Known issues

[Christen Fleming]

raster 3.0-7

Newer versions of raster have broken something in projection() and projection()<-. I am currently working on a fix. For the moment, it looks like everything works fine if you don't have the raster package loaded.

ctmm 0.5.5

occurrence() has a bug that distorts the density estimate if the mean location is far from the origin of the coordinate system. This bug is fixed in version 0.5.6.

ctmm 0.5.3 & 0.5.4

Step-wise model selection in ctmm.select() can proceed from OUF to the new OUf/OUO models introduced in 0.5.3 and terminate without attempting an OU model, even though an OU/IID model should ultimately be selected. This behavior is counter intuitive, because the OUf/OUO models are more autocorrelated than the OUF model and the OU model is less autocorrelated, but this does happen with real data. This issue is fixed in ctmm 0.5.4.

But, in pushing out the quick fix to 0.5.3's ctmm.select() bug, I slipped in a different bug to summary() to where all CIs on model-fit timescales are reported to run from 0 to infinity. The model fits are actually fine. The bug is isolated to summary()'s timescale CIs when applied to model fits. This issue is fixed in ctmm 0.5.5.

ctmm 0.5.10+

An update in either R package manipulate or RStudio has caused manipulate functions like zoom() and ctmm.guess() to be unreliable on some systems. I have not been able to successfully isolate this behavior. Running the functions multiple times often works.