(i).chanlocs) == size(tmp,1)
if strcmpi(LIMO.Analysis,'Time-Frequency')
if size(selected_channels,2) == 1
data(1,:,:,j,i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,reshape(tmp,LIMO.data.size4D(1:3)));
else
out = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,reshape(tmp,LIMO.data.size4D(1:3)));
data(1,:,:,j,i) = out(i,:,:);
end
else
if size(selected_channels,2) == 1
data(1,:,j,i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,tmp);
else
out = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,tmp); % out is for all expected chanlocs, ie across subjects
data(1,:,j,i) = out(i,:,:); % matches the expected chanloc of the subject
end
end
end
clear tmp
else
if max(j) > size(LIMO.design.X,2)
warning('subject %g, parameter %g not computed: \n the design only includes %g regressors plus the constant',Paths{i},size(LIMO.design.X,2));
else
warning('subject %g, \n parameter %g not computed - continuous regressor',Paths{i},j);
end
end
end
elseif strcmpi(Q,'Pool Conditions')
if max(parameters) <= sum(LIMO.design.nb_conditions)+sum(LIMO.design.nb_interactions) || ...
max(parameters) == size(LIMO.design.X)
index = find(sum(LIMO.design.X(:,parameters)==1,2)); % find all trials from selected columns
if strcmpi(weighted_mean,'yes')
for channel=size(Yr,1):-1:1
if strcmpi(LIMO.Analysis,'Time-Frequency')
for f=size(Yr,2):-1:1
fw(1,f,:,:) = squeeze(Yr(channel,f,:,index)).*repmat(squeeze(LIMO.design.weights(channel,f,index))',size(Yr,3),1);
end
tmp(channel,:,:) = limo_tf_4d_reshape(fw,LIMO.data.size3D);
clear fw;
else
tmp(channel,:,:) = squeeze(Yr(channel,:,index)).*repmat(LIMO.design.weights(channel,index),size(Yr,2),1);
end
end
else
if strcmpi(LIMO.Analysis,'Time-Frequency')
tmp = limo_tf_4d_reshape(squeeze(Yr(:,:,:,index)),LIMO.data.size3D);
else
tmp = squeeze(Yr(:,:,index)); % retain those trials only
end
end
% 1st level analysis
% --------------------
if strcmpi(Estimator1,'Trimmed mean') % trim raw data @ 20%
tmp=limo_trimmed_mean(tmp,20);
elseif strcmpi(Estimator1,'Median') % median raw data
tmp = nanmedian(tmp,3);
elseif strcmpi(Estimator1,'HD') % mid-decile Harrell-Davis of raw data
tmp = limo_harrell_davis(tmp,0.5);
elseif strcmpi(Estimator1,'Mean') % mean of raw data on which we do across subjects TM, HD and Median
tmp = nanmean(tmp,3);
end
if strcmpi(Analysis_type,'Full brain analysis') && length(subj_chanlocs(i).chanlocs) == size(tmp,1)
if strcmpi(LIMO.Analysis,'Time-Frequency')
data(:,:,:,i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,reshape(tmp,LIMO.data.size4D(1:3)));
else
data(:,:,i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,tmp);
end
elseif strcmpi(Analysis_type,'1 channel only') && length(subj_chanlocs(i).chanlocs) == size(tmp,1)
if strcmpi(LIMO.Analysis,'Time-Frequency')
if size(selected_channels,2) == 1
data(1,:,:,i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,reshape(tmp,LIMO.data.size4D(1:3)));
else
out = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,reshape(tmp,LIMO.data.size4D(1:3))); % out is for all expected chanlocs, ie across subjects
data(1,:,:,i) = out(i,:,:,:); % matches the expected chanloc of the subject
end
else
if size(selected_channels,2) == 1
data(1,:,i) = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,tmp);
else
out = limo_match_elec(subj_chanlocs(i).chanlocs,expected_chanlocs,begins_at,ends_at,tmp); % out is for all expected chanlocs, ie across subjects
data(1,:,i) = out(i,:,:); % matches the expected chanloc of the subject
end
end
end
clear tmp
else
fprintf('pooling not computed - one or more continuous regressor selected \n');
end
end
clear Yr
end
% update estimator1 name
if strcmpi(weighted_mean,'yes')
Estimator1 = 'Weighted mean';
end
end
else
limo_errordlg('nb of arguments incorrect');
return
end % closes varargin
%% Analysis part
% --------------
cd(current_dir)
if ~isempty(data)
% Data is either [channel, frame, trials/subject] or [channel,
% frame, conditions (from parameters), subjects] but we always want 4D
% or 5D data with 1 or more conditions
if ~strcmpi(limo.Analysis,'Time-Frequency') && ndims(data) == 3
tmp = data; clear data
for i=size(tmp,1):-1:1
for j=size(tmp,2):-1:1
data(i,j,1,:) = tmp(i,j,:); % now data is 4D
end
end
elseif strcmpi(limo.Analysis,'Time-Frequency') && ndims(data) == 4
tmp = data; clear data
for i=size(tmp,1):-1:1
for j=size(tmp,2):-1:1
for k=size(tmp,3):-1:1
data(i,j,k,1,:) = tmp(i,j,k,:); % now data is 5D
end
end
end
end
n = size(data,ndims(data)); % number of subjects always last
if ndims(data) < 4
limo_errordlg('an unexpected issue occured, the number of dimensions is too low, likely caused by selected only 1 subject')
return
elseif n < 3
limo_errordlg('LIMO cannot do group bootrap estimates - too few subjects')
return
end
if n<=10 && strcmpi(Estimator2,'HD')
msgbox('CI of the Harell Davis estimates cannot be computed for less than 11 observations - switched to median','Computation info');
Estimator2 = 'Median';
end
% save as
if nargout ==0
if exist('savename','var')
name = savename;
else
name = cell2mat(limo_inputdlg('save as [?]','name option'));
if isempty(name)
disp('no name selected - aborded'); return
end
end
if exist('Estimator1','var')
newname = sprintf('%s_single_subjects_%s',name,Estimator1);
if ~strcmpi(limo.Analysis,'Time-Frequency')
Data.data = data; Data.limo = limo;
save (newname,'Data'); clear Data
elseif strcmpi(limo.Analysis,'Time-Frequency')
Data.data = data; Data.limo = limo;
save (newname,'Data'); clear Data
end
end
else
result.subjects = data;
if exist('limo','var')
result.limo = limo;
end
end
disp('processing data across subjects ..')
% --------------------------------------------------------------
if nargout == 1 && exist('limo','var')
result.limo = limo;
end
if strcmpi(Estimator2,'Mean') || strcmpi(Estimator2,'All')
disp('Compute the Mean estimator and 95% CI ...')
index = 1; h = waitbar(0,'computing','name','% done');
if strcmpi(limo.Analysis,'Time-Frequency')
M = NaN(size(data,1),size(data,2),size(data,3),size(data,4),3);
for k = 1:size(data,4)
for channel =1:size(data,1)
waitbar(index/(size(data,4)*size(data,1)));
index = index+1;
if strcmpi(Analysis_type,'1 channel only')
for f=size(data,2):-1:1
tmp(1,f,:,:) = data(1,f,:,k,:);
end
tmp = limo_tf_4d_reshape(tmp,...
[size(data,1) size(data,2)*size(data,3) size(data,5)]);
else
tmp = limo_tf_4d_reshape(squeeze(data(:,:,:,k,:)),...
[size(data,1) size(data,2)*size(data,3) size(data,5)]);
end
tmp = squeeze(tmp(channel,:,:));
Y = tmp(:,~isnan(tmp(1,:)));
[est,ci] = limo_central_estimator(Y,'mean');
M(channel,:,:,k,1) = reshape(ci(1,:),size(data,2),size(data,3));
M(channel,:,:,k,2) = reshape(est,size(data,2),size(data,3));
M(channel,:,:,k,3) = reshape(ci(2,:),size(data,2),size(data,3));
end
end
else
M = NaN(size(data,1),size(data,2),size(data,3),3);
for k = 1:size(data,3)
for channel =1:size(data,1)
waitbar(index/(size(data,3)*size(data,1)));
index = index+1;
tmp = squeeze(data(channel,:,k,:));
Y = tmp(:,~isnan(tmp(1,:)));
[est,ci] = limo_central_estimator(Y,'mean');
M(channel,:,k,1) = ci(1,:);
M(channel,:,k,2) = est;
M(channel,:,k,3) = ci(2,:);
end
end
end
close(h);
if nargout ==0
if nargin == 3 || nargin == 4
newname = sprintf('%s_Mean',name);
else
newname = sprintf('%s_Mean_of_%s',name,Estimator1);
end
Data.mean = M;
if exist('limo','var')
Data.limo = limo;
end
save (newname,'Data');
else
result.mean = M;
end
end
% --------------------------------------------------------------
if strcmpi(Estimator2,'Trimmed mean') || strcmpi(Estimator2,'All')
disp('Compute 20% Trimmed Mean estimator and 95% CI ...')
index = 1; h = waitbar(0,'computing','name','% done');
if strcmpi(limo.Analysis,'Time-Frequency')
TM = NaN(size(data,1),size(data,2),size(data,3),size(data,4),3);
for k = 1:size(data,4)
for channel =1:size(data,1)
waitbar(index/(size(data,4)*size(data,1)));
index = index+1;
if strcmpi(Analysis_type,'1 channel only')
for f=size(data,2):-1:1
tmp(1,f,:,:) = data(1,f,:,k,:);
end
tmp = limo_tf_4d_reshape(tmp,...
[size(data,1) size(data,2)*size(data,3) size(data,5)]);
else