"binning" data with 2D coordinates

I have a set of data points that are stored in a 3x1000 matrix. Each data point has (x, y, z) coordinates. How can I collapse this data into 2D coordinates, and then "bin" the data into a 3D graph?

Answer 1

[Walter Roberson]

[ux, ~, xidx] = unique(data(1,:));
[uy, ~, yidx] = unique(data(2,:));

%count the number of points at each unique x/y combination
counts = accumarray([xidx(:), yidx(:)], 1);  

%average the z that fall into each unique x/y combination
avgs = accumarray([xidx(:), yidx(:)], data(3,:).');

%create a list of the z that fall into each unique x/y combination
zs = accumarray([xidx(:), yidx(:)], data(3,:).', [], @(V) {V}, {});

But perhaps what you want is instead a 2D histogram:

https://www.mathworks.com/help/matlab/ref/histogram2.html

or if you have an older version:

https://www.mathworks.com/videos/making-a-2d-or-3d-histogram-to-visualize-data-density-97260.html

https://www.mathworks.com/matlabcentral/fileexchange/45325-efficient-2d-histogram--no-toolboxes-needed