This can probably be simplified, but eh.
r=22.5;
R=linspace(0,10,10);
x=repelem(R,3);
z=-(r-sqrt(r^2-R.^2));
Z=repelem(z,3);
Z(2:3:end)= Z(2:3:end)-1.04;
plot(R,z,'.');hold on
plot(x,Z,'-');
Z0 = Z;
tol = 1E-15; % or pick some tolerance
uz = unique(Z);
for k = 1:numel(uz)
idx = find(abs(Z-uz(k))<tol);
if numel(idx)>1
Z(idx(2:end)) = 0; % get rid of the rightmost instances
%Z(idx(1:end-1)) = 0; % get rid of the leftmost instances
end
end
[Z0' Z']
Z now only contains unique values.
If the duplicate values are created with a known period, it may suffice to directly remove them without the need for comparison.
