distToNearest errors

Issue #10 resolved
Former user created an issue

Following the distToNearest vignette, I tried the following:

> file <- system.file("extdata", "Influenza.tab", package="shm")
> db <- readChangeoDb(file)
> dim(db)
[1] 1359   35
> dist_m1n <- distToNearest(db, model="m1n", first=FALSE, normalize="length")
Warning messages:
1: In db$V == g :
  longer object length is not a multiple of shorter object length
2: In db$V == g :
  longer object length is not a multiple of shorter object length
3: In db$V == g :
  longer object length is not a multiple of shorter object length
4: In setup_parallel() : No parallel backend registered
5: <anonymous>: ... may be used in an incorrect context:.fun(piece, ...)6: <anonymous>: ... may be used in an incorrect context:.fun(piece, ...)

I understand that error4 might be related to our cluster, but the rest seem to be independent ? I also tried another small sample dataset, and got the same errors, as well as a similar one:

8: In db$J == g : longer object length is not a multiple of shorter object length

Comments (10)

  1. Jason Vander Heiden

    Warnings 4-6 are a problem with how we have the multiprocessing setup. Should run fine with warnings 4-6.

    Warnings 1-3 are a problem that I thought we fixed yesterday. Is this a build from today?

  2. Jason Vander Heiden

    Okay, that might solve the problem. Unless you need it urgently, I'd probably wait until Monday though. I'm cleaning up the vignettes now, so hopefully all the little errors (that we know about) should be resolved by Monday.

  3. Former user Account Deleted

    Warnings 1-3 are a problem that I thought we fixed yesterday. Is this a build from today?

    mea culpa . I was watching the issues but i suppose it's just good code etiquette to pull the bleeding edge before posting an issue?

    How do you prefer we operate with these repos ?

    LIke, would you like us to stick to tagged releases or pull from master after each commit ?

  4. Jason Vander Heiden

    Well, changeo, alakazam and shm don't have any tagged releases yet... They are all still prerelease, and I think it'll probably still be about two weeks before we have an official release. shm, in particular, will go through a few more changes.

    At this point, I would just pull the current version and rebuild before doing any analysis with them.

  5. Former user Account Deleted

    errors 1-6 gone on cluster with latest code.

    errors 1-3 gone on my local computer (parallel errors still reported, fwiw, but runs fine).

    Resolved ?

    Thanks guys.

  6. Namita Gupta

    Parallel warnings were fixed last night, I think. So the very latest version should run without any warnings at all :)

  7. Jason Vander Heiden

    This looks fixed. Along with all the other immediate stuff we needed to do, so the shm package should be good to use at this point!

  8. Log in to comment