Convert list of points into string and group it in tens

Hello,

I have a list of points which I want to convert to string and format it in a certain way.

I have manged to do this, but the code looks a bit complicated and wonder how this could be rewrite in a more efficient way.

points = randi(10,36,2);
points_len = size(points,1);
sPoints = '';
for kk = 1:points_len 
    if rem(kk+9, 10) == 0 
        if kk > points_len - rem(points_len,10)
            endPoint = points_len ;
        else
            endPoint = kk+9;
        end
        sPoints = [sPoints, sprintf('Points %0.0f to %0.0f\n', kk, endPoint)];
    end
    sPoints = [sPoints, num2str(points(kk,:)), newline];
end
sPoints

sPoints = 
    'Points 1 to 10
     9  7
     10  10
     4  10
     8  9
     8  4
     6  1
     2  9
     7  1
     4  2
     9  9
     Points 11 to 20
     9  8
     10   1
     5  6
     4  9
     1  8
     1  7
     5  3
     10   7
     2  5
     9  3
     Points 21 to 30
     2  4
     7  4
     6  10
     7  7
     6  6
     8  10
     2  6
     1  2
     8  5
     5  1
     Points 31 to 36
     6  1
     3  4
     2  1
     6  1
     2  9
     5  2
     '

Answer 1

[Rik]

I would suggest putting each section in a cell, so you can easilly add the headers:

points = randi(10,36,2);
PointsPerSection = 10;
NumRows = ceil(size(points,1)/PointsPerSection);
rows = zeros(NumRows,1);
rows(1:(end-1)) = PointsPerSection;
rows(end) = size(points,1)-sum(rows); % Determine last element dynamically
data = mat2cell(points,rows,2)

data = 4×1 cell array
    {10×2 double}
    {10×2 double}
    {10×2 double}
    { 6×2 double}

for n=1:numel(data)
    tmp = data{n};
    HeaderPart = sprintf('Points %d to %d\n'   ,...
        (n-1)*PointsPerSection + 1           ,...
        (n-1)*PointsPerSection + size(tmp,1) );
    % The sprintf function allows providing arrays, but will process them
    % by column, while humans will read the text by row. That is why we
    % need to use .' to transpose the data.
    DataPart = sprintf('%.0f %.0f\n',tmp.');
    data{n} = [HeaderPart DataPart];
end
sPoints = horzcat(data{:});
disp(sPoints)

Points 1 to 10
5 2
5 4
1 4
4 4
6 10
10 7
9 9
9 5
10 9
5 9
Points 11 to 20
8 10
10 5
4 9
2 8
4 10
2 9
2 8
8 1
5 4
4 9
Points 21 to 30
2 7
3 9
4 3
5 4
8 2
10 10
9 10
10 6
1 10
7 4
Points 31 to 36
4 6
5 2
2 9
7 3
9 8
5 4

[Askic V]

Thank you rik, it is very elegant solution. Can you just clarify why you used tmp.' instead for example tmp' if you needed to transpose the data?

[Rik]

You're welcome.

As to your question: habit. A single quote is not just the transpose, but the conjugate transpose. For real numbers there is no difference, but if you forget when working with complex number you will have a very difficult problem to debug. I try to learn myself good habits to avoid such bugs from ever occuring. For similar reasons I no longer use length when I actually mean numel (or I use size with an input argument).

a = 1+i;
a' , a.'

ans = 1.0000 - 1.0000i

ans = 1.0000 + 1.0000i

[Askic V]

Cool, I didn't know this. Thanks again!