Contents
Dependencies:
% Matlab statistics toolbox % Matlab signal processing toolbox % Statistical Parametric Mapping (SPM) software https://www.fil.ion.ucl.ac.uk/spm/ % For full functionality, the full suite of CANlab toolboxes is recommended. See here: Installing Tools
Canlab dataset basic usage
Shows how to load a sample canlab_dataset object, extract variables, and make some plots.
% Instructions to include in help files as appropriate: % ----------------------------------------------------------------------- % % Before you start, the CANlab_Core_Tools must be added to your path with % subfolders. Otherwise, you will get errors. % % Sample datasets are in the "Sample_datasets" folder in CANlab_Core_Tools. % % This example will use behavioral data in the file: % 'Jepma_IE2_single_trial_canlab_dataset.mat' % The dataset is a multi-level dataset with 34 subjects, 2 subject-level vars, 10 event-level vars % consisting of a time series of trial-by-trial ratings of pain, expectations, and cue values for each subject % % These data have been submitted for publication (M. Jepma, Wager et al.) % Please do not reuse without permission.
Load the sample dataset and print a summary
% Load sample data file (this is in Sample_datasets in Core - make sure it's on your path) load('Jepma_IE2_single_trial_canlab_dataset.mat'); % loads object called DAT % Print a summary: print_summary(DAT); % see also list_variables(DAT)
--------- DATASET VARS --------
Experiment name:
IE2 : Marieke Jepma
34 subjects, 2 subject-level vars, 10 event-level vars
Warning: Missing variable type for variable MeanPain. Type assumed to be
numeric.
Warning: Missing variable type for variable MeanNPS. Type assumed to be numeric.
--------- SUBJECT LEVEL VARS: data_obj.Subj_Level --------
Var_name Description
__________ ___________
'MeanPain' 'MeanPain'
'MeanNPS' 'MeanNPS'
Warning: Missing variable type for variable trial. Type assumed to be numeric.
Warning: Missing variable type for variable temp. Type assumed to be numeric.
Warning: Missing variable type for variable cue. Type assumed to be numeric.
Warning: Missing variable type for variable cue valence. Type assumed to be
numeric.
Warning: Missing variable type for variable expected pain. Type assumed to be
numeric.
Warning: Missing variable type for variable reported pain. Type assumed to be
numeric.
Warning: Missing variable type for variable vif. Type assumed to be numeric.
Warning: Missing variable type for variable NPS. Type assumed to be numeric.
Warning: Missing variable type for variable Expected pain next trial. Type
assumed to be numeric.
Warning: Missing variable type for variable Pain next trial. Type assumed to be
numeric.
--------- EVENT LEVEL VARS: data_obj.Event_Level --------
Var_name Description
__________________________ __________________________________________________________
'trial' 'Trial number'
'temp' 'Temperature: -0.5 = 48 degrees, 0.5 = 49 degrees'
'cue' 'Cue identity: 1-3 = CSlow, Cue 4-6 = CShigh, 7 = neutral'
'cue valence' 'Cue valence: -.5 low +.5 high'
'expected pain' 'Expected pain'
'reported pain' 'Reported pain'
'vif' 'Variance inflation factor for trial'
'NPS' 'NPS response'
'Expected pain next trial' 'Expected pain on next trial'
'Pain next trial' 'Reported pain on next trial'
Warning: Missing variable type for variable MeanPain. Type assumed to be
numeric.
Warning: Missing variable type for variable MeanNPS. Type assumed to be numeric.
Warning: Missing variable type for variable trial. Type assumed to be numeric.
Warning: Missing variable type for variable temp. Type assumed to be numeric.
Warning: Missing variable type for variable cue. Type assumed to be numeric.
Warning: Missing variable type for variable cue valence. Type assumed to be
numeric.
Warning: Missing variable type for variable expected pain. Type assumed to be
numeric.
Warning: Missing variable type for variable reported pain. Type assumed to be
numeric.
Warning: Missing variable type for variable vif. Type assumed to be numeric.
Warning: Missing variable type for variable NPS. Type assumed to be numeric.
Warning: Missing variable type for variable Expected pain next trial. Type
assumed to be numeric.
Warning: Missing variable type for variable Pain next trial. Type assumed to be
numeric.
--------- SUBJECT LEVEL VARS: data_obj.Subj_Level --------
Var_name Min_value Max_value Mean_value St_Dev IQR NaN_count
__________ _________ _________ __________ ______ ______ _________
'MeanPain' 5.8015 46.911 27.45 9.5331 12.429 0
'MeanNPS' -0.72369 87.081 32.079 20.682 27.863 0
--------- EVENT LEVEL VARS: data_obj.Event_Level --------
Var_name Min_value Max_value Mean_value St_Dev IQR NaN_count
__________________________ _________ _________ __________ _______ _______ _________
'trial' 1 70 35.4 20.231 35 48
'temp' -0.54286 0.52857 -0.0025484 0.49995 1 48
'cue' 1 7 4.0039 2.0048 4 48
'cue valence' -1 1 0 0.92536 2 48
'expected pain' 0 81.333 31.975 16.78 23.533 48
'reported pain' 0 67.8 27.503 13.149 18.333 48
'vif' 1.0589 2.9802 1.3221 0.2779 0.19197 48
'NPS' -187.75 273.47 32.208 44.535 52.665 78
'Expected pain next trial' 0 81.333 31.976 16.755 23.533 82
'Pain next trial' 0 67.8 27.549 13.166 18.467 82
Extract data two variables and make a plot
% Extract expectation and pain ratings, two continuous event-level variables from the object [~, expect] = get_var(DAT, 'expected pain'); [~, pain] = get_var(DAT, 'reported pain'); % Make a multi-level scatterplot create_figure('lines_expect_pain', 1, 2); [han, Xbin, Ybin] = line_plot_multisubject(expect, pain, 'n_bins', 5, 'group_avg_ref_line', 'MarkerTypes', 'o', 'colors', custom_colors([1 .7 .4], [1 .7 .4], 100), 'gcolors', {[.4 .2 0] [.8 .4 0]}); axis tight xlabel('Expectations'); ylabel('Pain'); set(gca, 'FontSize', 24);
Warning: Missing variable type for variable expected pain. Type assumed to be numeric. Warning: Missing variable type for variable reported pain. Type assumed to be numeric. r = 0.76, t( 33) = 12.53, p = 0.000000, num. missing: 2
Extract data from one variable conditional on another variable, and plot
% Now extract expectation and pain variables conditional on cue value: [~, expect_cshigh] = get_var(DAT, 'expected pain', 'conditional', {'cue valence' 1}); [~, expect_csmed] = get_var(DAT, 'expected pain', 'conditional', {'cue valence' 0}); [~, expect_cslow] = get_var(DAT, 'expected pain', 'conditional', {'cue valence' -1}); [~, pain_cshigh] = get_var(DAT, 'reported pain', 'conditional', {'cue valence' 1}); [~, pain_csmed] = get_var(DAT, 'reported pain', 'conditional', {'cue valence' 0}); [~, pain_cslow] = get_var(DAT, 'reported pain', 'conditional', {'cue valence' -1}); % Now make the plot, without individual subject lines % First set colors and condition names color1 = {[.9 .4 .2] [.6 .3 0]}; color2 = {[.5 .5 .5] [.2 .2 .2]}; color3 = {[.4 .8 .4] [.2 .7 .2]}; condition_names = {'High Cue' 'Medium Cue' 'Low Cue'}; % Now plot: subplot(1, 2, 2) [han1, Xbin, Ybin] = line_plot_multisubject(expect_cshigh, pain_cshigh, 'n_bins', 5, 'group_avg_ref_line', 'MarkerTypes', 'o', 'colors', color1, 'center', 'noind', 'nolines'); [han2, Xbin, Ybin] = line_plot_multisubject(expect_csmed, pain_csmed, 'n_bins', 5, 'group_avg_ref_line', 'MarkerTypes', 'o', 'colors',color2, 'center', 'noind', 'nolines'); [han3, Xbin, Ybin] = line_plot_multisubject(expect_cslow, pain_cslow, 'n_bins', 5, 'group_avg_ref_line', 'MarkerTypes', 'o', 'colors',color3, 'center', 'noind', 'nolines'); xlabel('Expectations'); ylabel('Pain'); set(gca, 'FontSize', 24); legend([han1.grpline_handle(1) han2.grpline_handle(1) han3.grpline_handle(1)], condition_names); drawnow, snapnow
Warning: Missing variable type for variable expected pain. Type assumed to be numeric. Warning: Missing variable type for variable expected pain. Type assumed to be numeric. Warning: Missing variable type for variable expected pain. Type assumed to be numeric. Warning: Missing variable type for variable reported pain. Type assumed to be numeric. Warning: Missing variable type for variable reported pain. Type assumed to be numeric. Warning: Missing variable type for variable reported pain. Type assumed to be numeric. r = 0.54, t( 33) = 3.64, p = 0.000919, num. missing: 3 r = 0.24, t( 33) = 1.43, p = 0.161345, num. missing: 3 r = 0.59, t( 33) = 5.64, p = 0.000003, num. missing: 2
