[R] [EXTERNAL] Re: Very strange behavior of 'rep'

Thu Aug 15 21:00:32 CEST 2024

Ok � to be fair, it looks like I need to load everything and reproduce exactly as its occuring. I suspect its a weird memory hole in one of the loaded packages.

Here � this should do it.

"%,%" <- paste0

"factorial.design" <-
function(...)
{
  m <- match.call()
  cc <- m
  cc[[1]] <- as.name("c")
  arg.vals <- eval(cc, sys.parent())
  n.args <- length(arg.vals)

  n.l <- 0
  n.r <- n.args-1
  fwd.cpd <- c(1, cumprod(arg.vals))
  rev.cpd <- c(cumprod(arg.vals[n.args:2])[(n.args-1):1],1)
  "%,%" <- paste0
  main.call <- as.call(expression(cbind))
  for(k in 1:n.args)
  {
      arg.k <- eval(m[[1+k]], sys.parent())
      rep.call <- as.call(expression(rep))
      rep.call$x <- 1:arg.k
      if(n.l>0) rep.call$times <- fwd.cpd[k]
      if(n.r>0) rep.call$each <- rev.cpd[k]
      main.call[["x" %,% k]] <- rep.call
      n.l <- n.l + 1
      n.r <- n.r - 1
  }
  eval(main.call)
}

add.zeros <-
function(x,B)
{
    do.one <- function(x,B)
    {
      delta <- floor(logb(B, 10)) - floor(logb(x, 10))
      z <- ""
      if(delta > 0) z <- paste(rep("0", delta), collapse="")
      z%,%x
    }
    if(length(x)==1) ans <- do.one(x,B)
    if(length(x)>1) ans <- sapply(x, FUN=do.one, B=B)
    ans
}

library(lme4)

rho.ICC <- 0.5
s2.b <- 0.5*rho.ICC/(1-rho.ICC)
s2.e <- 0.5
delta <- 0.27

n.per.grp.lst <- (264/2*(1-450/9900*(5:0)))

## I will different values of 'h' and at these contrasts as well as flat
## I want to compare the AUC approach with just the test of the main effects constant arm coefficient

tt.seq <- c(0,3,6,12,18)
n.tt <- length(tt.seq)
cntr <- c(0.5,1,1,1,0.5)

## (delta1 + 2*delta2 + 2*delta3 + 2*delta4 + delta5)/2 = 0.27
## delta1 + 2*delta2 + 2*delta3 + 2*delta4 + delta5 = 2*0.27
## (delta1 + 2*delta2 + 2*delta3 + 2*delta4 + delta5)/8 = 2*0.27/8 = 0.27/4

H <- 0.08+0.005*(0:6)
shapes <- list(INC= (-floor(n.tt/2)):floor(n.tt/2),
               DEC= floor(n.tt/2):(-floor(n.tt/2)),
               FLT= rep(0,n.tt),
               PLT= c(-floor(n.tt/2):0,rep(floor(n.tt/2)+1, n.tt-floor(n.tt/2)-1)))
trnds <- names(shapes)

delta0.j <- rep(5, n.tt)

n.n.lst <- length(n.per.grp.lst)
n.trnds <- length(trnds)
n.H <- length(H)

## We're swarming over conditions. It works like this. This is the
## full list of conditions as a factorial design list. You have to
## determine the total number of conditions and then set up that many
## directories in the swarm folder
conds <- factorial.design(n.n.lst, n.trnds, n.H)

## currently with n.n.lst=6, n.trnds=4, n.H=7 that gives
## length(conds)=6x4x7=168

## then one copy of this file sits in each of the subdirectories.
## Parameters for this run are determined by getting the subdirectory
## number and then pulling the parameter indices from that element
## in the factorial design.
node <- 29

n.per.grp <- n.per.grp.lst[conds[node, 1]]
trnd <- trnds[conds[node, 2]]
h <- H[conds[node, 3]]

delta.j <- delta + h*shapes[[trnd]]

## covariates:
ID.i <- c(outer(rep(1:n.per.grp,each=n.tt),c(0,n.per.grp),FUN="+"))
arm.i <- rep(0:1, each=n.per.grp*n.tt)

length(arm.i)

________________________________
From: Rui Barradas <ruipbarradas using sapo.pt>
Sent: Thursday, August 15, 2024 2:51 PM
To: Izmirlian, Grant (NIH/NCI) [E] <izmirlig using mail.nih.gov>; r-help using r-project.org <r-help using r-project.org>
Subject: [EXTERNAL] Re: [R] Very strange behavior of 'rep'

�s 19:39 de 15/08/2024, Izmirlian, Grant (NIH/NCI) [E] via R-help escreveu:
> \n<<SEND AS PLAIN TEXT!>>\n\n \n<<
> This is very weird. I was running a swarm job on the cluster and it bombed
> only for n.per.grp=108, not for the other values. Even though
> n.per.grp*n.tt is 540, so that the length of the call to 'rep'
> should be 1080, I'm getting a vector of length 1078.
>      n.per.grp <- 108
>      n.tt <- 5
>      n.per.grp*n.tt
>      length(rep(0:1, each=n.per.grp*n.tt))
>      length(rep(0:1, each=108*5))
>>> \n<<Grant Izmirlian grant.izmirlian at nih dot gov>>\n\n\n\n
> --please do not edit the information below--
>
> R Version:
>   platform = x86_64-pc-linux-gnu
>   arch = x86_64
>   os = linux-gnu
>   system = x86_64, linux-gnu
>   status =
>   major = 4
>   minor = 4.1
>   year = 2024
>   month = 06
>   day = 14
>   svn rev = 86737
>   language = R
>   version.string = R version 4.4.1 (2024-06-14)
>   nickname = Race for Your Life
>
> Locale:
>   LC_CTYPE=C.UTF-8;LC_NUMERIC=C;LC_TIME=C.UTF-8;LC_COLLATE=C.UTF-8;LC_MONETARY=C.UTF-8;LC_MESSAGES=C.UTF-8;LC_PAPER=C.UTF-8;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=C.UTF-8;LC_IDENTIFICATION=C
>
> Search Path:
>   .GlobalEnv, package:lme4, package:Matrix, package:stats,
>   package:graphics, package:grDevices, package:utils, package:datasets,
>   package:showtext, package:showtextdb, package:sysfonts,
>   package:methods, Autoloads, package:base
>
>       [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://gcc02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fstat.ethz.ch%2Fmailman%2Flistinfo%2Fr-help&data=05%7C02%7Cizmirlig%40mail.nih.gov%7C4e8285dc4d984018750208dcbd5b9586%7C14b77578977342d58507251ca2dc2b06%7C0%7C0%7C638593448252923470%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C0%7C%7C%7C&sdata=ERmJnji4cVBWI36ESgUcMPY5E52l%2FTrb%2BF4s%2F1gGSVg%3D&reserved=0<https://stat.ethz.ch/mailman/listinfo/r-help>
> PLEASE do read the posting guide https://gcc02.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.r-project.org%2Fposting-guide.html&data=05%7C02%7Cizmirlig%40mail.nih.gov%7C4e8285dc4d984018750208dcbd5b9586%7C14b77578977342d58507251ca2dc2b06%7C0%7C0%7C638593448252932192%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C0%7C%7C%7C&sdata=UXsJuXSKTRbYPwcPnTwKQeARyE6AksBsXwNq4NICJGw%3D&reserved=0<http://www.r-project.org/posting-guide.html>
> and provide commented, minimal, self-contained, reproducible code.
Hello,

I cannot reproduce this behavior.

n.per.grp <- 108
n.tt <- 5
n.per.grp*n.tt
#> [1] 540
length(rep(0:1, each = n.per.grp*n.tt))
#> [1] 1080
length(rep(0:1, each = 108*5))
#> [1] 1080

But my version of R and my OS are different.
(I don't see how the error in the OP can be related to R version or OS.)

R.version
#>                _
#> platform       x86_64-w64-mingw32
#> arch           x86_64
#> os             mingw32
#> crt            ucrt
#> system         x86_64, mingw32
#> status
#> major          4
#> minor          4.1
#> year           2024
#> month          06
#> day            14
#> svn rev        86737
#> language       R
#> version.string R version 4.4.1 (2024-06-14 ucrt)
#> nickname       Race for Your Life

Hope this helps,

Rui Barradas

--
Este e-mail foi analisado pelo software antiv�rus AVG para verificar a presen�a de v�rus.
https://gcc02.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.avg.com%2F&data=05%7C02%7Cizmirlig%40mail.nih.gov%7C4e8285dc4d984018750208dcbd5b9586%7C14b77578977342d58507251ca2dc2b06%7C0%7C0%7C638593448252937455%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C0%7C%7C%7C&sdata=wlmoS3S2xNYFV9E%2FlVjvKw1XGhSQjx3KoQjgOBwNp4I%3D&reserved=0<http://www.avg.com/>
CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and are confident the content is safe.

	[[alternative HTML version deleted]]