Taking column mean over a list of data frames in R

Question

Here's what I'm trying to do. My data frame has a factor variable, "country", and I want to split the data frame based on country. Then, I want to take the column mean over every variable for every country's data frame.

Data here: https://github.com/pourque/country-data

I've done this so far...

myList <- split(df1, df1$country)
for(i in 1:length(myList)) {
aggregate <- mapply(myList[[i]][,-c(38:39)], colMeans)
}

(I'm not including the 38th and 39th columns because those are factors.)

I've read this (function over more than one list) , which makes me think mapply is the answer here...but I'm getting this error:

Error in match.fun(FUN) : 
'myList[[i]][, -c(38:39)]' is not a function, character or symbol 

Maybe I'm formatting it incorrectly?

Answer 1

It's straightforward in base R using aggregate without the need to split the data.frame into a list beforehand. Here's an example using the built-in iris data where you compute the mean of all variables except those in the first and second column by group of Species:

data(iris)
aggregate(. ~ Species, iris[-(1:2)], mean)
#     Species Petal.Length Petal.Width
#1     setosa        1.462       0.246
#2 versicolor        4.260       1.326
#3  virginica        5.552       2.026

The . inside aggregate is used to specify that you want to use all remaining columns of the data.frame except the grouping variable (Species in this case). And because you specify iris[-(1:2)] as input data, the first and second columns are not used either.

---

For your data, it should then be something like:

aggregate(. ~ country, df1[-c(38:39)], mean)

Answer 2

library(dplyr)

df1 %>%
    group_by(country) %>%
    select(-age, -gender) %>%
    summarise_each(funs(mean))