是否替换列出的值？在数据集中

Question

我正在尝试替换乳腺癌数据集中列出的Node_caps的一些值。根据最常见的答案选择yes或no的值。

我已经找出了最常见的答案，我想用来替换这些值，但似乎无法真正让这些值的替换起作用。以下是我尝试过的两个版本。

traindata$Node_caps[traindata$Node_caps == '?' && traindata$Inv_nodes == '0-2'] <- 'no'

和

for (i in 1:nrow(traindata)) {
  if (is.nan(traindata[i,6])) {
    if (traindata$Inv_nodes[i] == "0-2") {
      traindata$Node_caps[i] = "no"
        } else if (traindata$Inv_nodes[i] == "3-5") {
      traindata$Node_caps[i] = "yes"
    } else if (traindata$Inv_nodes[i] == "9-11") {
      traindata$Node_caps[i] = "yes"
    }else 
  }
}

> dput(head(traindata))
structure(list(Class = c("no-recurrence-events", "no-recurrence-events", 
"no-recurrence-events", "no-recurrence-events", "no-recurrence-events", 
"no-recurrence-events"), Age = c("30-39", "40-49", "40-49", "60-69", 
"40-49", "60-69"), Menopause = c("premeno", "premeno", "premeno", 
"ge40", "premeno", "ge40"), Tumor_size = c("30-34", "20-24", 
"20-24", "15-19", "0-4", "15-19"), Inv_nodes = c("0-2", "0-2", 
"0-2", "0-2", "0-2", "0-2"), Node_caps = c("no", "no", "no", 
"no", "no", "no"), Deg_malig = c(3L, 2L, 2L, 2L, 2L, 2L), Breast = c("left", 
"right", "left", "right", "right", "left"), Irradiate = c("no", 
"no", "no", "no", "no", "no")), row.names = c(NA, 6L), class = "data.frame")

> head(traindata)
                 Class   Age Menopause Tumor_size Inv_nodes Node_caps Deg_malig Breast Irradiate
1 no-recurrence-events 30-39   premeno      30-34       0-2        no         3   left        no
2 no-recurrence-events 40-49   premeno      20-24       0-2        no         2  right        no
3 no-recurrence-events 40-49   premeno      20-24       0-2        no         2   left        no
4 no-recurrence-events 60-69      ge40      15-19       0-2        no         2  right        no
5 no-recurrence-events 40-49   premeno        0-4       0-2        no         2  right        no
6 no-recurrence-events 60-69      ge40      15-19       0-2        no         2   left        no
> tail(traindata)
                Class   Age Menopause Tumor_size Inv_nodes Node_caps Deg_malig Breast Irradiate
281 recurrence-events 50-59      ge40      40-44       6-8       yes         3   left       yes
282 recurrence-events 30-39   premeno      30-34       0-2        no         2   left        no
283 recurrence-events 30-39   premeno      20-24       0-2        no         3   left       yes
284 recurrence-events 60-69      ge40      20-24       0-2        no         1  right        no
285 recurrence-events 40-49      ge40      30-34       3-5        no         3   left        no
286 recurrence-events 50-59      ge40      30-34       3-5        no         3   left        no
> traindata[c(traindata$Node_caps == '?'),]
                   Class   Age Menopause Tumor_size Inv_nodes Node_caps Deg_malig Breast
146 no-recurrence-events 40-49   premeno      25-29       0-2         ?         2   left
164 no-recurrence-events 60-69      ge40      25-29       3-5         ?         1  right
165 no-recurrence-events 60-69      ge40      25-29       3-5         ?         1  right
184 no-recurrence-events 50-59      ge40      30-34      9-11         ?         3   left
185 no-recurrence-events 50-59      ge40      30-34      9-11         ?         3   left
234    recurrence-events 70-79      ge40      15-19      9-11         ?         1   left
264    recurrence-events 50-59      lt40      20-24       0-2         ?         1   left
265    recurrence-events 50-59      lt40      20-24       0-2         ?         1   left
    Irradiate
146       yes
164       yes
165       yes
184       yes
185       yes
234       yes
264        no
265        no

我只想替换其中的？而不是其他任何人。下面的表格可能会让它更清晰。

> table(traindata$Inv_nodes,traindata$Node_caps)
       
          ?  no yes
  0-2     3 201   9
  12-14   0   1   2
  15-17   0   1   5
  24-26   0   0   1
  3-5     2  15  19
  6-8     0   3  14
  9-11    3   1   6

我需要一些帮助来找出我做错了什么

Answer 1

已更新

很抱歉，我没有意识到我的代码中有一个错误。我对您的样本数据做了一些细微的修改，以便我们可以应用您的条件。

library(dplyr)
library(janitor)

# First we extract the most common answers for every categroy

df %>% 
  tabyl(Inv_nodes, Node_caps) %>%
  select(!`?`) %>%
  rowwise() %>%
  mutate(most_common = names(cur_data())[which.max(c_across(no:yes)) + 1]) -> tbl1


# A tibble: 3 x 4
# Rowwise: 
  Inv_nodes    no   yes most_common
  <chr>     <dbl> <dbl> <chr>      
1 0-2           2     0 no         
2 3-5           0     1 yes        
3 9-11          0     0 no  

# Then we replace the ?s with the most common answers

df %>% 
  rowwise() %>%
  mutate(Node_caps = ifelse(Node_caps == "?", 
                            tbl1$most_common[tbl1$Inv_nodes == Inv_nodes], 
                            Node_caps)) %>%
  select(Inv_nodes, Node_caps)


# A tibble: 6 x 2
# Rowwise: 
  Inv_nodes Node_caps
  <chr>     <chr>    
1 0-2       no       
2 9-11      no       
3 0-2       no       
4 3-5       yes      
5 3-5       yes      
6 0-2       no 

Data

structure(list(Class = c("no-recurrence-events", "no-recurrence-events", 
"no-recurrence-events", "no-recurrence-events", "no-recurrence-events", 
"no-recurrence-events"), Age = c("30-39", "40-49", "40-49", "60-69", 
"40-49", "60-69"), Menopause = c("premeno", "premeno", "premeno", 
"ge40", "premeno", "ge40"), Tumor_size = c("30-34", "20-24", 
"20-24", "15-19", "0-4", "15-19"), Inv_nodes = c("0-2", "9-11", 
"0-2", "3-5", "3-5", "0-2"), Node_caps = c("?", "?", "no", "?", 
"yes", "no"), Deg_malig = c(3L, 2L, 2L, 2L, 2L, 2L), Breast = c("left", 
"right", "left", "right", "right", "left"), Irradiate = c("no", 
"no", "no", "no", "no", "no")), row.names = c(NA, 6L), class = "data.frame")