CODEWITHSUNDEEP
regexr
Oscar Benitez
Oscar Benitez

Reputation: 267

r delete rows with numbers

I want to delete the rows in my data frame where the var1 or var2 has a digit. I've tried with different variations of regex but none of them seem to work. Below is a simplified example looking for digits on var1.

var1<-c("000","1","1","1","1","000256","1","shall","1","1","1","1","the","1",
        "001","1","1","1","1","one")
var2<-c("people","0","00","000","1","2","3","begin","4","5","6","7","8","a",
        "and","billion","hour","in","is","million")
val<-c(1639,1703,655,3542,3273,9658,2562,1027,3340,2236,971,783,1057,673,1658,
       1367,843,1921,459,2589)
df<-data.frame(var1,var2,val)
df<-subset(df,df$var1!="[0-9]*")

What am I doing wrong?

Upvotes: 2

Views: 1953

Answers (2)

Steven Beaupr&#233;
Steven Beaupr&#233;

Reputation: 21621

Same idea as lukeA but using filter() from dplyr

library(dplyr)
df %>% filter(!grepl("[[:digit:]]", var1) & !grepl("[[:digit:]]", var2))

or slice()

df %>% slice(which(!grepl("[[:digit:]]", var1) & !grepl("[[:digit:]]", var2)))

or as per @akrun suggestion:

df %>% slice(intersect(grep("[^0-9]", var1), grep("[^0-9]", var2)))

Which gives:

#   var1    var2  val
#1 shall   begin 1027
#2   one million 2589

Upvotes: 2

lukeA
lukeA

Reputation: 54237

Try this

subset(df, !grepl("[0-9]", var1) & !grepl("[0-9]", var2))
# var1    var2  val
# 8  shall   begin 1027
# 20   one million 2589

Upvotes: 3

Related Questions