CODEWITHSUNDEEP
rtidyverse
user9964925
user9964925

Reputation: 23

could not sum up using map_dbl

I am trying to sum up all the numeric variables in iris, and not sum the non-numeric variables.

However, this does not work:

iris %>% map_dbl(function(x) if(is.numeric(x)) sum(x))

Why doesn't this work? It works if I remove the non-numeric variables first, as below

iris[1:4] %>% map_dbl(function(x) if(is.numeric(x)) sum(x))

Could we still use map_dbl for this case?

Upvotes: 2

Views: 408

Answers (2)

Ronak Shah
Ronak Shah

Reputation: 389135

You can also take a look at map_if where you can specify separate functions to be applied for TRUE and FALSE condition.

purrr::map_if(iris, is.numeric, sum, .else = function(x) return(NA))

#$Sepal.Length
#[1] 876.5

#$Sepal.Width
#[1] 458.6

#$Petal.Length
#[1] 563.7

#$Petal.Width
#[1] 179.9

#$Species
#[1] NA

Upvotes: 1

nyk
nyk

Reputation: 680

The following does not work because you did not specify what R should do if the variable is not a numeric variable. That is why you encountered an error.

iris %>% map_dbl(function(x) if(is.numeric(x)) sum(x))

However, if you do specify else, it will work:

iris %>% map_dbl(function(x) if(is.numeric(x)) sum(x) else NA)
Sepal.Length  Sepal.Width Petal.Length  Petal.Width      Species 
       876.5        458.6        563.7        179.9           NA

If you filter out non-numeric variable, it will work because there is no need for R to check for else. The if condition will always be met.

iris[1:4] %>% map_dbl(function(x) if(is.numeric(x)) sum(x))

Alternatively, you can use dplyr:

library(dplyr)
iris %>% summarize(across(is.numeric, sum))
  Sepal.Length Sepal.Width Petal.Length Petal.Width
1        876.5       458.6        563.7       179.9

Upvotes: 3

Related Questions