Regular expression vectors
Base R provides functions for parsing character vectors with a single regular expression. This package provides fast C code for vectors of regular expressions.
install.packages("devtools")
devtools::install_github("tdhock/revector")
library(revector)
example(str_match_each)stringi provides similar functionality:
> genotype <- c("GA", "TAATAAA")
> pat <- c("(?<A>[ATCG])(?<B>[ATCG])", "^(?<C>TAA|TAAA)(?<D>TAA|TAAA)")
> revector::str_match_each(genotype, pat)
A B
[1,] "G" "A"
[2,] "TAA" "TAA"
> stringi::stri_match_first_regex(genotype, pat)
[,1] [,2] [,3]
[1,] "GA" "G" "A"
[2,] "TAATAA" "TAA" "TAA"The code above shows that revector gives colnames using the first pattern (and ignores the others), whereas stringi ignores all of the names in the pattern.
> pat <- c("(?<A>[ATCG])", "^(?<C>TAA|TAAA)(?<D>TAA|TAAA)")
> revector::str_match_each(genotype, pat)
Error in revector::str_match_each(genotype, pat) :
first pattern has 1 group(s), pattern 2 has 2
> stringi::stri_match_first_regex(genotype, pat)
[,1] [,2] [,3]
[1,] "G" "G" NA
[2,] "TAATAA" "TAA" "TAA"The code above shows that revector stops with an error if there is not the same number of groups in each pattern, whereas stringi uses the max group number, and fills with NA when a pattern has less than the max number of groups.