Extract values after a word/alphabet from one column to a new column

Based on the data below how can I create another column based on the values after the alphabetU in every row of column GEOID?

Data:

id = c(1, 2, 3, 4, 5)
GEOID = c("1400000US42101000101", "1400000US42101000102", "1400000US42101000103",
"1400000US42101000104", "1400000US42101000105")

df = data.frame(id, GEOID)

Desired output:

id = c(1, 2, 3, 4, 5)
GEOID = c("1400000US42101000101", "1400000US42101000102", "1400000US42101000103",
"1400000US42101000104", "1400000US42101000105")
GEOID_Final = c("42101000101", "42101000102", "42101000103", "42101000104", "42101000105")
Desired_df = data.frame(id, GEOID, GEOID_Final)

Code:

library(stringr)
library(dplyr)

desired_df = df %>% word(?, sep = "US") # Stuck

I would use regular expressions to find every digits before "US" and US, and remove them from the string, only keeping the numbers after US.

library(stringr); library(dplyr)

id = c(1, 2, 3, 4, 5)
GEOID = c("1400000US42101000101", "1400000US42101000102", "1400000US42101000103",
          "1400000US42101000104", "1400000US42101000105")

df = data.frame(id, GEOID)


df %>%
  mutate(
    # replace every digit before US and US with empty string                       
    GEOID_final = str_replace(GEOID, pattern="\\d+US" , replacement ="")
  )
# id                GEOID GEOID_final
# 1  1 1400000US42101000101 42101000101
# 2  2 1400000US42101000102 42101000102
# 3  3 1400000US42101000103 42101000103
# 4  4 1400000US42101000104 42101000104
# 5  5 1400000US42101000105 42101000105

str_replace takes a string (or in this case, a vector of string, GEOID), and will replace with replacement what has been matched with the pattern. The pattern I use is \\d+US, which means 'match any digit one or more times, followed by US'. This is then replaced with nothing, only keeping the digits after US. You could also match it with .+US, which means 'match anything before US, and US'.

Hope this helps!

If you want to use word()

desired_df = df %>% word(?, sep = "US") # Stuck

try this with stringr::word

word(string = str_replace(GEOID,pattern = 'US',replacement = " "),start = 2,end = 2)

result
> word(string = str_replace(GEOID,pattern = 'US',replacement = " "),start = 2,end = 2)
[1] "42101000101" "42101000102" "42101000103" "42101000104" "42101000105"

save them in a 3rd column or use in mutate:

df%>%mutate(
GEOID_Final =  word(string = str_replace(GEOID,pattern = 'US',replacement = " "),start = 2,end = 2)
  )

result

    id                GEOID GEOID_Final
1  1 1400000US42101000101 42101000101
2  2 1400000US42101000102 42101000102
3  3 1400000US42101000103 42101000103
4  4 1400000US42101000104 42101000104
5  5 1400000US42101000105 42101000105