从 Qualtrics 调查中提取数据，但在 R 中提取计数数据和类别时保留行 ID

Question

我有一个数据框，是从 Qualtrics 中进行的一项调查中获得的，其中我们确定了每个钓鱼者收获的鱼的种类和数量（ID = 个体钓鱼者）。我正在尝试整理数据，以便可以使用调查数据构建收获数据框架，但保留调查 ID。

这是我的数据：

df<-structure(list(ID = 1:3, Q7 = c("Other species (please list species name; i.e. Tarpon),Other species (please list species name; i.e. Amberjack)", 
"Red Drum (a.k.a. Redfish or Red),Other species (please list species name; i.e. Tarpon),Other species (please list species name; i.e. Amberjack)", 
"Red Drum (a.k.a. Redfish or Red),Other species (please list species name; i.e. Tarpon)"
), Q7_7_TEXT = c("Tink", "Blue", "Blue"), Q7_8_TEXT = c("Chii", 
"Red", NA), Q8_1 = c(NA, "2", "7"), Q8_7 = c("1", "4", "9"), 
    Q8_8 = c("2", "5", NA)), class = "data.frame", row.names = 3:5)

我使用以下代码提取每个物种的物种和收获数量，以便将信息存储在包含物种和计数的数据框 (

harv

) 中：

library(tidyverse)
library(stringi)
harv<-data.frame(spp=unlist(strsplit(df$Q7, ",")))
oth<-na.omit(data.frame(spp=stri_remove_empty(c(unlist(t(df[,3:4]))))))

idx <- grep('Other', harv$spp)
n <- min(length(idx), nrow(oth))
harv[idx, 'spp'] <- oth

harv.num<-na.omit(as.numeric(stri_remove_empty(c(unlist(t(df[,5:7]))))))

harv$num<-harv.num

harv
                               spp num
1                             Tink   1
2                             Chii   2
3 Red Drum (a.k.a. Redfish or Red)   2
4                             Blue   4
5                              Red   5
6 Red Drum (a.k.a. Redfish or Red)   7
7                             Blue   9

如何从

df

中提取此数据，同时维护原始数据框中的调查 ID，以便

harv

看起来像这样：

                               spp num ID
1                             Tink   1  1
2                             Chii   2  1
3 Red Drum (a.k.a. Redfish or Red)   2  2
4                             Blue   4  2
5                              Red   5  2
6 Red Drum (a.k.a. Redfish or Red)   7  3
7                             Blue   9  3

Answer 1

您可以使用

pivot_longer

和对列名称进行一些初步重命名来实现您想要的结果，如下所示：

library(dplyr, warn = FALSE)
library(tidyr)

df |> 
  rename(Q7_1_TEXT = Q7) |> 
  rename_with(~gsub("_TEXT$", "", .x), ends_with("TEXT")) |> 
  pivot_longer(-ID, names_to = c(".value", "name"), names_sep = "_") |> 
  filter(!is.na(Q8)) |> 
  select(ID, spp = Q7, num = Q8)
#> # A tibble: 7 × 3
#>      ID spp                                                                num  
#>   <int> <chr>                                                              <chr>
#> 1     1 Tink                                                               1    
#> 2     1 Chii                                                               2    
#> 3     2 Red Drum (a.k.a. Redfish or Red),Other species (please list speci… 2    
#> 4     2 Blue                                                               4    
#> 5     2 Red                                                                5    
#> 6     3 Red Drum (a.k.a. Redfish or Red),Other species (please list speci… 7    
#> 7     3 Blue                                                               9