How can I filter dataframes from a list using Purrr?

Question

I have a list of dataframes. There are two types of dataframes, one listing names and ages, the other listing tests and grades (as in the 'combined list' below).

df1 <- data.frame(name = c("Alice", "Bob"),
                 age = c(23, 41))
df2 <- data.frame(test = c("Geography", "Science"),
                  grade = c("A", "B"))
df3 <- data.frame(name = c("Claire", "David"),
                  age= c(50, 32))
df4 <- data.frame(test = c("Geography", "Science"),
                  grade = c("B", "B"))

combined_list <- c(df1,df2,df3,d4)

I would like to subset the list into a dataframe consisting of only the names and the ages, so it would look something like this:

   name age
1 Alice  23
2   Bob  41
3 Claire 50
4 David  32

I found the Keep function of the Purrr package, which should be able to filter the list based on a condition, but I haven't found a way to make it work. This is what I've tried so far:

library(purrr)
purrr:keep(function(x) filter(!name=NULL))

How can I make this work? Are there other solutions?

Answer 1

We create the logical condition to Filter the elements of the list where it returns a single TRUE. It can be done by wrapping with all after creating a logical vector with %in% ie subsetting only those elements having both 'name', and 'age' as column names. Then, we bind the list elements to a single data.frame with bind_rows

library(dplyr)
library(purrr)
keep(combined_list, ~  all(c("name", 'age') %in% names(.x))) %>% 
     bind_rows

-output

#     name age
#1  Alice  23
#2    Bob  41
#3 Claire  50
#4  David  32

---

Or another option is to bind all the datasets together, then select only the relevant columns and remove the NA rows

bind_rows(combined_list) %>% 
    select(name, age) %>% 
    na.omit

---

In base R , we can use Filter with rbind inside do.call

do.call(rbind, Filter(function(x) 
      all(c("name", "age") %in% names(x)), combined_list))
#    name age
#1  Alice  23
#2    Bob  41
#3 Claire  50
#4  David  32

data

combined_list <- list(df1,df2,df3,df4)