Question

我有与此类似的代码，其中包含用于创建数据框“ full_tb”的管道，该管道失败，因为倒数第二行（产生ID列的突变）正在调用尚未创建的对象（“ full_tb”）

library(random)
library(dplyr)  

set.seed(1)
Codes <- as.vector(randomStrings(n = 10, len = 3, digits = TRUE, upperalpha = FALSE,
         unique = TRUE))

frame1 <- data.frame(
  A = sort(Codes),
  B = sample(x = c("Tree", "Shrub", "Fern"), size = 10, replace = TRUE))
)

frame2 <- data.frame(
  Row_no = sort(sample(x = 1:10)),
  C = sample(x = sample(x = c("Tree", "Shrub", "Fern"), size = 30, replace = TRUE))
)

# Here is where the problem begins

full_tb <- frame1 %>% mutate(Row_no = as.numeric(rownames(frame1))) %>%  
           inner_join(frame2) %>%  
           mutate(ID = as.numeric(rownames(full_tb))) %>%  
           select(ID, A, B, C)

# Joining by = "Row_no"  
# Error in mutate_impl(.data, dots):   
# Evaluation error: object 'full_tb' not found

但是，如果我将管道分成两个块，则运行正常。

full_tb <- frame1 %>% mutate(Row_no = as.numeric(rownames(frame1))) %>%  
           inner_join(frame2)

# Joining by = "Row_no"  

full_tb  <- full_tb %>% mutate(ID = as.numeric(rownames(full_tb))) %>%  
            select(ID, A, B, C)

是否存在一种解决方法，可以将所有内容都管道化为一个块，而不必将第一个代码块分为两部分？

Answer 1

通过在要传递的代码块中添加点作为行名的参数，在连接后为整个数据帧生成ID。因此，无需在行名中指定数据框的名称。

full_tb <- frame1 %>% mutate(Row_no = as.numeric(rownames(frame1))) %>%  
           inner_join(frame2) %>%  
           mutate(ID = as.numeric(rownames(.))) %>%  
           select(ID, A, B, C)

在dplyr中调用尚未创建的对象时出现“评估错误：找不到对象”

1 个答案: