subset of a subset of a list

问题

In R, I have a list, composed of 12 sublists, each themselves composed of 5 subsublists as follow

lists and sublists

In this example, I want to extract the info "MSD", for each of these sublists.

I can extract the level "statistics" for each using

lapply(letters, '[[', "statistics")

That worked well. It gave me all the values contained in the sublist "statistics", for each list However, I want to go one level down of that, as I am not interested in the other data such as MSerror, Df, ..... Only MSD

I tried

lapply(letters, '[[', "statistics", "MSD")

and many others without success.

If I wanted only the first sublist, it will work with

letters[[1]][["statistics"]][["MSD"]]

but then, I have to do:

letters[[1]][["statistics"]][["MSD"]]
letters[[2]][["statistics"]][["MSD"]]
letters[[3]][["statistics"]][["MSD"]]

which I want to avoid for a matter of time.

thanks for your help.

回答1:

We can use a lambda/anonymous function

lapply(letters, function(x) x[["statistics"]][["MSD"]])

The benefit of this function is that if we have multiple nested elements, we don't have to call n lapply and should be faster

Or use map

library(tidyverse)
map(letters, ~ .x[["statistics"]][["MSD"]])

Also, regarding the claim that this wouldn't work if there are not some elements in the list,

set.seed(24)
lst1 <- replicate(3,  list(statistics = list(MSD = rnorm(20))))
names(lst1)[2] <- "Hello"

It is true that it wouldn't work. However, it wouldn't work with the solution claimed to be work as well.

回答2:

One option is a nested lapply:

lapply(lapply(letters, '[[', "statistics"),`[[`,"MSD")

The benefit of this over @akrun's approach, which may not be relevant in this case, is that it will work even if some of the elements within the list don't have an $statistics$MSD item.

来源：https://stackoverflow.com/questions/53524769/subset-of-a-subset-of-a-list

标签

list

sublist