[Thread Prev][Thread Next][Index]

[ferret_users] lag correlations: time-lag products for all lags

Hi Ferreters -

I want to construct a lag-correlation matrix, ideally without a REPEAT loop (there are thousands of points in my time series).

Thus I need to find the time-lagged products, including all lags within a range.

Suppose I have a time series of values DATA[L=1:LL], and the lag range is from L-R to L+R.
If I put the lag range on a M axis, _E=1:2*R+1, each element would be the sum of lagged products:

LAG[M=1] is the sum of: DATA*DATA (over all L)

LAG[M=2] is the sum of: DATA*DATA[L=@shf:1]

LAG[M=3] is the sum of: DATA*DATA[L=@shf:2]


LAG[M=2*R+1] is the sum of: DATA*DATA[L=@shf:2*R]   

For the moment, let's not worry about the ones that run off the end of the L-axis.

Can I do this without a massive REPEAT loop? For instance, is there a way to promote the time series DATA to 2-d where the 2nd dimension is the time-shifted terms as above?

Billy K

[Thread Prev][Thread Next][Index]
Contact Us
Dept of Commerce / NOAA / OAR / PMEL / Ferret

Privacy Policy | Disclaimer | Accessibility Statement