将字符串列列表转换为数字并汇总

Question

我有一个数据框，其中一列包含一串数字。每行包含一组长度不同的数字。例如：

Month       Count
Jan         "[1.2445, 23.888883, 16.11208347]" 
Feb         "[2.6473, 400.6256]"
March       "[6723.1838282, 187.1212, 90.111, 75.1342899]" 

目标是将列表转换为数字格式，并对每行求和。结果类似于（为了方便起见，四舍五入小数）：

Month       Count
Jan         41.245 
Feb         403.2729
March       7075.500 

我使用了适用于单行的代码，但我无法将其逐行抽象到整个数据帧。

sum(as.numeric(strsplit(substr(Data$Count, 2, nchar(Data$Count) - 1), ',')[[1]]))

Answer 1

你就快到了 - 在基础 R 中，你可以将

lapply

与

strsplit

和

gsub

（而不是

substr

）一起使用：

unlist(lapply(strsplit(gsub("\\[|\\]", "", df$Count), ","), \(x) sum(as.numeric(x))))

# [1]   41.24547  403.27290 7075.55032

数据：

df <- read.table(text = 'Month       Count
                 Jan         "[1.2445, 23.888883, 16.11208347]" 
                 Feb         "[2.6473, 400.6256]"
                 March       "[6723.1838282, 187.1212, 90.111, 75.1342899]" ', h = TRUE)

Answer 2

Data <- read.table(text = 'Month       Count
Jan         "[1.2445, 23.888883, 16.11208347]" 
Feb         "[2.6473, 400.6256]"
March       "[6723.1838282, 187.1212, 90.111, 75.1342899]"', 
                   header = T, stringsAsFactors = F)

library(dplyr, warn.conflicts = F)

Data %>% 
  mutate(Count = strsplit(gsub("\\[|\\]", "", Count), ", ")) %>% 
  tidyr::unnest_longer(Count) %>% 
  summarise(Count = sum(as.numeric(Count)), .by = Month) %>% 
  as.data.frame()

#>   Month      Count
#> 1   Jan   41.24547
#> 2   Feb  403.27290
#> 3 March 7075.55032

^{创建于 2024-04-29，使用 reprex v2.0.2}