Remove string from a vector in R

Question

I have a vector that looks like

> inecodes
   [1] \"01001\" \"01002\" \"01049\" \"01003\" \"01006\" \"01037\" \"01008\" \"01004\" \"01009\" \"01010\" \"01         


        

Answer 1

Try this. Match should be much faster

pos<-which(!is.na(pob[match(sub('^([0-9]+)-.*$','\\1',pob),inecodes)]))
pob[pos]<-sub('^[0-9]+-(.*)$','\\1',pob[pos])

Please do post the timings if you manage to get this. Match usually solves many computational issues for large data sets lookup. Would like to see if there are any opposite scenarios.

Answer 2

Not sure why you needed inecodes at all, since you can use sub to remove all digits:

sub('^\\d+-', '', pob)

Result:

 [1] "Alegría-Dulantzi"           "Amurrio"                    "Añana"                     
 [4] "Aramaio"                    "Armiñón"                    "Arraia-Maeztu"             
 [7] "Arratzua-Ubarrundia"        "Artziniega"                 "Asparrena"                 
[10] "Ayala/Aiara"                "Baños de Ebro/Mañueta"      "Barrundia"                 
[13] "Berantevilla"               "Bernedo"                    "Campezo/Kanpezu"           
[16] "Elburgo/Burgelu"            "Elciego"                    "Elvillar/Bilar"            
[19] "Erriberagoitia/Ribera Alta"

One reason that you might need inecodes is that you have codes in pob that don't exist in inecodes, but that doesn't seem like the case here. If you insist on using inecodes to remove numbers from pob, you can use str_replace_all from stringr:

library(stringr)

str_replace_all(pob, setNames(rep("", length(inecodes)), paste0(inecodes, "-")))

This gives you the exact same result:

 [1] "Alegría-Dulantzi"           "Amurrio"                    "Añana"                     
 [4] "Aramaio"                    "Armiñón"                    "Arraia-Maeztu"             
 [7] "Arratzua-Ubarrundia"        "Artziniega"                 "Asparrena"                 
[10] "Ayala/Aiara"                "Baños de Ebro/Mañueta"      "Barrundia"                 
[13] "Berantevilla"               "Bernedo"                    "Campezo/Kanpezu"           
[16] "Elburgo/Burgelu"            "Elciego"                    "Elvillar/Bilar"            
[19] "Erriberagoitia/Ribera Alta"

Data:

inecodes = c("01001", "01002", "01049", "01003", "01006", "01037", "01008", 
"01004", "01009", "01010", "01011", "01013", "01014", "01016", 
"01017", "01021", "01022", "01023", "01046", "01056", "01901", 
"01027", "01019", "01020", "01028", "01030", "01031", "01032", 
"01902", "01033", "01036", "01058", "01034", "01039", "01041", 
"01042", "01043", "01044", "01047", "01051", "01052", "01053", 
"01054", "01055")

pob = c("01001-Alegría-Dulantzi", "01002-Amurrio", "01049-Añana", "01003-Aramaio", 
"01006-Armiñón", "01037-Arraia-Maeztu", "01008-Arratzua-Ubarrundia", 
"01004-Artziniega", "01009-Asparrena", "01010-Ayala/Aiara", "01011-Baños de Ebro/Mañueta", 
"01013-Barrundia", "01014-Berantevilla", "01016-Bernedo", "01017-Campezo/Kanpezu", 
"01021-Elburgo/Burgelu", "01022-Elciego", "01023-Elvillar/Bilar", 
"01046-Erriberagoitia/Ribera Alta")

Answer 3

library(stringr)

for(code in inecodes) {
  ix <- which(str_detect(pob, code))
  pob[ix] <- unlist(str_split(pob, "-", 2))[2]
}