crawl wrap does not fit all tracks

[SutorSW]

Hello all,

My name is Sean and I'm a student at Texas Tech University working on some tortoise movement and connectivity analyses. I am working with temporally irregular data (missing observations and irregular time stamps) collected from GPS loggers. I am attempting to regularize the data and impute missing locations for very small time gaps (< 1 hr), and so I have been using crawlWrap. This seems to be working very well at regularizing the data, but I am having an issue I have not encountered in any vignette walk-throughs and that I am struggling to resolve: crawlWrap is not fitting all my tracks (in the case below, it is predicting locations for 1193/1664 tracks). Any ideas what might be causing this?

**crw_out <- crawlWrap(obsData = data_split, timeStep = "30 min",
Time.name = "time", coord = c("x", "y"))

Fitting 1664 track(s) using crawl::crwMLE...
DONE

Predicting locations (and uncertainty) at 30 mins time steps for 1193 track(s) using crawl::crwPredict...**

Note that the many tracks are because I have split tracks that are greater than 2 hours apart and discarded any tracks shorter than 2 hours. I am awaiting permission to provide a small sample of the data so that I may be able to provide reproducible code.

Thank you in advance,

Sean

[bmcclintock]

It's likely that the failed tracks aren't converging due to poor initial values (theta) and/or insufficient data. If you search Discussions for "initial values" or "crawl", you'll find tips for dealing with this. For example, see #79. See also the theta argument in the crawlWrap help file (?crawlWrap).

[SutorSW]

Thank you for getting back to me this. I had been giving "retryFits" a go, but specifying theta has certainly helped.