解析r中的列(或其他语言,如SQL)

这是当前的数据帧：

baking_time <- c("20 to 30 min","20 to 30 min","40 to 50 min","10 to 30 min","60 to 90 min","40 to 50 min")
cake_type <- c("Chocolate","Chocolate","Lemon","Tart","German","Lemon")


recipes <- data.frame(baking_time,cake_type)

现在我正在尝试解析烘焙时间来得到这个：

baking_time <- c(25,25,45,20,75,45)

我尝试过使用解析但是我在解析这两个数字时遇到的问题比对它们执行操作有困难

mutate(avg_time = (parse_number(baking_time) + parse_number(baking_time))/2)

解决方法

我们提取列的数字部分并获得平均值

library(tidyverse)
recipes %>% 
   mutate(avg_time = str_extract_all(baking_time,"\\d+") %>%
           map(.,~ mean(as.numeric(.x))))
#   baking_time cake_type avg_time
#1 20 to 30 min Chocolate       25
#2 20 to 30 min Chocolate       25
#3 40 to 50 min     Lemon       45
#4 10 to 30 min      Tart       20
#5 60 to 90 min    German       75
#6 40 to 50 min     Lemon       45

注意：readr :: parse_number提取第一个数字部分.如果有多个元素,需要将其分解并应用parse_number

recipes %>% 
   separate(baking_time,into = c("first","second"),sep=" to ",remove = FALSE) %>% 
   transmute(baking_time,avg_time = (parse_number(first) + parse_number(second))/2)

使用基数R,一个选项是在使用gsub将非数字部分更改为分隔符后使用read.csv读取,获取rowMeans

rowMeans(read.csv(text=gsub("\\D+",",recipes$baking_time),header = FALSE)[-3])
#[1] 25 25 45 20 75 45