uniqindx#

Purpose#

Computes the sorted index of x, leaving out duplicate elements.

Format#

index = uniqindx(x, flag)#

Parameters:

x (Nx1 or 1xN vector) – data
flag (scalar) – 1 if numeric data, 0 if character.

Returns:

index (Mx1 vector) – indices corresponding to the elements of x sorted in ascending order with duplicates removed.

Examples#

x = { 5, 4, 4, 3, 3, 2, 1 };

// Create a sorted index of all the unique elements in 'x'
ind = uniqindx(x, 1);

// Use the index 'ind' to return all of the unique elements
// of 'x' in ascending order
y = x[ind];

After running the above code, ind and y are equal to:

       7.0000000         1.0000000
       6.0000000         2.0000000
ind =  4.0000000    y =  3.0000000
       3.0000000         4.0000000
       1.0000000         5.0000000

Remarks#

Among sets of duplicates it is unpredictable which elements will be indexed.