Create a New Variable Based on the First 7 Characters of Existing Variable

Create a new variable based on the first 7 characters of existing variable

You are passing wrong argument to substr

Try

variable2 <- substr(mydata$variable,1,7)

Create new variable based on another variable that contrains character strings of numbers?

You can use the following:-

 df$Relevant_Diagnoses <- ifelse(is.na(df$Diagnoses), NA,
                                 ifelse(grepl("2", df$Diagnoses), 1, 0))

Create a new variable based in conditions in R

The syntax of mutate is mutate(.data, new_col = something). Also, since you have multiple condition based on var1, you'll need to have a nested ifelse. Therefore, the correct way of doing it is:

library(dplyr)

df %>% mutate(new = ifelse(var1 == 1, 
                           var2, 
                           ifelse(var1 == 5, 
                                  var3, 
                                  ifelse(var1 == 10, 
                                         NA, 
                                         ifelse(var1 == 15, 
                                                var4, 
                                                NA)))))

Or use dplyr::case_when, which is way clearer in the syntax:

df %>% mutate(new = case_when(var1 == 1 ~ as.character(var2),
                              var1 == 5 ~ var3,
                              var1 == 15 ~ var4))

Output

The output of the above two methods are the same.

  var1 var2 var3 var4  new
1    1   45 AH67 A456   45
2    1   78 GH98 D788   78
3    5   46 GD94 M747 GD94
4    5   98 GF21 G589 GF21
5   10   47 GD09 I989 <NA>
6   10   54 KG32 U456 <NA>
7   15   48 FS89 C191 C191
8   15   66 GF23 Y198 Y198

In R: How to create new variables based on conditions on existing variables

You could create an index dataset ('df2') and merge that with the original dataset ('df1')

 merge(df1, df2)
 #  var1 var2           var3    var4
 #1   01    A       New York Buffalo
 #2   01    A       New York Buffalo
 #3   01    B       New York Cornell
 #4   02    C North Carolina Raleigh
 #5   02    C North Carolina Raleigh
 #6   03    D          Texas  Dallas

data

df1 <- structure(list(var1 = c("01", "01", "01", "02", "02", "03"), 
var2 = c("A", "B", "A", "C", "C", "D")), .Names = c("var1", 
"var2"), row.names = c(NA, -6L), class = "data.frame")

df2 <-  data.frame(var1=c('01', '01', '02', '03'), var2=LETTERS[1:4], 
 var3=c('New York', 'New York', 'North Carolina', 'Texas'),
 var4=c('Buffalo', 'Cornell', 'Raleigh', 'Dallas'))

creating a new variable according to existing variables using R

ifelse evaluates a condition (whether or not temp is NA) in an element by element fashion. We will test whether or not temp is NA and assign the resulting value as NA or surv1 depending on the outcome.

data$surv2 <- with(data, ifelse( is.na(temp), NA, surv1))

---

Related Topics