Mean of one vector of list of lists-CodePudding

I have following data set

myList <- split(x = ChickWeight, f = ChickWeight$Diet)

and i want to calculate the average of weights by list, i.e four different averages. one possible solution is

a<-lapply(myList, `[[`, 1)
lapply(a, mean)

But is it possible if i can have mean function in "a". i.e

a<-lapply(myList, `[[`, 1, mean)

CodePudding user response：

Use an anonymous function -

lapply(myList, function(x) mean(x$weight))
#Also by position of the column
#lapply(myList, function(x) mean(x[[1]]))

#$`1`
#[1] 102.6455

#$`2`
#[1] 122.6167

#$`3`
#[1] 142.95

#$`4`
#[1] 135.2627

If the dataset is not already splitted you can use aggregate.

aggregate(weight~Diet, ChickWeight, mean)

CodePudding user response：

You could define a function colMean to use in lapply.

colMean <- function(x, col) colMeans(x[, col, drop=FALSE])

lapply(myList, colMean, col='weight')  ## also: `col=1`
# $`1`
# weight 
# 102.6455 
# 
# $`2`
# weight 
# 122.6167 
# 
# $`3`
# weight 
# 142.95 
# 
# $`4`
# weight 
# 135.2627

Also works for multiple columns.

lapply(myList, colMean, col=c('weight', 'Time')
# $`1`
# weight      Time 
# 102.64545  10.48182 
# 
# $`2`
# weight      Time 
# 122.61667  10.91667 
# 
# $`3`
# weight      Time 
# 142.95000  10.91667 
# 
# $`4`
# weight      Time 
# 135.26271  10.75424