purrr-tutorial | A introduction to purrr | Learning library

Short Answer

No need for map here:

The recipe for this kind of problem is always the same:

Decompose the problem, solve it for an individual case, and then put it back together inside out.

As you observed, mtcars %>% split(.$cyl) gives you a list of lists (list of data.frames). You want to map mean over the inner lists.

So let’s do it for one list first:

We can use map2 and apply the pmap function for each row.

Yes, the problem is the different data classes. When you use c(), everything gets coerced to the same atomic type . So c(1, "full_name") is turned into c("1", "full_name") which is a character vector. And the first list doesn't have a named element with the name "1". If you want to have mixed classes in R, you generally use lists. And it appears that map supports lists. For example

Since you mention the purrr package, some other tidyverse packages are worth considering.

• dplyr for mutate(), when applying purrr::map() to a column of a data frame and storing the result as list-column.
• tidyr for unnest(), which expands a list-column so that each row inside a list-column becomes a row in the overall data frame.
• tibble for nicely printed nested data frames

Sample files are needed to demonstrate. This code uses the openxlsx package to create one file containing two sheets (the built-in iris and mtcars datasets), and another file containing three sheets (adding the built-in attitude dataset).

Not sure if there is a more purrry way of doing this, but it works.

I think you have a naming conflict in your environment. As a result, the map function you are calling is another function from another package (probably the maps package, which matches the arguments listed in your error code, see here). Try to call the function with explicit context as below. Also, slightly change your syntax because otherwise your return will be empty.