-
Notifications
You must be signed in to change notification settings - Fork 1
/
nemar_dataqual.m
346 lines (322 loc) · 15 KB
/
nemar_dataqual.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
function [STUDY, ALLEEG] = nemar_dataqual(dsnumber, STUDY, ALLEEG)
outpath = sprintf('/expanse/projects/nemar/openneuro/processed/%s', dsnumber);
eeglabroot = '/expanse/projects/nemar/eeglab';
pipelineroot = fullfile(eeglabroot, 'plugins', 'NEMAR-pipeline');
addpath(fullfile(pipelineroot,'JSONio'));
if nargin < 2
studyFile = fullfile(outpath, [dsnumber '.study']);
[STUDY, ALLEEG] = pop_loadstudy(studyFile);
end
nGoodData = [];
goodDataPercentRaw = [];
nGoodChans = [];
goodChansPercentRaw = [];
icaFail = [];
nICs = [];
nGoodICs = [];
goodICAPercentRaw = [];
linenoise_magn = [];
check_chanlocs = [];
maxCount = 0;
for i=1:numel(ALLEEG)
EEG = ALLEEG(i);
dataqual_path = fullfile(EEG.filepath, [EEG.filename(1:end-4) '_dataqual.json']);
cur_report = jsonread(dataqual_path);
if isfield(cur_report, 'nGoodData')
nGoodData = [nGoodData str2num(cur_report.nGoodData)];
[count, edges] = histcounts(nGoodData);
%maxCount = max([maxCount max(count)]);
end
if isfield(cur_report, 'goodDataPercentRaw')
goodDataPercentRaw = [goodDataPercentRaw str2num(cur_report.goodDataPercentRaw)];
[count, edges] = histcounts(goodDataPercentRaw);
maxCount = max([maxCount max(count)]);
end
if isfield(cur_report, 'nGoodChans')
nGoodChans = [nGoodChans cur_report.nGoodChans];
[count, edges] = histcounts(nGoodChans);
maxCount = max([maxCount max(count)]);
end
if isfield(cur_report, 'goodChansPercentRaw')
goodChansPercentRaw = [goodChansPercentRaw str2num(cur_report.goodChansPercentRaw)];
[count, edges] = histcounts(goodChansPercentRaw);
%maxCount = max([maxCount max(count)]);
end
if isfield(cur_report, 'icaFail')
icaFail = [icaFail cur_report.icaFail];
[count, edges] = histcounts(nICs);
%maxCount = max([maxCount max(count)]);
end
if isfield(cur_report, 'nICs')
nICs = [nICs cur_report.nICs];
[count, edges] = histcounts(nICs);
%maxCount = max([maxCount max(count)]);
end
if isfield(cur_report, 'nGoodICs')
nGoodICs = [nGoodICs cur_report.nGoodICs];
[count, edges] = histcounts(nGoodICs);
%maxCount = max([maxCount max(count)]);
end
if isfield(cur_report, 'goodICAPercentRaw')
goodICAPercentRaw = [goodICAPercentRaw str2num(cur_report.goodICAPercentRaw)];
[count, edges] = histcounts(goodICAPercentRaw);
maxCount = max([maxCount max(count)]);
end
if isfield(cur_report, 'linenoise_magn')
linenoise_magn = [linenoise_magn str2num(cur_report.linenoise_magn(1:numel(cur_report.linenoise_magn)-2))];
[count, edges] = histcounts(linenoise_magn);
maxCount = max([maxCount max(count)]);
end
end
% generate histogram figures
generate_figure(dsnumber, numel(ALLEEG), goodChansPercentRaw, goodDataPercentRaw, goodICAPercentRaw, linenoise_magn, maxCount);
% update nemar.json
nemarjson_path = fullfile(outpath, 'code', 'nemar.json');
nemarjson = jsonread(nemarjson_path);
if isfield(nemarjson, 'data_quality')
nemarjson.data_quality.good_channels_lower_bound = min(nGoodChans);
nemarjson.data_quality.good_channels_higher_bound = max(nGoodChans);
nemarjson.data_quality.good_data_lower_bound = min(nGoodData);
nemarjson.data_quality.good_data_higher_bound = max(nGoodData);
nemarjson.data_quality.good_brain_ica_lower_bound = min(nGoodICs);
nemarjson.data_quality.good_brain_ica_higher_bound = max(nGoodICs);
nemarjson.data_quality.pipeline_n_dataset = numel(ALLEEG);
nemarjson.data_quality.pipeline_import = numel(ALLEEG);
preprocess_path = fullfile(outpath, 'logs', 'ind_pipeline_status.csv');
if exist(preprocess_path, 'file')
status_tbl = readtable(preprocess_path);
nemarjson.data_quality.check_import = sum(status_tbl.check_import);
nemarjson.data_quality.check_chanloc = sum(status_tbl.check_chanloc);
end
end
plot = [];
plot.title = "Histogram of data quality measures";
plot.extension = [dsnumber '_histogram.png'];
nemarjson.plots = [plot];
jsonwrite(nemarjson_path, nemarjson); % it's collapsing the plots array right now. Not working
function generate_figure(dsnumber, total_count, goodChansPercentRaw, goodDataPercentRaw, goodICAPercentRaw, linenoise_magn, maxCount)
% figure('position', [629 759 896 578], 'color', 'w');
fig = figure('position', [629 759 896 878], 'color', 'w');
set(fig,'defaultAxesColorOrder',[[0 0 0]; [0 0 0]]);
fontsize = 15;
subplot(3,2,1)
cleanraw_count = numel(goodDataPercentRaw);
ica_count = numel(goodICAPercentRaw);
process_status = [cleanraw_count total_count-cleanraw_count; ica_count total_count-ica_count];
ba = bar(process_status, 'BarLayout', 'stacked', 'BarWidth', 0.4); %, 'FaceAlpha', 0.5);
ba(1).FaceColor = [8/255, 135/255, 1/255];
ba(2).FaceColor = [216/255, 44/255, 1/255];
title('Pipeline success','FontWeight','Normal');
ylabel(sprintf('Recordings (# of %d)', total_count));
set(gca, 'XTickLabel',{sprintf(" Data\\newlinecleaning") sprintf(" ICA\\newlinedecomp")});
legend({'Finished' 'Failed' }, 'Location', 'southeast');
lims = ylim;
ylim([lims(1) lims(2)+round(lims(2)/50)])
set(gca, 'fontsize', fontsize);
subplot(3,2,2)
bins = 5:10:95;
colorGood = [0.5 1 0.5]; % #088701
colorBad = [0.5 0 0]; % #d82c01
jetColor = jet;
halfJet = jetColor(end:-1:129,:);
[counts, edges] = histcounts(goodDataPercentRaw, bins)
yyaxis right
b = bar(edges(1:end-1), counts, 'BarWidth', 0.8, 'FaceColor', 'flat');%[0.4660 0.6740 0.1880]);
colorsIdx = round(linspace(1,size(halfJet,1),numel(counts)));
colorscheme = halfJet(colorsIdx, :) % interp1([0;1],[colorBad; colorGood],round(linspace(1,numel(halfJet),numel(bins))));
colormap(colorscheme)
colorbar('Location', 'westoutside', 'Ticks',[0, 1], 'TickLabels', {'Poor', 'Good'});
% colorscheme = interp1([0;1],[colorBad; colorGood],linspace(0,1,numel(counts)));
b.CData = colorscheme;
xlim([0 100])
xticks(0:20:100)
limits = ylim;
title('Data frames','FontWeight','Normal');
xlabel('Data frames retained (%)');
ylabel(sprintf('Recordings (# of %d)', total_count), 'Color', 'k');
set(gca, 'fontsize', fontsize);
yyaxis left
yticklabels({[]})
subplot(3,2,3)
bins = min(linenoise_magn):2:max(linenoise_magn);
[counts, edges] = histcounts(linenoise_magn, bins)
b = bar(edges(1:end-1), counts, 'BarWidth', 0.8, 'FaceColor', 'flat');%[0.4660 0.6740 0.1880]);
colorsIdxLinenoise = round(linspace(1,size(halfJet,1),numel(counts)));
colorschemeLinenoise = halfJet(flip(colorsIdxLinenoise), :) % interp1([0;1],[colorBad; colorGood],round(linspace(1,numel(halfJet),numel(bins))));
colormap(colorschemeLinenoise);
b.CData = colorschemeLinenoise;
title('Line noise','FontWeight','Normal');
ylabel(sprintf('Recordings (# of %d)', total_count));
xlabel('Channel-RMS line noise (dB)');
set(gca, 'fontsize', fontsize);
subplot(3,2,4)
% chans_below_threshold = goodChansPercentRaw(goodChansPercentRaw<90);
% chans_above_threshold = goodChansPercentRaw(goodChansPercentRaw>90);
% max_chans = max([chans_below_threshold, chans_above_threshold]);
bins = 5:10:95;
[counts, edges] = histcounts(goodChansPercentRaw, bins)
yyaxis right
b = bar(edges(1:end-1), counts, 'BarWidth', 0.8, 'FaceColor', 'flat');%[0.4660 0.6740 0.1880]);
% colorscheme = interp1([0;1],[colorBad; colorGood],linspace(0,1,numel(counts)));
colormap(colorscheme);
colorbar('Location', 'westoutside', 'Ticks',[0, 1], 'TickLabels', {'Poor', 'Good'});
b.CData = colorscheme;
% hist2(chans_above_threshold, chans_below_threshold, bins);
title('Data channels','FontWeight','Normal');
xlabel(sprintf('Data channels retained (%%)\n '));
ylabel(sprintf('Recordings (# of %d)', total_count), 'Color', 'k');
% legend({'Good' 'Problematic' },'Location','northeast');
xlim([0 100])
xticks(0:20:100)
% limits = ylim
% limits
% yticks([1:round((limits(2)-1)/5):limits(2)])
set(gca, 'fontsize', fontsize);
yyaxis left
yticklabels({[]})
ylabel('');
subplot(3,2,[5 6])
generateParticipantFig(dsnumber)
set(gca, 'fontsize', fontsize);
print(gcf,'-dsvg','-noui',fullfile(outpath, 'code', [ dsnumber '_histogram.svg' ]))
print(gcf,'-dpng','-noui',fullfile(outpath, 'code', [ dsnumber '_histogram.png' ]))
close
end
function res = has_file(dspath, fileext)
res = 0;
dsdir = dir(dspath);
curfile = [];
while ~isempty(dsdir)
curfile = dsdir(1);
dsdir(1) = [];
if curfile.isdir
if ~any(strcmp(curfile.name, {'.','..'}))
child_dir = dir(fullfile(curfile.folder, curfile.name));
dsdir = [dsdir; child_dir];
end
else
if endsWith(curfile.name, fileext)
res = 1;
return
end
end
end
end
function res = has_hed(dspath)
res = 0;
dsdir = dir(dspath);
curfile = [];
while ~isempty(dsdir)
curfile = dsdir(1);
dsdir(1) = [];
if curfile.isdir
if ~any(strcmp(curfile.name, {'.','..'}))
child_dir = dir(fullfile(curfile.folder, curfile.name));
dsdir = [dsdir; child_dir];
end
else
if endsWith(curfile.name, 'events.json')
eventjson = jsonread(fullfile(curfile.folder, curfile.name));
% if any of the event key has HED. Structure format: column -> HED
columns = fieldnames(eventjson);
for c=1:numel(columns)
col_struct = eventjson.(columns{c});
if isfield(col_struct, 'HED')
res = 1;
return
end
end
end
end
end
end
function hist2(data1, data2, bins);
% HIST2 - draw superimposed histograms
%
% Usage:
% >> hist2(data1, data2, bins);
%
% Inputs:
% data1 - data to plot first process
% data2 - data to plot second process
%
% Optional inputs:
% bins - vector of bin center
%
% Author: Arnaud Delorme (SCCN, UCSD)
if nargin < 1
help hist2;
return;
end
if nargin < 3
bins = linspace(min(min(data1), min(data2)), max(max(data1), max(data2)), 100);
elseif length(bins) == 1
bins = linspace(min(min(data1), min(data2)), max(max(data1), max(data2)), bins);
end
hist(data1, bins);
hold on; hist(data2, bins);
%figure; hist( [ measure{:,5} ], 20);
%hold on; hist([ measure{:,2} ], 20);
c = get(gca, 'children');
set(gca, 'fontname', 'arial');
numfaces = size(get(c(1), 'Vertices'),1);
set(c(1), 'FaceVertexCData', repmat([1 0 0], [numfaces 1]), 'Cdatamapping', 'direct', 'edgecolor', 'k', 'facecolor',[0.64 0.13 0.08]); %[0.8500 0.3250 0.0980]);
numfaces = size(get(c(2), 'Vertices'),1);
set(c(2), 'FaceVertexCData', repmat([0 0 1], [numfaces 1]), 'Cdatamapping', 'direct', 'edgecolor', 'k', 'facecolor', [0.4660 0.6740 0.1880]);
ylabel('Number of values');
% xlim([bins(1) bins(end)]);
% yl = ylim;
% xl = xlim;
% line([xl(1) xl(1)]+(xl(2)-xl(1))/2000, yl, 'color', 'k');
% line(xl, [yl(1) yl(1)]+(yl(2)-yl(1))/2000, 'color', 'k');
end
function generateParticipantFig(dsnumber)
participant_tsv = sprintf('/expanse/projects/nemar/openneuro/%s/participants.tsv', dsnumber);
% read the participants.tsv file
participant_tbl = readtable(participant_tsv, 'FileType', 'text', 'Delimiter', '\t');
% get the participant_id column
participant_id = participant_tbl.participant_id;
% get the age column
if ismember('age', participant_tbl.Properties.VariableNames)
age = participant_tbl.age;
elseif ismember('AGE', participant_tbl.Properties.VariableNames)
age = participant_tbl.AGE;
end
% get the sex column
if ismember('gender', participant_tbl.Properties.VariableNames)
gender = participant_tbl.gender;
elseif ismember('GENDER', participant_tbl.Properties.VariableNames)
gender = participant_tbl.GENDER;
end
age_male = age(strcmp(gender, 'M'));
age_female = age(strcmp(gender, 'F'));
% create a figure
gaps = 10;
age_range = max(age) - min(age);
age_range_buffer = round(0.1*age_range);
edges = min(age)-age_range_buffer:round(0.1*((max(age)+age_range_buffer)-(min(age)-age_range_buffer))):max(age)+age_range_buffer %round(linspace(min(age)-1, max(age)+1, gaps))
[~, edges] = histcounts(age, edges)
[female_age_counts,~] = histcounts(age_female, edges)
[male_age_counts,~] = histcounts(age_male, edges)
% center_female = 0.5*(edges(1:end-1)+edges(2:end));
% center_male = 0.5*(edges(1:end-1)+edges(2:end));
% edges = round(edges)
center = 0.5*(edges(1:end-1)+edges(2:end))
age_hist = [male_age_counts; female_age_counts];
b = bar(center, age_hist, 'BarLayout', 'stacked', 'BarWidth', 1);
title({'';'Age and gender'},'FontWeight','Normal');
b(1).FaceColor =[5/255 56/255 138/255]; % #05388a male
b(1).DisplayName = 'Male'; % #05388a male
b(2).FaceColor =[212/255 95/255 154/255]; % #D45F9A female
b(2).DisplayName = 'Female'; % #D45F9A female
%bar(center, male_age_counts, 'BarWidth', 0.8,'FaceColor',[0.64 0.13 0.08]); hold on; bar(center, female_age_counts, 'BarWidth', 0.8,'FaceColor', [0.4660 0.6740 0.1880]);
xticks(edges);
ylabel(sprintf('Subjects (# of %d)', numel(participant_id)));
xlabel('Subject age by gender (years)')
y = ylim;
ylim([0 y(2) + ceil(y(2)*0.2)]);
% legend({sprintf('%d Female', numel(age_female)) sprintf('%d Male', numel(age_male))});
legend([b(2), b(1)]);
% hist2(age_male, age_female)
end
end