CODEWITHSUNDEEP
rdplyrplyr
user5057431
user5057431

Reputation:

alternative to dlply in R

My dataframe contains the following:

data$Value   data$Name
774          Name1
770          Name1
778          Name1
804          Name1
804          Name1
802          Name1
804          Name1
900          Name2
905          Name2
805          Name2
900          Name2
950          Name2
860          Name2
870          Name2
etc...       etc... for 100 Names

So, each Name has 7 Values associated to it. I would like to group this data frame by variable "Name", split it and returning the 7 values for each of those names as a list. This is the format of my desired output:

my_list$Name1 = 
[[1]]
[1] 774 770 778 804 804 802 804 

my_list$Name2 = 
[[1]]
[1] 900, 905, 805, 900, 950, 860, 870  

etc...

The easiest solution is to use plyr's dlply function:

my_list <- dlply(data, "Name", function(x) list(x$Value))

However, I would like to avoid using plyr. What would be a good alternative? I have considered splitting my data the following way: 

splits <- function(x) { 
  y <- subset(data, select = c(Name, Value))
  splits <- split(y, y$Name)
  return(splits)
}

my_list <- splits(data)

However, this still returns me a list in the following format:

      Value   Name
      (dbl)   (chr)
1      774    Name1
2      770    Name1
3      778    Name1
4      804    Name1
5      804    Name1
6      802    Name1
7      804    Name1

UPDATE: SOLUTION:

my_list <- lapply(split(data$Value, data$Name), list)

Upvotes: 4

Views: 915

Answers (1)

Sotos
Sotos

Reputation: 51592

One way to do it,

y <- subset(data, select = c(Name, Value))
list <- split(y, y$Name)

unlist(sapply(list, '[', 1))

or

unlist(unname(sapply(list, '[', 1)))

or simply (compliments of @docendo discimus)

unlist(unname(split(y$Value, y$Name)))

Regarding your comment, then you can do,

unname(split(y$Value, y$Name))
#[[1]]
#[1] 774 770 778 804 804 802 804

#[[2]]
#[1] 900 905 805 900 950 860 870

Upvotes: 3

Related Questions