Question

我必须在一个数据帧中添加一个新行total，在其中尝试添加该行的值，其值类似于Mazda。下面是我正在使用的df。

df <- data.frame(month = c("mazda 3", "mazda cx5", "mazda 6","mazda miata","honda civic","honda accord"),
             april = c(.1,.2,.3,.3,.4,.5),
             may = c(.3,.4,.5,.2,.1,.5),
             june = c(.2,.1,.5,.1,.2,.3))


d2<- df %>% mutate(total == (rowsum(df[-1], df[rownames(month) %like% "Mazda"])))

输出应为：

df_out <- data.frame(month = c("mazda 3", "mazda cx5", "mazda 6","mazda miata","honda civic","honda accord","total_mazda"),
                 april = c(.1,.2,.3,.3,.4,.5,.8),
                 may = c(.3,.4,.5,.2,.1,.5,1.4),
                 june = c(.2,.1,.5,.1,.2,.3,.9))

Answer 1

我们可以获取sum中数字列的summarise_at，同时根据“ month”中的“ mazda”子字符串对值进行子集设置，创建“ month”列并与原始数据集绑定

library(tidyverse)
df %>% 
  summarise_at(2:4, funs(sum(.[str_detect(month, 'mazda')]))) %>% 
  mutate(month = 'Total') %>% 
  bind_rows(df, .)

Answer 2

尝试使用apply(df, 2, ...)在第一列中使用grepl遮罩mazda的掩码来循环浏览这些列。

我使用了一些bind_(rows|cols)技巧来获取正确格式的数据帧。

library(dplyr)

df <- data_frame(month = c("mazda 3", "mazda cx5", "mazda 6","mazda miata","honda civic","honda accord"),
             april = c(.1,.2,.3,.3,.4,.5),
             may = c(.3,.4,.5,.2,.1,.5),
             june = c(.2,.1,.5,.1,.2,.3))

df_out <- bind_rows(
  df %>% as_data_frame(),
  data_frame(month = "total_mazda") %>%
    bind_cols(
      apply(df[, 2:ncol(df)],
        2,
        function(x, y = grepl(".*(m|M)azda.*", df[[1]])) sum(x[y])
      ) %>%
        as.list() %>%
        as_data_frame()))

将选定的行汇总到R中的新行

2 个答案: