[R-meta] escalc error interpretation

Thu, Aug 26, 2021 11:47 PM

Just to confirm, this is indeed the issue:

escalc(measure="PLO", xi=0, mi=0)

Error in escalc(measure = "PLO", xi = 0, mi = 0) :                                                                    
  One or more group sizes are < 1.

If there are 0 people for whom the event of interest did occur (xi) and 0 people for whom the event did not occur (mi), then the group size (xi+mi) must be 0.

Best,
Wolfgang

-----Original Message-----
From: R-sig-meta-analysis [mailto:r-sig-meta-analysis-bounces at r-project.org] On
Behalf Of Michael Dewey
Sent: Thursday, 26 August, 2021 19:09
To: Philippe Tadger; r-sig-meta-analysis at r-project.org
Subject: Re: [R-meta] escalc error interpretation

Dear Philippe

Looking at out1 and out2 there seems to be a pair which are both zero.
Perhaps that is the issue.

Michael

On 26/08/2021 17:58, Philippe Tadger wrote:

Dear all,

I have an error with the function escalc,

Error in escalc(measure = "PLO", xi = out1, mi = out2, data = ddd, add =
1/2, : One or more group sizes are < 1.

that I don't know how to interpret what is wrong with my data. As follow
a reproducible example:

library(metafor)
ddd<-data.frame(
  ? tn = c(5535L,
  ???????? 5535L,1222L,1222L,
  ???????? 1032L,1032L,1392L,
  ???????? 1392L,990L,990L,
  ???????? 182L,182L,1406L,
  ???????? 1406L,21043L,21043L,
  ???????? 100L,100L,9647L,
  ???????? 9647L),
  ? tp = c(15L,15L,
  ???????? 2L,2L,2L,2L,
  ???????? 10L,10L,2L,2L,1L,
  ???????? 1L,9L,9L,28L,28L,
  ???????? 0L,0L,33L,33L),
  ? np = c(15L,15L,
  ???????? 2L,2L,2L,2L,
  ???????? 10L,10L,2L,2L,1L,
  ???????? 1L,9L,9L,28L,28L,
  ???????? 0L,0L,36L,36L),
  ? nn = c(5802L,
  ???????? 5802L,1234L,1234L,
  ???????? 1071L,1071L,1438L,
  ???????? 1438L,1005L,1005L,
  ???????? 194L,194L,1416L,
  ???????? 1416L,21900L,
  ???????? 21900L,105L,105L,
  ???????? 10121L,10121L),
  ? MA_num = c(10L,10L,
  ???????????? 10L,10L,10L,10L,
  ???????????? 10L,10L,10L,10L,
  ???????????? 10L,10L,10L,10L,
  ???????????? 10L,10L,10L,10L,
  ???????????? 10L,10L),
  ? SUBMA_num = c(45L,45L,
  ??????????????? 45L,45L,45L,45L,
  ??????????????? 45L,45L,45L,45L,
  ??????????????? 45L,45L,45L,45L,
  ??????????????? 45L,45L,45L,45L,
  ??????????????? 45L,45L),
  ? study = c(735L,
  ??????????? 735L,738L,738L,739L,
  ??????????? 739L,740L,740L,
  ??????????? 744L,744L,745L,
  ??????????? 745L,747L,747L,750L,
  ??????????? 750L,753L,753L,
  ??????????? 755L,755L),
  ? out1 = c(15L,
  ?????????? 5535L,2L,1222L,2L,
  ?????????? 1032L,10L,1392L,2L,
  ?????????? 990L,1L,182L,9L,
  ?????????? 1406L,28L,21043L,
  ?????????? 0L,100L,33L,
  ?????????? 9647L),
  ? out2 = c(0L,267L,
  ?????????? 0L,12L,0L,39L,
  ?????????? 0L,46L,0L,15L,0L,
  ?????????? 12L,0L,10L,0L,
  ?????????? 857L,0L,5L,3L,
  ?????????? 474L),
  ? study = as.factor(c("1","1","2","2",
  ????????????????????? "3","3","4","4",
  ????????????????????? "5","5","6","6",
  ????????????????????? "7","7","8","8",
  ????????????????????? "9","9","10",
  ????????????????????? "10")),
  ? group = as.factor(c("1","2","1","2",
  ????????????????????? "1","2","1","2",
  ????????????????????? "1","2","1","2",
  ????????????????????? "1","2","1","2",
  ????????????????????? "1","2","1","2"))
)
levels(ddd$group) <- c("sensitivity", "specificity")
ddd <- escalc(measure="PLO", xi=out1, mi=out2, data=ddd, add=1/2, to="all",
  ????????????? include=group=="sensitivity")

The error is triggered for the following conditional in the escalc function

if (any <https://rdrr.io/r/base/any.html>(ni < 1, na.rm=TRUE
<https://rdrr.io/r/base/logical.html>)) stop
<https://rdrr.io/r/base/stop.html>(mstyle$stop
<https://rdrr.io/r/base/stop.html>("One or more group sizes are < 1."))

Thanks in advance for your time

[R-meta] escalc error interpretation

Thread (3 messages)