How to find the center of matrix

Dear all helper,

I want like to calculate the center of force in 2 point that I draw in the red color by using matlab.

Here is attech file of the csv data.

[Dyuman Joshi]

Center of force? Like Center of Mass or different?

What exactly is your data?

[Mingde]

Hello @Dyuman Joshi@DGM

Center of force and center of mass are the same.

Here is the sample data(CSV file) for measuring the center of force.

[DGM]

Your data is 14925x10, with periodic rows of NaN every 17 rows. It's not clear what this data represents or how the NaNs should be treated. Is this a contiguous 2D region? Multiple independent 2D regions? A 3D volume? Are the NaN rows merely delimiters/breaks to be discarded, or do they have numerical meaning? If they are to be kept, how should they be handled?

Answer 1

[DGM]

This seems fairly straightforward.

% a test array
F = zeros(5);
F(2,2) = 10;
F(4,4) = 10;

% global sum of F
Ft = sum(F(:))

Ft = 20

% coordinates of F centroid
cy = sum((1:size(F,1)).'.*sum(F,2))/Ft

cy = 3

cx = sum((1:size(F,2)).*sum(F,1))/Ft

cx = 3

% visualize the result
imshow(F,[]); hold on
plot(cx,cy,'*','markersize',20)

[DGM]

Assuming the NaN rows are page breaks and that each page is to be treated independently:

F = readmatrix('20220427T131641-2d.csv');

% reshape and crop off NaNs 
F = reshape([F; F(17,:)],17,10,[]);
F = F(1:end-1,:,:);

% for sake of demonstration, just pick two pages
F = F(:,:,370:371)

F = 
F(:,:,1) =

     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     7     7     7     7     7     8     7     7     8     7
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0


F(:,:,2) =

     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     7     7     7     7     7     7     7     7     7     7
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0
     0     0     0     0     0     0     0     0     0     0

szf = size(F);
% global sum of F for all pages
Ft = sum(sum(F,1),2)

Ft = 
Ft(:,:,1) =

    72


Ft(:,:,2) =

    70

% coordinates of F centroid
cy = sum((1:szf(1)).'.*sum(F,2))./Ft

cy = 
cy(:,:,1) =

     6


cy(:,:,2) =

     6

cx = sum((1:szf(2)).*sum(F,1))./Ft

cx = 
cx(:,:,1) =

    5.5556


cx(:,:,2) =

    5.5000

You can always squeeze() or reorient the outputs as you choose.

[Mingde]

thank you for your answer. @DGM

I've used matlab R2015b. So the F = readmatrix('20220427T131641-2d.csv'); is Undefined function or variable.

[DGM]

Oh. If you're using R2015b, you'll need to do more than that. Here. I'm sure there are other ways to do it, but this was tested in R2015b.

% read the file
% this is being done by lazy array expansion, which could be improved
fid = fopen('20220427T131641-2d.csv');

columns = 10;

F = [];
t = [];
while ~feof(fid)
    thislineraw = fgetl(fid);
    
    % try to get a line
    thisline = sscanf(thislineraw,'%f,',[1 columns]);
    
    if isnumeric(thisline) && numel(thisline) == columns
        % if it's a numeric vector with 10 elements, it's data
        F = [F; thisline]; %#ok<*AGROW>
        continue;
    end
    
    % if it's not, then it's a block header
    % i don't know if it's needed, but for example, 
    % i'm going to get the timestamps in vector t
    % if timestamps aren't needed, this part can be omitted
    thisline = regexp(thislineraw,': (\d+\.\d+),+$','tokens');
    if ~isempty(thisline)
        t = [t; str2double(thisline{1}{1})];
    end	
end

% reshape F 
F = reshape(F,16,10,[]);

% for sake of demonstration, just pick two pages
F = F(:,:,370:371)

szf = size(F);
% global sum of F for all pages
Ft = sum(sum(F,1),2)

% coordinates of F centroid
% need to use bsxfun() for array expansion prior to R2016b
cy = sum(bsxfun(@times,(1:szf(1)).',sum(F,2)))./Ft
cx = sum(bsxfun(@times,(1:szf(2)),sum(F,1)))./Ft

[Mingde]

Thank you for your answer.

[Image Analyst]

You never answered his questions about whether your multiple regions are to be treated as a single region, or whether you want the location of the centroid for each region individually. But he gave you the centroid of all regions as if they were a single region and you accepted so I guess that's what you wanted.

Answer 2

[Image Analyst]

Did you try

data(isnan(data)) = 0;
mask = data ~= 0;
props = regionprops('table', mask, data, 'WeightedCentroid');

[Mingde]

Thank you for your answer.