Contents
Dataset Basics (setting sample attributes)
- For CoSMoMVPA's copyright information and license terms, #
- see the COPYING file distributed with CoSMoMVPA. #
% Set the targets and the chunks % % There are 10 runs with 6 volumes per run. The runs are vertically stacked one % above the other. The six volumes in each run correspond to the stimuli: % 'monkey','lemur','mallard','warbler','ladybug','lunamoth', in that order. Add % numeric targets labels (samples attribute) such that 1 corresponds to 'monkey', % 2 corresponds to 'lemur', etc. Then add numeric chunks (another samples % attribute) so that 1 corresponds to run1, 2 corresponds to run2, etc. config = cosmo_config(); data_path = fullfile(config.tutorial_data_path, 'ak6', 's01');
Load the dataset 'glm_T_stats_perrun.nii' masked with 'brain_mask.nii'
>@@>
mask_fn = fullfile(data_path, 'brain_mask.nii'); data_fn = fullfile(data_path, 'glm_T_stats_perrun.nii'); ds = cosmo_fmri_dataset(data_fn, 'mask', mask_fn); % <@@<
set targets
remember that targets are part of ds.sa and that they are stored in a column vector >@@>
ds.sa.targets = repmat(1:6, [1, 10])'; % 10 times labels 1 to 6, column vector % <@@<
set chunks
remember that chunks are part of ds.sa and that they are stored in a column vector >@@>
chunks = repmat(1:10, [6, 1]); chunks = chunks(:); % flatten matrix to a column vector ds.sa.chunks = chunks; % <@@<
Show the results
print the dataset
fprintf('\nDataset:\n');
cosmo_disp(ds);
Dataset:
.a
.vol
.mat
[ -3 0 0 121
0 3 0 -114
0 0 3 -11.1
0 0 0 1 ]
.xform
'scanner_anat'
.dim
[ 80 80 43 ]
.fdim
.labels
{ 'i'
'j'
'k' }
.values
{ [ 1 2 3 ... 78 79 80 ]@1x80
[ 1 2 3 ... 78 79 80 ]@1x80
[ 1 2 3 ... 41 42 43 ]@1x43 }
.sa
.targets
[ 1
2
3
:
4
5
6 ]@60x1
.chunks
[ 1
1
1
:
10
10
10 ]@60x1
.samples
[ 0.933 0.000495 2.1 ... 0.681 0.626 0.421
-0.772 0.117 1.73 ... 0.723 0.892 1.7
0.621 1.52 -0.25 ... -0.341 -0.832 -0.546
: : : : : :
-2.37 -1.92 -0.879 ... 0.784 0.67 0.522
-0.261 0.351 -0.0767 ... 2.81 2.27 -0.947
-1.29 -0.964 -0.966 ... 0.827 0.833 0.991 ]@60x43822
.fa
.i
[ 35 36 44 ... 46 47 47 ]@1x43822
.j
[ 17 17 17 ... 35 35 36 ]@1x43822
.k
[ 1 1 1 ... 39 39 39 ]@1x43822
print the sample attributes
fprintf('\nSample attributes (in full):\n'); cosmo_disp(ds.sa, 'edgeitems', Inf); % 'edgeitems determine how much of a % matrix is displayed. Try different values.
Sample attributes (in full):
.targets
[ 1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6 ]
.chunks
[ 1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3
4
4
4
4
4
4
5
5
5
5
5
5
6
6
6
6
6
6
7
7
7
7
7
7
8
8
8
8
8
8
9
9
9
9
9
9
10
10
10
10
10
10 ]
print targets and chunks next to each other
fprintf('\nTargets and chunks attributes (in full):\n'); nsamples = size(ds.samples, 1); fprintf('sample # target chunk\n'); index_target_chunks = [(1:nsamples)', ds.sa.targets, ds.sa.chunks]; cosmo_disp(index_target_chunks, 'edgeitems', Inf);
Targets and chunks attributes (in full): sample # target chunk [ 1 1 1 2 2 1 3 3 1 4 4 1 5 5 1 6 6 1 7 1 2 8 2 2 9 3 2 10 4 2 11 5 2 12 6 2 13 1 3 14 2 3 15 3 3 16 4 3 17 5 3 18 6 3 19 1 4 20 2 4 21 3 4 22 4 4 23 5 4 24 6 4 25 1 5 26 2 5 27 3 5 28 4 5 29 5 5 30 6 5 31 1 6 32 2 6 33 3 6 34 4 6 35 5 6 36 6 6 37 1 7 38 2 7 39 3 7 40 4 7 41 5 7 42 6 7 43 1 8 44 2 8 45 3 8 46 4 8 47 5 8 48 6 8 49 1 9 50 2 9 51 3 9 52 4 9 53 5 9 54 6 9 55 1 10 56 2 10 57 3 10 58 4 10 59 5 10 60 6 10 ]