has_rownames() glitch

[romainfrancois]

I believe has_rownames() suffers from a problem in .row_names_info(type = 1):

df <- data.frame(x = 1:10)
df2 <- structure(df, .drop = FALSE)
df3 <- `attr<-`(df, ".drop", FALSE)

.row_names_info(df, type = 0)
#> [1]  NA -10
.row_names_info(df2, type = 0)
#> [1] NA 10
.row_names_info(df3, type = 0)
#> [1]  NA -10

.row_names_info(df, type = 1)
#> [1] -10
.row_names_info(df2, type = 1)
#> [1] 10
.row_names_info(df3, type = 1)
#> [1] -10

.row_names_info(df, type = 2)
#> [1] 10
.row_names_info(df2, type = 2)
#> [1] 10
.row_names_info(df3, type = 2)
#> [1] 10

^{Created on 2021-02-11 by the reprex package (v0.3.0)}

Specifically:

.row_names_info(df2, type = 1)
#> [1] 10

In the end, I believe this is a bug in structure(), and I've added a fix to dplyr so that it does not use structure() in the problem from tidyverse/dplyr#5745 but I still believe has_rownames() could work around that quirk.

[krlmlr]

Thanks. I think the problem really is in structure(), though it might be difficult to fix (or even work as documented, haven't checked).

I also remember this worked differently in older versions of R. With current R 4.0, we could indeed tweak the detection of this case and check the class of .row_names_info() instead:

df <- data.frame(x = 1:10)
df2 <- structure(df, .drop = FALSE)

dput(df)
#> structure(list(x = 1:10), class = "data.frame", row.names = c(NA, 
#> -10L))
dput(df2)
#> structure(list(x = 1:10), class = "data.frame", row.names = c(NA, 
#> 10L), .drop = FALSE)

identical(row.names(df), row.names(df2))
#> [1] TRUE

.row_names_info(df, 0)
#> [1]  NA -10
.row_names_info(df2, 0)
#> [1] NA 10
.row_names_info(mtcars, 0)
#>  [1] "Mazda RX4"           "Mazda RX4 Wag"       "Datsun 710"         
#>  [4] "Hornet 4 Drive"      "Hornet Sportabout"   "Valiant"            
#>  [7] "Duster 360"          "Merc 240D"           "Merc 230"           
#> [10] "Merc 280"            "Merc 280C"           "Merc 450SE"         
#> [13] "Merc 450SL"          "Merc 450SLC"         "Cadillac Fleetwood" 
#> [16] "Lincoln Continental" "Chrysler Imperial"   "Fiat 128"           
#> [19] "Honda Civic"         "Toyota Corolla"      "Toyota Corona"      
#> [22] "Dodge Challenger"    "AMC Javelin"         "Camaro Z28"         
#> [25] "Pontiac Firebird"    "Fiat X1-9"           "Porsche 914-2"      
#> [28] "Lotus Europa"        "Ford Pantera L"      "Ferrari Dino"       
#> [31] "Maserati Bora"       "Volvo 142E"

^{Created on 2021-02-12 by the reprex package (v1.0.0)}

Need to review what the behavior looks like in R 3.3 .

[romainfrancois]

Yeah, the underlying problem is here: https://github.com/wch/r-source/blob/b30641d3f58703bbeafee101f983b6b263b7f27d/src/main/attrib.c#L71

Maybe it will be picked up eventually in R, but in the meantime, tibble could be defensive.

[github-actions]

This old thread has been automatically locked. If you think you have found something related to this, please open a new issue and link to this old issue if necessary.