CODEWITHSUNDEEP
r
flee
flee

Reputation: 1335

Apply conditional function to a dataframe

I want to apply a function to rows of a data frame. The function is conditional on the value of one column being greater than the value in another column. If the condition is met I take the element from two (other) columns and multiply them, the result is then added to a new column. If the initial condition is not met there is no multiplication and an original value is copied to the new column.

Create some data:

var0 <- c("A", "B", "C", "D", "E")
var1 <- rep(c(105,200), each = 5)
var2 <- c(110:114, 25:29)
var3 <- rep(c(560,135), each = 5)
var4 <- rep(c(0.5,0.2), each = 5)
my_df <- as.data.frame(cbind(var0, var1, var2, var3, var4))

Have a look at the data:

       var0 var1 var2 var3 var4
    1     A  105  110  560  0.5
    2     B  105  111  560  0.5
    3     C  105  112  560  0.5
    4     D  105  113  560  0.5
    5     E  105  114  560  0.5
    6     A  200   25  135  0.2
    7     B  200   26  135  0.2
    8     C  200   27  135  0.2
    9     D  200   28  135  0.2
    10    E  200   29  135  0.2

My attempt at writing the code:

apply(my_df, 1, function(x) {
  if(x$var3 > x$var1) {
    x$output <- x$var2 * x$var4
    } else {
      x$output <- x$var2
    }
      return(x)
  })

What the result should look like:

        var0 var1 var2 var3 var4 output
    1     A  105  110  560  0.5   55.0
    2     B  105  111  560  0.5   55.5
    3     C  105  112  560  0.5   56.0
    4     D  105  113  560  0.5   56.5
    5     E  105  114  560  0.5   57.0
    6     A  200   25  135  0.2   25.0
    7     B  200   26  135  0.2   26.0
    8     C  200   27  135  0.2   27.0
    9     D  200   28  135  0.2   28.0
    10    E  200   29  135  0.2   29.0

Because var3 is greater than var1 in the first 5 rows var2 * var4 occurs, in the last 5 rows the condition is not met so var2 is simply copied to the output column.

Upvotes: 1

Views: 8301

Answers (2)

Ajay Ohri
Ajay Ohri

Reputation: 3492

var0 <- c("A", "B", "C", "D", "E")
var1 <- rep(c(105,200), each = 5)
var2 <- c(110:114, 25:29)
var3 <- rep(560,135, 5)
var4 <- rep(c(0.5,0.2), each = 5)

to avoid numbers to be converted to factors I am using cbind.data.frame instead of as.data.frame of cbind

my_df <-cbind.data.frame(var0, var1, var2, var3, var4)

> str(my_df)
'data.frame':   10 obs. of  5 variables:
 $ var0: Factor w/ 5 levels "A","B","C","D",..: 1 2 3 4 5 1 2 3 4 5
 $ var1: num  105 105 105 105 105 200 200 200 200 200
 $ var2: int  110 111 112 113 114 25 26 27 28 29
 $ var3: num  560 560 560 560 560 560 560 560 560 560
 $ var4: num  0.5 0.5 0.5 0.5 0.5 0.2 0.2 0.2 0.2 0.2

I then use an ifelse condition to get the new column

>my_df$output=ifelse(my_df$var3>my_df$var1,my_df$var2*my_df$var4,my_df$var2)
> my_df
   var0 var1 var2 var3 var4 output
1     A  105  110  560  0.5   55.0
2     B  105  111  560  0.5   55.5
3     C  105  112  560  0.5   56.0
4     D  105  113  560  0.5   56.5
5     E  105  114  560  0.5   57.0
6     A  200   25  560  0.2    5.0
7     B  200   26  560  0.2    5.2
8     C  200   27  560  0.2    5.4
9     D  200   28  560  0.2    5.6
10    E  200   29  560  0.2    5.8

Note I was not getting the same values in var3 as yours. So I changed var3 to be the ones given

> var3 <- c(rep(560,5),rep(135,5))

> var3
 [1] 560 560 560 560 560 135 135 135 135 135
>  my_df <-cbind.data.frame(var0, var1, var2, var3, var4)
>  my_df$output=ifelse(my_df$var3>my_df$var1,my_df$var2*my_df$var4,my_df$var2)
> my_df
   var0 var1 var2 var3 var4 output
1     A  105  110  560  0.5   55.0
2     B  105  111  560  0.5   55.5
3     C  105  112  560  0.5   56.0
4     D  105  113  560  0.5   56.5
5     E  105  114  560  0.5   57.0
6     A  200   25  135  0.2   25.0
7     B  200   26  135  0.2   26.0
8     C  200   27  135  0.2   27.0
9     D  200   28  135  0.2   28.0
10    E  200   29  135  0.2   29.0

Upvotes: 1

Tim Biegeleisen
Tim Biegeleisen

Reputation: 521103

You don't need to use an apply() function here, you can just use ifelse():

df$output <- ifelse(df$var3 > df$var1, df$var2*df$var4, df$var2)

Upvotes: 2

Related Questions