Humble attempt to fix typos by adding codespell action, config etc.

.codespellrc

@@ -0,0 +1,29 @@
+[codespell]
+# *.gii files are GIFTI XML surface meshes (canlab_canonical_brains): treat as data.
+# docs_sphinx_old is legacy pre-built Sphinx output, External/ is vendored third-party code.
+skip = .git,.git-meta,.svn,*.pdf,*.svg,*.js,_build,*.html,*.rtf,*.rtf.svn-base,*.gii,*.nii,*.mat,*.png,*.jpg,*.jpeg,*.tif,*.tiff,*.ico,External,docs_sphinx_old
+check-hidden = true
+# Skip URLs (don't "fix" typos in third-party links) and base64-embedded image data
+# left behind by MATLAB live-script export in docs/markdown_tutorials/*.m.
+ignore-regex = https?://\S+|^%   data: \{"dataType":"image".*
+# Whitelist short variable names, MATLAB function/built-in names, proper names, and
+# domain abbreviations that codespell would otherwise flag. Case-insensitive.
+#   ons, ans, fpr, ttests, myu, whos, ser, htmp, fwe, dout -- pre-existing stats vars
+#   ttest -- MATLAB hypothesis-test function, used as identifier & in docs everywhere
+#   indx, alph, te, ist, selt, multy, shft, delt, dum, datin, fixin, runn, nneeded,
+#   figurestyle, cxan, losin, commun, pres, ues -- variable names confirmed in source
+#   assignin, evalin -- MATLAB built-in functions
+#   inout -- canlab_results_fmridisplay montage layout option ('leftright inout')
+#   ment -- hyphen-broken "displace-\nment" in comment (movement_regressors.m)
+#   sepulcre, shepard, shure, claus, lew -- proper names (Yeo atlas; Shepard interpolation; researchers)
+#   nd -- often 2nd, also matches Nd/ND chemical symbol
+#   ore -- short for "or" in some places; appears in surface .gii-derived comments
+#   fo -- variable name (`fo = gcf`) in make_figure_into_orthviews.m
+#   atmost -- variable name in OptimizeDesign11/GA3/Sobol/sobol.m
+#   sessons -- local "session ons" variable in Statistics_tools/Support_functions/spm2delta.m
+#   lightening -- real word (making lighter), used for white-matter brightness in mean_image.m
+#   prctiles -- struct field of descriptives() output (public API, documented in docs/)
+#   atleast, whch, whn, includ -- variable names (renaming would break code; "atleast" -> "at least" would be a syntax error)
+#   efficency -- legacy function file name OptimizeDesign11/core_functions/efficency.m (sibling efficiency.m is the newer version; both retained)
+#   ony, nam, tha, thre, firt, jus -- ambiguous variable/string fragments in context
+ignore-words-list = ons,ans,fpr,ttests,myu,whos,ser,htmp,fwe,dout,ttest,indx,alph,te,ist,selt,multy,shft,delt,dum,datin,fixin,runn,nneeded,figurestyle,cxan,losin,commun,pres,ues,assignin,evalin,inout,ment,sepulcre,shepard,shure,claus,lew,nd,ore,fo,atmost,sessons,lightening,prctiles,atleast,whch,whn,includ,efficency,ony,nam,tha,thre,firt,jus

.github/workflows/codespell.yml

@@ -0,0 +1,23 @@
+# Codespell configuration is within .codespellrc
+---
+name: Codespell
+
+on:
+  push:
+    branches: [master]
+  pull_request:
+    branches: [master]
+
+permissions:
+  contents: read
+
+jobs:
+  codespell:
+    name: Check for spelling errors
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v6
+      - name: Codespell
+        uses: codespell-project/actions-codespell@8f01853be192eb0f849a5c7d721450e7a467c579  # v2.2

CanlabCore/@atlas/atlas.m

@@ -225,7 +225,7 @@
         labels_4
         labels_5
         references
-        space_description = ''; % set this to something compatable with the allowed_sourcespace from render_on_surface for automatic projections to supported surfaces when plotting.
+        space_description = ''; % set this to something compatible with the allowed_sourcespace from render_on_surface for automatic projections to supported surfaces when plotting.
         property_descriptions = { ...
             'atlas_name: a short description or name of the atlas' ...
             'probability_maps: voxels x regions matrix with probability values for each region' ...

CanlabCore/@atlas/extract_data.m

@@ -77,7 +77,7 @@
 %         respectively. This is useful for visualizing and assessing pattern
 %         weights, i.e., when the input data image is a weight pattern map.
 %         It can tell you how strong the contribution of each region is to
-%         prediction (high weight values) and how homogenous it is (pos vs.
+%         prediction (high weight values) and how homogeneous it is (pos vs.
 %         negative energy/rmsvs). 
 %         RMSVs are expressed in weights / cm^3 of brain tissue, and volume is
 %         regularized by adding 1 cm^3 to avoid unstable results with small

CanlabCore/@atlas/merge_atlases.m

@@ -174,7 +174,7 @@
 
     % concatenating sparse matrices first is much faster than converting to full and the
     % concatenating full matrices so instead of doing a double conversion
-    % and concatenating, try concatenating sparse matricies if either of
+    % and concatenating, try concatenating sparse matrices if either of
     % them are sparse and then converting to full
     %atlas_obj.probability_maps = [full(atlas_obj.probability_maps) full(atlas_obj_to_add.probability_maps)];
     if issparse(atlas_obj.probability_maps) || issparse(atlas_obj_to_add.probability_maps)

CanlabCore/@atlas/montage.m

@@ -55,7 +55,7 @@
 %        fmridisplay or edit code for defaults as example).
 %
 %   **'montagetype':**
-%        'full' for full montages of axial and saggital slices.
+%        'full' for full montages of axial and sagittal slices.
 %        'full hcp' for full montage, but with surfaces and volumes from
 %        HCP data.
 %        'compact' [default] for single-figure parasagittal and axial slices.

CanlabCore/@atlas/probability_maps_to_region_index.m

@@ -53,7 +53,7 @@
     dropped_ind = find(~ismember(1:n_regions,n_maps));
 
     % note, labels must be last in this list or else it would end up being
-    % asigned to the labels variable which is needed subsequently
+    % assigned to the labels variable which is needed subsequently
     fnames = {'probability_maps', 'labels_2','labels_3','labels_4',...
         'labels_5','label_descriptions','labels'};
     assert(strcmp(fnames{end},'labels'));

CanlabCore/@atlas/reorder_atlas_regions.m

@@ -52,7 +52,7 @@
 % Programmers' notes:
 % Created by Tor Wager
 % 12/2019: Tor fixed bug in reordering; probability_maps field was not
-% preivously being considered, which resulted in wrong ordering when
+% previously being considered, which resulted in wrong ordering when
 % present. .dat is rebuilt from probability_maps if available.
 % 3/2025 Updated help and added labelgroups option
 

CanlabCore/@atlas/threshold.m

@@ -241,7 +241,7 @@
 
 % Identifies any fragmented parcels, splits them in two and postfixes
 % _fragN to the label name, retaining corresponding probabilities for
-% each regoin. N is the fragment index, frag1 is the largest and subsequent
+% each region. N is the fragment index, frag1 is the largest and subsequent
 % parcels are in order of size.
 function obj = spin_off_frag(obj)
     obj = obj.replace_empty();

CanlabCore/@brainpathway/brainpathway.m

@@ -584,7 +584,7 @@
          function [obj_subset, region_objects, region_indices] = cluster_region_subset_by_connectivity(obj, varargin)
             % [obj_subset, region_objects, region_indices] = cluster_region_subset_by_connectivity(obj, Group#, [# clusters])
             %
-            % Takes a subset of parcels idenfied based on obj.node_clusters 
+            % Takes a subset of parcels identified based on obj.node_clusters 
             % and clusters them based on profiles
             % of pairwise connectivity with other regions. Returns a new
             % object based on the target subset.

CanlabCore/@brainpathway/degree_calc.m

@@ -3,7 +3,7 @@
 % b2 = degree_calc(b2)
 %
 % Method for brainpathway
-% uses node_clusters to caculate overall, within-cluster(network), and
+% uses node_clusters to calculate overall, within-cluster(network), and
 % between-cluster degree
 %
 % NOTE: Currently uses region averages (.region), update to use nodes!

CanlabCore/@brainpathway_multisubject/compute_HDIs.m

@@ -93,7 +93,7 @@
     %% compute BCT measures on ref group  
     refgroup = bct_toolbox_undirected_graph_metrics(refgroup, thresh, varargin{:});
 
-    %% for each subject, regress on mean referece group to get an HDI (i.e., the slope)
+    %% for each subject, regress on mean reference group to get an HDI (i.e., the slope)
 
     for i=1:size(bs.connectivity.regions.r,3)
 

CanlabCore/@canlab_dataset/get_descriptives.m

@@ -118,7 +118,7 @@
 
 function [Min_value, Max_value, Mean_value, St_Dev, myIQR, NaN_count, text_vals] = get_descriptives_var(D, vname)
 
-% intialize
+% initialize
 [Min_value, Max_value, Mean_value, St_Dev, myIQR, NaN_count] = deal(NaN);
 text_vals = [];
 

CanlabCore/@canlab_dataset/plot_var.m

@@ -45,7 +45,7 @@
 %             plots across events with standard error shading
 %           - the grouping variable's description, if it exists, will
 %             be split along commas, and those values will be used as
-%             column lables
+%             column labels
 %
 %   **eventmeans:**
 %        calculate and plot subject means across event-level variables
@@ -64,7 +64,7 @@
 %
 %   **other:**
 %        other varargin are passed directly to barplot_columns.  So
-%        for example, '95CI' will make 95% confidence interals, instead
+%        for example, '95CI' will make 95% confidence intervals, instead
 %        of SE bars.
 %
 %

CanlabCore/@fmri_data/canlab_connectivity_preproc.m

@@ -178,7 +178,7 @@
 %               spikes., Consistent with principles in Lindquist 2019
 %               "sequential filtering" paper. See more notes below.
 %    08/01/19: Wani reversed Tor and Marianne's revision above 
-%               after dicussion with Tor and others. Spikes should be removed
+%               after discussion with Tor and others. Spikes should be removed
 %               before "low-pass filtering" the nuisance covariates, and
 %               added to the covariates later (see Ciric et al., 2018)
 %               Lindquist et al., 2019 did not address the issue related to 
@@ -571,7 +571,7 @@
 % time series data appropriately, accounting for filtering.
 %
 % 8/1/2019: Wani reversed this to the previous version after
-% dicussion with Tor and others. Spikes should be removed
+% discussion with Tor and others. Spikes should be removed
 % before "low-pass filtering" the nuisance covariates, and
 % added to the covariates later (see Ciric et al., 2018)
 

CanlabCore/@fmri_data/evaluate_spatial_scale.m

@@ -103,7 +103,7 @@
 %
 %   Phil: created, Dec 2018
 %
-%   TODOs: - change outcome metric to optionally incude accuracy/AUC/etc.
+%   TODOs: - change outcome metric to optionally include accuracy/AUC/etc.
 %          - change outcome to optionally reflect % of noise ceiling
 %          - account for regions that have small number of voxels
 % ..

CanlabCore/@fmri_data/fmri_data.m

@@ -60,7 +60,7 @@
 %   new_obj.volInfo = MC_Setup.volInfo;
 %   new_obj.dat = MC_Setup.unweighted_study_data(:, sym_cons);
 %
-% You can create an fmri_data object with extacted image data.
+% You can create an fmri_data object with extracted image data.
 % - Let "imgs" be a string array or cell array of image names
 % - This command creates an object with your (4-D) image data:
 % - fmri_dat = fmri_data(imgs);
@@ -453,7 +453,7 @@
                 end
 
                 % convert dat to single if needed
-                % this is esp relevant for converstion of atlas to fmri_data
+                % this is esp relevant for conversion of atlas to fmri_data
                 % because atlas has integer data type by default. this is
                 % to prevent a range of issues that can occur when using
                 % the fmri_data object later on and dat field is an

CanlabCore/@fmri_data/hrf_fit.m

@@ -51,7 +51,7 @@
 %
 % A. **Fit HRF using IL-function**
 %        Choose mode (deterministic/stochastic)
-%           - 0 - deterministic aproach 
+%           - 0 - deterministic approach 
 %           - 1 - simulated annealing approach
 %
 %        Please note that when using simulated annealing approach you

CanlabCore/@fmri_data/model_brain_pathway.m

@@ -178,7 +178,7 @@
 
 % ..
 %    Programmers' notes:
-% 10/10/2019 Wrote funcion (Phil Kragel)
+% 10/10/2019 Wrote function (Phil Kragel)
 % 4/14/2020 Major overhaul and documentation (Phil Kragel and Tor Wager)
 % 4/15/2020 Add more documentation (Phil Kragel)
 % 9/16/2020 Tor Wager - update documentation
@@ -307,7 +307,7 @@
 
 % Convert data into cell arrays for cross-validation
 % Each cell contains a Voxels x Images matrix for the corresponding fold
-% For 10-fold, images will be devided into 10 cells
+% For 10-fold, images will be divided into 10 cells
 
 % note by kodiweera: convert to a cell array before the cross validation loop. I took this loop out of the main loop below to make the the program little faster as this is repeatedlty run during the k-fold validation.
 % Convert to cell array, one cell per subject
@@ -766,17 +766,17 @@
 
 istat = stats.latent_correlation_interaction_ttest;
 
-stats_report{3} = sprintf('The interaction between on-targed (expected) and off-target pathways was %s signficant, t(%3.1f) = %3.2f, p = %3.6f.', ...
+stats_report{3} = sprintf('The interaction between on-targed (expected) and off-target pathways was %s significant, t(%3.1f) = %3.2f, p = %3.6f.', ...
     sigstr{(istat.p < 0.05) + 1}, istat.df(1), istat.tstat(1), istat.p(1));
 
 istat = stats.latent_correlation_pathway_one_ttest;
 
-stats_report{4} = sprintf('Pathway 1 on-target vs. off-target correlations were %s signficant, t(%3.1f) = %3.2f, p = %3.6f.', ...
+stats_report{4} = sprintf('Pathway 1 on-target vs. off-target correlations were %s significant, t(%3.1f) = %3.2f, p = %3.6f.', ...
     sigstr{(istat.p < 0.05) + 1}, istat.df(1), istat.tstat(1), istat.p(1));
 
 istat = stats.latent_correlation_pathway_two_ttest;
 
-stats_report{5} = sprintf('Pathway 2 on-target vs. off-target correlations were %s signficant, t(%3.1f) = %3.2f, p = %3.6f.', ...
+stats_report{5} = sprintf('Pathway 2 on-target vs. off-target correlations were %s significant, t(%3.1f) = %3.2f, p = %3.6f.', ...
     sigstr{(istat.p < 0.05) + 1}, istat.df(1), istat.tstat(1), istat.p(1));
 
 stats_report{6} = '       ';
@@ -803,17 +803,17 @@
 
 istat = stats.simple_correlation_interaction_ttest;
 
-stats_report{wh_last + 3} = sprintf('The interaction between on-targed (expected) and off-target pathways was %s signficant, t(%3.1f) = %3.2f, p = %3.6f.', ...
+stats_report{wh_last + 3} = sprintf('The interaction between on-targed (expected) and off-target pathways was %s significant, t(%3.1f) = %3.2f, p = %3.6f.', ...
     sigstr{(istat.p < 0.05) + 1}, istat.df(1), istat.tstat(1), istat.p(1));
 
 istat = stats.simple_correlation_pathway_one_ttest;
 
-stats_report{wh_last + 4} = sprintf('Pathway 1 on-target vs. off-target correlations were %s signficant, t(%3.1f) = %3.2f, p = %3.6f.', ...
+stats_report{wh_last + 4} = sprintf('Pathway 1 on-target vs. off-target correlations were %s significant, t(%3.1f) = %3.2f, p = %3.6f.', ...
     sigstr{(istat.p < 0.05) + 1}, istat.df(1), istat.tstat(1), istat.p(1));
 
 istat = stats.simple_correlation_pathway_two_ttest;
 
-stats_report{wh_last + 5} = sprintf('Pathway 2 on-target vs. off-target correlations were %s signficant, t(%3.1f) = %3.2f, p = %3.6f.', ...
+stats_report{wh_last + 5} = sprintf('Pathway 2 on-target vs. off-target correlations were %s significant, t(%3.1f) = %3.2f, p = %3.6f.', ...
     sigstr{(istat.p < 0.05) + 1}, istat.df(1), istat.tstat(1), istat.p(1));
 
 stats_report{wh_last + 6} = '       ';

CanlabCore/@fmri_data/model_mpathi.m

@@ -187,7 +187,7 @@
 
 % Convert data into cell arrays for cross-validation
 % Each cell contains a Voxels x Images matrix for the corresponding fold
-% For 10-fold, images will be devided into 10 cells
+% For 10-fold, images will be divided into 10 cells
 
 
 for xval_i=1:max(indices)

CanlabCore/@fmri_data/plot.m

@@ -1,5 +1,5 @@
 function [wh_outlier_uncorr, wh_outlier_corr, X] = plot(fmridat, varargin)
-% Custom plots for fmri_data object. Includes mulitple plots and outlier detection.
+% Custom plots for fmri_data object. Includes multiple plots and outlier detection.
 %
 % :Usage:
 % ::
@@ -79,7 +79,7 @@
 %         of each image are equal. If the variances are all of approximately the same scale,
 %         the diagonal will be a single color.
 %
-%        Correlation Matrix $\frac{{X}'X}{n-1}$, colums of X are Z-scored, n is number of rows in X
+%        Correlation Matrix $\frac{{X}'X}{n-1}$, columns of X are Z-scored, n is number of rows in X
 %        This plot shows the correlation between each image and the others.
 %        Now, the main diagonal will always be one because each image is
 %        perfectly correlated with itself. The off-diagonals should be positive