Using Autocorrelation with Finite Data Sets

[rzzzzzzz]

I am trying to autocorrelate data that has been collected in a text file.  However, every attempt at autocorrelation results in a triangular-shaped curve with a spike on the top.  I would like to know what I could add in my VI to eliminate the triangular-shaped curve, which arises because I am examining a finite data set.

Cheers,
Chris G.

[Zador]

When using the Fourier method, the standard approach is to pad the data with zeros (on both ends if you are interested in both positive and negative lags). If you have a copy of the Numerical Recipes book, check Chapter 13. If you don't have one, get it asap!

Zador

[rzzzzzzz]

But I expect my autocorrelation function will decay much more rapidly than my total acquisition time, so I dont believe "zero-buffing" will solve my problem.

The shape of my autocorrelation function is triangular because my raw data acts like a square pulse with noise.  Since this triangular shape dominates my resulting autocorrelation function, it makes it very difficult to analyze.  Any suggestions on how I can resolve this issue?

Thanks,

Chris G.

[rzzzzzzz]

Ah!

I believe I need to subtract off the mean value from my raw data.  This forces the average value of my raw data to be zero.  Thanks for your help, as it directed me to a great reference for my research.

Cheers,

Chris G.