convert documentation to roxygen

[chad-klumb]

closes #11

[chad-klumb]

build errors appear to be related to .Rd links to package ergm

not sure why this is showing up now, since those were in the existing versions of the .Rd files as well

[chad-klumb]

looks like the same issue was present in build #57 but as a note rather than a warning

[chad-klumb]

removed two links to package ergm (other option was making network suggest ergm)

build failures now appear related to CRAN connectivity issues

[skyebend]

this looks amazing! (and a huge amount of work!). how will we test the reverse dependencies to make sure they are not broken by subtle changes to namespace?

[chad-klumb]

I believe the two NAMESPACEs are equivalent; we are currently running revdeps on statnet2

closed/re-opened the PR to get travis to run again, but it now appears to be failing due to a different connectivity issue:

 ERROR
Running the tests in ‘tests/read.paj_tests.R’ failed.
Last 13 lines of output:
  > 
  > #issue with read.table and number.cols and fill...SanJuanSur_deathmessage.net has one row with 8 all the rest (including the first 5 have 5)
  > 
  > 
  > 
  > # this file has character encoding issues
  > scotland<-tempfile('scotland',fileext='.zip')
  > download.file('http://vlado.fmf.uni-lj.si/pub/networks/data/esna/scotland.zip',scotland)
  trying URL 'http://vlado.fmf.uni-lj.si/pub/networks/data/esna/scotland.zip'
  Error in download.file("http://vlado.fmf.uni-lj.si/pub/networks/data/esna/scotland.zip",  : 
    cannot open URL 'http://vlado.fmf.uni-lj.si/pub/networks/data/esna/scotland.zip'
  In addition: Warning message:
  In download.file("http://vlado.fmf.uni-lj.si/pub/networks/data/esna/scotland.zip",  :
    URL 'http://vlado.fmf.uni-lj.si/pub/networks/data/esna/scotland.zip': status was 'Couldn't connect to server'
  Execution halted

[chad-klumb]

I hope to return to this after #14 (which is of more immediate importance) is merged. I can compare revdeps for this against master as I did for #14, so that even if we can't test every last one, we can see that there are no new errors among those that we can test.

[krivit]

Great work, Chad!

Regarding the download test, we could condition it on environment variable _R_CHECK_CRAN_INCOMING_REMOTE_ not being set to "off" (or similar).

@CarterButts , since you are the maintainer, do you want to give this a once-over before merging?

[skyebend]

I looked at diffs for a bunch of the .rd files, and they look much better than the original.

[krivit]

Looks like merging PR #14 broke this PR. Meanwhile, there haven't been any objections from Carter. @chad-klumb , can you reconcile these changes, and then we'll merge?

[chad-klumb]

believe merge conflicts have been resolved

will try running revdeps tonight; results hopefully tomorrow

[chad-klumb]

results were the same as for current master, with the exception that collpcm was removed from CRAN at the time the revdeps were run (apparently it's now back up), so that particular package wasn't tested this time around

I say merge it