一行上的R-最小值和最大值，定义为许多列

Question

I have 4 variables, each in triplicates, and I need to find the average for each triplicate, in each row, and on top of that find minimum and maximum for each 5th row, in order to plot error bars. Here is an example with two variables:

  x  a1  a2  a3  b1  b2  b3
220   2   7  71  28  53  31
221   5  13  85  33  51  34

So I need the average of a1-a3 when x=220, and so on, and then I need minimum and maximum for row 1, row 6 and so on. I assume I need to do this in order to get the data I need to make error bars, but I am new to R, so there might be a better way.

Answer 1

May be this helps

library(dplyr)
library(tidyr)
add_rownames(df1) %>% 
        gather(Var, Val, -x, -rowname) %>% 
        extract(Var, into='Var', '([^0-9]*).*') %>%
        group_by(rowname, x, Var) %>% 
        summarise(Mean=mean(Val), Min=min(Val), Max=max(Val))

Or using base R

 res <-  do.call(rbind, 
             lapply(split(names(df1)[-1],
                sub('\\d+$', '', names(df1)[-1])), function(x) {
              x1 <- df1[x]
              data.frame(x=df1[1],Mean=rowMeans(x1), Max=do.call(pmax, x1),
                 Min=do.call(pmin, x1))}))

data

df1 <- structure(list(x = 220:221, a1 = c(2L, 5L), a2 = c(7L, 13L), 
a3 = c(71L, 85L), b1 = c(28L, 33L), b2 = c(53L, 51L), b3 = c(31L, 
34L)), .Names = c("x", "a1", "a2", "a3", "b1", "b2", "b3"
 ), class = "data.frame", row.names = c(NA, -2L))