R &引用；错误：行的标识符重复；在数据帧及其某些子集中，但不在其他子集中

我遇到了一个相当令人费解的tidyr:：spread（）错误。当我试图在完整的数据帧中运行代码（下面的示例）时，我得到了“行的重复标识符”错误

我对（非常大的）数据帧进行子集，以调查并重新运行代码。这一次它起了作用（见下文第1DF dput小节）。然后我用另一个子集（subset2df）再次尝试，再次收到错误消息。老实说，我不知道如何理解这一点，并将非常感谢任何帮助

可复制代码如下：

第1df分节：

structure(list(v1 = structure(c(NA, NA, NA, NA, NA, NA, NA, NA, 
NA, NA, NA, 1L, NA, NA, NA, NA), .Label = "2", class = "factor"), 
    v2 = structure(c(1L, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, 1L, NA, NA), .Label = "2", class = "factor"), 
    v3 = structure(c(NA_integer_, NA_integer_, NA_integer_, NA_integer_, 
    NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_, 
    NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_, 
    NA_integer_, NA_integer_), .Label = character(0), class = "factor"), 
    v4 = structure(c(NA, 1L, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA), .Label = "2", class = "factor"), 
    v5 = structure(c(NA, NA, NA, NA, NA, NA, NA, NA, 1L, NA, 
    NA, NA, NA, NA, NA, NA), .Label = "2", class = "factor"), 
    v6 = structure(c(NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, 1L), .Label = "2", class = "factor"), 
    v7 = structure(c(NA_integer_, NA_integer_, NA_integer_, NA_integer_, 
    NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_, 
    NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_, 
    NA_integer_, NA_integer_), .Label = character(0), class = "factor"), 
    v8 = structure(c(NA_integer_, NA_integer_, NA_integer_, NA_integer_, 
    NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_, 
    NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_, 
    NA_integer_, NA_integer_), .Label = character(0), class = "factor"), 
    v9 = structure(c(NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    1L, NA, NA, NA, NA, NA), .Label = "1", class = "factor"), 
    v10 = structure(c(NA_integer_, NA_integer_, NA_integer_, 
    NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_, 
    NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_, 
    NA_integer_, NA_integer_, NA_integer_), .Label = character(0), class = "factor"), 
    v11 = structure(c(NA, NA, NA, NA, NA, NA, NA, 1L, NA, NA, 
    NA, NA, NA, NA, NA, NA), .Label = "2", class = "factor"), 
    v12 = structure(c(NA, NA, NA, NA, 1L, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA), .Label = "2", class = "factor"), 
    v13 = structure(c(NA, NA, NA, NA, NA, NA, NA, NA, NA, 1L, 
    NA, NA, NA, NA, NA, NA), .Label = "1", class = "factor"), 
    v14 = structure(c(NA, NA, NA, NA, NA, NA, 2L, NA, NA, NA, 
    NA, NA, 1L, NA, NA, NA), .Label = c("1", "2"), class = "factor"), 
    v15 = structure(c(NA, NA, NA, 1L, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA), .Label = "2", class = "factor"), 
    v16 = structure(c(NA, NA, 2L, NA, NA, 2L, NA, NA, NA, NA, 
    NA, NA, NA, NA, 1L, NA), .Label = c("1", "2"), class = "factor"), 
    respondentID = structure(c(7L, 7L, 7L, 5L, 6L, 6L, 4L, 4L, 
    4L, 3L, 3L, 3L, 2L, 2L, 2L, 1L), .Label = c("EO15", "EO17", 
    "EO19", "EO21", "Eo23", "EO23", "EO24"), class = "factor")), .Names = c("v1", 
"v2", "v3", "v4", "v5", "v6", "v7", "v8", "v9", "v10", "v11", 
"v12", "v13", "v14", "v15", "v16", "respondentID"), row.names = c(NA, 
-16L), class = "data.frame")

第2df子类：

structure(list(v2 = structure(c(NA, NA, NA, NA, NA, NA, NA, NA, 
NA, NA, NA, 1L, NA, NA, 1L, NA, 1L, NA, 1L, NA, NA, NA, NA, NA, 
NA, NA, NA, NA, NA, 1L, NA, NA, NA, NA, NA, NA, NA, NA, NA), .Label = "2", class = "factor"), 
    v4 = structure(c(NA, NA, NA, NA, NA, NA, 1L, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA), .Label = "2", class = "factor"), 
    v5 = structure(c(NA, NA, NA, 2L, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, 1L, NA, NA, NA, NA, NA), .Label = c("1", 
    "2"), class = "factor"), v6 = structure(c(NA, 1L, NA, NA, 
    2L, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA), .Label = c("1", "2"), class = "factor"), 
    v9 = structure(c(NA, NA, NA, NA, NA, 1L, NA, NA, NA, 1L, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    1L, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA), .Label = "2", class = "factor"), 
    v11 = structure(c(NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 1L, NA, 
    NA, 1L, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA), .Label = "1", class = "factor"), 
    v12 = structure(c(NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, 1L, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 2L, NA), .Label = c("1", 
    "2"), class = "factor"), v13 = structure(c(NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 1L, NA, NA, NA, 
    NA, NA, 2L, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, 2L, NA, NA, NA), .Label = c("1", "2"), class = "factor"), 
    v14 = structure(c(NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 1L, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA), .Label = "2", class = "factor"), 
    v15 = structure(c(NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    1L, NA, 1L, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, 1L, NA, NA, NA, NA, NA, NA, NA, NA), .Label = "2", class = "factor"), 
    v16 = structure(c(NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, 1L, NA, NA, NA, NA, 1L, NA, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA), .Label = "2", class = "factor"), 
    respondentID = structure(c(21L, 20L, 20L, 19L, 18L, 18L, 
    1L, 1L, 16L, 16L, 16L, 10L, 10L, 17L, 15L, 15L, 15L, 14L, 
    14L, 14L, 13L, 12L, 12L, 11L, 11L, 11L, 8L, 9L, 9L, 6L, 7L, 
    7L, 3L, 2L, 3L, 4L, 4L, 4L, 5L), .Label = c("EO11", "Eo14", 
    "EO14", "EO16", "EO18", "EO26", "EO27", "Eo28", "EO28", "EO3", 
    "Eo30", "EO32", "EO331", "EO35", "EO37", "EO4", "EO41", "EO6", 
    "EO6 ", "EO7", "EO7 "), class = "factor")), .Names = c("v2", 
"v4", "v5", "v6", "v9", "v11", "v12", "v13", "v14", "v15", "v16", 
"respondentID"), row.names = c(NA, -39L), class = "data.frame")

combodf_id（执行代码所需）：

代码：

结果\u df%
聚集（键=组合，值=val，-响应ID）%>%
过滤器（！is.na（val））%>%
左联合（combodf\u id，by=“combo”）%>%
安排（响应ID）%>%
重命名（rose_color1=color1，rose_color2=color2，
郁金香颜色1=颜色3，郁金香颜色2=颜色4）%>%
聚集（颜色、价值、玫瑰色1：郁金香色2）%>%
分离（颜色，分为=c（'flower'，'color'））%>%
排列（颜色、值）%>%
变异（val=if_else（val==1，‘玫瑰’、‘郁金香’））%>%
变异（val=if_else（val==flower，1，0））%>%
选择（响应ID、花、颜色1、颜色2、选项=val）

---

下面@Tung的解决方案与中的非常相似，但两者都不能完全解决问题

在您的

子类t2df中

您已经删除了一些响应列（例如v1、v3、v7等）。。。这可能就是为什么有一堆行/响应者没有任何响应（全部NA）

您的

respondentID

列在某些值中具有空格级别，这些值可能会在以后把事情搞砸（例如“EO7”和“EO7”）

有重复的行，例如

subset2df[c（15,17），]

你所有的专栏都是因素。。。特别是对于带有整数值的响应列，我发现这很奇怪

tidyr

函数

GARGET

和

SPRIDE

将考虑因子的级别，当您有因子变量的子集且并非所有级别都在数据中表示时，这看起来特别奇怪

您可能应该首先解决这些问题，因为它们可能会导致您以后的问题，但是。。。出现错误“行的重复标识符”的原因是，要传递到

spread（color，value）

的数据帧具有重复的行。您可以通过在其前面添加一行来强制执行此操作，但请注意，必须这样做的唯一原因是因为之前的其他问题

subset2df %>%
  as_tibble() %>% 
  mutate_at(vars(v2:v16), as.integer) %>% 
  gather(key = combo, value = val, -respondentID, na.rm = T) %>%
  filter(!is.na(val)) %>%
  left_join(combodf_id, by = "combo") %>%
  arrange(respondentID) %>% 
  rename(rose_color1 = color1, rose_color2 = color2,
         tulip_color1 = color3, tulip_color2 = color4) %>% 
  gather(color, value, rose_color1:tulip_color2) %>% 
  separate(color, into = c('flower', 'color')) %>% 
  distinct() %>% 
  spread(color, value) %>% 
  mutate(val = if_else(val == 1, 'rose', 'tulip')) %>% 
  mutate(val = if_else(val == flower, 1, 0)) %>% 
  select(respondentID, flower, color1, color2, choice = val)

不过，我强烈建议您首先解决上述所有问题，就像这样（请注意，您不需要在链的更深处使用

distinct

命令，因为您已经将该命令应用于原始数据）

这两个选项都可能重复（插入

group\u by_at（vars（-val））%%>%mutate（row\u id=1:n（））%%>%。
result_df <- subset1df %>%
  gather(key = combo, value = val, -respondentID) %>%
  filter(!is.na(val)) %>%
  left_join(combodf_id, by = "combo") %>%
  arrange(respondentID) %>% 
  rename(rose_color1 = color1, rose_color2 = color2,
         tulip_color1 = color3, tulip_color2 = color4) %>% 
  gather(color, value, rose_color1:tulip_color2) %>% 
  separate(color, into = c('flower', 'color')) %>% 
  spread(color, value) %>% 
  mutate(val = if_else(val == 1, 'rose', 'tulip')) %>% 
  mutate(val = if_else(val == flower, 1, 0)) %>% 
  select(respondentID, flower, color1, color2, choice = val)

subset2df %>%
  as_tibble() %>% 
  mutate_at(vars(v2:v16), as.integer) %>% 
  gather(key = combo, value = val, -respondentID, na.rm = T) %>%
  filter(!is.na(val)) %>%
  left_join(combodf_id, by = "combo") %>%
  arrange(respondentID) %>% 
  rename(rose_color1 = color1, rose_color2 = color2,
         tulip_color1 = color3, tulip_color2 = color4) %>% 
  gather(color, value, rose_color1:tulip_color2) %>% 
  separate(color, into = c('flower', 'color')) %>% 
  distinct() %>% 
  spread(color, value) %>% 
  mutate(val = if_else(val == 1, 'rose', 'tulip')) %>% 
  mutate(val = if_else(val == flower, 1, 0)) %>% 
  select(respondentID, flower, color1, color2, choice = val)

subset2df %>%
  as_tibble() %>%  # tibble has better printing methods
  mutate_at(vars(-respondentID), as.integer) %>%  # convert response to numeric
  mutate(respondentID = as.character(respondentID)) %>%  # convert to char
  mutate(respondentID = trimws(respondentID)) %>%  # remove whitespace
  distinct() %>%  # remove duplicate rows
  gather(key = combo, value = val, -respondentID, na.rm = T) %>%
  left_join(combodf_id, by = "combo") %>%
  mutate_at(vars(color1:color4), as.character) %>%  # convert colors to char
  rename(rose_color1 = color1, rose_color2 = color2, 
         tulip_color1 = color3, tulip_color2 = color4) %>% 
  gather(color, value, rose_color1:tulip_color2) %>% 
  separate(color, into = c('flower', 'color')) %>% 
  spread(color, value) %>% 
  mutate(val = if_else(val == 1, 'rose', 'tulip')) %>% 
  mutate(val = if_else(val == flower, 1L, 0L)) %>% 
  select(respondentID, flower, color1, color2, choice = val)