Inconsistency when cbind tibble and matrix in R - r

Question

Unfortunately, I have spotted a weird inconsistency in the colnames when cbind different 2 particular objects: tibbles that has been by_group()ed and matrix. I writing this here because I would understand what is going on under the hood with the cbind operation and these 2 objects.
Consider the following objects:
Simple tibble
library(tidyverse)
tbl <- tibble(tbl_name = seq(1,8))
# # A tibble: 8 x 1
# tbl_name
# <int>
# 1 1
# 2 2
# 3 3
# 4 4
# 5 5
# 6 6
# 7 7
# 8 8
Simple data.frame
df <- data.frame(df_name = seq(1,8))
df
# df_name
# 1 1
# 2 2
# 3 3
# 4 4
# 5 5
# 6 6
# 7 7
# 8 8
Simple matrix
mtx <- matrix(seq(1,8), nrow = 8)
colnames(mtx) <- "mtx_name"
# mtx_name
# [1,] 1
# [2,] 2
# [3,] 3
# [4,] 4
# [5,] 5
# [6,] 6
# [7,] 7
# [8,] 8
by_grouped tibble
tb2 <- tibble(tbl2_name = seq(1,8),
tbl_group_by = c("a","b","b","c","d","d","d","d"))
tb2 <- tb2 %>%
group_by(tbl_group_by) %>%
mutate(N_by_group = n())
# A tibble: 8 x 3
# Groups: tbl_group_by [4]
# tbl2_name tbl_group_by N_by_group
# <int> <chr> <int>
# 1 1 a 1
# 2 2 b 2
# 3 3 b 2
# 4 4 c 1
# 5 5 d 4
# 6 6 d 4
# 7 7 d 4
# 8 8 d 4
When cbind them:
>This works (a.k.a: keeps the correct names)
# Comparison
# tibble & data.frame: OK
cbind(tbl,df)
# tbl_name df_name
# 1 1 1
# 2 2 2
# 3 3 3
# 4 4 4
# 5 5 5
# 6 6 6
# 7 7 7
# 8 8 8
# matrix & data.frame: OK
cbind(mtx,df)
# mtx_name df_name
# 1 1 1
# 2 2 2
# 3 3 3
# 4 4 4
# 5 5 5
# 6 6 6
# 7 7 7
# 8 8 8
# tibble & matrix: OK
cbind(tbl,mtx)
# tbl_name mtx_name
# 1 1 1
# 2 2 2
# 3 3 3
# 4 4 4
# 5 5 5
# 6 6 6
# 7 7 7
# 8 8 8
This doesn't work (a.k.a: destroyed the colname of the matrix)
# tibble(group_by()) & matrix: oops!!!!
cbind(tb2,mtx)
# New names:
# * NA -> ...4
# # A tibble: 8 x 4
# # Groups: tbl_group_by [4]
# tbl2_name tbl_group_by N_by_group ...4[,"mtx_name"]
# <int> <chr> <int> <int>
# 1 1 a 1 1
# 2 2 b 2 2
# 3 3 b 2 3
# 4 4 c 1 4
# 5 5 d 4 5
# 6 6 d 4 6
# 7 7 d 4 7
# 8 8 d 4 8
Any intuition of what's happening or how to prevent it, is very welcome. Thank you in advance.

Answer 1

We can remove the group attributes with ungroup and now cbind should work
library(dplyr)
cbind(ungroup(tb2), mtx)
-output
# tbl2_name tbl_group_by N_by_group mtx_name
#1 1 a 1 1
#2 2 b 2 2
#3 3 b 2 3
#4 4 c 1 4
#5 5 d 4 5
#6 6 d 4 6
#7 7 d 4 7
#8 8 d 4 8
Or specifically use cbind.data.frame because by default it may use cbind.matrix
cbind.data.frame(tb2, mtx)
When we create the 'tb2', after grouping, make sure to ungroup to prevent this kind of issues
tb2 <- tb2 %>%
group_by(tbl_group_by) %>%
mutate(N_by_group = n()) %>%
ungroup
Or make use of is_grouped_df to find if the data is grouped or not and then ungroup
f1 <- function(dat) {
if(dplyr::is_grouped_df(dat)) {
dat <- ungroup(dat)
}
dat
}
cbind(f1(tb2), mtx)