Problem with arrival/departure dates for stationary periods

[Beata]

Hi,

I am performing gls analysis of Intigeo devices in FlightR. When I use stationary.migration.summary function I get  overlapping dates for stationary period. So I decided to use find.times.distribution to find arrival/departure dates to/from stationary periods calculated by stationary.migration.summary. Now I am wondering what key should I use to unify this approach across all my birds? Any advise?

Here is a piece of my results of two stationary periods.

Meanlat         SDlat                  Meanlon           SDlon            Dist2               Arrival.Q.50                      Departure.Q.50

 25.595678  2.4360796      -12.919237         1.0270114   453.1756       2019-09-28 07:40:58       2019-10-11 20:53:39

19.598708    2.1146732        -3.262447         0.6437934    1310.7202    2019-10-12 00:57:31      2020-03-16 08:01:10

 

[Eldar]

 Hi Beata,

 What kind of overlapping dates you get? It is likely that dates will overlap at the ends - around 90-95% percentile. Try adjusting prob.cutoff to e.g. 0.5 in the stationary.migration.summary. I think you should get the same results as from find.times.distribution then.

   Hope this helps,

     Eldar

[Beata]

Hi Eldar, 

thank you for a reply. I get overlaping median dates or they seem quite unrealistic (like those in first post where departure date is 11. October and arrival date to next stopover (over 1000 km) is 12. October.

I tried adjusting the probability cutoff. The maximum I was able to set was 0.3. If I set 0.5 I get message that the bird didn't move.