unexpected behavior when intersecting three tibbles

[sheridar]

When passing a list of tibbles to bed_intersect an unexpected error is thrown:

x <- tibble::tribble(
  ~chrom, ~start, ~end,
  'chr1', 100,    500,
  'chr2', 200,    400,
  'chr2', 300,    500,
  'chr2', 800,    900
)

y <- tibble::tribble(
  ~chrom, ~start, ~end, ~value,
  'chr1', 150,    400,  100,
  'chr1', 500,    550,  100,
  'chr2', 230,    430,  200,
  'chr2', 350,    430,  300
)

z <- tibble::tribble(
  ~chrom, ~start, ~end, ~value,
  'chr1', 150,    400,  100,
  'chr1', 500,    550,  100,
  'chr2', 230,    430,  200,
  'chr2', 750,    900,  400
)

lst <- list(y, z)

# These variations work:
bed_intersect(x, y, z)
bed_intersect(x, list(y, z))
bed_intersect(x, lst[1:2])

# VARIATION 1: If a list is passed, error is thrown
# this seems to be due to line 91:
# https://github.com/rnabioco/valr/blob/799dd742ebd681bced0ce03f294d55487729aeaf/R/bed_intersect.r#L91
bed_intersect(x, lst)

# this produces different error
bed_intersect(x, lst[1])

# VARIATION 2: This ignores the third tibble
bed_intersect(x, lst[[1]], lst[[2]])

[kriemo]

thanks for the detailed reprex. I worried that code might not be robust for detecting certain inputs. I'll work on a solution.

[kriemo]

With the changes introduced by #381 these errors should be resolved. Github actions are not running the build checks for ubuntu for some reason. Once these complete successfully I'll merge the changes and notify you.

library(valr)
library(tibble)
x <- tibble::tribble(
  ~chrom, ~start, ~end,
  'chr1', 100,    500,
  'chr2', 200,    400,
  'chr2', 300,    500,
  'chr2', 800,    900
)

y <- tibble::tribble(
  ~chrom, ~start, ~end, ~value,
  'chr1', 150,    400,  100,
  'chr1', 500,    550,  100,
  'chr2', 230,    430,  200,
  'chr2', 350,    430,  300
)

z <- tibble::tribble(
  ~chrom, ~start, ~end, ~value,
  'chr1', 150,    400,  100,
  'chr1', 500,    550,  100,
  'chr2', 230,    430,  200,
  'chr2', 750,    900,  400
)

lst <- list(y, z)

bed_intersect(x, y, z)
#> # A tibble: 11 × 8
#>    chrom start.x end.x start.y end.y value.y .source .overlap
#>    <chr>   <dbl> <dbl>   <dbl> <dbl>   <dbl> <chr>      <int>
#>  1 chr1      100   500     150   400     100 y            250
#>  2 chr1      100   500     150   400     100 z            250
#>  3 chr1      100   500     500   550     100 y              0
#>  4 chr1      100   500     500   550     100 z              0
#>  5 chr2      200   400     230   430     200 y            170
#>  6 chr2      200   400     230   430     200 z            170
#>  7 chr2      200   400     350   430     300 y             50
#>  8 chr2      300   500     230   430     200 y            130
#>  9 chr2      300   500     230   430     200 z            130
#> 10 chr2      300   500     350   430     300 y             80
#> 11 chr2      800   900     750   900     400 z            100
bed_intersect(x, list(y, z))
#> # A tibble: 11 × 8
#>    chrom start.x end.x start.y end.y value.y .source .overlap
#>    <chr>   <dbl> <dbl>   <dbl> <dbl>   <dbl> <chr>      <int>
#>  1 chr1      100   500     150   400     100 y            250
#>  2 chr1      100   500     150   400     100 z            250
#>  3 chr1      100   500     500   550     100 y              0
#>  4 chr1      100   500     500   550     100 z              0
#>  5 chr2      200   400     230   430     200 y            170
#>  6 chr2      200   400     230   430     200 z            170
#>  7 chr2      200   400     350   430     300 y             50
#>  8 chr2      300   500     230   430     200 y            130
#>  9 chr2      300   500     230   430     200 z            130
#> 10 chr2      300   500     350   430     300 y             80
#> 11 chr2      800   900     750   900     400 z            100
bed_intersect(x, lst[1:2])
#> # A tibble: 11 × 8
#>    chrom start.x end.x start.y end.y value.y .source .overlap
#>    <chr>   <dbl> <dbl>   <dbl> <dbl>   <dbl> <chr>      <int>
#>  1 chr1      100   500     150   400     100 1            250
#>  2 chr1      100   500     150   400     100 2            250
#>  3 chr1      100   500     500   550     100 1              0
#>  4 chr1      100   500     500   550     100 2              0
#>  5 chr2      200   400     230   430     200 1            170
#>  6 chr2      200   400     230   430     200 2            170
#>  7 chr2      200   400     350   430     300 1             50
#>  8 chr2      300   500     230   430     200 1            130
#>  9 chr2      300   500     230   430     200 2            130
#> 10 chr2      300   500     350   430     300 1             80
#> 11 chr2      800   900     750   900     400 2            100

bed_intersect(x, lst)
#> # A tibble: 11 × 8
#>    chrom start.x end.x start.y end.y value.y .source .overlap
#>    <chr>   <dbl> <dbl>   <dbl> <dbl>   <dbl> <chr>      <int>
#>  1 chr1      100   500     150   400     100 1            250
#>  2 chr1      100   500     150   400     100 2            250
#>  3 chr1      100   500     500   550     100 1              0
#>  4 chr1      100   500     500   550     100 2              0
#>  5 chr2      200   400     230   430     200 1            170
#>  6 chr2      200   400     230   430     200 2            170
#>  7 chr2      200   400     350   430     300 1             50
#>  8 chr2      300   500     230   430     200 1            130
#>  9 chr2      300   500     230   430     200 2            130
#> 10 chr2      300   500     350   430     300 1             80
#> 11 chr2      800   900     750   900     400 2            100

bed_intersect(x, lst[1])
#> # A tibble: 6 × 7
#>   chrom start.x end.x start.y end.y value.y .overlap
#>   <chr>   <dbl> <dbl>   <dbl> <dbl>   <dbl>    <int>
#> 1 chr1      100   500     150   400     100      250
#> 2 chr1      100   500     500   550     100        0
#> 3 chr2      200   400     230   430     200      170
#> 4 chr2      200   400     350   430     300       50
#> 5 chr2      300   500     230   430     200      130
#> 6 chr2      300   500     350   430     300       80

bed_intersect(x, lst[[1]], lst[[2]])
#> # A tibble: 11 × 8
#>    chrom start.x end.x start.y end.y value.y .source .overlap
#>    <chr>   <dbl> <dbl>   <dbl> <dbl>   <dbl> <chr>      <int>
#>  1 chr1      100   500     150   400     100 1            250
#>  2 chr1      100   500     150   400     100 2            250
#>  3 chr1      100   500     500   550     100 1              0
#>  4 chr1      100   500     500   550     100 2              0
#>  5 chr2      200   400     230   430     200 1            170
#>  6 chr2      200   400     230   430     200 2            170
#>  7 chr2      200   400     350   430     300 1             50
#>  8 chr2      300   500     230   430     200 1            130
#>  9 chr2      300   500     230   430     200 2            130
#> 10 chr2      300   500     350   430     300 1             80
#> 11 chr2      800   900     750   900     400 2            100

^{Created on 2021-10-17 by the reprex package (v2.0.0)}