Variable marker size scatter plot

I am trying to create an scatter plot as shown below from a regular scatter plot.

The objetive is that the size of the marker corresponding to a certain area becomes dependant of the number of points in that area.

Does anybody know how to do this?

Answer 1

[Steven Lord]

The size input argument to the scatter function can be a scalar or an array of the same size as the coordinate data. In the example below the marker at coordinates (10, 100) is much larger than the marker at coordinates (1, 1).

x = 1:10;
y = x.^2;
scatter(x, y, 2*y)

[Xavier Simón Montero]

Thank you Steven.

I am aware there can be different sizes, I am trying to link them to the relative number of points of the certain area containing those points.

Actually triying with a bivariant normal distribution, but my computer rans out of memory for my number of points (x and y of 20300 points each)

[Steven Lord]

Ah, so the problem you're trying to solve is how to bin the data to generate the array of marker sizes? I think the histcounts2 function will help you with that task.

[Xavier Simón Montero]

Thank you, this has worked