r - 根据级别中的分数 "_"将列拆分为三列

Question

我有一个 data.frame，我想使用 ggplot2 来绘制它

df <- read.table(text = c("
scenario    value 
measured_raw21  55
measured_raw22  60
simulated_raw21 54
measured_raw21_drain    55
measured_raw22_drain    60
measured_adj21  23
simulated_raw22 59
simulated_raw21_drain   54.5
simulated_raw22_drain   60.2
measured_adj21_drain    23
measured_adj22  27
simulated_adj21 22
measured_adj22_drain    27
simulated_adj21_drain   23.4
simulated_adj22 27.5
simulated_adj22_drain   27.2
measured_0.5    40
measured_0.8    55
measured_0.5_drain  40
measured_0.8_drain  55
simulated_0.5   41
simulated_0.8   56
simulated_0.5_drain 39.9
simulated_0.8_drain 55.3"), header =T)

scenario 列包含有关是模拟还是测量、scenarioID(raw21、raw22)以及是否有的信息>排水或nodrain

我想将场景列分成 3 列，如下所示

                scenario       cat  scenario1   drain value
1         measured_raw21  measured      raw21 nodrain  55.0
2         measured_raw22  measured      raw22 nodrain  60.0
3        simulated_raw21 simulated      raw21 nodrain  54.0
4   measured_raw21_drain  measured      raw21   drain  55.0
5   measured_raw22_drain  measured      raw22   drain  60.0
6         measured_adj21  measured      adj21 nodrain  23.0
7        simulated_raw22 simulated      raw22 nodrain  59.0
8  simulated_raw21_drain simulated      raw21   drain  54.5
9  simulated_raw22_drain simulated      raw22   drain  60.2
10  measured_adj21_drain  measured      adj21   drain  23.0
11        measured_adj22  measured      adj22 nodrain  27.0
12       simulated_adj21 simulated      adj21 nodrain  22.0
13  measured_adj22_drain   meaured      adj22   drain  27.0
14 simulated_adj21_drain simulated      adj21   drain  23.4
15       simulated_adj22 simulated      adj22 nodrain  27.5
16 simulated_adj22_drain simulated      adj22   drain  27.2
17           meaured_0.5  measured        0.5 nodrain  40.0
18           meaured_0.8  measured        0.8 nodrain  55.0
19     meaured_0.5_drain  measured        0.5   drain  40.0
20     meaured_0.8_drain  measured        0.8   drain  55.0
21         simulated_0.5 simulated        0.5 nodrain  41.0
22         simulated_0.8 simulated        0.8 nodrain  56.0
23   simulated_0.5_drain simulated        0.5   drain  39.9
24   simulated_0.8_drain simulated        0.8   drain  55.3

我是这样做的

df$cat <- c("measured", "measured", "simulated","measured", "measured", "measured","simulated", 
            "simulated", "simulated", "measured", "measured", "simulated",  "measured", "simulated", "simulated", "simulated", 
            "measured", "measured", "measured", "measured", "simulated",  "simulated", "simulated", "simulated")     

df$scenario1 <- c("raw21","raw22","raw21","raw21","raw22","adj21","raw22",'raw21',"raw22","adj21","adj22",
"adj21","adj22",'adj21',"adj22", 'adj22', "0.5", "0.8", "0.5", "0.8", "0.5", "0.8", "0.5", "0.8")

df$drain <- c("nodrain", "nodrain", "nodrain", "drain", "drain", "nodrain", "nodrain" ,"drain", 
"drain", "drain", "nodrain", "nodrain", "drain", "drain", "nodrain", "drain", "nodrain", "nodrain", 
"drain", "drain", "nodrain", "nodrain", "drain", "drain")

这是我想要的最终情节

library(tidyr)
library(dplyr)
library(ggplot2)
df_fin <-  df %>% 
  select(scenario1, drain, cat, value) %>% 
  spread(cat, value)


ggplot(df_fin, aes(measured, simulated, col = scenario1))+
  geom_point()+
  facet_wrap(~drain)

我手动将 df 中的 scenario 列拆分为 3 列。然而，在可重现的示例中只有 24 个观察结果，但在我的原始 data.frame 中，我有很多观察结果，这将花费时间按照我上面的方式分割场景列。

如果您能提供将场景列分成 3 列的有效方法，我将不胜感激？

Answer 1

您可以使用单独的:

library(tidyr)
df %>% 
    separate(scenario, into = c("cat", "scenario1", "drain"), sep = "_", remove = FALSE) %>% 
    replace_na(list(drain = "nodrain"))        
    # if there is no drain, the cell will be <NA>, repalce it with nodrain

#                scenario       cat scenario1   drain value
#1         measured_raw21  measured     raw21 nodrain  55.0
#2         measured_raw22  measured     raw22 nodrain  60.0
#3        simulated_raw21 simulated     raw21 nodrain  54.0
#4   measured_raw21_drain  measured     raw21   drain  55.0
#5   measured_raw22_drain  measured     raw22   drain  60.0
#6         measured_adj21  measured     adj21 nodrain  23.0
#7        simulated_raw22 simulated     raw22 nodrain  59.0
#8  simulated_raw21_drain simulated     raw21   drain  54.5
#9  simulated_raw22_drain simulated     raw22   drain  60.2
#10  measured_adj21_drain  measured     adj21   drain  23.0
#11        measured_adj22  measured     adj22 nodrain  27.0
#12       simulated_adj21 simulated     adj21 nodrain  22.0
#13  measured_adj22_drain  measured     adj22   drain  27.0
# ...