为什么 tidyverse group_by 在 R 更新后表现出意外

Question

It used to work fine and then I updated R, After the updates. the group_by function considers every row as a group. In the following example dataset dtt if I filter the dataset to only one group and run the code, it works as expected. However if run the same code for all groups, it does not work as expected.

Here are working and not working codes and below is data.

#Filter dtt to only one group (x,y) and run the code then it works as expected as below

dtt_xy<-dtt%>%
        filter(x==-121 & y == 65)
dtt_xy

dtt_output <- dtt_xy%>%
  group_by(x, y) %>%
  group_by(grp = cumsum(c(TRUE, diff(Date) != 1)), .add = TRUE)
dtt_output #Expected output

#Now if run the same code for the whole dataset ie, dtt it does not work

dtt_output <- dtt%>%
  group_by(x, y) %>%
  group_by(grp = cumsum(c(TRUE, diff(Date) != 1)), .add = TRUE)
dtt_output #Not expected output . expectation is 35 groups 

Sample Data

dtt<-structure(list(x = c(-121, -120, -121, -120, -121, -120, -121, 
-120, -121, -120, -121, -120, -121, -120, -121, -120, -121, -120, 
-121, -120, -121, -121, -121, -120, -121, -120, -121, -120, -121, 
-120, -121, -120, -121, -120, -121, -120, -121, -120, -121, -120, 
-121, -120, -121, -120, -121, -120, -121, -120, -121, -120, -121, 
-120, -121, -120, -121, -120, -121, -120, -121, -120, -121, -120, 
-121, -120, -121, -120, -121, -120, -121, -120, -121, -120, -121, 
-120, -121, -120, -121, -120, -121, -120, -121, -120, -121, -120, 
-121, -120, -120, -120, -121, -120, -121, -120, -121, -120, -121, 
-120, -121, -120, -121, -120, -121, -120, -121, -120, -121, -120, 
-121, -120, -121, -120, -121, -120, -121, -120, -121, -120, -121, 
-120, -121, -120, -121, -120, -121, -120, -121, -120, -121, -120, 
-121, -120, -121, -120, -121, -120, -121, -120, -121, -120, -121, 
-120, -121, -120, -121, -120, -121, -120, -121, -120, -121, -120
), y = c(65, 65, 63, 63, 65, 65, 63, 63, 65, 65, 63, 63, 65, 
65, 63, 63, 65, 65, 63, 63, 65, 63, 65, 65, 63, 63, 65, 65, 63, 
63, 65, 65, 63, 63, 65, 65, 63, 63, 65, 65, 63, 63, 65, 65, 63, 
63, 65, 65, 63, 63, 65, 65, 63, 63, 65, 65, 63, 63, 65, 65, 63, 
63, 65, 65, 63, 63, 65, 65, 63, 63, 65, 65, 65, 65, 65, 65, 65, 
65, 63, 63, 65, 65, 63, 63, 63, 63, 65, 63, 63, 63, 65, 65, 63, 
63, 65, 65, 63, 63, 65, 65, 63, 63, 65, 65, 63, 63, 63, 63, 65, 
65, 65, 65, 65, 65, 65, 65, 65, 65, 63, 63, 65, 65, 63, 63, 65, 
65, 65, 65, 63, 63, 65, 65, 63, 63, 65, 65, 63, 63, 65, 65, 63, 
63, 65, 65, 63, 63, 65, 65, 63, 63), Date = structure(c(5123, 
5123, 5123, 5123, 5124, 5124, 5124, 5124, 5125, 5125, 5125, 5125, 
5126, 5126, 5126, 5126, 5127, 5127, 5127, 5127, 5128, 5128, 5177, 
5177, 5177, 5177, 5178, 5178, 5178, 5178, 5179, 5179, 5179, 5179, 
5180, 5180, 5180, 5180, 5181, 5181, 5181, 5181, 5200, 5200, 5200, 
5200, 5201, 5201, 5201, 5201, 5202, 5202, 5202, 5202, 5203, 5203, 
5203, 5203, 5204, 5204, 5204, 5204, 5205, 5205, 5205, 5205, 5206, 
5206, 5206, 5206, 5238, 5238, 5239, 5239, 5240, 5240, 5273, 5273, 
5273, 5273, 5274, 5274, 5274, 5274, 5319, 5319, 5320, 5325, 5326, 
5326, 5327, 5327, 5327, 5327, 5328, 5328, 5328, 5328, 5329, 5329, 
5329, 5329, 5330, 5330, 5330, 5330, 5331, 5331, 5344, 5344, 5345, 
5345, 5381, 5381, 5382, 5382, 5383, 5383, 5383, 5383, 5384, 5384, 
5384, 5384, 5401, 5401, 5402, 5402, 5402, 5402, 5403, 5403, 5403, 
5403, 5404, 5404, 5404, 5404, 5405, 5405, 5405, 5405, 5406, 5406, 
5406, 5406, 5407, 5407, 5407, 5407), class = "Date")), row.names = c(NA, 
-150L), class = c("tbl_df", "tbl", "data.frame"))

Session Info

R version 4.0.4 (2021-02-15)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 19042)

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.1252  LC_CTYPE=English_United States.1252   
[3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C                          
[5] LC_TIME=English_United States.1252    

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

other attached packages:
[1] forcats_0.5.1   stringr_1.4.0   dplyr_1.0.5     purrr_0.3.4     readr_1.4.0     tidyr_1.1.3    
[7] tibble_3.1.0    ggplot2_3.3.3   tidyverse_1.3.0

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.6       cellranger_1.1.0 pillar_1.5.1     compiler_4.0.4   dbplyr_2.1.0     tools_4.0.4     
 [7] jsonlite_1.7.2   lubridate_1.7.10 lifecycle_1.0.0  gtable_0.3.0     pkgconfig_2.0.3  rlang_0.4.10    
[13] reprex_1.0.0     cli_2.3.1        rstudioapi_0.13  DBI_1.1.1        haven_2.3.1      withr_2.4.1     
[19] xml2_1.3.2       httr_1.4.2       fs_1.5.0         generics_0.1.0   vctrs_0.3.6      hms_1.0.0       
[25] grid_4.0.4       tidyselect_1.1.0 glue_1.4.2       R6_2.5.0         fansi_0.4.2      readxl_1.3.1    
[31] modelr_0.1.8     magrittr_2.0.1   backports_1.2.1  scales_1.1.1     ellipsis_0.3.1   rvest_1.0.0     
[37] assertthat_0.2.1 colorspace_2.0-0 utf8_1.2.1       stringi_1.5.3    munsell_0.5.0    broom_0.7.5     
[43] crayon_1.4.1 

Answer 1

Yes, this has been one of the recent change in dplyr when you do a nested group_by . An issue was created earlier for this but it was closed and it doesn't seem that this behaviour is going to change.

Solution is to use mutate to create the new column and then use it in group_by .

library(dplyr)

dtt%>%
  group_by(x, y) %>%
  mutate(grp = cumsum(c(TRUE, diff(Date) != 1))) %>%
  group_by(grp, .add = TRUE)

# A tibble: 150 x 4
# Groups:   x, y, grp [35]
#       x     y Date         grp
#   <dbl> <dbl> <date>     <int>
# 1  -121    65 1984-01-11     1
# 2  -120    65 1984-01-11     1
# 3  -121    63 1984-01-11     1
# 4  -120    63 1984-01-11     1
# 5  -121    65 1984-01-12     1
# 6  -120    65 1984-01-12     1
# 7  -121    63 1984-01-12     1
# 8  -120    63 1984-01-12     1
# 9  -121    65 1984-01-13     1
#10  -120    65 1984-01-13     1
# … with 140 more rows

Answer 2

Does this answer your question?

dt %>%
  group_by(x, y) %>%
  mutate(grp = cumsum(c(TRUE, diff(Date) != 1)), .add = TRUE) %>%
  mutate(event = if (n() >= 5)
    cur_group_id()[n() >= 5]
    else    NA)