r - 为什么 group_by() 会影响范围外的数据帧？

Question

如果我将分组数据框传递给函数，然后更改分组变量的名称，原始数据框的分组将更改为新名称。当函数返回时(我不返回更改后的数据框)，原始数据框的名称不变，但分组更改为不存在的名称。

# test scoping of group_by() which appears to change groups
library(dplyr)

muck_up_group<-function(mydf){
  mydf<-mydf %>% rename(UhOh=Species)
}

dont_muck_up_group<-function(mydf){
  mydf<-mydf %>% ungroup()
  mydf<-mydf %>% rename(UhOh=Species)
}

data("iris")
iris<-as_tibble(iris) %>% group_by(Species)
iris
# A tibble: 150 x 5
# Groups:   Species [3]
#   Sepal.Length Sepal.Width Petal.Length Petal.Width Species
#          <dbl>       <dbl>        <dbl>       <dbl>  <fctr>
# 1          5.1         3.5          1.4         0.2  setosa

muck_up_group(iris) # original grouping changed to column name that doesn't exist
iris
# A tibble: 150 x 5
# Groups:   UhOh [3]
#   Sepal.Length Sepal.Width Petal.Length Petal.Width Species
#          <dbl>       <dbl>        <dbl>       <dbl>  <fctr>
# 1          5.1         3.5          1.4         0.2  setosa

#restore original state
iris<-as_tibble(iris) %>% group_by(Species)
dont_muck_up_group(iris) # original grouping preserved
iris
# A tibble: 150 x 5
# Groups:   Species [3]
#   Sepal.Length Sepal.Width Petal.Length Petal.Width Species
#          <dbl>       <dbl>        <dbl>       <dbl>  <fctr>
# 1          5.1         3.5          1.4         0.2  setosa

我可以理解为什么更改分组变量的名称可能是不好的做法，但这是允许的。这似乎是当内容通过值传递时变量的属性通过引用传递的示例(正如我们理解的 R 通常所做的那样)。

> sessionInfo()
R version 3.4.0 (2017-04-21)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 7 x64 (build 7601) Service Pack 1

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] graphics  grDevices utils     datasets  stats     methods   base     

other attached packages:
 [1] lubridate_1.6.0               bindrcpp_0.2                  mFilter_0.1-3                
 [4] ggrepel_0.6.5                 reshape2_1.4.2                scales_0.4.1                 
 [7] purrr_0.2.3                   readr_1.1.1                   tidyr_0.7.0                  
[10] tibble_1.3.4                  tidyverse_1.1.1               knitr_1.17                   
[13] Rblpapi_0.3.6                 stringr_1.2.0                 rvest_0.3.2                  
[16] xml2_1.1.1                    devtools_1.13.3               dplyr_0.7.2                  
[19] plyr_1.8.4                    ggplot2_2.2.1                 PerformanceAnalytics_1.4.3541
[22] xts_0.10-0                    zoo_1.8-0                    

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.12       lattice_0.20-35    assertthat_0.2.0   rprojroot_1.2      digest_0.6.12     
 [6] psych_1.7.5        R6_2.2.2           cellranger_1.1.0   backports_1.1.0    evaluate_0.10.1   
[11] httr_1.3.1         highr_0.6          rlang_0.1.2        curl_2.8.1         lazyeval_0.2.0    
[16] readxl_1.0.0       TTR_0.23-2         tidyquant_0.5.3    rmarkdown_1.6      labeling_0.3      
[21] foreign_0.8-67     munsell_0.4.3      broom_0.4.2        compiler_3.4.0     modelr_0.1.1      
[26] pkgconfig_2.0.1    base64enc_0.1-3    mnormt_1.5-5       htmltools_0.3.6    tidyselect_0.1.1  
[31] withr_2.0.0        Quandl_2.8.0       grid_3.4.0         nlme_3.1-131       jsonlite_1.5      
[36] gtable_0.2.0       magrittr_1.5       quantmod_0.4-10    stringi_1.1.5      RColorBrewer_1.1-2
[41] tools_3.4.0        forcats_0.2.0      glue_1.1.1         hms_0.3            rsconnect_0.8.5   
[46] parallel_3.4.0     yaml_2.1.14        colorspace_1.3-2   memoise_1.1.0      bindr_0.1         
[51] haven_1.1.0       
>

漏洞？谢谢。

Answer 1

请参阅上面@aosmith 的评论。 Dplyr 关闭问题。