CRAN submission

[xhdong-umd]

I did a R CMD Check, the list is quite long, include numerous small but cumbersome details. This could be a very time consuming process. I may also need to run check on 3 platforms: windows, mac, and linux.

I think this is an important direction but it will take time. I'll check on this when I got time and no other higher priority tasks on hand.

[chfleming]

CRAN check is very tedious. Fortunately, though, there is very little variation between platforms. Most of the effort is the upfront cost, and then there are small changes with every development release of R. If you run into anything that stumps you, please let me know.

[xhdong-umd]

I went through them and tried to fix most of them. There are two issues still remained:

• I called some ctmm internal function which are discouraged. @chfleming Are there exported function alternative? Or can you export them?

  'ctmm:::assign_speeds' 'ctmm:::distanceMLE' 'ctmm:::get.error' 'ctmm:::is.calibrated' 'ctmm:::unit' 'ctmm:::variogram.fit.backend'

• another main problem came from the package usage vignette. I'm not sure the check is reporting problems when the RMD can be knitted without any problem.
• At last, the knitted package usage vignette is about 6M, while the package size need to be less than 5M. I need either to reduce the size or justify it.

[chfleming]

That warning is really for packages that aren't maintained by the same group. I would try something like this:

FN <- c("assign_speeds","distanceMLE","get.error","is.calibrated","unit","variogram.fit.backend")
for(fn in FN) { assign(fn,eval(parse(text=paste0("ctmm:::",fn)))) }