CODEWITHSUNDEEP
r
Sociopath
Sociopath

Reputation: 13401

How to fetch numbers from a character vectors in R?

I have a character vector and want to fetch a numbers from it both integer as well as Floats.

    a <- c("Sam is 5 years old", "Ram is 3.7 years old" , "John is 17 years 2 months old")

output should be:

      [1]  5  3.7 17.2

Upvotes: 3

Views: 126

Answers (2)

Sotos
Sotos

Reputation: 51592

Here is a one liner,

as.numeric(gsub(' ', '.', trimws(gsub('\\D+', ' ', a))))
#[1]  5.0  3.7 17.2

Upvotes: 3

akrun
akrun

Reputation: 887128

We can use parse_number from readr

readr::parse_number(a)
#[1]  5.0  3.7 17.0

Update

Based on the OP's new example

library(stringr)
sapply(str_extract_all(a, "[0-9]+\\s+(years|months)"), function(x) {
       x1 <- readr::parse_number(x)
       head(if(length(x1)==2)   x1 + round(x1[2]/12, 1) else x1, 1)})
#[1]  5.0  7.0 17.2

If we don't need to be concerned about dividing by 12 for the 'months, another option is

as.numeric(sapply(regmatches(a, gregexpr('[0-9]+', a)), paste, collapse="."))
#[1]  5.0  3.7 17.2

Upvotes: 4

Related Questions