使用dplyr的mutate函数生成错误列名问题求助

问题场景 #

尝试用dplyr的mutate()给数据集添加total_boxoffice列，两种写法都得到了列名为Domestic_boxoffice的结果，而非预期的total_boxoffice：

• Option 1（基础写法）：

starwars <- mutate(starwars, total_boxoffice = Domestic_boxoffice + Worldwide_boxoffice, .after=Worldwide_boxoffice)

• Option 2（管道写法）：

starwars %>% mutate(total_boxoffice = Domestic_boxoffice + Worldwide_boxoffice)

完整代码：

homeDir <- getwd()
csvPath <- paste(homeDir, "/starwars.csv", sep = "")
starwars <- read.csv(csvPath)
starwars <- starwars %>% mutate(total_boxoffice = Domestic_boxoffice + Worldwide_boxoffice, .after=Worldwide_boxoffice)
starwars

数据集结构（dput输出）：

structure(list(Release_date = c("Dec 20, 2019", "May 25, 2018", 
"Dec 15, 2017", "Dec 16, 2016", "Dec 18, 2015", "Aug 15, 2008", 
"May 19, 2005", "May 16, 2002", "May 19, 1999", "May 25, 1983", 
"May 21, 1980", "May 25, 1977"), Movie = c("Star Wars: The Rise of Skywalker", 
"Solo: A Star Wars Story", "Star Wars Ep. VIII: The Last Jedi", 
"Rogue One: A Star Wars Story", "Star Wars Ep. VII: The Force Awakens", 
"Star Wars: The Clone Wars", "Star Wars Ep. III: Revenge of the Sith", 
"Star Wars Ep. II: Attack of the Clones", "Star Wars Ep. I: The Phantom Menace", 
"Star Wars Ep. VI: Return of the Jedi", "Star Wars Ep. V: The Empire Strikes Again", 
"Star Wars Ep. IV: A New Hope"), Production_budget = structure(c(275, 
275, 200, 200, 306, 8.5, 115, 115, 115, 32.5, 23, 11), dim = c(12L, 
1L), dimnames = list(NULL, "Production_budget")), Opening_weekend = structure(c(177.383864, 
84.420489, 220.009584, 155.081681, 247.966675, 14.611273, 108.435841, 
80.027814, 64.81097, 23.019618, 4.910483, 1.554475), dim = c(12L, 
1L), dimnames = list(NULL, "Opening_weekend")), Domestic_boxoffice = structure(c(515.202542, 
213.767512, 620.181382, 532.177324, 936.662225, 35.161554, 380.270577, 
310.67674, 474.544677, 309.205079, 291.73896, 460.998007), dim = c(12L, 
1L), dimnames = list(NULL, "Domestic_boxoffice")), Worldwide_boxoffice = structure(c(1072.848487, 
393.151347, 1331.635141, 1055.135598, 2064.615817, 68.695443, 
848.998877, 656.695615, 1027.044677, 475.106177, 549.001242, 
775.398007), dim = c(12L, 1L), dimnames = list(NULL, "Worldwide_boxoffice")), 
    total_boxoffice = structure(c(1588.051029, 606.918859, 1951.816523, 
    1587.312922, 3001.278042, 103.856997, 1229.269454, 967.372355, 
    1501.589354, 784.311256, 840.740202, 1236.396014), dim = c(12L, 
    1L), dimnames = list(NULL, "Domestic_boxoffice")), US_avg_ticket_price_in_USD = c(9.16, 
    9.11, 8.97, 8.65, 8.43, 7.18, 6.41, 5.81, 5.08, 3.15, 2.69, 
    2.23)), row.names = c(NA, -12L), class = "data.frame")

问题原因 #

从dput输出可以看到，Domestic_boxoffice和Worldwide_boxoffice并非普通数值向量，而是单列数据框（带有dim = c(12L, 1L)和dimnames属性）。当两个单列数据框相加时，结果仍为单列数据框，且会保留第一个操作数（Domestic_boxoffice）的列名。

dplyr的mutate()在处理这种“数据框类型”的赋值时，会直接使用该数据框自带的列名，从而覆盖了你指定的total_boxoffice。

解决方案 #

将单列数据框转换为普通向量后再计算，常用方式有三种：

1. 使用[[1]]提取数据框的第一列（向量形式）：

starwars <- starwars %>% 
  mutate(total_boxoffice = Domestic_boxoffice[[1]] + Worldwide_boxoffice[[1]], .after=Worldwide_boxoffice)

2. 使用pull()函数提取向量：

starwars <- starwars %>% 
  mutate(total_boxoffice = pull(Domestic_boxoffice) + pull(Worldwide_boxoffice), .after=Worldwide_boxoffice)

3. 读取数据时避免生成单列数据框：如果是read.csv导致的问题，可检查CSV文件格式，或使用read_csv()替代read.csv()（read_csv()默认会将单列解析为向量）。

内容的提问来源于stack exchange，提问作者Moris