CODEWITHSUNDEEP
rstringdata.tablena
velvetrock
velvetrock

Reputation: 585

R: Transfer numeric data and NA into string

The variable Recence_Connexion_Jrs is composed of "NA" and some figures (from 1 to 100), and I'm trying to make the "NA" equal to "No Open", and also cut the figures into 3 groups, and the labels of the groups are in the type of string. For example:

Recence_Connexion_Jrs           Connexion
      NA                         No open
      NA                         No open
      NA                         No open
      1                        connexion 0-7
      10                       connexion 7-30

The code that I used:

setDT(newdata)[!duplicated(newdata),Connexion:=ifelse(is.na(Recence_Connexion_jrs),
"No Open",cut(Recence_Connexion_jrs,breaks=c(-Inf,7,30,+Inf),
labels=c("connexion 0-7","connexion 7-30","connexion 30+")))]

But it gave the wrong result.

How can I solve that?

Upvotes: 1

Views: 59

Answers (1)

A5C1D2H2I1M1N2O1R2T1
A5C1D2H2I1M1N2O1R2T1

Reputation: 193687

I would suggest doing this in two steps: (1) using cut as you've done, and (2) using replace for the NA values.

Here's an example on a vector. No reason you can't do something similar in data.table:

set.seed(123)
vec <- sample(c(1:40, NA), 100, TRUE)
new <- cut(vec, c(-Inf, 7, 30, +Inf), 
           labels=c("connexion 0-7", "connexion 7-30", "connexion 30+"))
new <- replace(as.character(new), is.na(new), "No open")

Or, on a small data.table for you to see more easily:

library(data.table)
DT <- data.table(vec = c(1, NA, 8, 20, NA, 32))
DT[, new := as.character(
  cut(vec, c(-Inf, 7, 30, +Inf),
      labels = c("connexion 0-7", "connexion 7-30", "connexion 30+")))][
        , new := replace(new, is.na(new), "No open")
      ][]    
#    vec            new
# 1:   1  connexion 0-7
# 2:  NA        No open
# 3:   8 connexion 7-30
# 4:  20 connexion 7-30
# 5:  NA        No open
# 6:  32  connexion 30+

Upvotes: 3

Related Questions