r - 从宽到长，列名不一致

Question

我有以下数据框，您可以在此处获取 CSV 格式的数据:http://www.sharecsv.com/s/c4d94fd3a7dc43e3ec249bff373a0082/data.csv

我的昵称如下。正如您所看到的，名称不是连续的(例如，缺少 Q1、没有 Q3 或 Q4 等等)。我需要保持这种状态。检查某些案例是否有额外的不相关的“.1”。

  [1] "Q2_1"     "Q2_2"     "Q5_1"     "Q5_2"     "Q6_1"     "Q6_2"     "Q9_1"     "Q9_2"     "Q11_1"    "Q11_2"    "Q8_1"     "Q8_2"     "Q14_1"   
 [14] "Q14_2"    "Q16_1"    "Q16_2"    "Q10_1"    "Q10_2"    "Q11_1.1"  "Q11_2.1"  "Q19_1"    "Q19_2"    "Q20_1"    "Q20_2"    "Q21_1"    "Q21_2"   
 [27] "Q15_1"    "Q15_2"    "Q23_1"    "Q23_2"    "Q24_1"    "Q24_2"    "Q25_1"    "Q25_2"    "Q26_1"    "Q26_2"    "Q20_1.1"  "Q20_2.1"  "Q21_1.1" 
 [40] "Q21_2.1"  "Q29_1"    "Q29_2"    "Q30_1"    "Q30_2"    "Q35_1"    "Q35_2"    "Q36_1"    "Q36_2"    "Q37_1"    "Q37_2"    "Q38_1"    "Q38_2"   
 [53] "Q39_1"    "Q39_2"    "Q41_1"    "Q41_2"    "Q30_1.1"  "Q30_2.1"  "Q43_1"    "Q43_2"    "Q44_1"    "Q44_2"    "Q45_1"    "Q45_2"    "Q47_1"   
 [66] "Q47_2"    "Q48_1"    "Q48_2"    "Q36_1.1"  "Q36_2.1"  "Q37_1.1"  "Q37_2.1"  "Q51_1"    "Q51_2"    "Q52_1"    "Q52_2"    "Q53_1"    "Q53_2"   
 [79] "Q41_1.1"  "Q41_2.1"  "Q42_1"    "Q42_2"    "Q56_1"    "Q56_2"    "Q57_1"    "Q57_2"    "Q58_1"    "Q58_2"    "Q59_1"    "Q59_2"    "Q60_1"   
 [92] "Q60_2"    "Q61_1"    "Q61_2"    "Q62_1"    "Q62_2"    "Q63_1"    "Q63_2"    "Q64_1"    "Q64_2"    "Q65_1"    "Q65_2"    "Q53_1.1"  "Q53_2.1" 
[105] "Q54_1"    "Q54_2"    "Q68_1"    "Q68_2"    "Q75_1"    "Q75_2"    "Q57_1.1"  "Q57_2.1"  "Q58_1.1"  "Q58_2.1"  "Q59_1.1"  "Q59_2.1"  "Q60_1.1" 
[118] "Q60_2.1"  "Q61_1.1"  "Q61_2.1"  "Q81_1"    "Q81_2"    "Q82_1"    "Q82_2"    "Q83_1"    "Q83_2"    "Q87_1"    "Q87_2"    "Q88_1"    "Q88_2"   
[131] "Q89_1"    "Q89_2"    "Q90_1"    "Q90_2"    "Q91_1"    "Q91_2"    "Q92_1"    "Q92_2"    "Q93_1"    "Q93_2"    "Q94_1"    "Q94_2"    "Q95_1"   
[144] "Q95_2"    "Q74_1"    "Q74_2"    "Q75_1.1"  "Q75_2.1"  "Q76_1"    "Q76_2"    "Q77_1"    "Q77_2"    "Q100_1"   "Q100_2"   "Q101_1"   "Q101_2"  
[157] "Q102_1"   "Q102_2"   "Q103_1"   "Q103_2"   "Q104_1"   "Q104_2"   "Q105_1"   "Q105_2"   "Q106_1"   "Q106_2"   "Q107_1"   "Q107_2"   "Q108_1"  
[170] "Q108_2"   "Q113_1"   "Q113_2"   "Q114_1"   "Q114_2"   "Q117_1"   "Q117_2"   "Q96_1"    "Q96_2"    "Q97_1"    "Q97_2"    "Q98_1"    "Q98_2"   
[183] "Q121_1"   "Q121_2"   "Q103_1.1" "Q103_2.1" "Q104_1.1" "Q104_2.1" "Q127_1"   "Q127_2"   "Q128_1"   "Q128_2"   "Q129_1"   "Q129_2"  

问题:我需要将其从 WIDE 转换为 LONG，以获得如下内容:

QUESTION CASE VALUE
Q2        1     5
Q2        2     5
Q5        1     1
Q5        2     2

我尝试按如下方式 reshape ，但不断收到不同的错误，而且我也不确定是否正确分割了它:

test <- reshape(data, sep = "_", times = c(1, 2), direction = "long", varying = colnames(data))

Answer 1

您可以使用 pivot_longer 形式 tidyr，以“_”分隔:

library(tidyr)

pivot_longer(data, cols=everything(), names_sep="_",
             names_to=c("Question","Case"))

# A tibble: 194 x 3
   Question Case  value
   <chr>    <chr> <int>
 1 Q2       1         5
 2 Q2       2         5
 3 Q5       1         1
 4 Q5       2         2
 5 Q6       1         4
 6 Q6       2         4
 7 Q9       1         4
 8 Q9       2         4
 9 Q11      1         5
10 Q11      2         3
# ... with 184 more rows

---

尝试使用reshape 函数会因宽格式中的变量名称不一致而导致错误。例如，正如您提到的，某些名称有一个附加的“.1”，您说这是无关紧要的。 tidyr 包同意这一点，因为它只是收集分隔符之后的所有内容，并将找到的所有内容放入“Case”变量(names_to 参数中的第二项)。 reshape 函数更加严格。也就是说，附加的“.1”不是不相关(根据函数)，因为函数将尝试猜测名称中包含的值(“_”之后)，并查看长度不相等，并且失败并显示'variing'参数必须具有相同的长度。

下面给出了使用 sep="" 的替代方案...