CODEWITHSUNDEEP
rcsv
user3188390
user3188390

Reputation: 613

how to replace all rows in a particular column with another string in r?

I have a csv file like

data.csv

Identity,CampaignName,Adgroupname
12,abc1,asterix
13,abc2,asterix1
frere,abc3,asterix2
xrert,dasa,trete
14,tytyt,ababa

Where the first row are the headers of the CSV file.

In the above CSV file if 'rer' appears in a string in 'Identiity' column then replace the whole string with '17' . for example 'frere' will be '17'. I can use 'gsub' to replace 'rer' , but how to replace the whole string with another string? New to R, any help is appreciated.

My desired output 

Identity,CampaignName,Adgroupname
12,abc1,asterix
13,abc2,asterix1
17,abc3,asterix2
17,dasa,trete
14,tytyt,ababa

Upvotes: 1

Views: 618

Answers (1)

Ben Bolker
Ben Bolker

Reputation: 226087

dat <- read.csv(
   text=
"Identity,CampaignName,Adgroupname
12,abc1,asterix
13,abc2,asterix1
frere,abc3,asterix2
xrert,dasa,trete
14,tytyt,ababa")

By default the Identity column is a factor, and you can't introduce new strings into a factor:

 dat$Identity[grepl("rer",dat$Identity)] <- "17"
 ## Warning message:
 ## In `[<-.factor`(`*tmp*`, grepl("rer", dat$Identity), value = c(1L,  :
 ##   invalid factor level, NA generated

You can either use stringsAsFactors=FALSE when you read the data, or (starting from the original data set)

 dat$Identity <- as.character(dat$Identity)
 dat$Identity[grepl("rer",dat$Identity)] <- "17"

You may then want to convert the column back into a numeric variable:

 dat$Identity <- as.numeric(dat$Identity)

Upvotes: 2

Related Questions