如何將行和列都添加到R中的for回圈中？

我正在嘗試在 R 中創建一個空引導程式。資料集有四個物種和相關的樣本值。我想計算mean一個特殊的poisson函式，一次從與每個物種相關的分布中采樣 5 個值。這是我對代碼的嘗試。我采樣了 5 個樣本，并嘗試使用 summarise 從這五個值中計算均值和泊松函式。我收到一條錯誤訊息，! New rows can't add columns.顯示有關如何更正此問題以獲得所需輸出的任何建議（附加）。

set.seed(111)
library(truncnorm)
sample <- rtruncnorm(n = 1440,a = 0,b = 10,mean = 5,sd = 2)
sp <- rep(c("A","B","C","D"), each = 360)

df <- data.frame(sample, sp)

output <- tibble(mean.set = numeric(), 
                 poisson.set = numeric(), 
                 sp = character(),
                 set = numeric()

set.seed(42)                    
for(i in 1:1440){
  samp1 <- df %>% filter(sp == 'A') %>% sample_n(5, replace = TRUE) %>% summarise(mean.set = mean(sample, na.rm=TRUE), possion.set = ((var(sample, na.rm=TRUE)/ mean(sample, na.rm=TRUE)^2) - (1/mean(sample, na.rm=TRUE)))) %>% mutate(set = i)
  samp2 <- df %>% filter(sp == 'B') %>% sample_n(5, replace = TRUE) %>% summarise(mean.set = mean(sample, na.rm=TRUE), possion.set = ((var(sample, na.rm=TRUE)/ mean(sample, na.rm=TRUE)^2) - (1/mean(sample, na.rm=TRUE))))%>% mutate(set = i)
  samp3 <- df %>% filter(sp == 'C') %>% sample_n(5, replace = TRUE) %>% summarise(mean.set = mean(sample, na.rm=TRUE), possion.set = ((var(sample, na.rm=TRUE)/ mean(sample, na.rm=TRUE)^2) - (1/mean(sample, na.rm=TRUE))))%>% mutate(set = i)
  samp4 <- df %>% filter(sp == 'D') %>% sample_n(5, replace = TRUE) %>% summarise(mean.set = mean(sample, na.rm=TRUE), possion.set = ((var(sample, na.rm=TRUE)/ mean(sample, na.rm=TRUE)^2) - (1/mean(sample, na.rm=TRUE))))%>% mutate(set = i)
  output %>% add_row(bind_rows(samp1, samp2, samp3, samp4))  -> output
}


Error:
! New rows can't add columns.
? Can't find columns `possion.set` and `set` in `.data`.
Run `rlang::last_error()` to see where the error occurred

#Expected output
 set mean.set poisson.set sp
  1    5          2        A
  2    4          9        A
  ....
  48   12          0        A 
  1    5          2        B
  2    4          9        B
  ....
  48   22          0        B  
   .....

uj5u.com熱心網友回復：

錯誤問題是基于“輸出”中列名的差異，即poisson.setfor 回圈中的列名possion.set。通過糾正其中任何一個，它可以作業（此外，sp未在summarise結果中創建NA）

output <- tibble(mean.set = numeric(), 
                   possion.set = numeric(), 
                   sp = character(),
                   set = numeric()
  
  )
> set.seed(42)                    
> for(i in 1:1440){
    samp1 <- df %>% filter(sp == 'A') %>% sample_n(5, replace = TRUE) %>% summarise(sp = "A", mean.set = mean(sample, na.rm=TRUE), possion.set = ((var(sample, na.rm=TRUE)/ mean(sample, na.rm=TRUE)^2) - (1/mean(sample, na.rm=TRUE)))) %>% mutate(set = i)
    samp2 <- df %>% filter(sp == 'B') %>% sample_n(5, replace = TRUE) %>% summarise(sp = "B", mean.set = mean(sample, na.rm=TRUE), possion.set = ((var(sample, na.rm=TRUE)/ mean(sample, na.rm=TRUE)^2) - (1/mean(sample, na.rm=TRUE))))%>% mutate(set = i)
    samp3 <- df %>% filter(sp == 'C') %>% sample_n(5, replace = TRUE) %>% summarise(sp = "C", mean.set = mean(sample, na.rm=TRUE), possion.set = ((var(sample, na.rm=TRUE)/ mean(sample, na.rm=TRUE)^2) - (1/mean(sample, na.rm=TRUE))))%>% mutate(set = i)
    samp4 <- df %>% filter(sp == 'D') %>% sample_n(5, replace = TRUE) %>% summarise(sp = "D", mean.set = mean(sample, na.rm=TRUE), possion.set = ((var(sample, na.rm=TRUE)/ mean(sample, na.rm=TRUE)^2) - (1/mean(sample, na.rm=TRUE))))%>% mutate(set = i)
    output %>% add_row(bind_rows(samp1, B = samp2,samp3, samp4))  -> output
  }

此外，我們也可以在沒有for回圈的情況下執行此操作，即group_by“sp”列并replicate/rerun用于多次迭代

library(dplyr)
library(purrr)
output2 <- rerun(1440, df %>%
           group_by(sp) %>%
           slice_sample(n = 5, replace = TRUE) %>% 
           summarise(mean.set = mean(sample, na.rm=TRUE), 
           possion.set = ((var(sample, na.rm=TRUE)/ 
     mean(sample, na.rm=TRUE)^2) - (1/mean(sample, na.rm=TRUE))))) %>% 
    bind_rows(.id = 'set')

標籤：r for循环 dplyr 引导程序 4

上一篇：沒有特定模式的檔案夾中檔案的Bash回圈

下一篇：Python，在哪里添加第二個回圈