# Making Cumulative Difference Calculation

104 views (last 30 days)
Hi,
I have a parameter with a column of data like this:
RunTime:
3779024.8; 3779025.8; 3779026.8; 3779027.8; 3779028.8; 3779029.8;
I want to know how I can make another parameter which calculates the cumulative difference between the values and presents it like this:
Elapsed Time:
0; 1; 2; 3; 4; 5; 6;
Thanks,

José-Luis on 16 Aug 2012
Not entirely sure if that's what you meant but here goes:
data = sort(rand(10,1));
cumDiff = [0;cumsum(diff(data))];
Cheers!

Friedrich on 16 Aug 2012
Small note:
The perfomance of this pretty bad and it doesnt make any sense doing it like this. Simply subtract the first value from all others (assuming data is sorted ascending). This is approx. 20-30 times faster and give the exact same result.
>> data = (1:1000)';
>> tic,cumDiff = [0;cumsum(diff(data))];toc
Elapsed time is 0.000560 seconds.
>> tic,cumDiff = [0;cumsum(diff(data))];toc
Elapsed time is 0.000549 seconds.
>> tic,cumDiff = [0;cumsum(diff(data))];toc
Elapsed time is 0.000611 seconds.
>> tic,cumDiff = data-data(1);toc
Elapsed time is 0.000021 seconds.
>> tic,cumDiff = data-data(1);toc
Elapsed time is 0.000024 seconds.
>> tic,cumDiff = data-data(1);toc
Elapsed time is 0.000023 seconds.
>> tic,cumDiff = data-data(1);toc
Elapsed time is 0.000024 seconds.
José-Luis on 17 Aug 2012
You are right. I guess i was blinded by the wording of the question.
Cheers!

Azzi Abdelmalek on 16 Aug 2012
Edited: Azzi Abdelmalek on 16 Aug 2012
A=[3779024.8; 3779025.8; 3779026.8; 3779027.8; 3779028.8; 3779029.8];
result=[0 ;cumsum(diff(A))];
disp('elapsed time');disp(sprintf('%d;',result))

#### 1 Comment

Thanks,
I accept this answer as well