search

swihart / repeated

Non-normal repeated measurements models
GNU General Public License v2.0
0 stars 0 forks source link

kalcount.R examples (local-mac) #9

Closed swihart closed 7 years ago

swihart commented 7 years ago

kalcount.R examples (local-mac) skipped for now and left in \dontrun; continued on to do namespace stuff

swihart commented 7 years ago

Not complaining, but it appears the other fixes fixed this. Just removed \dontrun and everything is dandy on all 3 platforms.

swihart commented 7 years ago

removing dontrun dontmeanathing if you dontrerundevtools.

Turns out I'm still getting error local mac. I think it is becasue there is double in the .C call when it should be as.double. I'm going to fix these and rerun.

Result:

still bombs.

base::assign(".ptime", proc.time(), pos = "CheckExEnv")
> ### Name: kalcount
> ### Title: Repeated Measurements Models for Counts with Frailty or Serial
> ###   Dependence
> ### Aliases: deviance.kalcount fitted.kalcount kalcount print.kalcount
> ###   residuals.kalcount
... 13 lines ...
+   restovec(matrix(c(1,1,1,1,0,1,0,1,0,5,3,3,4,1,4,4,2,3,2,5),
+       ncol=5,byrow=TRUE))
> reps <- rmna(y, ccov=tr, tvcov=dd)
> #
> # log normal intensity, independence model
> kalcount(reps, intensity="log normal", dep="independence", preg=1,
+   pshape=0.1)
Error in kalcount(reps, intensity = "log normal", dep = "independence",  : 
  invalid mode (NULL) to pass to C or Fortran (arg 16)
Calls: kalcount -> .C
Execution halted

Which .C call is it? What argument is 16?

swihart commented 7 years ago

https://github.com/beechung/Latent-Factor-Models/issues/7

Freezing god in link above just made sure everything was as.double or as .integer. Double chekc that every arg is one of those...

swihart commented 7 years ago

Try

    tvcov=resp$tvcov$tvcov,

to


    tvcov=as.double(resp$tvcov$tvcov),

local-mac error removed!

Now test on Travis CI and Win-builder....results:

Solved like a boss.