Jash Shah Jash Shah - 1 month ago 12
R Question

Generate n distinct values x times in R

I would like to create a vector that has distinct values from 1 to 20 thirty times but not uniformly.

For example:
There can be four counts of 1, one count of 2, two counts of 3 etc. But the counts of each number must add up to thirty and there must be 20 distinct values.

I tried:

set.seed(3)
sample(x = 1:20, size = 30, replace = TRUE)


But it does not always give all the values from 1 to 20. Some values are returned a higher number of times and some values are not returned at all.

I would like to create a vector that has all distinct values and the numbers have to necessarily be integers.

Answer

You can do it in three times:

  • generate a size-20 sample without replacements : you have every values 1 time

  • generate a size-10 sample with replacements

  • sample the two samples

Here is the result

a <- sample(1:20, 20)
b <- sample(1:20, 10, replace = TRUE)
result <- sample(c(a, b), 30)

# result
#  [1]  1 10 20 11 16 12  9  8 20  4 15  2  7  5 19 18  6 13 14 17 11  5  1  7  4 19  6 16  3  3

# table(result) # every value appear at least one time
# result
#  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 
#  2  1  2  2  2  2  2  1  1  1  2  1  1  1  1  2  1  1  2  2 

Note that you can do it with a one-liner :

sample(c(sample(1:20, 20), sample(1:20, 10, replace = TRUE)), 30)

# [1]  4 13 15 20  6  5  9 11 11 14 17  1 10  9  3 10 11 12 18 17  8  7 18 12 19 16  2 13 13  4

Thanks to James's comment, you can use a faster solution:

sample(c(1:20,sample(20,10,replace=TRUE)))

Here is the microbenchmark comparison:

# Unit: relative
#     expr      min       lq     mean   median     uq       max neval
#  etienne 1.727202 1.538411 1.529077 1.571341 1.5998 0.6855444  1000
#    james 1.000000 1.000000 1.000000 1.000000 1.0000 1.0000000  1000
Comments