使用向量中的最后一个非空值填充空值

Question

I would like to fill in the missing values (not NA , just '' !) in a vector with the value before it. For example, if I have a vector defined as

vec <- c('Titanic', '', '', '', 'Donnie Darko', '', '', 'Twin Peaks', 
         'American Hustle', '')

my output vector would be

'Titanic', 'Titanic', 'Titanic', 'Titanic', 'Donnie Darko', 'Donnie Darko', 
'Donnie Darko', 'Twin Peaks', 'American Hustle', 'American Hustle'

How can I achieve this?

Answer 1

Here is a two-liner with nzchar and subsetting that should be quite efficient.

# get logical vector of elements with non-empty character elements
notMissings <- nzchar(movies)
# fill in missing values
movies[notMissings][cumsum(notMissings)]
 [1] "Titanic"         "Titanic"         "Titanic"         "Titanic"        
 [5] "Donnie Darko"    "Donnie Darko"    "Donnie Darko"    "Twin Peaks"     
 [9] "American Hustle" "American Hustle"

---

Here is a second method using rle .

# get run length encodings
temp <- rle(movies)
# get missing values    
missings <- nchar(temp$values) == 0
# fill in missing values
temp$values[missings] <- temp$values[which(missings) - 1]

# expand
inverse.rle(temp)
 [1] "Titanic"         "Titanic"         "Titanic"         "Titanic"        
 [5] "Donnie Darko"    "Donnie Darko"    "Donnie Darko"    "Twin Peaks"     
 [9] "American Hustle" "American Hustle"

Note that this second method will throw an error if the first element is the empty character, ''.

data

movies <- c('Titanic', '', '', '', 'Donnie Darko', '', '', 'Twin Peaks',
            'American Hustle', '')

Answer 2

Using Reduce in base R where vec is your vector:

Reduce(function(x,y) ifelse(y=="", x, y), vec, accumulate=TRUE)

#[1] "Titanic"       "Titanic"       "Titanic"       "Titanic"         #"Donnie Darko"   
#[6] "Donnie Darko"  "Donnie Darko"  "Twin Peaks"    "American Hustle" #"American Hustle"

Answer 3

Or we can use na.locf from zoo :

library(zoo)
vec <- c('Titanic', '', '', '', 'Donnie Darko', '', '', 'Twin Peaks', 'American Hustle', '')
vec[which(vec == "")] <- NA
na.locf(vec)

#  [1] "Titanic"         "Titanic"         "Titanic"         "Titanic" "Donnie Darko"    "Donnie Darko" 
#  [7] "Donnie Darko"    "Twin Peaks"      "American Hustle" "American Hustle"

Answer 4

We can also use

unlist(tapply(movies, cumsum(movies !=""), FUN = 
      function(x) rep(x[1], length(x))), use.names = FALSE)
#[1] "Titanic"         "Titanic"         "Titanic"         "Titanic"         "Donnie Darko"    "Donnie Darko"    "Donnie Darko"    "Twin Peaks"     
#[9] "American Hustle" "American Hustle"

data

movies <- c('Titanic', '', '', '', 'Donnie Darko', '', '', 'Twin Peaks', 
          'American Hustle', '')