Convert character variable of height (5ft6in) into numeric variable in inches?

Question

I would like to convert a height variable I have from character type to numeric. for context, this is so I can use the values to calculate body mass index.

Looking at the below example data frame, I would like to convert Height_1 into Height_2 (whereby Height_2 is in inches):

# Height_1    Height_2
# 5ft6in      66 
# XftXin      XXXX
# XftXin      XXXX 
# XftXin      XXXX
# XftXin      XXXX

I have tried a few things using the "tidyverse" and "measurements" packages but have not been able to create a variable like Height_2 above. For example:

library(dplyr)
library(tidyr)

df %>%
  separate(Height_1,c('feet', 'inches'), sep = 'ft', convert = TRUE, remove = FALSE) %>%
  mutate(Height_2 = 12*feet + inches)

I think this is because the above doesn't address the fact that there is "in" at the end of the values.

Answer 1

You can use regex to extract feet and inches data from Height_1 and then perform the calculation.

library(dplyr)
library(tidyr)

df %>%
  extract(Height_1, c('feet', 'inches'), '(\\d+)ft(\\d+)in', convert = TRUE, remove = FALSE) %>%
  transmute(Height_1, 
            Height_2 = 12*feet + inches)

#  Height_1 Height_2
#1   5ft6in       66
#2   4ft9in       57
#3  5ft12in       72
#4   4ft9in       57
#5   6ft2in       74

In base R -

transform(strcapture('(\\d+)ft(\\d+)in', df$Height_1, 
           proto = list(feet = numeric(), inches = numeric())), 
          Height_2 = 12*feet + inches)

data

df <- structure(list(Height_1 = c("5ft6in", "4ft9in", "5ft12in", "4ft9in", "6ft2in")), row.names = c(NA, -5L), class = "data.frame")