Joining a keyed table on a non-keyed table is not working sometimes

[symbalex]

I am joining two data.table objects: dt_tbl (which has a key automatically created by dcast) on Y (which does not have a key), on a column called ROLE_TYPE. I am expecting the NumTxns column in the final object to have value 86 for ROLE_TYPE == "A", but instead I get NA.

Interestingly, the first join on ROLE_TYPE (dt_tbl on the dcast-ed object) works fine.

Reproducible example

library(data.table)

dt_tbl <- data.table(
  ROLE_TYPE = c("D", "A"), 
  CountCases = c(16L, 25L)
)

X <- data.table(
  outlier = c(FALSE, TRUE), 
  ROLE_TYPE = c("A", "A"),
  N = c(220L, 29L)
  )

# a dcast-ed table is now keyed
str(dcast(X, ROLE_TYPE ~ outlier, value.var = "N", fill = 0)) 

# cast and join
dt_tbl <- dcast(X, ROLE_TYPE ~ outlier, value.var = "N", fill = 0)[
  dt_tbl,
  on = "ROLE_TYPE"
  ]
# this is correct
dt_tbl
str(dt_tbl)

Y <- data.table(ROLE_TYPE = "A", NumTxns = 86L)

dt_tbl <- Y[
  dt_tbl,
  on = "ROLE_TYPE"
  ]
# why is NumTxns NA?
dt_tbl
# ROLE_TYPE NumTxns FALSE TRUE CountCases
# 1:         D      NA    NA   NA         16
# 2:         A      NA   220   29         25

Output of sessionInfo()

> sessionInfo()
R version 3.5.1 (2018-07-02)
Platform: x86_64-apple-darwin15.6.0 (64-bit)
Running under: macOS Sierra 10.12.5

Matrix products: default
BLAS: /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/3.5/Resources/lib/libRlapack.dylib

locale:
[1] en_AU.UTF-8/en_AU.UTF-8/en_AU.UTF-8/C/en_AU.UTF-8/en_AU.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] data.table_1.12.1

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.0        rstudioapi_0.7    magrittr_1.5      usethis_1.4.0     devtools_2.0.1    pkgload_1.0.2     R6_2.3.0          rlang_0.3.1      
 [9] tools_3.5.1       pkgbuild_1.0.2    sessioninfo_1.1.1 cli_1.0.1         withr_2.1.2       remotes_2.0.2     yaml_2.2.0        assertthat_0.2.0 
[17] digest_0.6.18     rprojroot_1.3-2   crayon_1.3.4      processx_3.2.0    callr_3.0.0       base64enc_0.1-3   fs_1.2.6          ps_1.2.1         
[25] curl_3.3          testthat_2.0.0    glue_1.3.0        memoise_1.1.0     compiler_3.5.1    desc_1.2.0        backports_1.1.2   prettyunits_1.0.2

[franknarf1]

Yeah, the key for x should not be preserved after x[i, on=key(x)], so the first join is also incorrect and is where the problem started.

library(data.table)
dx = data.table(id = "A", key = "id")
di = list(c("D", "A"))
(res <- dx[di])
#    id
# 1:  D
# 2:  A
key(res)
# [1] "id"

It should be sorted by its key.

Btw, overwriting objects / reusing names makes the example more confusing than it needs to be.

[symbalex]

Thanks - sorry for the confusing example :)

[jangorecki]

patch submitted