英文:
Create new column based of subset of strings in another column in R?
问题
我正在尝试在我的数据框中创建一个新列，该列基于另一个字符串子集。以下是您的数据框和要求：
- 如果字符串中包含"Acoustic"、"radio"或"PIT"，新列的值应为"receiver based"。
- 如果字符串中同时包含"acoustic"和"satellite"，新列的值应为"Both"。
- 对于其他情况，新列的值应为"non receiver based"。
您可以使用以下代码实现这些要求：
df$Type <- ifelse(grepl("acoustic", df$Combination, ignore.case = TRUE) & grepl("satellite", df$Combination, ignore.case = TRUE), "Both",
ifelse(grepl("Acoustic", df$Combination), "Receiver Based",
ifelse(grepl("radio|PIT", df$Combination), "Receiver Based", "Non Receiver Based")))
这段代码首先检查是否同时包含"acoustic"和"satellite"，如果是，则设置为"Both"。接下来，它检查是否包含"Acoustic"，如果是，则设置为"Receiver Based"。最后，它检查是否包含"radio"或"PIT"，如果是，则设置为"Receiver Based"，否则设置为"Non Receiver Based"。
这样，新列"Type"将根据您的要求进行分类。
英文:
I am trying to create new a column in my dataframe, based off a subset of strings in another
This is my dataframe
df =structure(list(Combination = c("BRUV_Acoustic_Satellite", "BRUV_Acoustic_Satellite",
"BRUV_Acoustic_Satellite", "BRUV_Acoustic_Satellite", "BRUV_Acoustic_Satellite",
"BRUV_Acoustic_Satellite", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Controlled_Acoustic", "Controlled_Acoustic",
"Controlled_Acoustic", "Controlled_Acoustic", "Controlled_Acoustic",
"Controlled_Acoustic", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Stationary_Radio", "Stationary_Radio",
"Stationary_Radio", "Animalborne_Satellite_Archival", "Animalborne_Satellite_Archival",
"Animalborne_Satellite_Archival", "Animalborne_Satellite_Archival",
"Stationary_Acoustic", "Stationary_Acoustic", "Stationary_Acoustic",
"Stationary_Acoustic", "Stationary_Acoustic", "Stationary_Acoustic",
"BRUV_Acoustic_Satellite", "BRUV_Acoustic_Satellite", "BRUV_Acoustic_Satellite",
"Stationary_Archival", "Stationary_Archival", "Stationary_Archival",
"Stationary_Archival", "Stationary_Acoustic_Radio_PIT", "Stationary_Acoustic_Radio_PIT",
"Stationary_Acoustic_Radio_PIT", "Controlled_Acoustic", "Controlled_Acoustic",
"Stationary_PIT", "Stationary_PIT", "Stationary_Acousitc_PIT",
"Stationary_Acousitc_PIT", "Stationary_Acousitc_PIT", "BRUV_Acoustic",
"BRUV_Acoustic", "BRUV_Acoustic", "Stationary_Acoustic", "Stationary_Acoustic",
"Stationary_Acoustic", "Stationary_Acoustic", "Stationary_Acoustic",
"Stationary_Acoustic", "Stationary_Acoustic", "Stationary_Acoustic",
"Stationary_Acoustic", "Stationary_Acoustic", "Stationary_Archival",
"Stationary_Archival", "Stationary_Archival", "Stationary_Archival",
"Stationary_Satellite", "Controlled_Acoustic", "Controlled_Acoustic",
"Controlled_Acoustic", "Controlled_Acoustic", "BRUV_Acoustic",
"BRUV_Acoustic", "BRUV_Acoustic", "Animalborne_Satellite", "Animalborne_Satellite",
"Stationary_Archival", "Stationary_Archival", "Stationary_Archival",
"Stationary_Radio_PIT", "Stationary_Radio_PIT", "Controlled_Acoustic",
"Controlled_Acoustic", "Controlled_Acoustic", "Controlled_Acoustic",
"Controlled_Satellite", "Controlled_Satellite", "Controlled_Satellite",
"Controlled_Satellite", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Stationary_Acoustic", "Stationary_Acoustic", "Stationary_Acoustic",
"Animalborne_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival_PIT", "Animalborne_Archival_PIT", "Animalborne_Archival_PIT",
"Animalborne_Acoustic_Archival", "Animalborne_Acoustic_Archival",
"Animalborne_Acoustic_Archival", "Stationary_Acoustic", "Stationary_Acoustic",
"Stationary_Acoustic", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Animalborne_Archival", "Stationary_Acoustic_Archival",
"Stationary_Acoustic_Archival", "Stationary_Acoustic_Archival",
"Stationary_Acoustic", "Stationary_Acoustic", "Stationary_Acoustic",
"Stationary_Acoustic", "Stationary_Acoustic", "Stationary_Acoustic",
"Stationary_Acoustic", "Stationary_Acoustic", "Stationary_Acoustic",
"Stationary_Acoustic", "Stationary_Acoustic", "Animalborne_Archival",
"Animalborne_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Stationary_Acoustic_Archival", "Stationary_Acoustic_Archival",
"Stationary_Acoustic_Archival", "Stationary_Acoustic_Archival",
"Animalborne_Acoustic", "Animalborne_Acoustic", "Animalborne_Acoustic",
"Animalborne_Archival", "Animalborne_Archival", "Stationary_Acoustic_PIT",
"Stationary_Acoustic_PIT", "Stationary_Acoustic_PIT", "BRUV_Acoustic",
"BRUV_Acoustic", "BRUV_Acoustic", "BRUV_Acoustic", "BRUV_Acoustic",
"BRUV_Acoustic", "Controlled_Archival", "Controlled_Archival",
"Controlled_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Animalborne_Archival", "Stationary_Radio",
"Stationary_Acoustic_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Stationary_Acoustic_Archival", "Stationary_Acoustic_Archival",
"Stationary_Acoustic_Archival", "Controlled_Acoustic", "Controlled_Acoustic",
"Animalborne_Archival", "Animalborne_Archival", "Stationary_Acoustic",
"Stationary_Acoustic", "Animalborne_Satellite_Archival", "Animalborne_Satellite_Archival",
"Animalborne_Satellite_Archival", "Animalborne_Satellite_Archival",
"Stationary_Acoustic", "Stationary_Acoustic", "Stationary_Acoustic",
"Stationary_Acoustic", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Stationary_Satellite", "Stationary_Satellite",
"Stationary_Satellite", "Stationary_Satellite", "Stationary_Satellite",
"Animalborne_Archival", "Animalborne_Archival", "Stationary_Acoustic_Radio",
"Stationary_Acoustic_Radio", "Stationary_Acoustic_Radio", "Animalborne_Archival",
"Animalborne_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Stationary_Acoustic", "Stationary_Acoustic", "Animalborne_Archival",
"Animalborne_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "Animalborne_Archival", "Animalborne_Archival",
"Animalborne_Archival", "BRUV_Acoustic", "BRUV_Acoustic", "BRUV_Acoustic",
"BRUV_Acoustic", "BRUV_Acoustic", "Stationary_Acoustic", "Stationary_Acoustic",
"Stationary_Acoustic", "Stationary_Acoustic", "Stationary_Acoustic",
"Stationary_Acoustic", "Stationary_Acoustic", "Stationary_Acoustic",
"Stationary_Acoustic", "Animalborne_Satellite_Archival", "Animalborne_Satellite_Archival",
"Animalborne_Satellite_Archival", "Controlled_Acoustic", "Controlled_Acoustic",
"Controlled_Acoustic")), class = "data.frame", row.names = c(NA,
-245L))
I want a new column where the values relate to some strings so for every value that has a Acoustic, radio, or PIT the value in the new column should be receiver based and for everything else it should be non receiver based. But for columns that have acoustic and satellite in the column I need the value in the new column to say Both.
I have tried the ifelse command using this code
df$Type = ifelse(df$Combination == "Acoustic", 'Non Receiver Based', 'Receiver Based')
But it's calling all of them receiver based and I don't know how to incorporate all the arguments I am suggesting above.
答案1
得分: 1
使用grepl来检查字符串中是否包含其中一个单词。模式由用|分隔的单词组成，以表示检查字符串中是否出现这些单词中的任何一个：
min_string = c("BRUV_Acoustic_Satellite", "Stationary_Radio_PIT", "Animalborne_Satellite_Archival")
pattern = paste0(c("Acoustic", "radio", "PIT"), collapse = "|")
ifelse(!grepl(pattern, min_string), '非接收器',
ifelse(grepl("Acoustic", min_string) & grepl("Satellite", min_string), "两者",
"接收器"))
#[1] "两者" "接收器" "非接收器"
dplyr的替代方法是使用case_when，这可能更容易理解一些：
library(dplyr)
case_when(!grepl(pattern, min_string) ~ "非接收器",
grepl("Acoustic", min_string) & grepl("Satellite", min_string) ~ "两者",
grepl(pattern, min_string) ~ "接收器")
英文:
Use grepl to check if one of the words appears in the string. The pattern consists of the words separated by |, to say check if either of this words appears in the string:
min_string = c("BRUV_Acoustic_Satellite", "Stationary_Radio_PIT", "Animalborne_Satellite_Archival")
pattern = paste0(c("Acoustic", "radio", "PIT"), collapse = "|")
ifelse(!grepl(pattern, min_string), 'Non Receiver Based',
ifelse(grepl("Acoustic", min_string) & grepl("Satellite", min_string), "Both",
"Receiver Based"))
#[1] "Both" "Receiver Based" "Non Receiver Based"
A dplyr alternative is to use case_when, which might be a bit friendlier to understand:
library(dplyr)
case_when(!grepl(pattern, min_string) ~ "Non Receiver",
grepl("Acoustic", min_string) & grepl("Satellite", min_string) ~ "Both",
grepl(pattern, min_string) ~ "Receiver")
答案2
得分: 1
你可以使用 grepl() 来检查字符串中是否出现了特定模式，然后使用 if-else 语句来决定不同的情况。由于 if() 不支持矢量化操作，你需要将其包装在 Vectorize() 中，以便在 mutate() 中使用。
library(tidyverse)
match <- Vectorize(function(string) {
if (!grepl("Acoustic|radio|PIT", string)) {
"非接收器基础"
} else if (grepl("Acoustic", string) && grepl("Satellite", string)) {
"两者都有"
} else "接收器基础"
})
df %>% mutate(new_var = match(Combination))
英文:
You can use grepl() to check if a certain pattern occurs in a string, and then use if-else statements to decide on your cases. Since if() is not vectorized, you need to wrap it in Vectorize() to use it in mutate().
library(tidyverse)
match <- Vectorize(function(string) {
if (!grepl("Acoustic|radio|PIT", string)) {
"non receiver based"
} else if ((grepl("Acoustic", string)) &
(grepl("Satellite", string))) {
"Both"
} else "receiver based"
})
df %>% mutate(new_var = match(Combination))
答案3
得分: 0
你可以使用tidyverse中的一些函数。
下面，我创建一个名为new_col的新列，这是您期望的输出：
library(tidyverse)
df %>%
mutate(new_col = if_else(str_detect(Combination, "Acoustic")==T | str_detect(Combination, "Radio")==T | str_detect(Combination, "PIT"), "receiver based", "non receiver based")) %>%
mutate(new_col = if_else(str_detect(Combination, "Acoustic")==T & str_detect(Combination, "Satellite")==T, "Both", new_col))
英文:
You can use some functions from tidyverse.
Below, I create a new column called new_col which is your desired output:
library(tidyverse)
df %>%
mutate(new_col = if_else(str_detect(Combination, "Acoustic")==T | str_detect(Combination, "Radio")==T | str_detect(Combination, "PIT"), "receiver based", "non receiver based")) %>%
mutate(new_col = if_else(str_detect(Combination, "Acoustic")==T & str_detect(Combination, "Satellite")==T, "Both", new_col))
答案4
得分: 0
这个翻译如下：
这样的内容对你有帮助吗？
df1 <- df %>% mutate(new_col = case_when(
str_detect(Combination, "Acoustic_Satellite") ~ "both",
str_detect(Combination, "Acoustic") ~ "Receiver Based",
str_detect(Combination, "Radio") ~ "Receiver Based",
str_detect(Combination, "PIT") ~ "Receiver Based"))
df2 <- replace(df1, is.na(df1), "non receiver based")
这只是一个比较粗糙的解决方案，我相信有人会找到更好的方法。只有在你的数据集中"Acoustic_Satellite"是一个单词时才有效。
英文:
would something like this be helpfull to you?
df1 <- df %>% mutate(new_col = case_when(
str_detect(Combination, "Acoustic_Satellite") ~ "both",
str_detect(Combination, "Acoustic") ~ "Receiver Based",
str_detect(Combination, "Radio") ~ "Receiver Based",
str_detect(Combination, "PIT") ~ "Receiver Based"))
df2 <- replace(df1, is.na(df1), "non receiver based")
It is a rather dirty solution, I am sure someone will find a better one and only works if Acoustic_Satelite is a single word in your dataset but it does the job.
答案5
得分: 0
一个简单的逐步方法可以是：
df$Type <- "非接收器型"
i <- grepl("Acoustic", df$Combination)
df$Type[i | grepl("radio|PIT", df$Combination)] <- "接收器型"
df$Type[i & grepl("Satellite", df$Combination)] <- "两者"
rm(i)
df
# Combination Type
#1 BRUV_Acoustic_Satellite 两者
#2 BRUV_Acoustic_Satellite 两者
#3 BRUV_Acoustic_Satellite 两者
#4 BRUV_Acoustic_Satellite 两者
#5 BRUV_Acoustic_Satellite 两者
#6 BRUV_Acoustic_Satellite 两者
#7 Animalborne_Archival 非接收器型
#8 Animalborne_Archival 非接收器型
#9 Animalborne_Archival 非接收器型
#10 Controlled_Acoustic 接收器型
#...
只是为了好玩的一个基准测试：
library(tidyverse)
match <- Vectorize(function(string) {
if (!grepl("Acoustic|radio|PIT", string)) {
"非接收器型"
} else if ((grepl("Acoustic", string)) &
(grepl("Satellite", string))) {
"两者"
} else "接收器型"
})
bench::mark(check = FALSE,
Maël1 = local({pattern = paste0(c("Acoustic", "radio", "PIT"), collapse = "|")
cbind(df, Type=ifelse(!grepl(pattern, df$Combination), '非接收器型',
ifelse(grepl("Acoustic", df$Combination) & grepl("Satellite", df$Combination), "两者",
"接收器型"))) }),
Maël2 = local({pattern = paste0(c("Acoustic", "radio", "PIT"), collapse = "|")
cbind(df, Type=case_when(!grepl(pattern, df$Combination) ~ "非接收器",
grepl("Acoustic", df$Combination) & grepl("Satellite", df$Combination) ~ "两者",
grepl(pattern, df$Combination) ~ "接收器")) }),
"Lukas Unterschuetz" = local({df %>% mutate(new_var = match(Combination))}),
Leonardo19 = local({df %>%
mutate(new_col = if_else(str_detect(Combination, "Acoustic")==T | str_detect(Combination, "Radio")==T | str_detect(Combination, "PIT"), "接收器型", "非接收器型")) %>%
mutate(new_col = if_else(str_detect(Combination, "Acoustic")==T & str_detect(Combination, "Satellite")==T, "两者", new_col)) }),
procerus = local({df1 <- df %>% mutate(new_col = case_when(
str_detect(Combination, "Acoustic_Satellite") ~ "两者",
str_detect(Combination, "Acoustic") ~ "接收器型",
str_detect(Combination, "Radio") ~ "接收器型",
str_detect(Combination, "PIT") ~ "接收器型"))
replace(df1, is.na(df1), "非接收器型")}),
GKi = local({df$Type <- "非接收器型"
i <- grepl("Acoustic", df$Combination, fixed=TRUE)
df$Type[i | grepl("radio|PIT", df$Combination)] <- "接收器型"
df$Type[i & grepl("Satellite", df$Combination, fixed=TRUE)] <- "两者"
rm(i)
df})
)
结果：
expression 最小时间 中位时间 `每秒迭代次数` 内存分配量 `每秒垃圾回收次数` 迭代次数 垃圾回收次数
1 Maël1 531.63微秒 562.63微秒 1751. 31.1KB 2.02 865 1
2 Maël2 776.25微秒 820.19微秒 1204. 44.4KB 2.02 596 1
3 Lukas Unterschuetz 5.52毫秒 5.76毫秒 172. 11.8KB 4.13 83 2
4 Leonardo19 2.91毫秒 3.07毫秒 307. 50.6KB 6.26 147 3
5 procerus 1.89毫秒 2.02毫秒 457. 66.9KB 6.24 220 3
6 GKi 214.9微秒 231.51微秒 4241. 11.8KB 2.02 2099 1
英文:
A simple step by step way could be:
df$Type <- "non receiver based"
i <- grepl("Acoustic", df$Combination)
df$Type[i | grepl("radio|PIT", df$Combination)] <- "receiver based"
df$Type[i & grepl("Satellite", df$Combination)] <- "Both"
rm(i)
df
# Combination Type
#1 BRUV_Acoustic_Satellite Both
#2 BRUV_Acoustic_Satellite Both
#3 BRUV_Acoustic_Satellite Both
#4 BRUV_Acoustic_Satellite Both
#5 BRUV_Acoustic_Satellite Both
#6 BRUV_Acoustic_Satellite Both
#7 Animalborne_Archival non receiver based
#8 Animalborne_Archival non receiver based
#9 Animalborne_Archival non receiver based
#10 Controlled_Acoustic receiver based
#...
Just for fun a Benchmark:
library(tidyverse)
match <- Vectorize(function(string) { #Maybe another name would be better
if (!grepl("Acoustic|radio|PIT", string)) {
"non receiver based"
} else if ((grepl("Acoustic", string)) &
(grepl("Satellite", string))) {
"Both"
} else "receiver based"
})
bench::mark(check = FALSE,
Maël1 = local({pattern = paste0(c("Acoustic", "radio", "PIT"), collapse = "|")
cbind(df, Type=ifelse(!grepl(pattern, df$Combination), 'Non Receiver Based',
ifelse(grepl("Acoustic", df$Combination) & grepl("Satellite", df$Combination), "Both",
"Receiver Based"))) }),
Maël2 = local({pattern = paste0(c("Acoustic", "radio", "PIT"), collapse = "|")
cbind(df, Type=case_when(!grepl(pattern, df$Combination) ~ "Non Receiver",
grepl("Acoustic", df$Combination) & grepl("Satellite", df$Combination) ~ "Both",
grepl(pattern, df$Combination) ~ "Receiver")) }),
"Lukas Unterschuetz" = local({df %>% mutate(new_var = match(Combination))}),
Leonardo19 = local({df %>%
mutate(new_col = if_else(str_detect(Combination, "Acoustic")==T | str_detect(Combination, "Radio")==T | str_detect(Combination, "PIT"), "receiver based", "non receiver based")) %>%
mutate(new_col = if_else(str_detect(Combination, "Acoustic")==T & str_detect(Combination, "Satellite")==T, "Both", new_col)) }),
procerus = local({df1 <- df %>% mutate(new_col = case_when(
str_detect(Combination, "Acoustic_Satellite") ~ "both",
str_detect(Combination, "Acoustic") ~ "Receiver Based",
str_detect(Combination, "Radio") ~ "Receiver Based",
str_detect(Combination, "PIT") ~ "Receiver Based"))
replace(df1, is.na(df1), "non receiver based")}),
GKi = local({df$Type <- "non receiver based"
i <- grepl("Acoustic", df$Combination, fixed=TRUE)
df$Type[i | grepl("radio|PIT", df$Combination)] <- "receiver based"
df$Type[i & grepl("Satellite", df$Combination, fixed=TRUE)] <- "Both"
rm(i)
df})
)
Result
expression min median `itr/sec` mem_alloc `gc/sec` n_itr n_gc
<bch:expr> <bch:tm> <bch:tm> <dbl> <bch:byt> <dbl> <int> <dbl>
1 Maël1 531.63µs 562.63µs 1751. 31.1KB 2.02 865 1
2 Maël2 776.25µs 820.19µs 1204. 44.4KB 2.02 596 1
3 Lukas Unterschuetz 5.52ms 5.76ms 172. 11.8KB 4.13 83 2
4 Leonardo19 2.91ms 3.07ms 307. 50.6KB 6.26 147 3
5 procerus 1.89ms 2.02ms 457. 66.9KB 6.24 220 3
6 GKi 214.9µs 231.51µs 4241. 11.8KB 2.02 2099 1
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论