Question

I have two dataframes, looking sort of like:

Source_name <- c("name1", "name2", "name3", "name4", "name5")
Target_name <- c("name10", "name11", "name12", "name13", "name14")
values <- c("asd", "213", "kahsd", "a9u", "oau92")
values2 <- c("asdd", "oau892", "kahsd", "213", "213")
dat <- cbind(Source_name, values)
daf <- cbind(Target_name, values2)

dat
Source_name     values 
[1,] "name1"     "asd"  
[2,] "name2"     "213"  
[3,] "name3"     "kahsd"
[4,] "name4"     "a9u"  
[5,] "name5"     "oau92"

daf
Target_name     values2  
[1,] "name10"    "asdd"   
[2,] "name11"    "oau892"  
[3,] "name12"    "kahsd"
[4,] "name13"    "213"   
[5,] "name14"    "213"    

Each value only occurs once in dat, but may occur more than once in daf (or not at all). I would like to record those values in dat that occur at most once in daf, as per the desired_output data.frame.

unique_values <- c( "asd", "kahsd", "a9u", "oau92")
Source_name <- c( "name1", "name3", "name4", "name5")
Target_name <- c( "NA", "name12", "NA", "NA")
desired_output <- data.frame(cbind(unique_values, Source_name, Target_name))

desired_output
     unique_values    Source_name Target_name   
 1         asd         name1       
 3         kahsd       name3       name12   
 2         a9u         name4    
 4         oau92       name5   

I imagine there’s an easy way to do this using apply or something, but Im stumped.

Answer 1

You could merge your two data.frames:

dd <- merge(dat, daf, all.x = TRUE, by.x = "values", by.y = "values2")
dd
#   values Source_name Target_name
# 1    213       name2      name13
# 2    213       name2      name14
# 3    a9u       name4        <NA>
# 4    asd       name1        <NA>
# 5  kahsd       name3      name12
# 6  oau92       name5        <NA>

Then remove rows with values that show up twice or more:

dd[unlist(Filter(function(x)length(x)<2, split(seq_len(nrow(dd)), dd$values))), ]
#   values Source_name Target_name
# 3    a9u       name4        <NA>
# 4    asd       name1        <NA>
# 5  kahsd       name3      name12
# 6  oau92       name5        <NA>

Or as @hadley pointed out in the comments (thanks!):

dd[ave(dd$values, dd$values, FUN = length) < 2, ]