How to remove repeating elements but maintain occurrences in an array?

Is there a way that I can get the first occurrence of consecutively repeating values, even if the same value occurs at few different places in the matrix?

Say I have a matrix

X=[2 2 2 2 -1 -1 -1 6 6 6 6 6 5 5 5 2 2 2 2 7 7 6 6]

I want the result to be

ans=[2 -1 6 5 2 7 6].

I've checked a similar question How to remove repeating elements from an array but using unique and keeping the same order as original matrix gives

ans=[2 -1 6 5 7].

I know I can use a loop like below. But I was wondering if there's a function for this.

for i=1:(length(X)-1)
    if (X(i+1,6)-X(i,6))~=0
        ans(i,1)=X(i+1,6);
    else
        ans(i,1)=NaN;
    end
end

A function to do this will be great.

[Stephen23]

Why do this in a loop when you can use the vectorized solution that I posted? Loops are a poor way to solve this problem in MATLAB, vectorization is much neater and faster .

[Image Analyst]

I would not use ans as the name of a variable - that has a special meeting in MATLAB.

[Vignesh]

Thanks Stephen! That helped. Image Analyst, it was just an example for the question. Thanks for pointing out anyway.

Answer 1

[Stephen23]

You could do this with diff:

>> X=[2 2 2 2 -1 -1 -1 6 6 6 6 6 5 5 5 2 2 2 2 7 7 6 6];
>> Y = [true,diff(X)~=0];
>> X(Y)

ans = [2,-1,6,5,2,7,6]

This works well with integers. If you are working with floating point numbers, then you might need to include some kind of tolerance in the diff:

>> Y = [true,abs(diff(X))>tol];

[Stephen23]

If you want this as a function, then you can easily define your own:

>> start = @(v)v([true,diff(v)~=0]);
>> start(X)

ans =    2    -1     6     5     2     7     6

Answer 2

[Prashanth]

ans=unique(X);

[Walter Roberson]

1. This has the side effect of sorting the output, which was not desired
2. This as the side effect of only outputing any one value once. We see from the example input and output that 2 is expected as the output twice.

[DGM]

For example:

% our input vector
X = [2 2 2 2 -1 -1 -1 6 6 6 6 6 5 5 5 2 2 2 2 7 7 6 6];

% Stephen's answer
start = @(v)v([true,diff(v)~=0]);
start(X)

ans = 1x7

2 -1 6 5 2 7 6

<mw-icon class=""></mw-icon>

<mw-icon class=""></mw-icon>

% sorted output from unique()
usort = unique(X)

usort = 1x5

-1 2 5 6 7

<mw-icon class=""></mw-icon>

<mw-icon class=""></mw-icon>

% stable output from unique()
usort = unique(X,'stable')

usort = 1x5

2 -1 6 5 7

<mw-icon class=""></mw-icon>

<mw-icon class=""></mw-icon>

While the ordering can be resolved by using the 'stable' option, we're not actually after unique values at all. The goal is to reduce each run of a given value to a single instance of that value. Since there are multiple runs of 2 and 6, the output from unique(...,'stable') still results in a loss of information.

As an aside, using 'ans' as a variable name is a problem waiting to happen.