Nb weighted sg test (#126)

* Add weighted stimgen

* Create r2 viz function

* Fix typo

* Variable touchup

* Bold doc header

* Only viz if comparing trial fractions

code/scripts/pilot_reconstructions.m

@@ -6,7 +6,7 @@
 %  - A figure of reconstructions plotted against the target signal and simulated answers.
 
 DATA_DIR = ['/Users/nelsonbarnett/Desktop/Prof. Lammert Research/' ...
-    'Tinnitus/tinnitus-project/code/experiment/Data/data-paper-bugfixes'];
+    'Tinnitus/tinnitus-project/code/experiment/Data/data-weighted-test'];
 PROJECT_DIR = pathlib.strip(mfilename('fullpath'), 3);
 PUBLISH = false;
 
@@ -97,7 +97,7 @@
 
 %% Compute the reconstructions
 
-trial_fractions = 1; %linspace(0.1, 1, 10);
+trial_fractions = 0.1:0.1:1;
 
 % Container for r^2 values
 r2_cs_bins = zeros(height(T), length(trial_fractions));
@@ -228,6 +228,11 @@
     end
 end
 
+%% Visualize results if comparing trial fractions
+if length(trial_fractions) > 1
+    r2_bar(T)
+end
+
 %% Saving Results
 
 % Save the reconstruction waveforms

code/utils/cs.m

@@ -13,7 +13,7 @@
 %       where `n` is the number of trials/samples
 %       and `m` is the dimensionality of the stimuli/spectrum/bins
 % 
-% **OUTPUTS:*
+% **OUTPUTS:**
 %   - x: compressed sensing reconstruction of the signal.
 
 function x = cs(responses, Phi, Gamma)

code/utils/r2_bar.m

@@ -0,0 +1,78 @@
+% ### r2_bar 
+% 
+% Plots bar charts of r^2 values from table data. 
+% A separate figure is made for each subject.
+% 
+% **ARGUMENTS:**
+% 
+%   - T: `table` that includes r^2 values of interest
+% 
+% **OUTPUTS:**
+%   - n figures, where n is the number of subjects included
+%       in the table.
+% 
+% See Also:
+% pilot_reconstructions
+
+function r2_bar(T)
+
+    arguments
+        T (:,:) {mustBeUnderlyingType(T, "table")}
+    end
+
+    % Get indices of r^2 data from table
+    r2_lr_inds = contains(T.Properties.VariableNames,'r2_lr');
+    r2_cs_inds = contains(T.Properties.VariableNames,'r2_cs');
+
+    % Determine unique subjects
+    subjects = unique(T.subject_ID);
+
+    % Get column names with trial fraction information
+    fracs = T.Properties.VariableNames(r2_lr_inds);
+
+    % Create nice x-labels from column name info
+    frac_labels = cell(1, sum(r2_lr_inds));
+    for i = 1:length(fracs)
+        temp = regexprep(fracs{i}, {'\D*([\d\.]+\d)[^\d]*', ...
+            '[^\d\.]*'}, {'$1 ', ''}); % Parse for only numbers
+        frac_labels{i} = strcat(temp(2),'.',temp(3:end));
+    end
+
+    % Create a figure for each subject
+    for i = 1:length(subjects)
+        figure
+        t = tiledlayout('flow');
+        exp_rows = find(contains(T.subject_ID, subjects{i}));
+
+        % Plot two tiles (lr & cs) for each experiment within subject
+        for j = 1:length(exp_rows)
+            target_signal_name = T.target_signal_name{exp_rows(j)};
+            total_trials = T.total_trials(exp_rows(j));
+
+            nexttile
+            bar(table2array(T(exp_rows(j), r2_lr_inds)))
+            grid on
+            set(gca, 'XTickLabel', frac_labels)
+            xlabel(['Fraction of ', num2str(total_trials),...
+                ' trials'], 'FontSize', 14)    
+            ylabel('r^2', 'FontSize',14)
+            title(['Linear ', target_signal_name], ...
+                'FontSize', 14)
+
+            nexttile
+            bar(table2array(T(exp_rows(j), r2_cs_inds)))
+            grid on
+            set(gca, 'XTickLabel', frac_labels)
+            xlabel(['Fraction of ', num2str(total_trials),...
+                ' trials'], 'FontSize', 14)
+            ylabel('r^2', 'FontSize',14)
+            title(['CS ', target_signal_name], ...
+                'FontSize', 14)
+        end
+        title(t, ['Subject: ', subjects{i}], 'FontSize', 18)
+    end
+
+end %function
+
+
+

docs/utils.md

@@ -318,7 +318,7 @@ Write the stimuli into the stimuli file.
 where `n` is the number of trials/samples
 and `m` is the dimensionality of the stimuli/spectrum/bins
 
-**OUTPUTS:*
+**OUTPUTS:**
 - x: compressed sensing reconstruction of the signal.
 
 
@@ -522,6 +522,30 @@ stringified_properties = prop2str(obj, [], '&&')
 
 
 
+-------
+
+### r2_bar 
+
+Plots bar charts of r^2 values from table data. 
+A separate figure is made for each subject.
+
+**ARGUMENTS:**
+
+- T: `table` that includes r^2 values of interest
+
+**OUTPUTS:**
+- n figures, where n is the number of subjects included
+in the table.
+
+
+
+!!! info "See Also"
+    * [pilot_reconstructions](../scripts/#pilot_reconstructions)
+
+
+
+
+
 -------
 
 ### spect2binnedrepr