Creating subplots based on a cell array

Hi,

If I have two cells like this:

cell1 = {'AA' 'AA' 'BC' 'BC' 'BC' 'DD' 'DD'; 1 2 3 4 5 6 7; 7 6 5 4 3 2 1; 5 9 8 7 6 5 4}

cell1 = 4×7 cell array
    {'AA'}    {'AA'}    {'BC'}    {'BC'}    {'BC'}    {'DD'}    {'DD'}
    {[ 1]}    {[ 2]}    {[ 3]}    {[ 4]}    {[ 5]}    {[ 6]}    {[ 7]}
    {[ 7]}    {[ 6]}    {[ 5]}    {[ 4]}    {[ 3]}    {[ 2]}    {[ 1]}
    {[ 5]}    {[ 9]}    {[ 8]}    {[ 7]}    {[ 6]}    {[ 5]}    {[ 4]}

cell2 = {'AA' 'AA' 'BC' 'BC' 'BC' 'DD' 'DD'; 3 4 5 6 7 8 9; 2 3 4 5 6 7 8; 9 8 7 6 5 4 3}

cell2 = 4×7 cell array
    {'AA'}    {'AA'}    {'BC'}    {'BC'}    {'BC'}    {'DD'}    {'DD'}
    {[ 3]}    {[ 4]}    {[ 5]}    {[ 6]}    {[ 7]}    {[ 8]}    {[ 9]}
    {[ 2]}    {[ 3]}    {[ 4]}    {[ 5]}    {[ 6]}    {[ 7]}    {[ 8]}
    {[ 9]}    {[ 8]}    {[ 7]}    {[ 6]}    {[ 5]}    {[ 4]}    {[ 3]}

Is there a way to create subplots based on matching the strings from the first row, where there is not necessarily an equal number of strings for all?

The final figures would look something like this:

Thank you!

[the cyclist]

Is the pattern of alphabetic characters (e.g. 'AA') guaranteed to be identical between the two cell arrays?

[Stephen23]

Storing lots of scalar numeric inside cell arrays is very inefficient data design, and complicates the code required to process it. Why are you not using two arrays (i.e. a text vector and a numeric matrix) or a simple table instead?

[Liam]

Yes, they will be sorted relative to each other already in a prior step

[Liam]

Currently, the "titles" and data are actually stored how you said - text vector and a numeric matrix. I just assumed appending them together into a cell array would maybe make this subplot sorting easier.

Answer 1

[Voss]

cell1 = {'AA' 'AA' 'BC' 'BC' 'BC' 'DD' 'DD'; 1 2 3 4 5 6 7; 7 6 5 4 3 2 1; 5 9 8 7 6 5 4};
cell2 = {'AA' 'AA' 'BC' 'BC' 'BC' 'DD' 'DD'; 3 4 5 6 7 8 9; 2 3 4 5 6 7 8; 9 8 7 6 5 4 3};

assert(isequal(cell1(1,:),cell2(1,:)))
figure('Position',[50 50 500 1200]); % make a tall figure to see the subplots clearly

[g,g_id] = findgroups(cell1(1,:));
N = numel(g_id);
for ii = 1:N
    idx = find(g == ii);
    n = numel(idx);
    for jj = 1:n
        subplot(N*n,1,(ii-1)*n+jj)
        plot([cell1{2:end,idx(jj)}],[cell2{2:end,idx(jj)}]);
        title(cell1{1,idx(jj)});
    end
end

[Liam]

Modifying this a little worked perfectly for me, thanks!

Answer 2

[the cyclist]

cell1 = {'AA' 'AA' 'BC' 'BC' 'BC' 'DD' 'DD'; 1 2 3 4 5 6 7; 7 6 5 4 3 2 1; 5 9 8 7 6 5 4};
cell2 = {'AA' 'AA' 'BC' 'BC' 'BC' 'DD' 'DD'; 3 4 5 6 7 8 9; 2 3 4 5 6 7 8; 9 8 7 6 5 4 3};

numbers1 = cell2mat(cell1(2:end,:));
numbers2 = cell2mat(cell2(2:end,:));

[uniqueAlpha,~,indexFromUniqueToAll] = unique(cell1(1,:));
numberUniqueAlpha = numel(uniqueAlpha);

for na = 1:numberUniqueAlpha
    
    columnsThisAlpha = find(indexFromUniqueToAll==na);
    numberColumnsThisAlpha = numel(columnsThisAlpha);
    
    figure
    tt = tiledlayout(numberColumnsThisAlpha,1);
    title(tt,uniqueAlpha(na))
    for nc = 1:numberColumnsThisAlpha
        nexttile
        plot(numbers1(:,nc),numbers2(:,nc))
    end

end