Extract unique strings from a factor string variable

Question

I have a variable which contains the actor names.

(actor=structure(c(4L, 1L, 6L, 2L, 5L, 3L), .Label = c("Christian Bale, Tom Hardy, Anne Hathaway, Gary Oldman", 
"Jamie Foxx, Christoph Waltz, Leonardo DiCaprio, Kerry Washington", 
"Jennifer Lawrence, Josh Hutcherson, Liam Hemsworth, Stanley Tucci", 
"Leonardo DiCaprio, Joseph Gordon-Levitt, Ellen Page, Ken Watanabe", 
"Leonardo DiCaprio, Mark Ruffalo, Ben Kingsley, Max von Sydow", 
"Robert Downey Jr., Chris Evans, Scarlett Johansson, Jeremy Renner"
), class = "factor"))
# [1] Leonardo DiCaprio, Joseph Gordon-Levitt, Ellen Page, Ken Watanabe
# [2] Christian Bale, Tom Hardy, Anne Hathaway, Gary Oldman            
# [3] Robert Downey Jr., Chris Evans, Scarlett Johansson, Jeremy Renner
# [4] Jamie Foxx, Christoph Waltz, Leonardo DiCaprio, Kerry Washington 
# [5] Leonardo DiCaprio, Mark Ruffalo, Ben Kingsley, Max von Sydow     
# [6] Jennifer Lawrence, Josh Hutcherson, Liam Hemsworth, Stanley Tucci
# 6 Levels: Christian Bale, Tom Hardy, Anne Hathaway, Gary Oldman ...

I want to extract all the complete actor names from it (name + surname) and make them columns in an output matrix.

Answer 1

If you wanted to extract the unique names of actors, you can get the indicated actors with the as.character function, split it on the commas with strsplit, combine together all vectors in the resulting list with unlist, and grab the unique names with unique:

(all.actors <- unique(unlist(strsplit(as.character(actor), ", "))))
#  [1] "Leonardo DiCaprio"    "Joseph Gordon-Levitt" "Ellen Page"           "Ken Watanabe"        
#  [5] "Christian Bale"       "Tom Hardy"            "Anne Hathaway"        "Gary Oldman"         
#  [9] "Robert Downey Jr."    "Chris Evans"          "Scarlett Johansson"   "Jeremy Renner"       
# [13] "Jamie Foxx"           "Christoph Waltz"      "Kerry Washington"     "Mark Ruffalo"        
# [17] "Ben Kingsley"         "Max von Sydow"        "Jennifer Lawrence"    "Josh Hutcherson"     
# [21] "Liam Hemsworth"       "Stanley Tucci"    

By using as.character(actor), this code uses only the actors that show up in the the factor actor, even if that factor has many more levels that are unused. If you use levels(actor) instead, you will get all the actors in the factor's levels, regardless of whether they are used in actors. You can use whichever you prefer when defining all.actors.

If you wanted a matrix indicating the inclusion of each actor in each element of actor, you could then do

mat <- sapply(strsplit(as.character(actor), ", "), function(x) all.actors %in% x)
row.names(mat) <- all.actors
mat
#                       [,1]  [,2]  [,3]  [,4]  [,5]  [,6]
# Leonardo DiCaprio     TRUE FALSE FALSE  TRUE  TRUE FALSE
# Joseph Gordon-Levitt  TRUE FALSE FALSE FALSE FALSE FALSE
# Ellen Page            TRUE FALSE FALSE FALSE FALSE FALSE
# Ken Watanabe          TRUE FALSE FALSE FALSE FALSE FALSE
# Christian Bale       FALSE  TRUE FALSE FALSE FALSE FALSE
# Tom Hardy            FALSE  TRUE FALSE FALSE FALSE FALSE
# Anne Hathaway        FALSE  TRUE FALSE FALSE FALSE FALSE
# Gary Oldman          FALSE  TRUE FALSE FALSE FALSE FALSE
# Robert Downey Jr.    FALSE FALSE  TRUE FALSE FALSE FALSE
# Chris Evans          FALSE FALSE  TRUE FALSE FALSE FALSE
# Scarlett Johansson   FALSE FALSE  TRUE FALSE FALSE FALSE
# Jeremy Renner        FALSE FALSE  TRUE FALSE FALSE FALSE
# Jamie Foxx           FALSE FALSE FALSE  TRUE FALSE FALSE
# Christoph Waltz      FALSE FALSE FALSE  TRUE FALSE FALSE
# Kerry Washington     FALSE FALSE FALSE  TRUE FALSE FALSE
# Mark Ruffalo         FALSE FALSE FALSE FALSE  TRUE FALSE
# Ben Kingsley         FALSE FALSE FALSE FALSE  TRUE FALSE
# Max von Sydow        FALSE FALSE FALSE FALSE  TRUE FALSE
# Jennifer Lawrence    FALSE FALSE FALSE FALSE FALSE  TRUE
# Josh Hutcherson      FALSE FALSE FALSE FALSE FALSE  TRUE
# Liam Hemsworth       FALSE FALSE FALSE FALSE FALSE  TRUE
# Stanley Tucci        FALSE FALSE FALSE FALSE FALSE  TRUE