[Rd] pmatch bug?

[Duncan Murdoch]

I think your example is overly complicated.  Wouldn't it be enough to 
show one vector that gives a bad result?  For example:

   x <- c(1:51, 1:51)
   pmatch(x, x)

which gives

   [1]   1   2   3   4   5   6   7   8   9  10  11  12  13  14  15  16 
17  18  19  20  21  22  23  24  25  26  27  28  29
  [30]  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45 
46  47  48  49  50  51  NA  NA  NA  NA  NA  57  58
  [59]  59  60  61  62  63  64  65  66  67  68  69  70  71  72  73  74 
75  76  77  78  79  80  81  82  83  84  85  86  87
  [88]  88  89  90  91  92  93  94  95  96  97  98  99 100 101 102

(with the NAs showing up at locations 52 to 56).

I'm not 100% sure that's a bug, since the documentation for pmatch 
doesn't discuss what should happen if table (the second argument) 
contains duplicates.  I think I'd agree with you that you should get 
1:102 as the output, but maybe that was never intended to be supported.

Duncan Murdoch

On 2026-06-15 3:12 p.m., François Rousset via R-devel wrote:
> Dear R-devel list,
> 
> the following code shows NA's appearing in the result of pmatch() when
> comparing a vector to itself when the length of the vector is more than 100.
> The main specificity of this example is that elements are repeated in
> the vector which is matched to itself.
> The results when they do not include any NA (i.e. for argument of length
> <=100) are exactly as I expect from the documentation.
> I see that the source C code uses distinct algorithms whether n_input <=
> 100 || n_target <= 100 or not. Could there by a problem in the source
> code for larger values ?
> The NA's correspond to the first positions of the second replicate of
> the integer sequence, e.g. positions 52 to 56 if n=51 in the example below.
> 
> But as shown below, the NA's do not appear when the sequence is reversed
> before being compared to itself.
> 
> This was first detected with R 4.5.3 and is reproducible with a
> just-downloaded, virgin R-devel installation.
> 
> Thanks in advance for any feedback,
> 
> F.
> 
> ======================
> 
> countNAs <- function(n, rev.=FALSE) {
>     seqn <- seq(n)
>     if (rev.) seqn <- rev(seqn)
>     seqn <- rep(seqn,2)                         # of  length 2 n: NA's
> appear when 2 n > 100
>     chk <- pmatch(seqn, seqn)             # I expect the result to be
> seq(2*n)
>     sum(is.na(chk))
> }
> 
> sapply(1:100, countNAs )
> sapply(1:100, countNAs , rev.=TRUE)
> 
> =======================
> 
> Results:
>   > sapply(1:100, countNAs )
>     [1]  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0 0  0
> 0  0  0  0  0  0
>    [29]  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0 0  0
> 5  5  5  5  5  5
>    [57]  5  5  5  6  6  6  6  6  6  6  6  6  6  7  7  7  7  7  7  7 7  7
> 7  8  8  8  8  8
>    [85]  8  8  8  8  8  9  9  9  9  9  9  9  9  9  9 10
>   > sapply(1:100, countNAs , rev.=TRUE)
>     [1] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
> 0 0 0 0 0 0 0 0 0
>    [43] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
> 0 0 0 0 0 0 0 0 0
>    [85] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
> 
>   > sessionInfo()
> R Under development (unstable) (2026-06-14 r90150 ucrt)
> Platform: x86_64-w64-mingw32/x64
> Running under: Windows 10 x64 (build 19045)
> 
> Matrix products: default
>     LAPACK version 3.12.1
> 
> locale:
> [1] LC_COLLATE=French_France.utf8  LC_CTYPE=French_France.utf8
> [3] LC_MONETARY=French_France.utf8 LC_NUMERIC=C
> [5] LC_TIME=French_France.utf8
> 
> time zone: Europe/Paris
> tzcode source: internal
> 
> attached base packages:
> [1] stats     graphics  grDevices utils     datasets  methods  base
> 
> loaded via a namespace (and not attached):
> [1] compiler_4.7.0 tools_4.7.0
> 
> ______________________________________________
> R-devel using r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel