efficient way to create a list of lists

Question

I am using lapply and mapply to create a bunch of operations. My data is composed of lists.

For one of the calculations I am having some problems on how to use any of the apply functions. This is an example of what I want to do:

#list with a collection of objects, each object has a vector with the their 2 closest neighbors identification. 
list_a<-list(c(2,4),c(1,3),c(1,4),c(3,2))

#list with the obs of each point
list_obs<-list(runif(1000,1,2),runif(1000,0,2),runif(1000,0.5,2),runif(1000,0.1,1.5))

position n in list_a corresponds to the neigh. points of point n

position n in list_obs corresponds to observations of point n

What I want to do is to create a new list that will store in each position n the list of observations of the neighb. points:

#output
 out=list(list(list_obs[[2]],list_obs[[4]]),list(list_obs[[1]],list_obs[[3]]),list(list_obs[[1]],list_obs[[4]]),list(list_obs[[3]],list_obs[[2]]))
> str(out)
List of 4
 $ :List of 2
  ..$ : num [1:1000] 1.673 1.423 0.228 1.758 1.65 ...
  ..$ : num [1:1000] 0.679 1.341 0.148 0.867 0.724 ...
 $ :List of 2
  ..$ : num [1:1000] 1.25 1.5 1.58 1.54 1.6 ...
  ..$ : num [1:1000] 0.526 1.545 1.848 0.711 0.697 ...
 $ :List of 2
  ..$ : num [1:1000] 1.25 1.5 1.58 1.54 1.6 ...
  ..$ : num [1:1000] 0.679 1.341 0.148 0.867 0.724 ...
 $ :List of 2
  ..$ : num [1:1000] 0.526 1.545 1.848 0.711 0.697 ...
  ..$ : num [1:1000] 1.673 1.423 0.228 1.758 1.65 ...

Can anyone advise me on the best way to do this?

My data is huge, so a loop will take 2 much time to run. The data is composed of 3000 points (with 3000-1 neigh.), each point with +20.000 observations.

As you describe your data it could be in a matrix, 2 columns on left for neighbours, and 1000 columns of observations. That might be simpler. On loops it's a myth that for loops are (implicitly) slower than built-in like apply, mapply although the way you do it may lead to inefficiencies. some things like plyr maybe quicker.. example below.. you might try dput(head(somedata, 50)) for a reproducible example.. — Stephen Henderson
– Stephen Henderson, Commented Nov 27, 2013 at 7:46
I'm not aware there is a maximum matrix size other than memory, and I don't see why a list uses less memory or is easier to handle than a matrix of the same elements. If the data really is huge you might consider putting it in rsqlite and/or using the dplyr package. — Stephen Henderson
– Stephen Henderson, Commented Nov 27, 2013 at 8:07

Sven Hohenstein · Accepted Answer · 2013-11-27 08:21:48Z

1

Have you tried the following command?

lapply(list_a, function(x) list_obs[x])

answered Nov 27, 2013 at 8:21

Sven Hohenstein

82k17 gold badges150 silver badges173 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Troy · Accepted Answer · 2013-11-27 08:27:28Z

1

lapply(list_a,FUN=function(x)lapply(x,FUN=function(y)list_obs[y]))

but, as per comments, there might be a better way to do it

sorry saw that was just answered!

lapply(list_a,FUN=function(x)list_obs[x])

is better!

answered Nov 27, 2013 at 8:27

Troy

8,70134 silver badges33 bronze badges

Collectives™ on Stack Overflow

efficient way to create a list of lists

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related