Means multiple columns by multiple groups [duplicate]

Question

I am trying to find the means, not including NAs, for multiple columns withing a dataframe by multiple groups

airquality <- data.frame(City = c("CityA", "CityA","CityA",
                                  "CityB","CityB","CityB",
                                  "CityC", "CityC"),
                         year = c("1990", "2000", "2010", "1990", 
                                  "2000", "2010", "2000", "2010"),
                         month = c("June", "July", "August",
                                   "June", "July", "August",
                                   "June", "August"),
                         PM10 = c(runif(3), rnorm(5)),
                         PM25 = c(runif(3), rnorm(5)),
                         Ozone = c(runif(3), rnorm(5)),
                         CO2 = c(runif(3), rnorm(5)))
airquality

So I get a list of the names with the number so I know which columns to select:

nam<-names(airquality)
namelist <- data.frame(matrix(t(nam)));namelist

I want to calculate the mean by City and Year for PM25, Ozone, and CO2. That means I need columns 1,2,4,6:7)

acast(datadf, year ~ city, mean, na.rm=TRUE)

But this is not really what I want because it includes the mean of something I do not need and it is not in a data frame format. I could convert it and then drop, but that seems like a very inefficient way to do it.

Is there a better way?

Answer 1

We can use dplyr with summarise_at to get mean of the concerned columns after grouping by the column of interest

library(dplyr)
airquality %>%
   group_by(City, year) %>% 
   summarise_at(vars("PM25", "Ozone", "CO2"), mean)

Or using the devel version of dplyr (version - ‘0.8.99.9000’)

airquality %>%
     group_by(City, year) %>%
     summarise(across(PM25:CO2, mean))