This may do what you want:
% DEFINE: Variable Density Vector —
% s: 3-element vector of start values
% e: 3-element vector of end values
% d: 3-element vector of density values (number of entries between start and end)
vdv = @(s,e,d) sort([linspace(s(1),e(1),d(1)) linspace(s(2),e(2),d(2)) linspace(s(3),e(3),d(3))]);
s = [1 4 6];
e = [3.9 5.9 10];
d = [3 5 3];
V = vdv(s,e,d);
It does no error checking (for instance to correct for overlapping start and end values), and it simply sorts the vector (in ascending order) to eliminate obvious problems, but it’s the only sort of ‘function’ I can imagine that will do what you want. The total length is the sum of the ‘d’ vector. Its construction and operation is straightforward, so you can alter it as you wish to make it more functional in your application.
