-
Notifications
You must be signed in to change notification settings - Fork 34
/
mv_plot_result.m
245 lines (221 loc) · 9.99 KB
/
mv_plot_result.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
function h = mv_plot_result(result, varargin)
%Provides a simple visual representation of the results obtained with the
%functions mv_classify_across_time, mv_classify_timextime, mv_classify,
%and mv_regress.
%
%The type of plot depends on which of these functions was used and on the
%dimensionality of the data.
%
%Usage:
% h = mv_plot_result(result,<...>)
%
%Parameters:
% result - results struct obtained from one of the
% classification functions above. A cell array of
% results can be provided (e.g. results for different
% subjects); in this case, all results need to be
% created with the same function using the same metric.
% If multiple metrics have been used, a separate plot
% is generated for each metric.
%
% ADDITIONAL KEY-VALUE ARGUMENTS:
% new_figure - if 1, results are plotted in a new figure. If 0, results
% are plotted in the current axes instead (default 1)
% combine - if multiple results are given, specifies how to combined
% them (see mv_combine_results for details)
% mask - binary mask (eg statistical significance). Masks out
% data in images. For lines, produces bold lines.
%
% RETURNS:
% h - struct with handles to the graphical elements
% (c) matthias treder
% Parse any key-value pairs
opt = mv_parse_key_value_pairs(varargin{:});
if ~isfield(opt,'new_figure'), opt.new_figure = 1; end
if ~isfield(opt,'combine'), opt.combine = 'merge'; end
if ~isfield(opt,'mask'), opt.mask = []; end
% if multiple results are provided, they must be combined first
if iscell(result) && numel(result)>1
result = mv_combine_results(result, opt.combine);
if isempty(result.perf), error('Could not combine results'), end
end
% prepare plotting instructions if there's none yet
if ~isfield(result, 'plot')
result = mv_prepare_plot(result);
end
n_metrics = result.n_metrics;
if n_metrics == 1
if ~iscell(result.metric), result.metric = {result.metric}; end
if ~iscell(result.perf) || (strcmp(result.metric{1},'none') && ~iscell(result.perf{1})), result.perf = {result.perf}; end
if ~iscell(result.perf_std), result.perf_std = {result.perf_std}; end
if ~iscell(result.perf_dimension_names), result.perf_dimension_names = {result.perf_dimension_names}; end
end
for mm=1:n_metrics
if opt.new_figure, figure, else, clf; end
p = result.plot{mm};
metric = result.metric{mm};
perf = result.perf{mm};
perf_std = result.perf_std{mm};
if ~iscell(p.title), p.title = {p.title}; end
switch(p.plot_type)
case 'confusion_matrix'
% ---------- CONFUSION MATRIX ----------
n_classes = result.n_classes;
imagesc(perf)
colorbar
h.xlabel(mm) = xlabel(p.xlabel, p.label_options{:});
h.ylabel(mm) = ylabel(p.ylabel, p.label_options{:});
set(gca,'Xtick',1:n_classes,'Ytick',1:n_classes)
for rr=1:n_classes
for cc=1:n_classes
if perf(rr,cc) < 0.005
% if it would appear as a "0.00" we just plot a "0"
text(cc,rr, '0', p.text_options{:})
else
text(cc,rr, sprintf('%0.2f',perf(rr,cc)), p.text_options{:})
end
end
end
h.title(mm) = title(p.title, p.title_options{:});
case 'bar'
% ---------- BAR PLOT ----------
n_bars = p.n_bars;
h.bar = bar(1:n_bars, perf);
hold on
if p.combined % multiple results combined
if ~isvector(perf)
% grouped bar graph
create_legend(p.legend_labels, p.legend_options);
% find centers of grouped bars
offset = [h.bar.XOffset];
xd = h.bar(1).XData;
centers = zeros(size(perf));
for bb = 1:n_bars
centers(bb,:) = offset + xd(bb);
end
% need to place the errorbars over the grouped bars now
errorbar(centers(:), perf(:), perf_std(:), p.errorbar_options{:})
else
errorbar(1:n_bars, perf, perf_std, p.errorbar_options{:})
end
else
% Indicate SEM if the bars are not grouped
errorbar(1:n_bars, perf, perf_std, p.errorbar_options{:})
end
set(gca,'XTick',1:n_bars, 'XTickLabel', p.xticklabel)
h.ylabel(mm) = ylabel(p.ylabel, p.label_options{:});
h.title = title(p.title);
case 'line'
% ---------- LINE PLOT ----------
if (nargin > 1) && ~ischar(varargin{1}), x = varargin{1};
else, x = 1:length(perf);
end
cfg = [];
cfg.label_options = p.label_options;
cfg.title_options = p.title_options;
cfg.hor = p.hor;
cfg.mark_bold = opt.mask;
for ii=1:size(perf,3) % in case there's multiple results to be plotted
subplot(1, size(perf,3), ii)
h.plot(mm) = mv_plot_1D(cfg, x, perf(:,:,ii), perf_std(:,:,ii));
h.xlabel(mm) = xlabel(p.xlabel, p.label_options{:});
h.ylabel(mm) = ylabel(p.ylabel, p.label_options{:});
if p.add_legend, create_legend(p.legend_labels, p.legend_options), end
if iscell(p.title)
title(p.title{ii}, p.title_options{:});
else
h.title(mm) = title(p.title, p.title_options{:});
end
end
case 'dots'
% ---------- DOTS PLOT for 'none' metric ----------
if (nargin > 1) && ~ischar(varargin{1}), x = varargin{1};
else, x = 1:size(perf, 3);
end
cfg = [];
cfg.label_options = p.label_options;
cfg.title_options = p.title_options;
cfg.hor = p.hor;
cfg.mark_bold = opt.mask;
if strcmp(result.task, 'classification')
testlabel = result.testlabel;
if ~iscell(testlabel), testlabel = {testlabel}; end
for rep = 1:p.n_repetitions
for fold = 1:p.n_folds
subplot(p.n_repetitions, p.n_folds, (rep-1)*p.n_folds + fold)
cla
for c = 1:result.n_classes % each class separately
ix_class = testlabel{rep,fold}==c;
% need to repeat x for number of instances in each class
xx = mv_repelem(x, sum(ix_class));
% get values corresponding to class
vals = cellfun(@(dat) dat(ix_class), squeeze(perf(rep, fold,:,:,:)), 'Un', 0);
vals = cat(1, vals{:});
plot(xx, vals,'.')
hold all
end
create_legend(p.legend_labels, p.legend_options);
title(p.title{rep,fold}, p.title_options{:})
ylabel(p.ylabel, p.label_options{:})
grid on
end
end
else % regression
for rep = 1:p.n_repetitions
for fold = 1:p.n_folds
subplot(p.n_repetitions, p.n_folds, (rep-1)*p.n_folds + fold)
cla
vals = squeeze(perf(rep, fold,:,:,:));
xx = mv_repelem(x, numel(vals{1}));
vals = cat(1, vals{:});
plot(xx, vals,'.')
title(p.title{rep,fold}, p.title_options{:})
ylabel(p.ylabel, p.label_options{:})
grid on
end
end
end
case 'image'
% ---------- IMAGE ----------
% apply mask
if ~isempty(opt.mask)
perf = bsxfun(@times, perf, opt.mask);
perf(perf==0) = nan;
end
% settings for 2d plot
cfg= [];
if (nargin > 1) && ~ischar(varargin{1}), cfg.x = varargin{1};
else, cfg.x = 1:size(perf,2);
end
if (nargin > 2) && ~ischar(varargin{1}) && ~ischar(varargin{2}), cfg.y = varargin{2};
else, cfg.y = 1:size(perf,1);
end
cfg.climzero = p.climzero;
cfg.global_clim = p.global_clim;
cfg.xlabel = p.xlabel;
cfg.ylabel = p.ylabel;
cfg.colorbar_title = metric;
cfg.colorbar_location = p.colorbar_location;
cfg.label_options = p.label_options;
cfg.title_options = p.title_options;
for ix=1:size(perf,3) % in case there's multiple results to be plotted
subplot(1, size(perf,3), ix)
cfg.title = p.title(ix);
cfg.title_options = p.title_options;
if p.size_metric_dimension == 1
h.ax(mm) = mv_plot_2D(cfg, perf(:,:,ix));
else
h.ax(mm) = mv_plot_2D(cfg, squeeze(permute(perf(:,:,ix), [p.plot_dimensions, setdiff(1:3, p.plot_dimensions)])));
end
end
end
grid on
end
% --- helper functions ---
function leg = create_legend(lab, opt)
leg = legend(lab, opt{:});
if isprop(leg,'AutoUpdate')
set(leg, 'AutoUpdate', 0)
end
end
end