Hi all, I’ve just started to learn working with R. I wonder if it is possible to make a ‘temporal correlogram’ in R and how to do it. Actually what I have are movement tracks of birds, existing as XY-positions every 5 minutes for a duration of at least 60 minutes (but up to 8 hours). From these tracks, I calculated step lengths (ie distances travelled within each 5 min-step) and turning angles (also per 5 min-step) for each ‘tracking session’ (I tracked each individual bird several times). I manually calculated the correlation between step-legths at time t and t+5min and they are strongluy correlated. I want to know for how many time steps this correlation holds. Therefore,I want to calculate autocorrelations (between step lengths, between turning angles - and eventually also between step lengths and turning angles, but that’s more complex) for subsequent steps (so I need a moving window frame for delta time/step). On the Y-axis, I would then have rho (autocorrelation coefficient for step length or turning angle), and on the X-axis, delta time (as time steps 0, 5, 10,15, ... minutes). Ideal would be to have the average rho + CI (or SD) displayed over all individuals (or tracking sessions). Is there a package to do this i R? I searched the web but found only correlograms for spatial data (sncf-package), not temporal. Thank you very much for your kind assistance! Valérie ---------------------------------------------------------------------------- -------------------------------------------------- Valérie Lehouck Assistant Professor Research Unit Terrestrial Ecology Ghent University KL Ledeganckstraat 35 9000 Ghent Belgium Tel.: +32(0)9/264 52 58 website: http://www.ecology.ugent.be/terec/personal.php?pers=vl [[alternative HTML version deleted]]