多个 if else 语句对 R 中新列中的现有值进行分类

Question

我已经尝试复制这篇文章中描述的方法（Trying to create a new column using multiple if else statements in R）

我想对患者血液检查的严重程度进行分类。我的目的是为每个患者的血液工作值给一个已经存在的值一个特定的分数（即 0、1、2、3）。之后，我想将这些新值保存到新列中。 截止值是：

if value is >=150000, score = 0
if value is <150000, score = 1
if value is <100000, score = 2
if value is <50000, score = 3
if value is <20000, score = 4

输入是

> dput (platelets_v1)
structure(list(ID = c(13055908, 13059026, 13154920, 13201107, 
13207119, 13207948, 13234892, 13261022, 13082943, 13193903, 13259391, 
13283776, 13262499, 13154288, 13207315, 13269178, 13135316, 13055690, 
13207670, 13220627, 13233898, 13055009, 13044947, 13181075, 13261607, 
13186960, 13240091, 13060589, 13201616, 13260671, 13302375, 13021555, 
13054278, 13062360, 13035346, 13077712, 13128769, 13267480, 13160156, 
13040172, 13160971, 13239318, 12977871, 13090190, 13321288, 13040530, 
13100979, 13124511, 13192142, 13289317, 13315577, 13154966, 13044653, 
13079694, 13128639, 13165362, 13207352, 13049409, 12999835, 13210994, 
13283675, 13223721, 13064865, 13104602, 13036280, 13040507, 12964437, 
13029805, 13029001, 12993036, 13072516, 13060586, 13119819, 13040632
), platelets = c("469.000", "NA", "NA", "243.000", "NA", "NA", 
"NA", "334.000", "522.000", "NA", "NA", "NA", "NA", "312.000", 
"421.000", "NA", "321.000", "NA", "NA", "NA", "298.000", "263.000", 
"109.000", "280.000", "NA", "NA", "430.000", "288.000", "159.000", 
"528.000", "NA", "163.000", "NA", "439.000", "NA", "477.000", 
"NA", "473.000", "NA", "459.000", "183.000", "343.000", "285.000", 
"459.000", "253.000", "NA", "227.000", "NA", "569.000", "NA", 
"NA", "NA", "239.000", "382.000", "270.000", "NA", "362.000", 
"NA", "146.000", "367.000", "NA", "531.000", "NA", "363000", 
"NA", "257000", "158000", "56000", "417", "NA", "171000", "NA", 
"NA", "NA")), row.names = c(NA, -74L), class = c("tbl_df", "tbl", 
"data.frame"))

我尝试了以下方法：

> labels <- c('0', '1', '2','3', '4')
> breaks <- c(500000, 150000, 100000, 50000, 20000)
> teste01 <- platelets_v1 %>% mutate(platelets_v1 = cut(platelets_v1, breaks = breaks, labels = labels, include.lowest = TRUE))

想要的结果：

ID platelets score
13055908 469000 0
13059026 NA NA
13154920 NA NA
13201107 243000 0

等等

任何灯光都将不胜感激。

Answer 1

platelets_v1 %>%
  mutate(
    platelets = suppressWarnings(as.numeric(platelets)),
    bin = 5L - cut(platelets,
                   c(0, 20000, 50000, 100000, 150000, Inf), labels = FALSE)
  ) %>%
  slice(c(1:5, n() - 0:4))
# # A tibble: 10 x 3
#          ID platelets   bin
#       <dbl> <chr>     <int>
#  1 13055908 469.000       4
#  2 13059026 NA           NA
#  3 13154920 NA           NA
#  4 13201107 243.000       4
#  5 13207119 NA           NA
#  6 13040632 NA           NA
#  7 13119819 NA           NA
#  8 13060586 NA           NA
#  9 13072516 171000        0
# 10 12993036 NA           NA

或

platelets_v1 %>%
  mutate(
    platelets = suppressWarnings(as.numeric(platelets)),
    bin = 5L - findInterval(platelets,
                            c(0, 20000, 50000, 100000, 150000, Inf))
  )

但是，如果您想要找到可能不完全对齐的范围（左闭右开）的通用能力，那么

platelets_v1 %>%
  mutate(
    platelets = suppressWarnings(as.numeric(platelets)),
    bin = case_when(
      platelets < 20000 ~ 4, 
      platelets < 50000 ~ 3, 
      platelets < 100000 ~ 2, 
      platelets < 150000 ~ 1, 
      platelets >= 150000 ~ 0)
    )

这里的顺序很重要，因为如果你把它们颠倒过来，一切都会是0 或1（和NA）。此外，您可能会想使用between ...意识到这在两侧都是“封闭的”，因此between(platelets, 20000, 50000) 相当于20000 <= platelets & platelets <= 50000，您的逻辑表明您更喜欢... & platelets < 50000。

此外，人们可能会想用TRUE ~ 0 替换platelets >= 150000 ~ 0，因为假设所有剩余的值都必须属于该类别。由于您的数据包含NA，因此我建议您不要这样做，而对于未满足的条件，更愿意保留NA 的默认值。