How do I get exact position of a fluctuation in a time series?

Suppose I have a very long row vector, say, of size 1*n for n>10000. I plot the vector as a time series. The time series has some large fluctuations, and I want to locate the exact time when the first fluctuation occurs. Is there any Matlab function, or any simple way to do this?

Answer 1

[Image Analyst]

You can use find(). For example, to find the first index where your signal surpasses a value of 50, do

index = find(yourSignal >= 50, 1, 'first');

[Vic]

What about if I don't know the value of "50", and I just know there will be a "relatively large" fluctuation?

[Image Analyst]

You can use outlier detection. For example look at the Median Absolute Deviation.

Or you can use http://www.mathworks.com/matlabcentral/fileexchange/3961-deleteoutliers from Brett at the Mathworks.

[Vic]

I posted a more precise question, plz see if u can help. thx a lot.

[Image Analyst]

I didn't notice any difference in the question. My answer stays the same.