Convert a Binary Sample in Number

Question

I have a binary sample like this :

Z = c(0,0,0,1,0,1,1,1,0,1,0,0,1,0,1,0,1,1,1,0,1,0,1,0)

I would like to convert all the sequences of the length 4 in number, i.e :

I need to read my original binary sample and convert all the possible sequence of length 4 into numbers.

Example: The sequence 0000 will be 1, the sequence 0001 will be 2, the sequence 0010 will be 3, ..., the sequence 1111 will be 16.

The expected output should be a new sample formed by the numbers 1,2,3,...16 having the same length as the original sample :

Z = c(0,0,0,1,0,1,1,1,0,1,0,0,1,0,1,0,1,1,1,0,1,0,1,0)

Z1 = c(2,3,6,12,8,15,14,11,5,10,3,11,6,12,8,15,14,11,6,11)

How can I do that in R ?

Answer 1

Try :

z<-c(0,0,0,1,0,1,1,1,0,1,0,0,1,0,1,0,1,1,1,0,1,0,1,0)
y<-as.character(z)
z1<-sapply(1:(length(y)-3),function(x){strtoi(paste(y[x:(x+3)],collapse=''),2)+1})

[1]  2  3  6 12  8 15 14 11  5 10  3  6 11  6 12  8 15 14 11  6 11

The code works like this :

• convert the vector z as a character vector (y)
• for the length of y minus 3 (because each number requires 4 digits), apply the strtoi function

The strtoi function convert the number by specifying the base of the input number (here, 2 because it's binary). We add 1 because in binary 0000 equals 0 and not 1.

Note: the conversion to character is optional, you can directly do

sapply(1:(length(z)-3),function(x){strtoi(paste(z[x:(x+3)],collapse=''),2)+1})

it will also be faster to use vapply :

vapply(1:(length(z)-3),function(x){strtoi(paste(z[x:(x+3)],collapse=''),2)+1},FUN.VALUE=1)

Unit: microseconds
   expr     min      lq     mean   median      uq     max neval cld
 vapply 206.866 209.111 214.3936 210.0735 211.356 338.362   100  a 
 sapply 230.278 231.882 234.0249 232.8440 234.128 273.897   100   b

Answer 2

Here is another approach:

Z <- c(0,0,0,1,0,1,1,1,0,1,0,0,1,0,1,0,1,1,1,0,1,0,1,0)
Z.tmp <- embed(Z,4)
Z1 <- as.vector(Z.tmp %*% c(1,2,4,8) + 1)