purrr list evaluation strangeness

Question

It seems that purrr functions evaluate differently if lists have titles or not, but why? Is this just "one of those things"?

Example:

func_b <- function(x,y,z) paste(x,y,z)

## Works as expected
pmap(list(iris$Sepal.Length, iris$Sepal.Width, iris$Petal.Length), func_b) %>% head(2)

[[1]] [1] "5.1 3.5 1.4"

[[2]] [1] "4.9 3 1.4"

## Doesn't work
pmap(map(iris[,1:3],list),func_b)

Error in .f(Sepal.Length = .l[[c(1L, 1L)]], Sepal.Width = .l[[c(2L, 1L)]], : unused arguments (Sepal.Length = .l[[c(1, 1)]], Sepal.Width = .l[[c(2, 1)]], Petal.Length = .l[[c(3, 1)]])

But the only difference seems to be that one of the lists retains its titles while the other one doesn't?

list(iris$Sepal.Length, iris$Sepal.Width, iris$Petal.Length) %>% map(head)

[[1]] [1] 5.1 4.9 4.7 4.6 5.0 5.4

[[2]] [1] 3.5 3.0 3.2 3.1 3.6 3.9

[[3]] [1] 1.4 1.4 1.3 1.5 1.4 1.7

as.list(iris[,1:3]) %>% map(head)

$Sepal.Length [1] 5.1 4.9 4.7 4.6 5.0 5.4

$Sepal.Width [1] 3.5 3.0 3.2 3.1 3.6 3.9

$Petal.Length [1] 1.4 1.4 1.3 1.5 1.4 1.7

class(list(iris$Sepal.Length, iris$Sepal.Width, iris$Petal.Length)) == class(as.list(iris[,1:3]))

[1] TRUE

str(list(iris$Sepal.Length, iris$Sepal.Width, iris$Petal.Length)); str(as.list(iris[,1:3]))

List of 3
 $ : num [1:150] 5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
 $ : num [1:150] 3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
 $ : num [1:150] 1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...
List of 3
 $ Sepal.Length: num [1:150] 5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
 $ Sepal.Width : num [1:150] 3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
 $ Petal.Length: num [1:150] 1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...

If we just get rid of the names of the second list, it works without a problem.

aa <- as.list(iris[,1:3]) 
names(aa) <- NULL      
pmap(aa,func_b) %>% head(2)

[[1]] [1] "5.1 3.5 1.4"

[[2]] [1] "4.9 3 1.4"

So my specific question: WHY do the titles affect the evaluation method? Is there any way to convert things without breaking the dplyr pipe to eliminate the names?

Unnamed lists are assigned to the function arguments by position, named lists by name. That means purrr is trying to do func_b(Sepal.Length = ....., etc). You can always pipe into unname if necessary. — Axeman
Nice, the unname works beautifully: map(iris[,1:3],list) %>% unname() %>% pmap(func_b) %>% map(head) — Amit Kohli

hrbrmstr hrbrmstr · Accepted Answer · 2017-06-20T12:59:19

Axeman is spot-on but you can defensibly & dynamically program around it:

library(purrr)

func_b <- function(...) {
  args <- list(...)
  paste(args[[1]], args[[2]], args[[3]])
}

list(iris$Sepal.Length, iris$Sepal.Width, iris$Petal.Length) %>%
  pmap(func_b) %>%
  head(2)
## [[1]]
## [1] "5.1 3.5 1.4"
## 
## [[2]]
## [1] "4.9 3 1.4"

iris[,1:3] %>% pmap(func_b) %>% head(3)
## [[1]]
## [1] "5.1 3.5 1.4"
## 
## [[2]]
## [1] "4.9 3 1.4"
## 
## [[3]]
## [1] "4.7 3.2 1.3"

func_c <- function(...) {
  args <- list(...)
  paste0(args, collapse = " ")
}

list(iris$Sepal.Length, iris$Sepal.Width, iris$Petal.Length) %>%
  pmap(func_c) %>%
  head(2)
## [[1]]
## [1] "5.1 3.5 1.4"
## 
## [[2]]
## [1] "4.9 3 1.4"

iris[,1:3] %>% pmap(func_c) %>% head(3)
## [[1]]
## [1] "5.1 3.5 1.4"
## 
## [[2]]
## [1] "4.9 3 1.4"
## 
## [[3]]
## [1] "4.7 3.2 1.3"

iris[,1:2] %>% pmap(func_c) %>% head(3)
## [[1]]
## [1] "5.1 3.5"
## 
## [[2]]
## [1] "4.9 3"
## 
## [[3]]
## [1] "4.7 3.2"

purrr list evaluation strangeness

1 Answers